Journal Club with Dr. Peter Attia | Effects of Light & Dark on Mental Health & Treatments for Cancer

00:00:00.000 | - Welcome to the Huberman Lab Podcast,

00:00:02.280 | where we discuss science and science-based tools

00:00:04.880 | for everyday life.

00:00:05.920 | I'm Andrew Huberman,

00:00:10.360 | and I'm a professor of neurobiology and ophthalmology

00:00:13.580 | at Stanford School of Medicine.

00:00:15.820 | Today marks the second episode in our Journal Club series

00:00:18.960 | with myself and Dr. Peter Attia.

00:00:21.500 | Dr. Peter Attia, as many of you know,

00:00:23.320 | is a medical doctor who is a world expert

00:00:26.460 | in all things healthspan and lifespan.

00:00:29.280 | He is the author of the bestselling book, "Outlive,"

00:00:32.080 | as well as the host of his own terrific podcast, "The Drive."

00:00:35.520 | For today's episode,

00:00:36.460 | Peter and I each select a different paper to share with you.

00:00:39.800 | We selected these papers

00:00:40.840 | because we feel they are both extremely interesting

00:00:44.180 | and extremely actionable.

00:00:45.880 | First, I present a paper that is about how light exposure

00:00:49.840 | during the morning and daytime,

00:00:51.540 | as well as dark exposure at night,

00:00:54.240 | each have independent and positive effects on mental health,

00:00:57.800 | as well as the ability to reduce the symptoms

00:01:00.840 | of many different mental health disorders.

00:01:03.600 | Now, I've talked before on this podcast and elsewhere

00:01:05.980 | about the key importance of seeing morning sunlight,

00:01:09.160 | as well as trying to be in dim light at night.

00:01:11.680 | However, the data presented in the paper today

00:01:14.160 | really expands on that by identifying the key importance

00:01:17.720 | of not just morning sunlight,

00:01:19.480 | but getting bright light in one's eyes

00:01:21.440 | as much as is safely possible throughout the entire day

00:01:24.200 | and a separate additive effect of being in as much darkness

00:01:29.200 | at night as possible.

00:01:30.880 | I describe the data in a lot of detail,

00:01:32.720 | although you do not need a background in biology

00:01:34.880 | in order to understand that discussion.

00:01:36.880 | And there's a key takeaway,

00:01:38.840 | which is that if you can't get enough light in your eyes

00:01:41.760 | during the daytime, you would be well-advised

00:01:44.540 | to get as much darkness exposure at night.

00:01:47.720 | In other words, light and dark have independent

00:01:50.120 | and additive effects on mental health.

00:01:52.600 | And during today's discussion,

00:01:53.700 | you'll learn exactly how to apply light exposure

00:01:56.160 | and dark exposure in order to get those benefits.

00:01:59.780 | Then Peter presents a paper

00:02:01.080 | about novel treatments for cancer.

00:02:03.620 | I must say it's an extremely important conversation

00:02:05.880 | that everybody, regardless of whether or not

00:02:08.520 | you may have had cancer or know somebody who's had cancer,

00:02:11.820 | ought to listen to.

00:02:13.160 | He highlights the current technology of cancer treatments,

00:02:16.100 | as well as the future technology of cancer treatments,

00:02:19.320 | and the key role that the immune system

00:02:21.880 | and the autoimmune system play in treatments for cancer.

00:02:26.120 | I assure you that by the end

00:02:27.300 | of today's Journal Club episode,

00:02:29.140 | you will have learned a ton of new information

00:02:32.680 | about light and dark and mental health,

00:02:35.280 | as well as cancer and the immune system

00:02:37.620 | and treatments for curing cancer.

00:02:39.800 | Before we begin, I'd like to emphasize that this podcast

00:02:42.300 | is separate from my teaching and research roles at Stanford.

00:02:45.000 | It is, however, part of my desire and effort

00:02:47.060 | to bring zero cost to consumer information

00:02:48.940 | about science and science-related tools

00:02:50.940 | to the general public.

00:02:52.420 | In keeping with that theme,

00:02:53.540 | I'd like to thank the sponsors of today's podcast.

00:02:56.560 | Our first sponsor is Eight Sleep.

00:02:58.560 | Eight Sleep makes smart mattress covers

00:03:00.220 | with cooling, heating, and sleep tracking capacity.

00:03:03.100 | I've spoken many times before in this podcast

00:03:05.100 | about the fact that sleep is the foundation

00:03:07.160 | of mental health, physical health, and performance.

00:03:09.760 | Now, a key component of getting a great night's sleep

00:03:12.420 | is that in order to fall and stay deeply asleep,

00:03:15.020 | your body temperature actually has to drop

00:03:16.840 | by about one to three degrees.

00:03:18.460 | And in order to wake up feeling refreshed and energized,

00:03:21.500 | your body temperature actually has to increase

00:03:23.520 | by about one to three degrees.

00:03:25.300 | One of the best ways to make sure

00:03:26.600 | that those temperature changes occur

00:03:28.120 | at the appropriate times, at the beginning and throughout,

00:03:30.880 | and at the end of your night when you wake up,

00:03:33.020 | is to control the temperature of your sleeping environment.

00:03:35.580 | And that's what Eight Sleep allows you to do.

00:03:37.680 | It allows you to program the temperature of your mattress

00:03:40.040 | and sleeping environment such that you fall

00:03:42.140 | and stay deeply asleep easily and wake up each morning

00:03:45.540 | feeling incredibly refreshed and energized.

00:03:47.980 | I've been sleeping on an Eight Sleep mattress cover

00:03:49.740 | for almost three years now,

00:03:51.100 | and it has dramatically improved the quality of my sleep.

00:03:53.940 | So much so that when I travel and I'm at a hotel

00:03:56.740 | or an Airbnb and I don't have access to my Eight Sleep,

00:03:59.300 | I very much look forward to getting home

00:04:00.740 | because my sleep is always better

00:04:02.600 | when I sleep on my Eight Sleep mattress cover.

00:04:04.660 | If you'd like to try Eight Sleep,

00:04:05.980 | you can go to eightsleep.com/huberman

00:04:08.920 | to get $150 off their Pod3 mattress cover.

00:04:12.080 | Eight Sleep currently ships in the USA, Canada, UK,

00:04:15.040 | select countries in the EU, and Australia.

00:04:17.260 | Again, that's eightsleep.com/huberman.

00:04:20.680 | Today's episode is also brought to us by BetterHelp.

00:04:23.800 | BetterHelp offers professional therapy

00:04:25.580 | with a licensed therapist carried out online.

00:04:28.480 | I've been going to therapy for well over 30 years.

00:04:31.180 | Initially, I didn't have a choice.

00:04:32.660 | It was a condition of being allowed to stay in school,

00:04:35.100 | but pretty soon I realized

00:04:36.700 | that therapy is extremely valuable.

00:04:38.520 | In fact, I consider doing regular therapy

00:04:40.820 | just as important as getting regular exercise,

00:04:43.580 | including cardiovascular exercise and resistance training,

00:04:45.900 | which of course I also do every week.

00:04:47.980 | The reason I know therapy is so valuable

00:04:49.940 | is that if you can find a therapist

00:04:51.680 | with whom you can develop a really good rapport,

00:04:54.280 | you not only get terrific support

00:04:56.580 | for some of the challenges in your life,

00:04:58.520 | but you also can derive tremendous insights

00:05:01.140 | from that therapy.

00:05:02.260 | Insights that can allow you to better

00:05:04.020 | not just your emotional life and your relationship life,

00:05:06.700 | but of course also the relationship to yourself

00:05:08.980 | and to your professional life, to all sorts of career goals.

00:05:12.300 | In fact, I see therapy as one of the key components

00:05:14.540 | for meshing together all aspects of one's life

00:05:17.260 | and being able to really direct one's focus and attention

00:05:20.180 | toward what really matters.

00:05:21.900 | If you'd like to try BetterHelp,

00:05:23.100 | go to betterhelp.com/huberman

00:05:25.580 | to get 10% off your first month.

00:05:27.220 | Again, that's betterhelp.com/huberman.

00:05:30.140 | Today's episode is also brought to us by Juve.

00:05:33.020 | Juve makes medical grade red light therapy devices.

00:05:36.340 | Now, if there's one thing I've consistently emphasized

00:05:38.480 | on this podcast, it's the incredible role

00:05:40.780 | that light can have on our biology.

00:05:43.140 | Of course, I'm always telling people

00:05:44.580 | that they should get sunlight in their eyes

00:05:46.180 | as soon as possible after waking on as many days

00:05:48.620 | of their life as possible

00:05:50.100 | for sake of setting circadian rhythm,

00:05:51.600 | daytime mood focus and alertness, and improve sleep.

00:05:54.740 | Now, in addition to sunlight,

00:05:56.520 | red light and near infrared light

00:05:58.300 | has been shown to have positive effects

00:05:59.740 | on improving numerous aspects of cellular and organ health,

00:06:03.100 | including faster muscle recovery,

00:06:05.500 | improved skin health and wound healing,

00:06:07.620 | even improvements in acne, or that is removal of acne,

00:06:11.360 | reducing pain and inflammation,

00:06:13.120 | improving mitochondrial function,

00:06:14.980 | and even improving vision itself.

00:06:17.500 | What sets Juve apart

00:06:18.500 | and why it's my preferred red light therapy device

00:06:20.860 | is that it has clinically proven wavelengths,

00:06:23.760 | meaning it uses specific wavelengths of red light

00:06:26.540 | and near infrared light in combination

00:06:28.800 | that trigger the optimal cellular adaptations.

00:06:31.620 | Personally, I use the handheld Juve every day.

00:06:33.760 | The handheld Juve is about the size

00:06:35.020 | of a thick piece of toast.

00:06:36.500 | And I also own a Juve panel

00:06:38.080 | that allows for full body exposure.

00:06:39.460 | And I use that one approximately five times per week

00:06:42.780 | for about 10 to 15 minutes per session.

00:06:45.040 | If you'd like to try Juve,

00:06:46.160 | you can go to Juve, spelled J-O-O-V-V.com/huberman.

00:06:51.040 | Again, that's juve.com/huberman.

00:06:53.840 | For this month only, January, 2024,

00:06:56.880 | Juve is offering exclusive discounts

00:06:59.080 | to Huberman Lab listeners

00:07:00.600 | with up to $500 off select Juve products.

00:07:03.500 | Again, that's juve, spelled J-O-O-V-V.com/huberman

00:07:07.680 | to get up to $500 off select Juve products.

00:07:10.720 | And now for my discussion with Dr. Peter Attia.

00:07:13.960 | Andrew, great to have you here for Journal Club number two.

00:07:17.900 | I'm already confident this is gonna become a regular for us.

00:07:21.040 | I'm excited.

00:07:21.880 | I really enjoy this because I get to pick papers.

00:07:25.720 | I'm really excited about it.

00:07:26.720 | I get to hear papers that you're excited about.

00:07:29.000 | And we get to sharpen our skills

00:07:31.960 | at reading and sharing data.

00:07:33.700 | And people listening can do that as well.

00:07:37.160 | So last time I went first,

00:07:38.640 | so I think I'm gonna put you on the hot seat first

00:07:41.740 | and have you go first and I'll follow you.

00:07:44.180 | Okay.

00:07:45.800 | Well, I'm really excited about this paper

00:07:47.880 | for a number of reasons.

00:07:50.160 | First of all, it, at least by my read,

00:07:53.480 | is a very powerful paper in the sense

00:07:55.960 | that it examined light exposure behavior

00:07:59.840 | as well as dark exposure behavior.

00:08:02.260 | And that's gonna be an important point

00:08:03.880 | in more than 85,000 people as part of this cohort in the UK.

00:08:08.880 | I'll just mention a couple of things

00:08:11.900 | to give people background,

00:08:12.880 | and I'll keep this relatively brief.

00:08:15.360 | First of all, there's a longstanding interest

00:08:18.000 | in the relationship between light and mental health

00:08:20.680 | and physical health.

00:08:21.800 | And we can throw up some very well-agreed-upon bullet points.

00:08:26.800 | First of all, there is such a thing

00:08:28.640 | as seasonal affective disorder.

00:08:30.440 | It doesn't just impact people

00:08:32.180 | living at really northern locations,

00:08:34.560 | but basically there's a correlation

00:08:36.480 | between day length and mood and mental health,

00:08:40.380 | such that for many people, not all, but for many people,

00:08:44.600 | when days are longer in the spring and summer,

00:08:46.960 | they feel better.

00:08:48.240 | They report fewer depressive symptoms.

00:08:50.940 | And conversely, when days are shorter,

00:08:53.940 | significantly more people report

00:08:55.920 | feeling lower mood and affect.

00:08:58.000 | Okay, so there's a longstanding treatment

00:09:01.080 | for seasonal affective disorder,

00:09:02.900 | which is to give people exposure to very bright light,

00:09:06.600 | especially in the morning.

00:09:07.900 | The way that that's normally accomplished

00:09:10.900 | is with these SAD lamps, seasonal affective disorder lamps.

00:09:14.520 | And those lamps are basically bright,

00:09:16.860 | meaning more than 10,000 lux lights

00:09:20.340 | that they place on their kitchen counter

00:09:22.420 | or at their table in the morning or in their office.

00:09:26.060 | So they're getting a lot of bright light.

00:09:28.500 | That has proven to be fairly effective

00:09:30.620 | for the treatment of seasonal affective disorder.

00:09:33.120 | What's less understood is how light exposure

00:09:35.360 | in the middle of the night

00:09:37.200 | can negatively impact mood and health.

00:09:40.400 | And so where we are headed with this

00:09:43.000 | is that there seems to be,

00:09:44.300 | based on the conclusions of this new study,

00:09:46.720 | a powerful and independent role

00:09:50.900 | of both daytime light exposure

00:09:52.800 | and nighttime dark exposure for mental health.

00:09:56.840 | Now, a couple of other key points,

00:09:58.960 | the biological mechanisms for all this

00:10:00.680 | are really well-established.

00:10:02.520 | There's a set of cells in the neural retina,

00:10:04.980 | which aligns the back of your eye.

00:10:06.500 | They're sometimes called

00:10:07.340 | intrinsically photosensitive retinal ganglion cells,

00:10:10.160 | or sometimes called melanopsin retinal ganglion cells.

00:10:13.560 | We'll talk about those in a bit of detail in a moment.

00:10:16.600 | It's well-known that those cells are the ones

00:10:18.900 | that respond to two different types of light input,

00:10:23.480 | not one, but two different types of light input,

00:10:25.260 | and send information to the hypothalamus,

00:10:27.320 | where your master circadian clock resides,

00:10:29.400 | and then your master circadian clock

00:10:31.080 | sends out secretory signals, so peptides, hormones,

00:10:35.340 | but also neural signals to the brain and body,

00:10:37.320 | and say, "Hey, now it's daytime, now it's nighttime.

00:10:40.980 | Be awake, be asleep."

00:10:42.500 | But it goes way beyond that.

00:10:43.940 | These melanopsin intrinsically photosensitive

00:10:46.240 | retinal ganglion cells, we know,

00:10:48.000 | also project to areas of the brain like the habenula,

00:10:51.000 | which can trigger negative affect, negative mood.

00:10:54.080 | They can trigger the release of dopamine

00:10:56.720 | or the suppression of dopamine, the release of serotonin,

00:10:59.320 | the suppression of serotonin.

00:11:01.400 | And so they're not just cells

00:11:03.620 | for setting your circadian clock.

00:11:05.000 | They also have a direct line, literally one synapse away,

00:11:09.340 | into the structures of the brain

00:11:10.800 | that we know powerfully control mood.

00:11:12.720 | So the mechanistic basis for all this is there.

00:11:15.320 | So there's just a couple of other key points to understand

00:11:17.760 | for people to really be able to digest

00:11:19.980 | the data in this paper fully.

00:11:22.040 | There are basically two types of stimuli

00:11:26.400 | that these cells respond to.

00:11:28.280 | One is very bright light, as we just talked about.

00:11:31.000 | That's why getting a lot of daytime sunlight

00:11:34.240 | is correlated with elevated mood.

00:11:36.340 | That's why looking at a 10,000 lux artificial lamp

00:11:39.060 | can offset seasonal affective disorder.

00:11:41.160 | - By the way, just a couple of questions on that.

00:11:43.540 | How many lux does the sun provide on a sunny day at noon?

00:11:47.440 | - Okay, great question.

00:11:48.760 | So if you're out in the sun with no cloud cover

00:11:53.160 | or minimal cloud cover in the middle of the day at noon,

00:11:56.840 | chances are it's over 100,000 lux.

00:12:00.320 | On a really bright day, could be 300,000 lux, okay?

00:12:06.880 | Most indoor environments,

00:12:09.980 | even though they might seem very bright,

00:12:12.300 | I like to think of your kind of like department store

00:12:15.800 | with the bright lights, believe it or not,

00:12:17.920 | that's probably only closer to 6,000 lux maximum

00:12:21.740 | and probably more like 4,000 lux.

00:12:24.520 | Most brightly lit indoor environments are not that bright

00:12:28.560 | when it comes down to total photon energy.

00:12:31.320 | Now, here's the interesting thing.

00:12:33.520 | On a cloudy day, when you're outside,

00:12:37.240 | it can be as bright as an average of 100,000 lux,

00:12:43.040 | but it won't seem that bright

00:12:45.020 | because you don't quote unquote see the sun,

00:12:47.700 | but it's also because when there's cloud cover,

00:12:50.640 | a lot of those long wavelengths of light,

00:12:53.540 | such as orange and red light, aren't coming through.

00:12:56.560 | However, and this is so important,

00:12:59.000 | the circadian clock, the suprachiasmatic nucleus,

00:13:02.420 | it sums photons.

00:13:04.900 | It's a photon summing system.

00:13:07.900 | So basically if you're outside in 8,000 lux,

00:13:12.080 | very overcast UK winter day,

00:13:15.540 | and you're walking around hopefully without sunglasses

00:13:18.440 | because sunglasses are gonna filter

00:13:19.760 | a lot of those photons out,

00:13:22.200 | your circadian clock is summing the photons.

00:13:25.400 | So it's an integration mechanism.

00:13:27.920 | It's not triggered in a moment.

00:13:29.760 | And actually the experiments of recording from these cells

00:13:32.120 | first done by David Berson at Brown

00:13:34.200 | were historic in the field of visual neuroscience

00:13:37.040 | when shown bright light

00:13:38.800 | on these intrinsically photosensitive cells,

00:13:41.400 | crank up the intensity of the light

00:13:42.800 | and the neurons would ramp up their membrane potential

00:13:47.900 | and then start spiking, firing action potentials,

00:13:51.040 | long trains of action potentials

00:13:52.600 | that have been shown to go on for hours.

00:13:55.440 | And so that's the signal that's propagating

00:13:57.520 | into the whole brain and body.

00:13:59.440 | Okay, so the important thing to understand

00:14:02.120 | is this is not a quick switch.

00:14:03.880 | That's why I suggest on non cloudy days, we'll call them,

00:14:08.560 | that people get 10 minutes or so of sunlight in their eyes

00:14:12.400 | in the early part of the day,

00:14:14.000 | another 10 minimum in the later part of the day,

00:14:16.980 | as much sunlight in their eyes

00:14:18.240 | as they safely can throughout the day.

00:14:19.720 | But since you're a physician, I should just,

00:14:22.680 | and you had a guest on talking about this recently,

00:14:24.480 | when the sun is low in the sky, low solar angle sunlight,

00:14:27.440 | that's really the key time

00:14:28.440 | for reasons we'll talk about in a moment.

00:14:30.000 | And when the sun is low in the sky,

00:14:31.900 | you run very, very little risk of inducing cataract

00:14:35.480 | by looking in the general direction of the sun.

00:14:37.200 | You should still blink as needed to protect the eyes.

00:14:39.660 | It's when the sun is overhead

00:14:41.320 | and there's all those photons coming in quickly

00:14:44.320 | in a short period of time that you do have to be concerned

00:14:47.840 | about cataract and macular degeneration

00:14:50.660 | if you're getting too much daytime sunlight.

00:14:52.940 | So the idea is sunglasses in the middle of the day are fine,

00:14:55.300 | but you really should avoid using them

00:14:56.660 | in the early and later part of the day,

00:14:58.460 | unless you're driving into the sun

00:15:00.020 | and you need, you know, for safety reasons.

00:15:01.620 | - Another question, Andrew. - Yeah.

00:15:02.700 | - If a person is indoors, but they have large windows,

00:15:06.960 | so they're getting tons of sunlight into their space,

00:15:11.340 | they don't even need ambient indoor light,

00:15:13.820 | how much of the photons are making it through the glass?

00:15:16.300 | And how does that compare to this effect?

00:15:19.480 | - Yeah, in general,

00:15:21.140 | unless the light is coming directly through the window,

00:15:24.080 | most of the relevant wavelengths are filtered out.

00:15:27.680 | - In other words, if you can't see the sun

00:15:29.740 | through the window,

00:15:30.860 | even if sufficient light is being provided,

00:15:33.580 | that's insufficient to trigger this phenomenon?

00:15:35.400 | - That's right.

00:15:36.240 | However, if you have, you know, windows on your roof,

00:15:39.820 | which some people do, skylights,

00:15:41.540 | that makes the situation much, much better.

00:15:44.260 | In fact, the neurons that, in the eye,

00:15:46.940 | that signal to the circadian clock

00:15:49.120 | and these mood centers in the brain reside mainly

00:15:51.540 | in the bottom two thirds of the neural retina

00:15:54.340 | and are responsible for looking up, basically.

00:15:57.020 | They're gathering light from above.

00:15:59.220 | These cells are also very low resolution,

00:16:01.420 | so think of them as big pixels.

00:16:03.940 | They're not interested in patterns and edges and movement.

00:16:06.540 | They're interested in how much ambient light

00:16:08.260 | there happens to be.

00:16:09.360 | Now, keep in mind that this mechanism is perhaps

00:16:11.560 | the most well-conserved mechanism in cellular organisms.

00:16:16.280 | So there, and I'll use that as a way to frame up

00:16:18.780 | the four types of light that one needs to see

00:16:22.180 | every 24 hours for optimal health.

00:16:24.500 | And when I say optimal health,

00:16:25.960 | I really mean mental health and physical health,

00:16:27.860 | but we're gonna talk about mental health mainly today

00:16:29.720 | in this paper.

00:16:30.560 | There's an absolutely beautiful evolutionary story

00:16:35.160 | whereby single cell organisms, all the way to humans,

00:16:39.300 | dogs, rabbits, and everything in between

00:16:42.580 | have at least two conopsins,

00:16:45.180 | one that responds to short wavelength light,

00:16:47.040 | AKA blue light, and another one that responds

00:16:49.360 | to longer wavelength light, orange and red.

00:16:51.580 | So your dogs have this, we have this,

00:16:56.400 | and it's a comparison mechanism in these cells of the eye,

00:17:00.520 | these neurons of the eye.

00:17:01.380 | They compare contrast between blues and orange,

00:17:04.440 | or sometimes blues and reds and pinks,

00:17:06.060 | which are also all long wavelength light.

00:17:09.340 | There are two times of day when the sky is enriched

00:17:13.540 | with blues, oranges, pinks, and reds,

00:17:16.980 | and that's low solar angle sunlight at sunrise

00:17:20.440 | and in the evening.

00:17:21.480 | These cells are uniquely available to trigger

00:17:27.420 | the existence of those wavelengths of light

00:17:30.740 | early in the day and in the evening,

00:17:32.460 | not in the middle of the day.

00:17:33.820 | So these cells have these two cone photopigments

00:17:35.980 | and they say, "How much blue light is there?

00:17:37.800 | How much red light is there?" or orange light.

00:17:39.940 | And the subtraction between those two

00:17:42.300 | triggers the signal for them to fire the signal off

00:17:45.540 | to the circadian clock of the brain.

00:17:47.460 | And that's why I say,

00:17:48.460 | "Look at low solar angle sunlight early in the day."

00:17:50.940 | What that does is it,

00:17:53.300 | what we call it is phase advances the clock.

00:17:55.480 | This can get a little technical

00:17:56.540 | and we don't want to get too technical here,

00:17:57.880 | but think about pushing your kid on a swing.

00:18:00.080 | The period of that swing, the duration of that swing

00:18:04.700 | is a little bit longer than 12 hours, okay?

00:18:09.480 | So when you stand closer to the kid,

00:18:12.000 | so your kid swings back and you give it a push,

00:18:15.340 | you're shortening the period, right?

00:18:16.980 | You're not allowing the swing to come all the way up.

00:18:19.620 | That's what happens when you look at morning sunlight.

00:18:21.720 | You're advancing your circadian clock.

00:18:23.520 | Translated to English or non nerd speak,

00:18:26.980 | you're making it such that you will want to go to bed

00:18:29.440 | a little bit earlier and wake up a little bit earlier

00:18:32.360 | the next day.

00:18:34.360 | In the evening, when you view low solar angle sunlight,

00:18:37.280 | so in the afternoon setting sun or evening setting sun,

00:18:41.960 | you do the exact opposite.

00:18:43.400 | You're phase delaying the clock.

00:18:45.000 | It's the equivalent of your kid

00:18:46.120 | being at the very top of the arc.

00:18:49.000 | And so it's gone maybe 12 and a half hour,

00:18:51.480 | let's say 12 and a half hours

00:18:53.640 | is the duration of that swing.

00:18:55.520 | And you run up and you push them from behind

00:18:57.920 | and give them a little more push.

00:18:59.400 | That's the equivalent of making yourself

00:19:00.960 | stay up a little later and wake up a little later.

00:19:03.580 | These two signals average so that your clock stays stable.

00:19:07.880 | You don't drift,

00:19:08.720 | meaning you're not waking up earlier every single day

00:19:11.520 | or going to sleep later every single day.

00:19:13.480 | This is why it's important to view low solar angle sunlight

00:19:17.320 | in the morning and again in the evening

00:19:19.500 | as often as possible.

00:19:21.600 | And it's done by that readout of those two photo pigments.

00:19:25.120 | Now, midday sun, which contains its bright light,

00:19:30.120 | but you see it as white light,

00:19:31.680 | contains all of those wavelengths at equal intensity.

00:19:34.640 | So the middle of the day

00:19:35.780 | is the so-called circadian dead zone.

00:19:38.300 | In the middle of the day,

00:19:39.620 | bright light triggers the activation of the other ops

00:19:43.260 | and the melanopsin, which increases mood,

00:19:46.760 | increases feelings of wellbeing,

00:19:48.200 | has some other consequences.

00:19:49.320 | But you can't shift your circadian clock

00:19:50.980 | by viewing the sun in the middle of the day

00:19:52.840 | because it's in the circadian dead zone.

00:19:54.540 | It's the equivalent of pushing your kid on the swing

00:19:56.820 | when they're at the bottom of the arc.

00:19:59.060 | You can get a little bit more, but not much.

00:20:02.180 | And in biological terms, you get nothing.

00:20:04.860 | So this is why looking at sunlight

00:20:06.140 | in the middle of the day is great,

00:20:07.180 | but it's not gonna help anchor your sleep-wake cycle.

00:20:10.340 | And if you think about it, this is incredible, right?

00:20:12.780 | Every organism from single cells to us has this mechanism

00:20:17.280 | to know when the sun is rising and when the sun is setting.

00:20:19.680 | And it's a color comparison mechanism,

00:20:23.040 | which tells us that actually color vision evolved first,

00:20:25.940 | not for pattern vision, not for seeing beautiful sunsets

00:20:29.160 | and recognizing that's beautiful

00:20:30.540 | or paintings or things of that sort,

00:20:32.460 | but rather for setting the circadian clock.

00:20:34.580 | - Now, what if you only do one of these, Andrew?

00:20:36.080 | So what if you've got constant exposure

00:20:38.580 | to low morning light, but your job prevents you

00:20:42.680 | from doing the same in the evening or vice versa?

00:20:45.460 | - Yeah, a great question.

00:20:46.860 | Better to get the morning light

00:20:48.860 | because if you have to pick between low solar angle light

00:20:52.140 | earlier or later in the day.

00:20:53.860 | And keep in mind, if you miss a day, no big deal.

00:20:56.080 | It's a slow integrative mechanism

00:20:57.780 | averaging across the previous two or three days.

00:21:00.620 | But if you miss a day,

00:21:01.460 | you'll wanna get twice as much light in your eyes

00:21:03.620 | that next morning.

00:21:05.780 | The reason it's better to do in the morning

00:21:07.580 | as opposed to the evening,

00:21:08.680 | although best would be to do both, excuse me,

00:21:13.620 | is that most people are getting

00:21:15.420 | some artificial light exposure in the evening anyway.

00:21:18.020 | And here's the diabolical thing.

00:21:20.220 | Your retina is very insensitive to light

00:21:23.220 | early in the day.

00:21:24.060 | You need a lot of photons to trigger this mechanism

00:21:26.260 | early in the day.

00:21:27.220 | As the day goes on, retinal sensitivity increases

00:21:29.760 | and it takes very little light

00:21:31.020 | to shift your circadian clock late in the day.

00:21:33.820 | Keep in mind also that if you do see afternoon

00:21:36.460 | and evening sunlight,

00:21:37.700 | there's a beautiful study published in Science Reports,

00:21:41.620 | yes, Science Reports, two years ago,

00:21:44.260 | showing that that can partially offset

00:21:46.580 | the negative effects of artificial light exposure at night.

00:21:49.420 | I think of this as your Netflix inoculation.

00:21:52.140 | The amount of melatonin suppression

00:21:53.940 | from nighttime light exposure

00:21:55.840 | is halved by viewing evening setting sun.

00:21:59.540 | Now, keep in mind,

00:22:00.380 | you don't need to see the sun cross the horizon.

00:22:02.020 | It can just be when it's low solar angle.

00:22:03.820 | So you're looking for those yellow, blue,

00:22:05.700 | or blue, pink, blue-red contrasts.

00:22:08.220 | And on cloudy days, believe it or not, they're still there.

00:22:11.180 | Just you don't perceive as much of it coming through.

00:22:14.300 | So that's three things that we should all strive to do.

00:22:17.660 | View low solar angle sunlight early in the day,

00:22:20.820 | view solar angle sunlight later in the day,

00:22:22.980 | and get as much bright light in our eyes as we safely can,

00:22:26.060 | ideally from sunlight throughout the day.

00:22:28.180 | And if you can't do that,

00:22:30.300 | perhaps invest in one of these satellites

00:22:33.820 | so that they can be a bit expensive.

00:22:35.580 | There are a couple of companies

00:22:36.420 | that are starting to design sunrise simulators

00:22:39.180 | and evening simulators that are actually good,

00:22:41.620 | that actually work.

00:22:43.400 | But right now, my read is that aside from one company

00:22:46.780 | out there, which by the way, I have no relationship to,

00:22:49.340 | it's called the TUO light, T-U-O.

00:22:51.420 | And that light bulb was developed by the biologists

00:22:55.300 | at the University of Washington who basically discovered

00:22:58.280 | these color opponent mechanisms.

00:23:00.180 | Those lights are not particularly expensive,

00:23:03.740 | but they do seem to work.

00:23:07.500 | In fact, the study that is emerging,

00:23:10.740 | again, unpublished data seems to indicate

00:23:12.620 | that if you look at it for more than five or six minutes,

00:23:15.300 | it can induce a mild euphoria.

00:23:17.260 | That's how powerful this contrast is.

00:23:19.460 | And what they did there in that light,

00:23:20.700 | I'll just tell you the mechanism is they figured out

00:23:23.140 | that when most people look at low solar angle sunlight

00:23:25.180 | in the morning, they're getting 19 reversals

00:23:28.100 | of blue-orange per second.

00:23:29.780 | So when you look at this light,

00:23:30.760 | it looks like a barely flashing white light,

00:23:34.300 | but it's reversals of orange and blue,

00:23:36.320 | orange and red and blue, and it's happening very fast.

00:23:39.860 | - And so what does the person looking at it perceive?

00:23:42.900 | - Well, I've used one of these.

00:23:44.460 | It just looks like a flickering light.

00:23:46.700 | And of course, there's always the potential

00:23:48.500 | of a placebo effect, but--

00:23:49.660 | - Well, that's what I was gonna say.

00:23:50.500 | Is there a way to control for that by having something

00:23:53.740 | that looks the same to the user,

00:23:56.420 | but of course is not producing the same photo effect?

00:23:59.340 | - Yeah, well, they've done that

00:24:00.320 | with the 10,000 lux sad lamps,

00:24:03.020 | and which most people use to try

00:24:05.720 | and induce sunrise simulation in their home.

00:24:08.760 | But keep in mind that sunrise gives you this comparison

00:24:13.340 | of short and long wavelength light.

00:24:15.660 | Just a bright 10,000 lux light triggers one of the options,

00:24:19.960 | but it won't set your circadian clock.

00:24:21.980 | So most of the sad lamps that are out there

00:24:24.340 | are activating only one of the mechanisms

00:24:27.320 | in these cells that's relevant,

00:24:29.060 | and not the one that's most relevant.

00:24:31.020 | So I'm excited about what TUO is doing.

00:24:32.920 | I think that, and again, I have no relation to them,

00:24:35.580 | except that I know the biologists who did the work

00:24:37.540 | that provide the mechanistic logic for that engineering.

00:24:41.400 | I still think we're in the really early days of this stuff.

00:24:46.400 | What should be done is to have this stuff

00:24:49.840 | built into your laptop, right?

00:24:51.960 | It should be built into your phone,

00:24:53.360 | and hopefully it will be.

00:24:54.680 | Now, I mentioned this color contrast thing

00:24:57.640 | in sunrise and sunset.

00:24:59.000 | I mentioned the bright light throughout the day,

00:25:01.000 | but there's a fourth light stimulus

00:25:03.120 | that turns out to be really important,

00:25:04.400 | and this will provide the segue into the paper.

00:25:08.100 | Turns out that dark exposure at night, independent

00:25:12.180 | of light exposure during the day,

00:25:13.660 | is important for mental health outcomes.

00:25:16.100 | Now, most people think dark exposure,

00:25:17.980 | how do I think about that?

00:25:18.880 | Well, it is dark yet- - Absence of light exposure.

00:25:20.740 | - It's the absence of light,

00:25:22.380 | but what this paper really drives home is that people

00:25:25.900 | who make it a point to get dark exposure at night,

00:25:28.860 | AKA the absence of light at night, actually benefit,

00:25:31.660 | even if they're not getting enough sunlight during the day,

00:25:33.900 | and this is especially true for people

00:25:35.540 | with certain mental health issues.

00:25:37.740 | So I don't think we can overstate the value

00:25:40.500 | of accurately timed light exposure to the eyes

00:25:45.500 | in the context of mental health.

00:25:47.220 | I think, you know, there's so much data by now.

00:25:50.940 | I will say, however, that some people seem more resilient

00:25:54.640 | to these light effects than others, meaning some people,

00:25:58.100 | you know, also don't suffer from jet lag too much.

00:26:00.260 | Some people can stay up late,

00:26:01.760 | get a lot of bright light exposure in the middle of the night

00:26:03.560 | and during the day, they've got their sunglasses on all day

00:26:06.100 | and they're in a great mood all the time.

00:26:07.780 | Other people are more susceptible to these sorts of things.

00:26:10.180 | And we don't know whether or not

00:26:11.620 | polymorphisms underlie that.

00:26:13.900 | I personally am very sensitive to sunlight

00:26:16.820 | in the sense that if I don't get enough sunlight,

00:26:19.140 | I don't feel well after a couple of days,

00:26:21.520 | but I'm less sensitive to light exposure at night,

00:26:24.900 | for instance.

00:26:25.860 | But I think it is perhaps, this is a big statement,

00:26:29.980 | but it is perhaps the most fundamental

00:26:33.260 | environmental stimulus for levels of arousal and alertness,

00:26:37.460 | which correlate with all sorts of, you know,

00:26:39.580 | neuromodulator and hormone outputs.

00:26:41.660 | And so none of this should come as any surprise.

00:26:44.860 | I will mention one last thing.

00:26:46.020 | There was a study published, gosh, over 10 years ago now

00:26:48.400 | from Chuck Zeisser's lab at Harvard Medical School.

00:26:51.020 | It's a phenomenal lab exploring

00:26:53.420 | circadian human health behavior.

00:26:56.140 | He's just considered, no pun, a luminary in the field.

00:27:00.000 | But there wasn't a study that was in error

00:27:03.300 | where they had published in Science Magazine

00:27:05.300 | that light shown behind the knee

00:27:07.440 | could shift circadian rhythms.

00:27:08.940 | And that paper was retracted.

00:27:11.300 | And a lot of people don't know that it was retracted.

00:27:13.160 | Light exposure to the eyes is what's relevant here.

00:27:15.920 | And as far as we know, the color of one's eyes,

00:27:18.020 | like the darkness or lightness of one's eyes,

00:27:19.920 | bears no relevance on their sensitivity

00:27:22.200 | to these types of mechanisms and on and on.

00:27:25.220 | - So one question, one comment.

00:27:26.660 | The question again is going back

00:27:28.040 | to the morning, evening light.

00:27:29.540 | And I spend a lot of time looking at those types of skies,

00:27:33.040 | for example, just 'cause of the nature of my hobbies,

00:27:36.260 | 'cause I'm always doing archery in the morning

00:27:37.980 | and rucking in the afternoon.

00:27:39.500 | So it's not uncommon that I'm seeing both of those.

00:27:42.420 | How relevant is it that the sun be above the horizon?

00:27:46.100 | So for example, it begins to get light

00:27:49.840 | about 30 minutes before sunrise.

00:27:52.860 | And then, so if sun rises at 7.30, first light is seven.

00:27:58.260 | And then, sort of 7.15 to 7.30 is actually quite bright.

00:28:03.240 | I mean, you can see anything and everything.

00:28:05.080 | And the same is true at sunset.

00:28:06.680 | So does that 30 minutes pre,

00:28:10.060 | or when sun is beneath the horizon,

00:28:12.320 | constitute part of that 10 minutes?

00:28:13.820 | - It does.

00:28:14.660 | I mean, in an ideal circumstance,

00:28:16.520 | you'd get outside and see the sunrise every day,

00:28:19.100 | and you'd see the sunset every day, even on cloudy days.

00:28:22.420 | Some people, like myself, wake up before the sun comes up.

00:28:25.500 | And I get this question all the time.

00:28:27.520 | Well, in the absence of powers to make the sunrise faster,

00:28:30.400 | which I'm not aware anyone has, certainly not me,

00:28:32.820 | I think the best thing to do is simply

00:28:34.380 | to turn on as many bright lights as you can indoors

00:28:37.500 | to trigger that melanopsin mechanism.

00:28:40.020 | If you wanna be awake, if you wanna stay asleep or sleepy,

00:28:42.500 | then keep them dim, and then get outside

00:28:45.060 | once the sun is starting to come out.

00:28:47.180 | Some people wake up after the sun has risen, right?

00:28:50.120 | In which case, get what you can.

00:28:52.020 | And some people wake up 10 a.m. or noon,

00:28:55.260 | in which case, you can still get the bright light exposure,

00:28:57.420 | but you won't shift your circadian clock.

00:28:59.900 | Now, in the evening, especially in the winter months,

00:29:03.380 | it's important to look west and try and get some sunlight

00:29:06.200 | in your eyes in the evening.

00:29:07.240 | If you've ever gone into the clinic, for instance,

00:29:10.060 | at two o'clock in the afternoon, after lunch,

00:29:12.640 | and then in the winter, and then come out and it's dark

00:29:16.440 | when you're walking to your car,

00:29:18.220 | it's a kind of eerie feeling.

00:29:20.960 | That sort of eerie feeling may correlate

00:29:22.880 | with the fact that you missed a signal.

00:29:24.820 | Your brain is trying to orient your brain and body in time,

00:29:27.980 | and that's what all of this is, right?

00:29:29.680 | It's trying to orient in time.

00:29:31.060 | And again, some people are more susceptible

00:29:32.820 | to that than others.

00:29:33.660 | Some people might like that feeling of,

00:29:35.920 | oh, I went in when it was bright,

00:29:37.320 | and I come out when it's dark.

00:29:38.820 | But the vast majority of people feel better

00:29:42.420 | when they're getting this morning

00:29:43.580 | and evening sunlight exposure.

00:29:44.820 | And this is especially important in kids, all right?

00:29:47.260 | This is one of the things that this paper points out,

00:29:49.740 | and there are good data, that people are spending

00:29:52.220 | approximately 90% of their time indoors nowadays.

00:29:56.160 | Daytime time, indoors.

00:29:58.260 | And those indoor environments are simply not bright enough.

00:30:01.700 | You think, oh, there's all these bright lights.

00:30:03.060 | And some people are putting blue blockers on

00:30:05.100 | in the middle of the day,

00:30:06.180 | which is the worst thing you could possibly do.

00:30:09.100 | If you're gonna wear blue blockers,

00:30:10.260 | and I don't think they're necessary,

00:30:11.420 | but if you're going to wear them,

00:30:12.260 | you'd wanna wear them at night.

00:30:14.180 | And in the evening, you don't need to wear blue blockers,

00:30:16.580 | you just simply should dim the lights,

00:30:18.700 | and ideally have lights that are set

00:30:20.660 | a little bit lower in your environment,

00:30:22.660 | which the Scandinavians have been doing for a long time.

00:30:25.160 | So kill the overhead lights.

00:30:27.580 | And don't obsess about bright light exposure

00:30:30.340 | in the middle of the night.

00:30:31.180 | In fact, for a long time,

00:30:32.600 | I and some other people were saying,

00:30:34.060 | oh, you know, even just a brief flash of light

00:30:36.180 | in the middle of the night can quash your melatonin.

00:30:38.700 | That's true, but the other time in which

00:30:42.100 | you're in this quote unquote circadian dead zone

00:30:44.420 | is in the middle of the night.

00:30:45.660 | You can't shift your circadian clock

00:30:47.120 | in the middle of the night.

00:30:48.780 | But all of this gets down to interweaving rhythms

00:30:53.180 | of light sensitivity, temperature, hormone output, cortisol.

00:30:57.940 | I mean, there's a whole landscape of circadian biology.

00:31:00.700 | This paper, which was published in a new journal

00:31:04.760 | I'm really excited about called "Nature Mental Health,"

00:31:07.380 | this journal was just launched recently,

00:31:09.380 | is entitled "Day and Night Light Exposure are Associated

00:31:14.060 | with Psychiatric Disorders and Objective Light Study

00:31:17.060 | in More Than 85,000 People."

00:31:19.100 | Now I have to say that I think the title

00:31:21.180 | of this paper is terrible.

00:31:23.020 | Sorry, folks at "Nature Mental Health,"

00:31:24.840 | because if one just read the title,

00:31:26.720 | it sounds like day and night light exposure

00:31:28.780 | are associated with psychiatric disorders, right?

00:31:31.440 | If this were a newspaper headline, you'd be like,

00:31:33.700 | oh my goodness, well, what are you supposed to do, right?

00:31:35.740 | But that's not the conclusion.

00:31:37.140 | The conclusion is that getting a lot of sunlight exposure

00:31:41.980 | during the day and getting a lot of dark exposure at night

00:31:46.080 | is immensely beneficial for psychiatric health

00:31:49.860 | and in a number of ways.

00:31:50.940 | Now, I'm not one to bring up another paper unannounced,

00:31:53.620 | but I will say that this paper built off a previous study

00:31:57.280 | entitled "Time spent in outdoor light is associated

00:31:59.800 | with mood, sleep, and circadian rhythm related outcomes."

00:32:03.300 | And that was a cross-sectional longitudinal study

00:32:05.500 | in 400,000 biobank participants.

00:32:09.060 | So this UK biobank is an incredibly valuable resource,

00:32:12.520 | and there are now multiple studies establishing

00:32:14.920 | that one's pattern of light exposure is extremely important.

00:32:19.060 | Now, the previous study in 400,000 participants

00:32:22.980 | basically nailed home the idea

00:32:24.580 | that the more time you spend outdoors,

00:32:27.620 | the better is your mood, the better is your sleep,

00:32:30.820 | the better is the rhythmicity of your sleep-wake cycles,

00:32:33.800 | and on and on.

00:32:34.820 | Something that I think, even though people will say,

00:32:37.140 | we've known that for thousands of years,

00:32:38.540 | needed scientific substantiation.

00:32:41.780 | This new study essentially looked at

00:32:45.000 | the relative contributions of daytime light exposure

00:32:48.600 | and nighttime dark exposure,

00:32:50.520 | and they did that on a background of looking,

00:32:53.460 | in particular, people who had major depressive disorder,

00:32:55.600 | generalized anxiety, PTSD, bipolar disorder.

00:32:58.680 | Here's the basic takeaway, and I'll quote them here,

00:33:04.540 | and then I'll tell you my interpretation.

00:33:06.860 | Here I'm quoting,

00:33:07.700 | "Avoiding night at light and seeking light during the day,"

00:33:10.440 | I love that word, seeking,

00:33:12.760 | "may be a simple and effective non-pharmacologic means

00:33:15.820 | for broadly improving mental health."

00:33:17.580 | So that's a pretty bold statement, right?

00:33:20.340 | And I love that they say seeking

00:33:21.620 | because it implies that people aren't reflexively

00:33:24.300 | getting the light exposure that they need,

00:33:26.040 | that this needs to be a practice,

00:33:27.540 | much like zone two cardio or resistance training.

00:33:30.540 | Okay, so what do they do in this study?

00:33:33.000 | So basically, they gathered up 100,000 people or so,

00:33:37.340 | it eventually was paired down to about 86,000 participants

00:33:42.200 | 'cause some just didn't qualify

00:33:43.840 | or didn't report their data back.

00:33:45.900 | They equipped them with accelerometers on their wrist,

00:33:49.200 | and those wrist devices also could measure ambient light.

00:33:53.580 | Now, that's not a perfect tool

00:33:55.140 | 'cause what you'd love to do is measure ambient light

00:33:56.800 | at the level of the eyes.

00:33:58.440 | By the way, will somebody design an eyeglass frame

00:34:01.380 | that changes color when you've gotten sufficient light

00:34:04.460 | from sunlight during the day,

00:34:05.660 | and then at night is a different color,

00:34:08.260 | and then if you're getting too much light exposure,

00:34:09.880 | will go to a different color frame?

00:34:11.420 | This has to be possible so that you don't have to wonder

00:34:14.340 | if you got enough light during the day.

00:34:16.340 | And of course, if it's at the level of the eyes,

00:34:19.220 | then you know that's what's landing at the eyes.

00:34:21.340 | - Ian, I mean, that's what I was gonna ask you about that.

00:34:23.700 | Do these wrist-based devices

00:34:26.380 | potentially get covered by clothing in some-

00:34:28.580 | - Turned over. - You have your sleeves down.

00:34:30.280 | I have my sleeves up.

00:34:31.340 | - They had it on the outside of the sleeve,

00:34:32.820 | but they asked that people just keep it

00:34:34.140 | on their dominant hand.

00:34:35.340 | It's not perfect,

00:34:36.820 | but in some ways it's kind of nice that it's not perfect.

00:34:39.300 | We could turn that disadvantage into advantage by thinking,

00:34:42.660 | when the person is out and about,

00:34:44.060 | they're not often looking right at the sun.

00:34:46.860 | If you're talking to a colleague

00:34:48.940 | under an overhang, for instance.

00:34:50.980 | So it's not perfect. - It's directionally.

00:34:53.420 | - It's directionally right.

00:34:54.860 | Okay, and then they had two hypotheses,

00:34:57.160 | two primary hypotheses.

00:34:58.240 | One, that greater light exposure in the day

00:34:59.780 | is associated with lower risk for psychiatric disorders,

00:35:02.140 | and two, second hypotheses,

00:35:04.180 | greater light exposure at night

00:35:05.620 | is associated with higher risk for psychiatric disorders

00:35:08.460 | and poorer mood.

00:35:09.580 | This is oh so relevant for the way we live now,

00:35:12.060 | people on screens and tablets in the middle of the night.

00:35:14.760 | Okay, then they collected information

00:35:16.980 | about how much light exposure people were getting,

00:35:19.800 | as well as their sleep and their activity and so on.

00:35:22.280 | I should mention this was done in males and females.

00:35:25.020 | It was a slightly older cohort

00:35:26.620 | than one is used to seeing people in their 50s and 60s.

00:35:29.840 | They had psychiatric diagnosis information,

00:35:32.180 | and then they divided people into essentially two groups,

00:35:35.500 | but they had a lower, so a Q1 and a Q2, a lower quartile.

00:35:39.420 | That meant people that were getting less daytime light

00:35:42.940 | as opposed to the third and fourth quartile,

00:35:45.180 | more daytime light.

00:35:46.540 | They also had a nighttime light exposure evaluation,

00:35:51.340 | and they had people in the low Q1 and Q2,

00:35:54.280 | so these people are getting less nighttime light

00:35:57.260 | versus Q3, Q4, more nighttime light.

00:36:00.600 | Nicely, they also looked at sleep duration,

00:36:04.020 | and they looked at photo period,

00:36:06.240 | meaning how long the days were for those individuals,

00:36:09.020 | how active they were, like 10 hours a day, 14 hours a day,

00:36:12.140 | because the more active you are,

00:36:13.940 | the more opportunity for light exposure you have

00:36:15.920 | during the day or night, for instance.

00:36:19.180 | Okay, so they had, I would say, fairly complete data sets.

00:36:24.180 | Then, and I'm just gonna hit the top contour

00:36:29.420 | of what they did in each.

00:36:30.260 | Sorry, sleep duration, sleep efficiency, et cetera,

00:36:33.060 | was determined off the accelerometer.

00:36:34.800 | That's right, as well as self-report.

00:36:36.900 | Yeah, not ideal, right?

00:36:38.500 | You'd love for people to be wearing a whoop bander

00:36:40.520 | or a ring or something of that sort,

00:36:42.540 | but this was initiated some time ago,

00:36:44.800 | so they either didn't have access to that technology

00:36:48.040 | or, for whatever reason, didn't select it.

00:36:50.080 | I'd like to take a brief moment

00:36:52.620 | and thank one of our sponsors, and that's AG1.

00:36:55.580 | AG1 is a vitamin mineral probiotic drink

00:36:58.200 | that also contains adaptogens.

00:37:00.480 | I started taking AG1 way back in 2012.

00:37:03.600 | The reason I started taking it

00:37:04.880 | and the reason I still take it every day

00:37:06.680 | is that it ensures that I meet all of my quotas

00:37:09.020 | for vitamins and minerals,

00:37:10.800 | and it ensures that I get enough prebiotic and probiotic

00:37:13.660 | to support gut health.

00:37:15.180 | Now, gut health is something that, over the last 10 years,

00:37:17.640 | we realized is not just important

00:37:19.620 | for the health of our gut, but also for our immune system

00:37:23.440 | and for the production of neurotransmitters

00:37:25.600 | and neuromodulators, things like dopamine and serotonin,

00:37:28.160 | in other words, gut health is critical

00:37:30.120 | for proper brain functioning.

00:37:32.000 | Now, of course, I strive to consume healthy whole foods

00:37:34.680 | for the majority of my nutritional intake every single day,

00:37:38.120 | but there are a number of things in AG1,

00:37:40.080 | including specific micronutrients

00:37:41.760 | that are hard to get from whole foods

00:37:43.320 | or at least in sufficient quantities.

00:37:45.200 | So AG1 allows me to get the vitamins and minerals

00:37:47.620 | that I need, probiotics, prebiotics, the adaptogens,

00:37:50.400 | and critical micronutrients.

00:37:52.400 | So anytime somebody asks me

00:37:53.960 | if they were to take just one supplement,

00:37:56.120 | what that supplement should be,

00:37:57.700 | I tell them AG1,

00:37:59.240 | because AG1 supports so many different systems

00:38:01.540 | within the body that are involved in mental health,

00:38:03.740 | physical health, and performance.

00:38:05.360 | To try AG1, go to drinkag1.com/huberman,

00:38:09.260 | and you'll get a year's supply of vitamin D3K2

00:38:12.400 | and five free travel packs of AG1.

00:38:14.500 | Again, that's drinkag1.com/huberman.

00:38:18.080 | Then what they did is they have information

00:38:22.860 | on who has major depressive disorder,

00:38:24.840 | who has PTSD, generalized anxiety, bipolar, psychosis,

00:38:28.740 | et cetera, and then they ran three models.

00:38:32.160 | And you can tell me what you think

00:38:34.300 | about the power of these models,

00:38:35.540 | but as somebody who thinks about the mechanistic aspect

00:38:39.480 | of all of this a lot,

00:38:40.640 | but not somebody who's ever run this type of study,

00:38:43.080 | I'd be really curious.

00:38:44.600 | Model one examined the unadjusted association

00:38:47.680 | between day and nighttime light exposure

00:38:50.820 | and psychiatric outcomes.

00:38:52.000 | So just basically asking, is there a relationship

00:38:54.880 | between how much light you get during the day

00:38:57.000 | and how much light you get at night

00:38:58.840 | and how bad your depression is or anxiety is, et cetera.

00:39:03.840 | Looking at just a standard ratio of the probability

00:39:08.520 | that you have a certain symptom or set of symptoms

00:39:12.320 | versus you don't given a certain amount of light exposure.

00:39:15.720 | Model two adjusted for the age of the person,

00:39:19.280 | their sex and ethnicity, and photo period.

00:39:23.080 | So they looked at how long the days were

00:39:25.000 | in that given person's region of the world.

00:39:27.520 | And then model three--

00:39:30.360 | - Were these people were all in the UK

00:39:32.080 | or were they around the world?

00:39:32.920 | - They were all in the UK, as far as I know.

00:39:35.680 | And then model three adjusted for employment.

00:39:38.260 | So employed versus unemployed,

00:39:39.920 | which if you think about it is pretty important.

00:39:41.600 | Like you say, well, an unemployed person

00:39:43.020 | has a lot more time to control these variables,

00:39:45.440 | but an employed person who's doing shift work does not.

00:39:49.160 | And they incorporated information

00:39:51.640 | about employed versus unemployed physical activity,

00:39:55.280 | which turns out to be very important.

00:39:57.440 | And then things like shift work, et cetera.

00:40:00.000 | And so these, what we can say very safely

00:40:03.160 | is that the outcomes with each of these models,

00:40:06.300 | the results were very similar.

00:40:08.780 | So we don't want to discard the differences

00:40:11.820 | between those models entirely.

00:40:13.180 | But in my read is in every figure of the paper,

00:40:16.480 | it doesn't seem like model one, two or three differ

00:40:20.080 | from one another in terms of total outcome.

00:40:22.160 | - Yeah, that's an unusual aspect of this paper.

00:40:25.220 | So these adjustments are very standard, right?

00:40:27.640 | So this is a classic tool that's used in most epidemiology

00:40:32.640 | because you don't have randomization.

00:40:35.300 | So once randomization is out the window,

00:40:37.300 | like so for example, the paper I'm going to present

00:40:39.920 | is based on an RCT, there will be no models.

00:40:42.700 | It's just here are the data.

00:40:44.040 | - Yeah, here they're asking people, what do you do?

00:40:46.520 | Report back to us, we're going to measure

00:40:47.940 | your light exposure, but no one was assigned

00:40:49.920 | to any groups or swap.

00:40:51.240 | Whatever quote unquote controls are there,

00:40:55.220 | they're really not there.

00:40:57.080 | It's just comparisons between groups.

00:40:59.000 | - So what is interesting to me is that

00:41:02.960 | it's exactly as you said, and we'll make all these figures

00:41:05.240 | available in addition to the papers.

00:41:06.960 | But I mean, it's very unusual that there's no difference

00:41:12.320 | between the unadjusted and the adjusted models.

00:41:15.520 | And as you say, there's probably two places out of 30

00:41:20.520 | when you look at all the different quartile comparisons

00:41:25.120 | where you might creep from statistically significant

00:41:29.760 | just out of it or just into it.

00:41:31.340 | But yeah, you could simplify this figure two completely

00:41:35.280 | by just showing one of the models and you would be

00:41:38.800 | getting 95% of the information, which is, you know.

00:41:42.480 | I mean, I think in one way that suggests

00:41:45.160 | that there's less dependency on those variables.

00:41:52.500 | Of course, it still doesn't address

00:41:54.880 | probably the greatest question I have here,

00:41:57.440 | which I'm sure we'll get to at some point as you continue.

00:42:01.000 | - Yeah, so I'm very curious what that question is,

00:42:03.160 | but I'll suppress my curiosity for the moment.

00:42:06.520 | You know, so if we look at figure two of this paper,

00:42:08.320 | and I realize a lot of people are listening

00:42:09.840 | and they're not able to look at this,

00:42:11.600 | although we have posted the figures

00:42:12.960 | on the YouTube versions of this,

00:42:14.900 | just want to make clear what's going on

00:42:19.080 | just for those that are listening.

00:42:20.340 | Essentially what they're looking at

00:42:21.720 | is what they call the odds ratio,

00:42:24.380 | which is the probability of something happening

00:42:27.260 | in one group divided by the probability

00:42:29.780 | of something happening in another group.

00:42:31.120 | I guess, by way of example would be, you know,

00:42:33.880 | if you were going to look at the odds ratio of, you know,

00:42:36.960 | the probability of somebody getting lung cancer

00:42:38.560 | if they smoke versus probability

00:42:40.080 | of somebody getting lung cancer if they don't smoke.

00:42:42.640 | - So odds ratios and hazard ratios are often confused.

00:42:46.160 | They're very similar,

00:42:47.840 | and odds ratios generally refer to a lifetime exposure,

00:42:51.160 | whereas a hazard ratio is defined

00:42:54.060 | over a specific period of time.

00:42:55.760 | But the math is still effectively the same.

00:42:58.060 | And using the example you gave,

00:43:00.080 | if you took the odds ratio of, you know, death,

00:43:05.000 | so let's talk all-cause mortality for a smoker

00:43:07.020 | versus a non-smoker, and the answer were 1.78.

00:43:09.840 | I'm making that up, but that's directionally correct.

00:43:12.440 | 1.78 as an odds ratio means there's a 78% chance greater

00:43:17.440 | of the outcome of interest, in this case,

00:43:21.000 | death by any cause in the affected group,

00:43:23.120 | which would be the smokers.

00:43:24.100 | So odds ratio of two is 100%,

00:43:27.120 | and odds ratio of three is 200%.

00:43:29.480 | So the math is take the number,

00:43:31.200 | subtract one, and that's the percent.

00:43:33.600 | You know, figure two of this paper

00:43:35.320 | is one of the key take-homes.

00:43:37.160 | They essentially look at the odds ratio

00:43:39.920 | of people who are in the, let's say that,

00:43:43.120 | let's just look at the nighttime light exposure.

00:43:47.000 | - And just remind me, Andrew,

00:43:48.280 | 'cause, and everybody else watching,

00:43:50.760 | every one of these is showing second, third, fourth

00:43:53.120 | as your x-axis, meaning they're all being compared

00:43:56.120 | to the first quartile, and the first quartile

00:43:59.880 | is lowest light exposure or highest light exposure?

00:44:03.120 | - Lowest, well, we have to--

00:44:05.400 | - With the differentiate between day and night.

00:44:06.840 | - That's right.

00:44:07.680 | - Okay, so restate it.

00:44:08.920 | - Sure, so if we look at, you know,

00:44:11.880 | what is your risk of a psychiatric challenge,

00:44:16.000 | broadly speaking?

00:44:16.840 | Well, panel A is major depressive disorder.

00:44:19.940 | If you are in the second quartile, third quartile,

00:44:22.840 | or fourth quartile of nighttime light exposure,

00:44:26.240 | so second being the least amount of nighttime light exposure,

00:44:29.540 | third being more nighttime light exposure,

00:44:32.240 | and fourth, the most nighttime light exposure

00:44:34.960 | relative to the first quartile.

00:44:36.540 | - This is just a stupid thing.

00:44:38.160 | Like, if I were doing this figure,

00:44:40.460 | if you were doing this in a lecture,

00:44:41.920 | you know what you would do to make it so easy?

00:44:43.940 | You would draw arrows on it

00:44:45.280 | that say increasing light exposure at night,

00:44:47.880 | decreasing light exposure in the day.

00:44:50.000 | It's the same information.

00:44:51.180 | It just makes it easier for the reader to understand.

00:44:53.360 | - Absolutely.

00:44:54.200 | - But maybe the teaching point, I think,

00:44:55.920 | is for people when they review articles,

00:44:57.900 | like, don't be afraid to do that and just kind of like,

00:45:01.600 | oh, I've got scribble all over this.

00:45:02.960 | - Yeah, exactly.

00:45:03.800 | So it's like, I draw the arrow.

00:45:05.260 | That's increasing light, that's decreasing light,

00:45:07.160 | and that's how I can pay attention

00:45:08.320 | to what's actually happening.

00:45:09.240 | - Right, and I'm actually in touch with the editorial staff

00:45:11.920 | at Nature Mental Health,

00:45:13.880 | although they don't know that I'm covering this paper

00:45:15.900 | until after this comes out.

00:45:17.000 | You know, I think one thing that scientific journals

00:45:18.840 | really, really need to do

00:45:20.400 | is start making the readability of the articles

00:45:24.400 | better for non-experts.

00:45:26.240 | I mean, chances are, if you can't understand a graph,

00:45:28.980 | and this is true for everybody,

00:45:31.060 | chances are there's a problem with the way it's presented.

00:45:34.840 | Put it on them, but then, of course, try and parse it,

00:45:37.620 | because, you know, rarely, if ever,

00:45:40.240 | is it all spelled out clearly.

00:45:41.460 | But anyway, that's what we're trying to do here.

00:45:42.860 | So yeah, the way I would have done is, say,

00:45:44.620 | second quartile is low amounts of nighttime light exposure

00:45:47.620 | and define what that is.

00:45:49.100 | You know, third quartile is more light exposure,

00:45:52.820 | and then fourth, maximum amount of light exposure at night.

00:45:55.420 | And basically what you see is that the probability

00:46:00.020 | of having worse major depressive symptoms

00:46:04.140 | linearly increases as you go

00:46:06.380 | from the second to third to fourth quartile.

00:46:08.740 | So more nighttime light exposure, worse for you,

00:46:13.500 | and there's a dose response, if you will, of the effect.

00:46:18.200 | Now, we can march through or describe figure two

00:46:22.540 | pretty quickly by saying the same thing is true,

00:46:25.620 | and now we're just talking about nighttime light exposure,

00:46:28.540 | for generalized anxiety disorder, so that's panel C.

00:46:32.060 | Bipolar disorder, although the difference

00:46:34.060 | between the second and third quartile in bipolar disorder

00:46:36.700 | isn't as dramatic, once you get up to the fourth quartile,

00:46:40.260 | bipolar symptoms get much worse

00:46:42.300 | when people are getting nighttime light exposure.

00:46:45.400 | I really want to emphasize that point,

00:46:47.420 | because they go on in the discussion of this paper

00:46:50.620 | to reemphasize that point several times.

00:46:52.660 | In fact, they say that while light exposure during the day,

00:46:56.900 | of course, we will go into the data,

00:46:59.620 | is beneficial for mental health.

00:47:01.600 | For people with bipolar disorder,

00:47:03.180 | it seems that light exposure at night

00:47:05.980 | is especially problematic,

00:47:07.660 | independent of how much sunlight

00:47:09.000 | they're getting during the day.

00:47:09.840 | So your bipolar, the person with bipolar disorder

00:47:13.660 | who's struggling with either a manic or depressive episode,

00:47:16.660 | who's making a point to get sunlight during the day,

00:47:18.980 | who's also getting light exposure at night,

00:47:21.780 | is making their symptoms worse.

00:47:23.140 | And keep in mind, they couldn't completely control this,

00:47:25.580 | but this is largely independent

00:47:27.960 | of things like sleep duration.

00:47:30.620 | So that doesn't necessarily mean

00:47:32.340 | that the person sleeping less,

00:47:33.420 | although in a manic episode, presumably they are,

00:47:36.700 | it's independent of exercise.

00:47:38.460 | It's independent of a bunch of other things,

00:47:40.160 | because any logical person will hear this and say,

00:47:43.320 | "Okay, well, they're getting more light at night

00:47:44.880 | because they're doing a bunch of other things."

00:47:46.500 | But it's largely independent of those other things.

00:47:49.760 | Likewise, the symptomology of PTSD gets far worse

00:47:53.860 | with increasing light exposure at night.

00:47:56.380 | Self-harm really takes a leap from being fairly,

00:48:01.380 | I don't want to say minimal,

00:48:02.620 | at the second and third quartile,

00:48:04.660 | so low in, let's say, medium.

00:48:06.380 | I'm using some, I'm taking some liberties here,

00:48:08.400 | but low and medium amounts

00:48:09.460 | of artificial light exposure at night.

00:48:11.580 | Then for people who get quite a lot

00:48:13.480 | of nighttime light exposure, self-harm goes up,

00:48:17.080 | and probability of psychotic episodes goes up,

00:48:21.420 | or psychotic symptoms.

00:48:22.680 | Now, what's nice is that the,

00:48:26.520 | what's nice about the data is that the exact inverse

00:48:29.080 | is basically true for daytime light exposure,

00:48:31.400 | although not across the board.

00:48:33.580 | We can generally say that for major depressive disorder,

00:48:36.920 | generalized anxiety, bipolar symptoms,

00:48:39.420 | there it's a little more scattered, PTSD, and self-harm,

00:48:42.820 | the more daytime light exposure, ideally from sunlight,

00:48:47.020 | 'cause that's actually what's being measured in most cases,

00:48:50.220 | we can talk about how we know that,

00:48:53.000 | is going to approximately linearly drop the probability

00:48:58.000 | or the severity of these symptoms.

00:49:00.380 | - And we could just explain again

00:49:01.620 | that the odds ratios now seem to be going down,

00:49:04.300 | so an odds ratio of 0.7 now refers to a 30% reduction

00:49:09.080 | in the variable of interest here.

00:49:10.880 | - Exactly.

00:49:11.980 | Now, the psychosis, panel F, which focuses on psychosis,

00:49:16.060 | I think is also worth mentioning in a bit more detail.

00:49:20.480 | There's a fairly dramatic reduction in psychotic symptoms

00:49:23.600 | as one gets more daytime light exposure,

00:49:27.060 | independent of nighttime light exposure.

00:49:29.280 | There's a well-known phenomenon called ICU psychosis,

00:49:33.800 | which is that people come into the hospital

00:49:36.440 | for a broken leg or a car accident,

00:49:38.580 | maybe they were getting surgery from Peter back when

00:49:41.900 | for something totally independent,

00:49:43.900 | they're housed in the hospital,

00:49:45.080 | and as anyone who's ever been in a hospital

00:49:46.940 | as a patient or visitor knows,

00:49:48.740 | the lighting environment of the hospital

00:49:51.080 | is absolutely dreadful for health, just dreadful.

00:49:53.980 | I mean, people often complain about the food

00:49:55.840 | in the cafeteria as being unhealthy,

00:49:58.180 | that's often not always true, not always true,

00:50:02.860 | but the lighting environments in hospitals

00:50:04.860 | is absolutely counter to health.

00:50:06.780 | - Especially in the intensive care unit, yeah.

00:50:09.140 | - Right, so--

00:50:10.060 | - I think the intensive care unit at Hopkins,

00:50:12.140 | the main one, the main ICU didn't have windows.

00:50:16.300 | People who go into the hospital with a brain injury

00:50:19.500 | or with a stroke or something,

00:50:21.280 | I get contacted all the time,

00:50:22.540 | even though I'm not a clinician,

00:50:23.540 | what should I do for my kid, my parent?

00:50:25.380 | I always say, get them near a window

00:50:27.060 | and start to the best of your abilities

00:50:28.860 | controlling their sleep-wake cycle.

00:50:30.740 | Now, oftentimes there are nurses coming in

00:50:33.480 | and taking blood tests and measuring pulses

00:50:35.580 | in the middle of the night, that's disruptive.

00:50:37.740 | There's bright light, not just blue light,

00:50:39.460 | that's disruptive, it's noisy, that's disruptive.

00:50:42.540 | ICU psychosis is when non-psychotic individuals

00:50:45.460 | start having psychotic episodes in the hospital

00:50:48.540 | because of nighttime light exposure

00:50:50.260 | and in some cases, lack of daytime sunlight.

00:50:53.920 | We can say that with some degree of confidence

00:50:57.100 | because when those people go home,

00:50:58.660 | even though sometimes their symptoms

00:51:00.540 | for what brought them to the hospital

00:51:02.500 | in the first place get worse, their psychosis goes away.

00:51:06.520 | Now, and it's independent of medication.

00:51:08.780 | So let's just be really direct.

00:51:12.860 | There is a possibility that we are all socially jet-lagged,

00:51:17.860 | that we are all disrupting these mood regulation symptoms,

00:51:23.860 | the systems, excuse me, by not getting enough daytime light

00:51:27.140 | and by getting too much nighttime light.

00:51:29.720 | If we want to look at just some of the bullet points

00:51:31.660 | or the takeaways, and Peter, thank you.

00:51:33.460 | You highlighted a few of these, but--

00:51:35.080 | Can we just go back to this figure too for a second?

00:51:36.940 | Oh, yeah, sure.

00:51:37.780 | There's a handful of things that really jumped out.

00:51:39.620 | I had a feeling Peter was gonna want to dig in the day.

00:51:40.760 | Yeah, yeah, I just-- Let's do it, let's do it.

00:51:42.140 | And again, I normally wouldn't make so much hay out of this

00:51:44.820 | except for the fact that they're so tight.

00:51:46.860 | But there are a few that really stand out.

00:51:50.940 | And again, I love this figure.

00:51:53.060 | I would have labeled it a little differently

00:51:54.600 | to make it completely user-friendly.

00:51:56.860 | But nevertheless, the increasing light at night

00:52:01.220 | and the impact on depression.

00:52:02.920 | Let me be really technical in what I say.

00:52:06.240 | And the relationship or correlation

00:52:07.940 | to depression is very strong.

00:52:11.180 | The relationship to light and self-harm

00:52:14.460 | in the upper quartile, so when you take those 25% of people

00:52:19.420 | with the most nighttime light,

00:52:21.860 | that relationship to self-harm is interesting

00:52:23.900 | and completely uncoupled from the other 75%.

00:52:26.820 | That's interesting.

00:52:28.140 | By uncoupled, you mean that at the lower levels

00:52:30.440 | of light exposure at night,

00:52:31.420 | you're not seeing an increase in self-harm.

00:52:33.460 | Not whatsoever.

00:52:34.300 | And then once you get to that fourth quartile--

00:52:36.080 | It's a big step.

00:52:36.920 | It's like a 30% greater risk of self-harm.

00:52:39.700 | Yeah, so it's totally flat.

00:52:40.860 | The first, second, third quartile, no different.

00:52:43.260 | And then fourth, big jump.

00:52:44.700 | And then the inverse relationship, right?

00:52:48.220 | As light increases during the daytime,

00:52:50.100 | you see this reduction in self-harm.

00:52:52.680 | Interesting.

00:52:53.520 | The PTSD relationship based on nighttime light

00:52:57.300 | and the psychosis relationship based on daytime light.

00:53:01.020 | Those are the ones that really jumped out to me.

00:53:03.500 | I think anxiety relatively less impressive here.

00:53:08.620 | And bipolar disorder didn't seem as strong as well.

00:53:13.060 | So I think those are the big ones that jumped out to me.

00:53:16.740 | Yeah, I agree.

00:53:17.580 | There's a bit more scatter on generalized anxiety

00:53:20.180 | and the degree of significant change is not as robust.

00:53:25.180 | In other words, getting a lot of daytime light,

00:53:28.540 | ideally from sunlight,

00:53:29.500 | is not necessarily going to reduce your levels of anxiety.

00:53:33.080 | Getting a lot of nighttime light exposure

00:53:35.340 | is not increasing nighttime anxiety that much.

00:53:38.620 | Although 20% is not nothing for nighttime light exposure.

00:53:43.620 | But yeah, the psychosis, major depression

00:53:49.580 | and self-harm are really, they leap out.

00:53:52.060 | Actually, maybe we can just drill a little bit deeper

00:53:54.360 | on major depression.

00:53:55.200 | And basically when you go from the second

00:53:56.460 | to third quartile of nighttime light exposure,

00:53:58.600 | so more than nighttime light exposure,

00:54:00.340 | you basically go from no significant increase

00:54:03.980 | to almost a 20% increase.

00:54:06.260 | And then as you get up to the fourth quartile,

00:54:09.780 | so the most nighttime light exposure,

00:54:11.940 | you're at about 25% increase in major depressive symptoms.

00:54:16.940 | That's no joke.

00:54:19.960 | And I think that we, if we were to,

00:54:24.240 | I mean, we don't have the data right here,

00:54:26.460 | but if we were to look at like standard SSRI treatment

00:54:29.060 | for major depression, people debate this pretty actively.

00:54:33.620 | But light is a very potent stimulus.

00:54:37.540 | And the timing of light is critical

00:54:38.940 | because on the inverse is also true.

00:54:41.300 | As you get to the fourth quartile

00:54:42.500 | of daytime light exposure,

00:54:44.180 | you get about a 20% reduction in major depressive disorder.

00:54:48.420 | What I like about a study like this

00:54:50.740 | is that it puts the error bars so easy to see on the data.

00:54:54.880 | And why is that interesting?

00:54:56.100 | Well, there's a belief that daytime light exposure

00:55:01.660 | there's a belief that bigger is always better in sample size.

00:55:06.580 | And we often talk about that

00:55:08.140 | through the lens of power analysis, right?

00:55:10.120 | So how many subjects do we need to reach a conclusion

00:55:15.120 | that is powered to this level?

00:55:18.820 | And that's true, but what I don't think gets discussed

00:55:22.340 | as often is the opposite of that,

00:55:25.060 | which is what if you overpower a study?

00:55:27.220 | In other words, what if the power analysis says

00:55:29.860 | to have a level of power at 90%, you need a thousand subjects

00:55:34.860 | and you say, great, we're going to do 10,000 subjects.

00:55:38.960 | Well, you're clearly powered for it,

00:55:40.700 | but you might be overpowered.

00:55:42.020 | And people might say, well, why would that be a bad thing?

00:55:44.420 | It could be a bad thing because it means you are very likely

00:55:47.500 | to reach statistical significance

00:55:49.900 | in things that might not be actually significant.

00:55:52.980 | And so one thing about this study that is just a quick back,

00:55:57.740 | like kind of a quick and dirty way to tell

00:55:59.780 | that it's probably not overpowered

00:56:02.100 | is that you have varying lengths of error bars.

00:56:06.540 | And what that tells me is that, and again,

00:56:10.080 | this is not like a formal statistical analysis.

00:56:12.660 | It's just kind of like a back

00:56:13.900 | of the envelope statistical analysis.

00:56:15.860 | If you look, for example, at self-harm in the top quartile,

00:56:18.660 | you actually have pretty big error bars.

00:56:19.940 | In fact, all the self-harm have sort

00:56:21.820 | of slightly bigger error bars.

00:56:23.380 | And yet, when you look at, for example, the depression,

00:56:25.860 | even though the error bars aren't all the same size,

00:56:27.700 | they're tighter.

00:56:28.540 | In fact, when you look at the relationship

00:56:30.140 | between depression and daytime light,

00:56:32.340 | the error bars are really, really small.

00:56:34.300 | So that just gives me confidence

00:56:36.680 | that there is variability in this,

00:56:38.460 | which paradoxically you kind of want to see

00:56:40.940 | 'cause it tells me that this wasn't just done,

00:56:43.900 | you know, there was, I think you said 8,000 subjects

00:56:46.120 | were in this, and I realized--

00:56:47.860 | - More than 86,000.

00:56:49.260 | - 86,000, sorry.

00:56:50.220 | Yeah, you realize that it wasn't that,

00:56:52.220 | oh, this should have been done with a 10th of that

00:56:54.540 | or a half of that.

00:56:55.380 | And we're picking up signal that is statistically relevant,

00:56:59.300 | but clinically irrelevant.

00:57:01.200 | - Yeah, thanks for that point.

00:57:03.140 | I didn't pay attention to that.

00:57:06.220 | I mean, I paid attention to the error bars,

00:57:07.660 | but I didn't know that.

00:57:08.900 | So thank you, I'm learning too.

00:57:11.760 | And I suppose for people that are listening,

00:57:13.620 | we can just give them a sense

00:57:14.500 | of what the error bar ranges are.

00:57:16.420 | For self-harm, they're running as much as 20%

00:57:21.100 | either side of the mean, the average.

00:57:23.420 | And for major depression,

00:57:24.380 | it looks like it's more like, let's say eight to 10%.

00:57:28.240 | - If that, yeah.

00:57:29.080 | - If that, maybe more like five, right?

00:57:31.120 | So, yeah, I see what you're saying.

00:57:33.420 | So when you get a very large sample size,

00:57:38.420 | you're going to have some outliers in there

00:57:40.980 | and you can mask those outliers

00:57:43.760 | just by having so many data points, right?

00:57:45.740 | - Yeah, because these error bars directly tell you

00:57:47.900 | whether or not you're statistically significant.

00:57:49.940 | So what's really nice about this type of graph,

00:57:53.180 | and you see these in,

00:57:54.500 | there's going to be a graph in my paper

00:57:55.900 | where you see the same analysis.

00:57:57.520 | They're always drawing the 95% confidence interval

00:58:01.740 | on the data point.

00:58:03.660 | And if the 95% confidence interval

00:58:06.700 | does not touch the line of unity,

00:58:09.740 | which in this case is the HODS ratio of 1.0 over the X-axis,

00:58:14.160 | then you know it's statistically significant

00:58:16.040 | to the confidence interval they've defined,

00:58:17.720 | which is almost assuredly 95%.

00:58:19.700 | Sometimes they'll make it tighter at 99.

00:58:21.780 | And so that's why you can just look right at these and go,

00:58:23.920 | oh, look, you know, in depression,

00:58:26.060 | the second quartile didn't reach statistical significance

00:58:29.280 | 'cause the error bars are touching the line,

00:58:31.460 | just as the case for the second and third quartile

00:58:34.420 | for self-harm.

00:58:35.940 | But when you look at the fourth quartile,

00:58:37.720 | you can see that the lower tip of the error bars

00:58:40.420 | isn't anywhere near unity.

00:58:42.140 | And so we know without having to look up the P value

00:58:45.060 | that it's smaller than either 0.05 or 0.1,

00:58:48.500 | however they've defined it.

00:58:49.900 | And it's really amazing when you see these

00:58:52.600 | overpowered studies,

00:58:53.860 | which are easier to do epidemiologically,

00:58:55.720 | where the P value ends up being microscopic.

00:59:00.180 | They can drive their P values down to anything low

00:59:03.100 | because sample size can be infinite,

00:59:04.700 | but you can see that it's just like the error bar

00:59:08.560 | is just skimming above the unity line,

00:59:13.240 | but it's so, so, so tight.

00:59:14.880 | - I'd like to take a quick break

00:59:16.980 | to acknowledge our sponsor, Element.

00:59:19.580 | Element is an electrolyte drink that has everything you need

00:59:22.240 | and nothing you don't.

00:59:23.160 | That means zero sugar

00:59:24.500 | and the appropriate ratios of the electrolyte,

00:59:26.860 | sodium, magnesium, and potassium.

00:59:29.160 | And that correct ratio of electrolytes

00:59:30.940 | is extremely important because every cell in your body,

00:59:34.100 | but especially your nerve cells, your neurons,

00:59:36.380 | relies on electrolytes in order to function properly.

00:59:39.440 | So when you're well hydrated

00:59:40.940 | and you have the appropriate amount of electrolytes

00:59:42.820 | in your system, your mental functioning

00:59:44.680 | and your physical functioning is improved.

00:59:46.760 | I drink one packet of Element dissolved

00:59:48.620 | in about 16 to 32 ounces of water

00:59:51.020 | when I wake up in the morning,

00:59:52.580 | as well as while I exercise.

00:59:54.540 | And if I've sweat a lot during that exercise,

00:59:56.620 | I often will drink a third Element packet dissolved

00:59:59.080 | in about 32 ounces of water after I exercise.

01:00:02.060 | Element comes in a variety of different flavors,

01:00:04.220 | all of which I find really tasty.

01:00:06.180 | I like the citrus, I like the watermelon,

01:00:07.820 | I like the raspberry.

01:00:08.900 | Frankly, I can't pick just one.

01:00:10.760 | It also comes in chocolate and chocolate mint,

01:00:12.840 | which I find tastes best if they are put into water,

01:00:15.300 | dissolved, and then heated up.

01:00:17.100 | I tend to do that in the winter months

01:00:18.560 | because, of course, you don't just need hydration

01:00:21.020 | on hot days and in the summer and spring months,

01:00:23.620 | but also in the winter when the temperatures are cold

01:00:26.700 | and the environment tends to be dry.

01:00:28.700 | If you'd like to try Element,

01:00:29.860 | you can go to Drink Element, spelled lment.com/huberman,

01:00:34.120 | to try a free sample pack.

01:00:35.340 | Again, that's drinkelement.com/huberman.

01:00:38.600 | One thing that I hope people are taking away from this study

01:00:43.060 | is that imagine you're somebody

01:00:45.500 | who has a very sensitive circadian mood system.

01:00:50.340 | Well, that would mean you need less daytime light exposure

01:00:54.660 | to feel good or less bad,

01:00:57.960 | but it also means that you might need

01:01:00.160 | very little light at night

01:01:01.960 | in order to negatively impact your mood systems.

01:01:06.960 | And in fact, they make this argument in the discussion

01:01:10.440 | as an interesting point that I think is worth mentioning,

01:01:14.100 | because here, again, what I like about this study

01:01:16.140 | is that they've separated day and nighttime light exposure.

01:01:19.100 | Turns out that many of the drugs

01:01:22.280 | that are used to treat bipolar disorder

01:01:24.500 | reduce the sense, are effective,

01:01:27.060 | perhaps in part because they reduce the sensitivity

01:01:30.820 | of the light-sensing circadian apparati.

01:01:35.220 | Now, that's interesting, right?

01:01:36.740 | If you think about this, okay,

01:01:38.180 | so these are drugs that can ameliorate

01:01:39.780 | some of the symptoms of bipolar,

01:01:42.200 | perhaps in part by reducing the extent

01:01:44.600 | to which nighttime light exposure

01:01:46.580 | can relieve bipolar symptoms,

01:01:48.760 | excuse me, can exacerbate bipolar symptoms.

01:01:51.180 | Conversely, there's evidence

01:01:53.720 | that people who take certain antidepressants

01:01:56.640 | may suppress the ability for daytime light

01:02:00.120 | to positively impact the mood systems of the brain.

01:02:02.780 | Now, of course, we don't want people halting their medication

01:02:05.400 | on the basis of that statement alone, please don't,

01:02:08.720 | talk to your psychiatrist.

01:02:09.840 | But if we know one thing for sure,

01:02:12.740 | it's that if you want a significant outcome

01:02:16.700 | and a paper as a scientist, give a drug, any drug,

01:02:21.700 | and look at the amount of rapid eye movement,

01:02:25.460 | sleep, or the circadian cycle,

01:02:27.200 | pretty much any drug alters the circadian rhythm

01:02:29.900 | for better or worse.

01:02:32.540 | But if we start to think about which medications

01:02:34.900 | might adjust our overall sensitivity to light,

01:02:37.540 | sometimes this could be a good thing.

01:02:38.780 | You think less sensitivity to light,

01:02:40.160 | well, for people who have bipolar disorder,

01:02:42.960 | the amount of daytime light exposure

01:02:44.640 | isn't that important for their overall mood regulation,

01:02:47.400 | but the amount of nighttime light exposure really is.

01:02:49.440 | In other words, darkness for eight hours every night

01:02:52.640 | should be viewed, in my opinion,

01:02:54.260 | as a treatment for bipolar disorder, not the only treatment.

01:02:57.560 | But it's also clear that we should all be avoiding

01:03:02.100 | really bright, extensive really bright

01:03:04.200 | nighttime light exposure.

01:03:05.280 | I mean, if anything, you know,

01:03:07.500 | my takeaway from this study is that darkness at night

01:03:10.260 | is the fourth key light stimulus.

01:03:12.620 | Now, a couple of things, very bright moonlight,

01:03:15.540 | very bright candlelight is probably only like,

01:03:19.380 | gosh, three to 50 lux.

01:03:24.460 | What?

01:03:25.300 | When you go outside on a brightly lit full moon night,

01:03:29.780 | I encourage people to download this free app,

01:03:31.780 | I have no relationship to it, called Light Meter,

01:03:33.860 | and it gives you a pretty good read

01:03:35.440 | of what the lux are in that environment.

01:03:37.460 | By the way, a lot of people don't realize this,

01:03:39.180 | they think you just tap the button

01:03:40.520 | and then it tells you how many lux, you hold it down,

01:03:42.420 | it's kind of fun, you can scan around the room

01:03:44.220 | and see how many lux are on average

01:03:46.080 | coming from that location or outside.

01:03:48.100 | Go out on a really bright moonlit night.

01:03:50.220 | - I mean, we should have a full moon tonight.

01:03:52.580 | - Yeah, let's do it.

01:03:53.900 | You're not gonna get above 100 lux.

01:03:55.620 | - That's incredible.

01:03:56.460 | - You're sitting at a candlelight dinner

01:03:58.100 | with your spouse or with friends,

01:03:59.920 | and it's clearly bright enough to see them.

01:04:02.100 | Put that lux meter right up, not too close to the flame,

01:04:06.860 | 50 to 200 lux.

01:04:08.860 | Maybe 400 lux. - I had an interesting

01:04:10.580 | experience a couple of months ago on an elk hunt

01:04:12.860 | where it was a full moon,

01:04:14.860 | which actually makes the hunting not so great,

01:04:17.700 | but it was the first time I've ever noticed

01:04:21.100 | my shadow in relation to the moon.

01:04:24.180 | That's how bright it seemed the light was.

01:04:26.580 | - This is Halloween appropriate.

01:04:28.100 | - Yeah, yeah, yeah. - Since we're coming up,

01:04:28.920 | we're recording this close to Halloween.

01:04:30.620 | - So, super interesting to think it could be that dim.

01:04:34.260 | - Campfire, you know, and firelight,

01:04:36.680 | you think, okay, gathering around a campfire,

01:04:38.320 | well then, okay, you know, everyone's circadian rhythm

01:04:40.500 | must have been disrupted for ages

01:04:42.060 | before the development of electricity.

01:04:44.980 | No, no, those campfires are extremely bright,

01:04:47.940 | but they're not that bright

01:04:51.340 | compared to a very densely overcast day.

01:04:53.900 | - And what is your phone

01:04:55.020 | if you don't use any sort of light mitigating tech on it?

01:05:00.020 | - Well, distance matters.

01:05:02.740 | - But at the distance we're holding it?

01:05:04.220 | - Yeah, so with all the wavelengths cranked up,

01:05:07.960 | so there is a nice feature intrinsic to the phone

01:05:11.360 | where you can eliminate the blues at night

01:05:12.920 | or, you know, this kind of thing,

01:05:14.320 | but if you crank it up to maximum light intensity,

01:05:16.280 | probably something like, you know, 500 to 1000 lux.

01:05:20.240 | Now, keep in mind though, it's additive, right?

01:05:23.560 | So it's over time.

01:05:24.800 | So lux is a measure of,

01:05:27.680 | I think it relates back to candelas

01:05:30.000 | is the amount of light shown,

01:05:32.120 | and I think it's like the one meter away

01:05:35.080 | and there's a squaring and a falling off of distance.

01:05:37.260 | We can look it up.

01:05:38.580 | These are old school measurements converted to lux,

01:05:41.780 | but keep in mind that if you're looking at your phone

01:05:43.740 | or tablet at 800 lux or 500 lux in the evening,

01:05:47.480 | and you do that for two hours,

01:05:49.080 | well, you're summing quite a lot of photons.

01:05:50.880 | Now it is true, and I do want to be fair to the biology,

01:05:55.680 | and it'd be dishonest to say anything different.

01:05:58.220 | You know, we've hammered on people

01:05:59.820 | about not shifting their circadian rhythm with light at night

01:06:03.100 | but we know that the middle of the day

01:06:04.260 | and the middle of the night are circadian dead zones.

01:06:06.020 | You can't shift your circadian rhythm that well

01:06:09.020 | in the middle of the day, in the middle of the night,

01:06:10.460 | but you can provide a wake up signal

01:06:13.220 | for your body and brain.

01:06:15.480 | It's really that sunrise and sunset that are critical.

01:06:17.660 | That's why I said there are four things,

01:06:18.720 | see sunrise or sun rising.

01:06:21.200 | You don't need to see across the horizon.

01:06:22.500 | Sunset, bright light during the day,

01:06:25.280 | minimize light exposure at night,

01:06:27.060 | and you don't need pitch black.

01:06:29.220 | In fact, pitch black probably just increases

01:06:31.060 | the frequency of injury.

01:06:32.060 | You know, I get up in the middle of the night,

01:06:33.220 | use the bathroom probably once, I think it's normal.

01:06:35.820 | I go back to sleep.

01:06:37.380 | You know, if it were pitch black, I'd probably injure myself.

01:06:40.140 | So just dim it down.

01:06:41.780 | Some people use red lights, you know,

01:06:43.580 | our friend Rick Rubin-- - Our mutual friend.

01:06:45.180 | - Yeah, Rick Rubin is--

01:06:46.780 | - Can we tell a funny story about Rick?

01:06:48.060 | - Yeah, of course, of course.

01:06:49.540 | - You know this story, but just in case Rick's listening,

01:06:52.380 | he'll appreciate this.

01:06:53.540 | You know, when Rick was here staying last summer,

01:06:58.020 | he's up in our guest house and he came down

01:07:00.300 | after the first night and he was like unacceptable.

01:07:03.300 | You know what I'm talking about, right?

01:07:04.140 | - Yeah, unacceptable accommodations.

01:07:05.580 | - What did he do?

01:07:06.780 | - He removed all the lighting that existed in that room

01:07:10.340 | and replaced it with red light bulbs,

01:07:12.340 | which I later used when I stayed here

01:07:14.380 | and then later stole when I left.

01:07:16.580 | The teenage me, I took them.

01:07:19.820 | I took them, that's why Derek, when he stayed here,

01:07:21.860 | Elaine or anyone else, he stayed here.

01:07:24.380 | Conti or anyone else didn't have those.

01:07:25.920 | I took them, I love them.

01:07:27.300 | - So funny.

01:07:28.320 | Jill is like, you know Rick changed every light

01:07:31.780 | in the guest house to red?

01:07:33.360 | I'm like, yeah, no, I didn't know that,

01:07:35.340 | but I'm not surprised.

01:07:36.300 | - Yeah, well in his place,

01:07:37.420 | he has mostly either no lighting or red lighting.

01:07:40.860 | So during the day, he just goes by ambient light

01:07:43.760 | and then red light in the evening or candlelight.

01:07:46.820 | And it's great.

01:07:47.700 | And you know, people hear red lights

01:07:49.060 | and they think they have to buy

01:07:49.940 | these expensive red light units,

01:07:51.580 | but that's not what we're talking about.

01:07:53.220 | You can literally buy red party lights or just a red bulb.

01:07:56.780 | Some people say, well, can I just use a red film

01:07:58.920 | or can I put a t-shirt over the lamp?

01:08:00.420 | I worry about people putting t-shirts over lamps

01:08:02.260 | because of the fire hazard.

01:08:03.620 | But I'll be honest, I dim the lights in my home at night.

01:08:07.820 | When I travel, sometimes I will bring one of the stolen

01:08:11.060 | from Rick Rubin red lights.

01:08:12.820 | - The Rick Rubin's dash.

01:08:14.380 | Here's something where I've sort of softened my tune.

01:08:16.780 | So I used to be kind of a hard liner,

01:08:19.500 | no blue light in the evening guy, you know,

01:08:22.980 | had the, you know, everything was red light at night

01:08:25.900 | as far as my phone using flux on the computer,

01:08:29.780 | you know, whatever it was.

01:08:31.020 | I suspect that that matters somewhat,

01:08:36.060 | but I think what matters more is the stimulation

01:08:38.860 | that may come from those things.

01:08:40.540 | And what I've come to realize, at least in me,

01:08:44.800 | which means it probably is true in others as well,

01:08:47.180 | and at least some others,

01:08:48.820 | is that what I'm doing on my phone matters more

01:08:53.620 | than how bright my phone is.

01:08:55.580 | In other words, if I've got the best blue light filter

01:08:58.460 | in the world on my phone,

01:08:59.780 | but I'm doom scrolling social media

01:09:02.460 | and getting lit up on email,

01:09:04.820 | that's way worse for me

01:09:07.020 | than if I've got my phone on maximum light

01:09:10.660 | and I'm like watching YouTube videos of F1 cars

01:09:15.660 | and driving around having fun.

01:09:17.280 | Like, it's a totally different experience.

01:09:20.260 | So the context matters.

01:09:22.380 | And I think for that reason,

01:09:24.880 | I would want people to be mindful of the whole picture.

01:09:28.220 | You know, going to bed under a period of, you know,

01:09:31.260 | intense duress brought on by something, you know,

01:09:35.440 | that's an equally dangerous component to all of this.

01:09:38.380 | That's distinct from what we're talking about.

01:09:40.480 | But, you know, I want people to be able to think of this

01:09:42.760 | in the context of everything.

01:09:43.920 | - Yeah, it's a really important point.

01:09:46.500 | One thing I'll say is that if you're going to stay up

01:09:50.580 | past your normal bedtime,

01:09:52.060 | if you're going to get a lot of light in your eyes,

01:09:54.020 | I would hope that it would be for fun reasons

01:09:55.820 | and for reasons you enjoy.

01:09:57.060 | You should definitely spend some nights out.

01:09:58.600 | You should definitely do some all-nighters studying

01:10:00.800 | if you really, you know,

01:10:01.720 | if it's going to help you get the grade that's permanent.

01:10:04.520 | Right, I'd certainly have done all-nighters studying

01:10:06.740 | and grant writing for years.

01:10:09.220 | You know, there are going to be the inevitable all-nighters

01:10:11.820 | due to, God forbid, a trip to the hospital

01:10:14.520 | or you heard something on the news that really amped you up

01:10:17.540 | or you just simply can't sleep.

01:10:18.740 | That stuff is going to happen.

01:10:20.620 | So I think the goal should be

01:10:22.020 | to minimize light exposure at night.

01:10:23.500 | And I think what you just said is especially true

01:10:26.420 | because we don't know, for instance,

01:10:28.460 | people talk about the negative impact of social media.

01:10:30.540 | Is it the fact that people are looking at this little box

01:10:33.220 | for so many hours per day?

01:10:34.540 | Is it all the things they're not doing?

01:10:36.340 | Is it what they're looking at per se?

01:10:39.100 | All of those things interact and are really important.

01:10:41.700 | We know based on studies from the Stanford Sleep Lab

01:10:44.980 | that if you wake up in the middle of the night,

01:10:46.780 | looking at what time it is can be very disruptive

01:10:49.780 | to your ability to fall back asleep

01:10:51.900 | and to your sense the next day.

01:10:53.780 | It's a placebo effect, but it's a powerful one

01:10:56.300 | of how tired you are the next day.

01:10:58.100 | They've done this where they wake people up

01:11:00.460 | in the middle of the night and then they say it's 4 a.m.

01:11:03.500 | versus 2 a.m. versus 6 a.m.

01:11:05.580 | And people's perceived levels of energy during the day

01:11:08.700 | in some ways correlate with what they think,

01:11:10.900 | how much sleep they think they got.

01:11:13.020 | Likewise, and this is one of the concerns,

01:11:14.940 | potential concerns with sleep trackers.

01:11:17.380 | Allie Crum talked about this when she came on our podcast.

01:11:21.020 | If people see a poor sleep score,

01:11:22.900 | they often feel worse than if they see a good sleep score.

01:11:26.180 | Now, of course, physiology matters.

01:11:27.900 | You can't lie to yourself and say you got a great night's

01:11:31.140 | sleep simply by virtue of a sleep score.

01:11:32.720 | But I worry more about the false,

01:11:36.460 | well, if it's a false negative,

01:11:38.020 | that we don't wanna put valence on this.

01:11:39.800 | Seeing a bad sleep score and then deciding

01:11:41.700 | that you're gonna have a terrible day.

01:11:43.900 | I think a bad sleep score is an indication

01:11:45.660 | that you might need to dial some things in a bit better.

01:11:48.580 | Getting a great sleep score is an indication

01:11:50.340 | that you might be doing a number of things right

01:11:52.060 | and start looking at these things as averages.

01:11:54.180 | Would you agree?

01:11:55.020 | - Yeah, completely.

01:11:55.860 | I don't think it's that different from CGM, right?

01:11:58.240 | Like I think that CGM is an amazing tool to provide insight.

01:12:02.260 | And you pretty much know the insights

01:12:04.140 | after a relatively short period of time.

01:12:06.380 | 30 days, maybe at the outside 90

01:12:08.980 | for a person with a very complicated life.

01:12:12.020 | And you know all you need to know

01:12:14.260 | about how the inputs affect the output.

01:12:16.860 | Thereafter, if you choose to use it, it's a behavioral tool.

01:12:20.040 | In other words, you're using this

01:12:21.580 | to build in a Hawthorne effect.

01:12:23.340 | I think the same is largely true with sleep trackers.

01:12:26.340 | Most people have this profound sense of learning

01:12:29.720 | when they first encounter one of these things.

01:12:32.180 | And it's again, you've heard it all a hundred times.

01:12:33.780 | Oh my God, I can't believe what alcohol does to my sleep.

01:12:37.020 | - Or caloric trackers.

01:12:38.140 | - Exactly.

01:12:38.980 | - Like I think a Lane Norton's app, Carbon,

01:12:41.200 | I have no financial relationship to it.

01:12:43.360 | I use it and it's taught me, wow,

01:12:47.060 | like I consume a lot of calories

01:12:48.880 | in the form of certain things at certain times of day.

01:12:50.980 | And there's just a lot of good learning in that.

01:12:52.860 | - But it's the act of tracking that helps you manage it.

01:12:55.560 | And similarly, I think it's the act of knowing

01:12:59.020 | you're gonna be looking at that score that gamifies it,

01:13:01.600 | that kind of helps people do the right things.

01:13:03.700 | Oh, you know what?

01:13:04.540 | I'm not gonna have that drink tonight,

01:13:05.340 | or I'm not gonna eat that snack before bed

01:13:07.620 | because I've now been conditioned

01:13:08.980 | to see how that impacts score.

01:13:10.420 | That said, I think that, you know,

01:13:13.680 | recovery scores and things like that

01:13:15.740 | are just notoriously poor at predicting performance.

01:13:20.260 | And I think there's a reason that serious athletes

01:13:23.060 | would never use things like that.

01:13:25.100 | They would tend to rely on the more tried and true methods

01:13:27.920 | of predicting behaviors, such as heart rate,

01:13:31.100 | maybe heart rate variability,

01:13:32.480 | but morning resting heart rate,

01:13:33.840 | probably more predictive than anything else.

01:13:35.700 | And then, you know, in workout things such as heart rate,

01:13:38.520 | heart rate recovery, lactate threshold, things like that.

01:13:40.680 | So yeah, I agree.

01:13:42.380 | I think we have to, and I say this as a guy

01:13:44.480 | who's generally perceived to be the most pro-device guy

01:13:48.420 | in the world, people would be surprised

01:13:50.180 | how sparingly I use things like that.

01:13:54.320 | - I mean, I do some tracking, not as much as you.

01:13:57.860 | I love things that seem to work the first time

01:14:02.760 | and every time in terms of our natural biology,

01:14:05.580 | based on a couple of criteria.

01:14:07.680 | There's an established mechanism.

01:14:09.380 | It's been explored in the context of pathology,

01:14:12.460 | like mental health disorders,

01:14:13.860 | as well as pro-health in healthy individuals.

01:14:17.940 | That it make really good sense

01:14:19.620 | at the level of kind of wellness

01:14:23.820 | and let's just say ancient health.

01:14:26.660 | You know, when you talk about getting a lot of sunlight

01:14:28.340 | during the day, like a lot of people will say,

01:14:30.080 | well, of course, get outside and play.

01:14:32.260 | Not getting too much light at night.

01:14:33.940 | Of course, this is just good old quote unquote,

01:14:35.740 | good old fashioned advice.

01:14:37.760 | People spend 90% of their time indoors now.

01:14:40.680 | Their daytime environments are too dim.

01:14:43.380 | Their nighttime environments are too bright.

01:14:45.860 | And this kind of misleading aspect of artificial light

01:14:50.820 | that when you see a bright bulb,

01:14:52.460 | you think I'm getting a lot of photons

01:14:55.320 | is part of the problem.

01:14:56.940 | And the fact that when you're out on an overcast day

01:14:59.420 | and it, you know, you think there's sun,

01:15:01.280 | quote unquote, isn't out.

01:15:02.940 | Well, it's hidden by cloud cover,

01:15:04.820 | but just think about how well you can navigate

01:15:07.460 | that environment without a flashlight

01:15:10.380 | versus at night where you would require a flashlight.

01:15:13.780 | We evolved under this traumatic difference

01:15:16.420 | in day/night availability of photons,

01:15:20.060 | independent of whether or not you can quote see the sun.

01:15:23.420 | And it's just very clear that all the mechanisms

01:15:27.540 | in our brain and body that regulate mood

01:15:29.380 | are just powerfully regulated by this stuff.

01:15:31.580 | So I've made it a point to really reduce

01:15:33.980 | the amount of nighttime light that I'm getting,

01:15:36.260 | but I'm less concerned about flipping on the light switch

01:15:39.060 | to use the bathroom as I used to be.

01:15:41.140 | I used to think, oh, I'm like quashing all my melatonin.

01:15:43.420 | This is terrible.

01:15:45.000 | I know I can't shift my circadian clock then.

01:15:46.980 | I know that that light, yes, while it's bright,

01:15:49.460 | if it's brief, I'm not gonna worry about it too much.

01:15:51.740 | Would it be better to have a, you know,

01:15:53.900 | a dim light on as opposed to a bright light?

01:15:55.840 | Sure, but I'm not gonna stress it in a hotel bathroom

01:15:58.340 | or something.

01:15:59.180 | I'm not gonna walk around, you know, shielding my eyes.

01:16:00.800 | So people sometimes ask me, by the way,

01:16:02.700 | is it different to look at the phone directly

01:16:05.140 | versus if you tilt the phone away?

01:16:07.060 | Well, it absolutely is.

01:16:08.400 | I mean, think about a flashlight shown on the ground

01:16:10.540 | in front of you.

01:16:11.580 | Very few photons getting in your eyes

01:16:13.860 | versus shown directly in your eyes.

01:16:15.600 | Think about ambient light from the sun going everywhere

01:16:17.980 | versus looking in the general direction of the sun.

01:16:20.480 | So east in the morning, west in the afternoon, of course.

01:16:23.820 | The directionality of the light matters.

01:16:25.980 | So I'm not saying it, you know,

01:16:28.300 | that you need to like peek at your phone

01:16:30.460 | as if you're looking, you know,

01:16:31.580 | over the edge of a bowl or something into it.

01:16:34.260 | But my friend, Samir Hattaru,

01:16:36.140 | who's head of the chronobiology unit at the National

01:16:38.320 | Institutes of Mental Health,

01:16:39.860 | we used to room together at meetings.

01:16:41.340 | We stopped 'cause he's a terrible snorer.

01:16:43.440 | So I just could, there were a few times

01:16:44.980 | when I considered suffocating him in the middle of the night

01:16:47.420 | since he was already suffocating himself.

01:16:49.000 | Now we just, we don't stay in the same rooms anymore.

01:16:51.960 | We're no longer postdocs.

01:16:53.600 | But I caught him looking at his phone

01:16:56.540 | in the middle of the night and he would tilt it like away,

01:16:59.440 | like he's holding a platter for those

01:17:01.040 | that are just listening and kind of like looking over

01:17:03.200 | at the screen there and like, what are you doing?

01:17:04.600 | This is ridiculous.

01:17:05.440 | I'm trying not to get so much light in my eyes.

01:17:07.320 | That's a little extreme, but I think it illustrates the point

01:17:09.960 | which is how much direct light exposure

01:17:11.640 | you get at night matters.

01:17:12.900 | How much direct sunlight exposure you get,

01:17:15.960 | especially early in late, early morning, late afternoon

01:17:19.000 | and throughout the day, it really matters.

01:17:20.980 | - Now remind me, Andrew,

01:17:21.960 | what is the wavelength of sunlight?

01:17:24.740 | - Great, so sunlight is going to include all visible,

01:17:27.520 | visible spectrum, right? - Which runs from how many?

01:17:29.400 | - Yeah, so, well, let's, let's,

01:17:31.040 | we can answer two questions there.

01:17:32.280 | This wrist sensor detected 470 degree,

01:17:36.080 | 70 nanometer to 650 nanometer light.

01:17:39.440 | So that's going to be blue and ultraviolet.

01:17:42.500 | Ultraviolet's kind of-

01:17:43.340 | - Yeah, that's kind of like blue to orange.

01:17:45.000 | - Yeah, blue, blue to orange.

01:17:46.600 | That's what this was measuring.

01:17:47.960 | So red light is going to be more like 680.

01:17:50.500 | Far red is getting out to 7720 and up, upwards of that.

01:17:55.320 | Blue light is going to fall somewhere in the low fours.

01:17:58.940 | Ultraviolet is getting down into the high threes

01:18:01.700 | and, and lower.

01:18:02.900 | And so these, these spectra of light.

01:18:04.860 | So during the day, you know, midday light,

01:18:06.720 | you're getting what looks like white light.

01:18:07.980 | You'll see, oh, the sky's blue

01:18:09.220 | and the sun is bright white light.

01:18:11.220 | It's not even yellow to your eye.

01:18:13.260 | And of course don't stare at it,

01:18:14.540 | especially in the middle of the day.

01:18:16.380 | You're getting all visible spectra.

01:18:18.800 | So you're getting everything from UV

01:18:20.360 | all the way out to red light.

01:18:22.020 | It's just coming in at equal intensities.

01:18:24.100 | - So is that a potential limitation of this study

01:18:26.280 | in that it didn't have a sensor that could pick up

01:18:28.940 | the full spectrum of light?

01:18:31.520 | - Potentially, especially since they're, you know,

01:18:34.300 | we don't think of humans as UV capable.

01:18:37.240 | Like we can't perceive UV light.

01:18:39.480 | Like a ground squirrel for instance can,

01:18:42.240 | has UV sensors in its eyes.

01:18:44.080 | Turns out, you know why they use this, it's crazy.

01:18:46.120 | They actually, you know, when the ground squirrels sit up

01:18:48.480 | on their haunches, they're actually signaling one another.

01:18:50.720 | They rub urine on their belly and it reflects UV.

01:18:53.840 | The New York Times for some reason has been running

01:18:57.220 | a series of papers or articles rather about

01:19:01.620 | naturally occurring fluorescence at night

01:19:04.120 | and all sorts of these scorpions and monotremes

01:19:07.180 | like the platypus.

01:19:08.480 | No one really knows the reason for these odd,

01:19:11.000 | odd wavelength of light emissions for all these animals.

01:19:14.100 | But you know, we view things in the blue, violet

01:19:17.820 | and up to red and you know, we're not pit vipers.

01:19:20.800 | We can't see far red, but we can see lower than 470 nanometers

01:19:25.580 | and we can see higher than 650.

01:19:29.160 | - Is there a technology reason why they had such a narrow

01:19:32.060 | band in these sensors?

01:19:33.120 | Is it not possible that they could have used a wrist sensor

01:19:35.180 | that was wider?

01:19:36.880 | - This study was initiated in 2013.

01:19:39.560 | The tech was probably far worse than it is now.

01:19:42.360 | Again, I would love for somebody to design an eyeglass

01:19:46.180 | where it's measuring how many photons you're getting

01:19:49.480 | across the day.

01:19:50.600 | I'm not a big fan of having everything be amplified.

01:19:53.520 | So I would love it if the frame would just shift color

01:19:56.220 | across the morning.

01:19:57.060 | Like you go outside on a cloudy day, you know,

01:19:58.700 | you wear these glasses and by the way,

01:20:01.240 | it's fine to wear eyeglasses or contacts

01:20:03.020 | for sunlight viewing for setting your circadian rhythm.

01:20:06.100 | People always say, well, why is that okay

01:20:08.000 | when a window's not?

01:20:08.840 | Well, corrective lenses are actually focusing the light

01:20:11.060 | onto your retina.

01:20:12.100 | The windows and windshields are scattering the light

01:20:15.000 | and filtering.

01:20:15.840 | - And how much are sunglasses filtering this out?

01:20:18.380 | - Way too much.

01:20:19.760 | We can safely say way too much.

01:20:21.100 | Probably causing a tenfold decrement in the total lux count

01:20:26.100 | that's landing on your retina.

01:20:28.060 | But of course, you know, sunglasses are important

01:20:30.200 | driving into sun and some people have very sensitive eyes.

01:20:34.140 | I can't sit at a cafe with a bright,

01:20:36.760 | brightly reflective table in the afternoon.

01:20:39.460 | I just squint like crazy.

01:20:41.100 | I can't do it.

01:20:41.940 | My dad who's, you know, darker eyed

01:20:43.560 | and, you know, he's South American descent.

01:20:45.780 | He, you know, he can just sit there just fine.

01:20:48.620 | My mom who's got light eyes like me, you know,

01:20:50.780 | we're like, ah, it's really tough.

01:20:53.140 | We just have a terrible time.

01:20:54.600 | You know, people differ in their light sensitivity.

01:20:58.900 | - So there's one other macro question I have here

01:21:02.300 | and it's not answerable because without randomization

01:21:05.820 | we can't know it.

01:21:07.380 | But it's the question of how much reverse causality

01:21:12.220 | can exist in these observations.

01:21:14.460 | So again, these observations demonstrate

01:21:18.100 | very tight correlations, very strong associations,

01:21:21.180 | especially in the five areas that we highlighted.

01:21:23.680 | But it's possible that part of what we're seeing

01:21:30.220 | is reverse causality brought on by both the treatments

01:21:35.180 | which you've already kind of alluded to

01:21:36.780 | and also the condition itself.

01:21:38.620 | - You want to explain reverse causality for people

01:21:40.580 | and maybe could you mention for those that missed

01:21:43.480 | the Hawthorne effect just 'cause you, yeah.

01:21:45.700 | - Yeah, the Hawthorne effect is an effect that is named

01:21:50.700 | after an observation of what took place in a factory

01:21:54.780 | where they were actually studying worker productivity

01:21:57.380 | with light of all things.

01:21:59.580 | But what it refers to is the idea that people

01:22:02.940 | will change their behavior when they are observed.

01:22:05.260 | So if I said, well, I really want to know

01:22:09.460 | what a day in the life is like for Andrew Huberman,

01:22:12.380 | I'm going to follow him around for a day.

01:22:14.720 | It's very unlikely that his behavior that day

01:22:17.620 | will be exactly as it was if I wasn't there.

01:22:20.500 | And so the reason-- - Which is why you probably

01:22:21.580 | will never see a day in the life of Andrew Huberman.

01:22:24.480 | Although it's pretty scripted unless I'm traveling.

01:22:27.380 | It's morning sunlight, hydration and some cardio

01:22:31.860 | or weight training and then a lot of time reading papers.

01:22:34.420 | It'd be the most boring video in the world

01:22:35.740 | 'cause it's mostly me reading and underlining things.

01:22:39.060 | - But it's why gamifying things can be beneficial, right?

01:22:43.920 | It's why a CGM can be beneficial

01:22:47.460 | because it's sort of like somebody's watching you

01:22:49.860 | and you're going to modify what you eat in response to it

01:22:52.080 | or why tracking can really be an effective way

01:22:55.040 | to reduce input because there's a sense of being monitored

01:22:59.060 | by doing that, especially if someone literally monitors it.

01:23:01.260 | In other words, you can set up an accountability partner

01:23:03.720 | where your health coach or someone

01:23:05.280 | is actually seeing the data.

01:23:07.060 | So that's what it is.

01:23:08.300 | Now, as far as reverse causality,

01:23:10.500 | when you look at variables,

01:23:12.960 | so let's just pick a common one that's unrelated to this.

01:23:17.960 | So there's an association that more diet soda consumption

01:23:23.240 | is associated with greater obesity.

01:23:26.380 | It's a bit paradoxical, right?

01:23:28.700 | - Is that true?

01:23:29.540 | - It is, yeah, it's been demonstrated in many series

01:23:32.780 | that the greater the consumption of diet soda,

01:23:36.100 | the greater the prevalence of obesity.

01:23:39.300 | And that has been postulated by some to suggest

01:23:42.740 | that non-nutritive sweeteners,

01:23:44.800 | such as aspartame or sucralose or things like that,

01:23:47.580 | are actually part of what's causing obesity.

01:23:50.300 | And while there are probably some arguments you could make

01:23:54.660 | around the impact that those things might have

01:23:56.420 | on the gut microbiome,

01:23:57.540 | and maybe there's some way and that's happening,

01:23:59.780 | it's also equally likely, if not probably more likely,

01:24:02.300 | that there's reverse causality there.

01:24:03.980 | That a person who is obese is therefore contemplating

01:24:08.640 | how much they're eating or thinking,

01:24:10.300 | hey, what's an easy way that I can reduce calories?

01:24:13.380 | How about instead of drinking a Coke, I drink a Diet Coke?

01:24:16.700 | And so there, the causality which you would impute to mean

01:24:20.780 | the drink is causing the obesity,

01:24:23.220 | it might be, no, the obesity is causing the choice of drink.

01:24:26.440 | So here, the question is,

01:24:28.560 | how much of the effect we're seeing

01:24:32.180 | is a result of the condition that's being studied, right?

01:24:36.120 | How much of the disruption in both day and night,

01:24:40.960 | light exposure is the result of the depression?

01:24:44.040 | It's dysregulating the sleep.

01:24:45.920 | Maybe they're sleeping more during the day

01:24:47.820 | and more awake at night because of depression.

01:24:50.280 | Again, these are, you can't know this.

01:24:52.600 | This is where epidemiology never allows us to determine this.

01:24:56.960 | And sadly, these questions can only be answered

01:24:59.800 | through either direct randomization

01:25:01.880 | or Mendelian randomization,

01:25:03.660 | which by the way, I was going to also ask you,

01:25:05.460 | do you know if anyone has examined this

01:25:07.780 | from a Mendelian standpoint?

01:25:09.860 | That would be very interesting

01:25:12.080 | 'cause I have to believe, well, it would be interesting.

01:25:15.940 | I don't know enough about the biology

01:25:17.400 | to know what SNPs would be studyable,

01:25:20.340 | but that would be interesting.

01:25:22.020 | - What Peter is saying is, you know,

01:25:23.500 | if you knew something about the genomes of these people,

01:25:27.000 | you would be in a great position to perhaps even link up

01:25:32.500 | light susceptibility genes or like sensitivity genes

01:25:36.800 | with genes for, you know,

01:25:38.480 | pathways involved in major depression, bipolar.

01:25:41.160 | I mean, getting to this issue of reverse causality,

01:25:43.200 | I mean, I think it's very straightforward to imagine

01:25:45.840 | that the person who's experiencing a manic episode

01:25:48.200 | is going to be up for two weeks at a time, sadly,

01:25:52.600 | and getting a lot of nighttime light exposure.

01:25:55.940 | You know, now nighttime dark exposure

01:25:59.200 | as a treatment for bipolar

01:26:00.280 | is something that people are starting to talk about.

01:26:02.440 | So making sure that even those people are awake,

01:26:04.300 | that they're at least blue blocking at night,

01:26:07.020 | reducing their online activities,

01:26:08.520 | but people with severe manic episodes

01:26:10.360 | have a hard time regulating their own behavior, of course.

01:26:13.800 | - And it's not one or the other.

01:26:14.880 | Like, I don't want the question to come across

01:26:16.900 | to the listener that it has to be one or the other.

01:26:19.640 | It's only A can cause B or B can cause A.

01:26:23.140 | No, it's actually,

01:26:24.360 | a lot of times these things feed off each other.

01:26:26.120 | Going back to the soda example,

01:26:28.120 | I actually think there's a bit of both, right?

01:26:30.440 | I actually think there's a real clear body habitus

01:26:34.020 | dictates beverage choice,

01:26:35.920 | but I also am starting to think

01:26:38.280 | that in susceptible individuals,

01:26:41.720 | non-nutritive sweeteners will alter the gut biome

01:26:44.200 | and that alters metabolism.

01:26:46.360 | - What about just hunger?

01:26:48.720 | I remember Lane telling me,

01:26:51.160 | and I've seen at least one of the studies that, you know,

01:26:54.580 | water is probably better for us than diet soda,

01:26:57.560 | but that for some people,

01:26:59.400 | diet soda is a great tool for reducing caloric intake.

01:27:03.600 | I also know some individuals, not me, who drink diet soda.

01:27:07.880 | I drink diet soda from time to time,

01:27:09.780 | mainly Stevia sweetened sodas,

01:27:11.320 | but what I'm referring to here are people besides myself

01:27:15.140 | who drink diet soda and it seems to stimulate their appetite.

01:27:18.560 | There's something about the perception of sweet

01:27:21.840 | as driving hunger, whereas not eating,

01:27:26.520 | or drinking anything with any sweetness

01:27:29.080 | doesn't seem to.

01:27:29.920 | This is one of the things I wonder if it impacts

01:27:33.000 | why some people like intermittent fasting,

01:27:35.520 | because for some people, you know,

01:27:37.160 | just even the perception,

01:27:38.720 | I wonder if the perception of sweetness

01:27:41.800 | or even just the smell of food we know

01:27:44.160 | can stimulate appetite.

01:27:45.280 | So you can mention the perception of sweetness in the mouth,

01:27:47.580 | even if there's no calories there.

01:27:49.520 | I don't think it necessarily makes people hypoglycemic,

01:27:52.680 | but perhaps it makes them think about,

01:27:54.800 | hmm, like sweet means food.

01:27:57.160 | For instance, for years, I loved the combination

01:27:59.440 | of a diet Coke and a slice of pizza

01:28:01.160 | whenever I was in New York, ideally two slices of pizza.

01:28:05.040 | So now every time I have a diet Coke,

01:28:07.000 | which isn't that often, but I like diet Coke,

01:28:08.720 | especially with a little bit of lemon in it,

01:28:10.720 | I just think about a slice of cheese

01:28:12.960 | or mushroom or pepperoni pizza.

01:28:14.800 | It's like, I want it, I crave it more.

01:28:16.440 | So there's a paired association there that I think is real.

01:28:19.020 | And we know based on Dana Small's lab at Yale,

01:28:22.080 | that there's this paired association

01:28:23.740 | between the sweetness from sucralose

01:28:26.860 | and that there's an insulin response.

01:28:28.800 | They actually had to cease the study in kids

01:28:30.620 | because they were becoming pre-diabetic,

01:28:33.980 | which unfortunately meant the study was never published.

01:28:36.260 | Have you talked to Dana on this podcast?

01:28:37.800 | No, we wrote a premium newsletter on this

01:28:41.760 | several months ago.

01:28:42.920 | It's gotta be like, I don't know, 10 to 20,000 words

01:28:47.920 | on all things related to sugar substitutes.

01:28:51.040 | Okay, I'll read that.

01:28:51.880 | So yeah, folks who are interested in this topic,

01:28:54.840 | I would refer them to the premium newsletter

01:28:56.700 | on sugar substitutes.

01:28:57.580 | I think it was our September edition.

01:29:00.120 | The short of it is the data are a little bit noisy,

01:29:03.700 | but there is indeed some sweeteners in some studies

01:29:08.560 | do result in that phenomenon.

01:29:10.740 | You described the cephalic insulin response.

01:29:13.040 | And I came away from the research that went into that,

01:29:17.820 | which was a Herculean effort on the part of the team.

01:29:23.360 | A little bit more confused than when I went in,

01:29:26.360 | but being even more cautious around artificial sweeteners

01:29:31.360 | than I was going in.

01:29:33.440 | And not for the reasons that I don't necessarily,

01:29:37.160 | I didn't find any evidence

01:29:39.440 | that these things are cancer causing, right?

01:29:40.980 | So that's the headline stuff people worry about.

01:29:42.720 | You have to ingest like 10 grams or something crazy.

01:29:46.560 | I came away more confident

01:29:48.560 | that from a long-term safety perspective,

01:29:51.760 | in terms of cancer and catastrophic outcomes like that,

01:29:55.000 | that wasn't the issue.

01:29:56.080 | But I came away much more cautious

01:29:58.480 | around these things can really be mucking around

01:30:00.920 | with both your brain chemistry and your gut chemistry,

01:30:04.320 | which can pertain to your metabolism.

01:30:07.320 | And therefore my takeaway was buyer beware,

01:30:11.120 | use limited amounts only.

01:30:12.980 | The one, by the way,

01:30:15.000 | that still emerged to me as a reasonable one,

01:30:17.760 | it's the only one that I use

01:30:19.120 | and I've talked about it a lot is Xylitol.

01:30:21.840 | Xylitol is the, pardon me, I shouldn't say it,

01:30:24.880 | Xylitol for chewing, so for gum and Allulose as an additive.

01:30:29.880 | So those were the-

01:30:31.800 | - Safer, you're saying?

01:30:32.640 | - Yes, those are basically the only two I will consume.

01:30:35.920 | - I'll drink a Diet Coke every now and again,

01:30:38.680 | if I'm on a plane or something.

01:30:40.300 | You know, this law that got passed a few years ago,

01:30:42.160 | you couldn't bring liquids of your own

01:30:44.680 | into the airport and onto the plane.

01:30:46.360 | Like what a great- - A few years ago.

01:30:47.960 | - What a great scheme.

01:30:49.440 | What a great scheme to get people

01:30:51.120 | to buy overpriced fluids in the airport.

01:30:54.320 | Like, I mean, there are more important issues in the world,

01:30:56.920 | but like this one really gets me.

01:30:58.560 | But yeah, I use a little bit,

01:31:01.720 | I drink things with a little bit of Stevia in it,

01:31:04.280 | the occasional Diet Coke and I generally avoid sucralose.

01:31:09.280 | I don't like the way it tastes, monk fruit's too sweet,

01:31:11.920 | but yeah, maybe we'll do a podcast on that in the future.

01:31:15.640 | Okay, so I think we can wrap this paper 'cause I really-

01:31:18.480 | - Well, but tell me what you think about that point, Andrew.

01:31:20.560 | Like how, I mean, you know more about this stuff than I do,

01:31:25.080 | but if you had to just lay on your judgment, right?

01:31:30.040 | So if it were 100 to zero, you would say,

01:31:33.960 | the light is 100% causal in the effects we're seeing.

01:31:38.960 | If it were zero to 100, you'd say, nope,

01:31:42.200 | the behavior is 100% causal of the exposure to light.

01:31:47.040 | Where do you, again, you can't know it.

01:31:49.680 | What does your intuition tell you?

01:31:50.880 | - Okay, there's my intuition

01:31:51.980 | and then there's my recognition of my own bias

01:31:54.800 | because I started working on these circadian pathways

01:31:59.800 | originating in the eye back in the year in '98

01:32:03.400 | as a graduate student at Berkeley.

01:32:04.760 | The cells, these melanopsin intrinsically sensitive

01:32:07.400 | retinal ganglion cells were discovered in the early 2000s

01:32:09.840 | by a guy named Iggy Provencio, Dave Burson,

01:32:12.120 | Sam Rattar, Sachin Pant, and others.

01:32:14.760 | And it was one of the most important discoveries

01:32:16.540 | in all of biology, clearly.

01:32:18.640 | So I've been very excited about these systems,

01:32:22.280 | but if I set that aside, so bias disclosure made,

01:32:27.200 | I think 65 to 75% of the effects

01:32:33.700 | are likely due to light directly.

01:32:36.280 | Now it's impossible to tease those apart, as you mentioned,

01:32:38.880 | but to play devil's advocate against myself,

01:32:43.120 | you know, you could imagine that the depressed individual

01:32:46.020 | is laying around indoors with the curtains drawn.

01:32:49.260 | They didn't sleep well the night before,

01:32:50.860 | which gives you a photo sensitivity that isn't pleasant.

01:32:53.480 | Like it sucks to have bright light in your eyes

01:32:55.320 | first thing in the morning, especially if you didn't sleep

01:32:58.560 | well, and then they're, you know, making their coffee

01:33:01.660 | in a dimly lit, what they think is brightly lit environment.

01:33:04.320 | And then they're, you're looking at their phone

01:33:06.040 | and the state of the world sucks

01:33:08.060 | and their state of their internal landscape is rough.

01:33:11.660 | And they're, maybe they're dealing with a pain or,

01:33:16.040 | or, you know, an injury or something.

01:33:18.560 | And their likelihood of getting outside is low.

01:33:21.080 | And when they do get outside, they're going to shuffle

01:33:22.960 | and not, you know, so I could see how the behaviors

01:33:25.440 | could really limit the amount of light exposure.

01:33:27.960 | And then evening rolls around, they've been tired all day

01:33:30.760 | and a common symptom of depression, you fall asleep.

01:33:32.980 | And then two or three in the morning, they're wide awake.

01:33:35.460 | What are you going to do at two or three

01:33:36.540 | when you're wide awake, sit in the dark?

01:33:38.520 | No, you're going to get online.

01:33:40.900 | You're going to listen to things.

01:33:41.960 | You might have, I'm not recommending this,

01:33:44.140 | but an alcoholic drink in order to try and fall asleep.

01:33:46.920 | I mean, this is the pattern.

01:33:48.280 | And so, you know, shaking up that pattern is really what,

01:33:51.800 | what so much of my public health work these days is about

01:33:56.380 | and trying to get people onto a more natural daylight,

01:33:58.860 | night, night, dark rhythm.

01:34:01.600 | But yeah, it's impossible to tease apart.

01:34:03.040 | We do know this, and this is really serious.

01:34:05.680 | We know that in almost every instance,

01:34:09.200 | almost every psychopathology report of suicide in the weeks,

01:34:14.200 | but especially in the days preceding suicide,

01:34:17.780 | that person's circadian rhythms looked almost inverted

01:34:21.220 | from their normal patterns.

01:34:22.980 | And that's true of non-bipolar individuals as well.

01:34:28.340 | You know, circadian disruption and disruption

01:34:30.500 | in psychiatric health are inextricable.

01:34:33.580 | Conversely, positive mood and affect

01:34:37.540 | and circadian behavior seem very correlated.

01:34:41.160 | I mean, I think it's clear that

01:34:44.400 | if you want to become an early riser,

01:34:45.660 | get light in your eyes and get activity

01:34:47.360 | in your body early in the day.

01:34:49.040 | You built, you entrained to those rhythms

01:34:51.240 | so that you start to anticipate that morning workout.

01:34:53.280 | You start to anticipate the morning sunlight.

01:34:55.500 | Just one more scientific point.

01:34:58.200 | We know that when you view bright sunlight in the morning

01:35:01.640 | or just sunlight that's illuminating your environment,

01:35:03.720 | as you said, you don't even have to see the sun itself,

01:35:06.200 | that there's a 50, five zero percent increase

01:35:09.440 | in the amplitude of the morning cortisol spike,

01:35:11.780 | which is a good thing.

01:35:12.960 | - Right, that's where you want it.

01:35:13.920 | - Because it's inversely,

01:35:15.600 | the amplitude of the morning cortisol spike

01:35:17.440 | is inversely related to the amplitude

01:35:19.120 | of the evening cortisol spike

01:35:20.480 | and high evening cortisol is associated

01:35:22.520 | with middle of the night waking and on and on.

01:35:26.080 | So, you know, I'm very bullish on these mechanisms.

01:35:29.540 | I also love that they're so deeply woven

01:35:32.280 | into our evolutionary history, you know,

01:35:34.000 | that we share with single cell organisms.

01:35:35.820 | It's so wild, but of course,

01:35:39.860 | there's going to be a bi-directionality there

01:35:41.700 | and it's impossible to see where one thing starts

01:35:44.380 | and the other one stops.

01:35:46.060 | - I mean, here's my take, Andrew.

01:35:47.640 | First of all, I actually, with far less authority than you,

01:35:52.000 | agree with your assessment

01:35:53.680 | and might even be a little bit more bullish,

01:35:55.400 | might even put it at 80/20.

01:35:57.460 | And here I'll give you my explanations,

01:36:00.500 | which stem more from my fastidious battles

01:36:05.740 | with epidemiology in general, right?

01:36:07.600 | Like, because so much of the world that I live in

01:36:10.640 | still has to rely on epidemiologic data.

01:36:13.520 | And so how do you make sense of it?

01:36:15.320 | And the truth of it is most of it is really pretty bad.

01:36:19.240 | But I tend to find myself looking

01:36:21.560 | at the Austin Bradford Hill criteria all the time.

01:36:24.760 | And for folks who don't know,

01:36:26.800 | he was a statistician who basically proposed

01:36:29.980 | a set of criteria, believe there are eight of them.

01:36:32.960 | And I can't believe I don't know

01:36:33.920 | every one of them off by heart.

01:36:35.320 | I certainly used to.

01:36:36.420 | But the more of these criteria that are met

01:36:40.740 | within your correlations,

01:36:42.820 | the more likelihood you will find causality.

01:36:46.760 | So when I think of your data here, the data in this paper,

01:36:49.780 | I'll tell you what makes these correlations

01:36:53.260 | seem to have causality within them

01:36:55.740 | in the direction that's being proposed.

01:36:57.700 | Look at the dose effect.

01:37:01.240 | So dose effect matters, and this is done in quartiles,

01:37:04.560 | and that's a very elegant thing.

01:37:05.780 | If they just did it as on/off, it would be harder.

01:37:09.000 | High, low.

01:37:09.840 | That's right, but the fact that they did it in quartiles

01:37:11.320 | allows you to see that every example in figure two,

01:37:14.380 | I don't believe there is an exception to this.

01:37:17.280 | No, and I think also in the other figure.

01:37:19.360 | There's only one exception to what I'm about to say.

01:37:21.620 | Sorry, two out of like, God knows how many,

01:37:24.240 | they're all monotonically increasing and decreasing.

01:37:27.200 | In other words, the dose effect is always present.

01:37:30.280 | Another thing is biologic plausibility.

01:37:33.080 | You've spoken at length about that today.

01:37:35.700 | So in other words, sometimes you have to look

01:37:37.720 | at epidemiology and ask, is there a biologic explanation?

01:37:40.860 | And here there is.

01:37:41.800 | You've added another one, which is evolutionary conservation

01:37:44.640 | to the biologic plausibility.

01:37:46.540 | Then you can talk about animal models or experiments

01:37:49.360 | in humans over short durations

01:37:51.700 | that generally support these findings.

01:37:54.180 | And so those are just a couple

01:37:55.440 | of the Bradford Hill criteria that lead to

01:38:01.840 | my belief that, yeah, there's reverse causality here,

01:38:05.360 | but it's not the full explanation.

01:38:08.820 | And that more of the explanation is probably

01:38:10.720 | the direction that's being proposed.

01:38:12.680 | And if that's true, 'cause then at the end of the day,

01:38:15.380 | like what's the purpose of the discussion?

01:38:16.840 | The purpose of the discussion is if you are

01:38:20.360 | under the influence of any of these psychiatric conditions,

01:38:24.140 | in addition to the treatments you're doing now,

01:38:26.280 | what else can you do?

01:38:27.620 | And to me, the takeaway is follow these light behaviors.

01:38:32.540 | I mean, it's a relatively low lift when you consider

01:38:35.740 | some of the other things.

01:38:36.580 | Like I'm over here asking people to do zone two

01:38:38.700 | for three hours a week and VO2 max workouts

01:38:41.380 | and all this other stuff.

01:38:42.800 | And like, I think all those things matter

01:38:45.020 | for mental health as much as physical health.

01:38:46.940 | But this strikes me as on the spectrum of low asks,

01:38:51.940 | if it shows, if it's only even 30% causality,

01:38:56.740 | 70% reverse causality, like I'll take those.

01:38:59.360 | I would still instate that.

01:39:01.180 | Yeah, and it's taking your coffee on the balcony.

01:39:03.860 | And people will often say,

01:39:05.860 | well, how do you do this with kids?

01:39:07.740 | The kids should be doing it too, right?

01:39:10.340 | It means popping your sunglasses off.

01:39:12.020 | It means getting out for just a few minutes.

01:39:13.680 | And the fact that it's additive,

01:39:15.260 | that these photon counting mechanisms, they sum, is great.

01:39:20.260 | And look, this paper also says,

01:39:24.580 | and I should have stated this earlier,

01:39:25.740 | if you missed your daytime light ration,

01:39:28.900 | get your nighttime dark ration.

01:39:30.960 | They are independent and additive.

01:39:33.720 | So that's, I mean, that's really something.

01:39:36.140 | But of course, ideally you get both.

01:39:37.740 | But I appreciate your take on it.

01:39:39.060 | And thanks for your expertise in parsing epidemiology.

01:39:44.060 | I look at fewer studies of that sort,

01:39:48.100 | but I learned from you.

01:39:49.500 | And that's one of the reasons

01:39:50.740 | I love doing these journal clubs is I learn.

01:39:52.940 | So along those lines,

01:39:54.940 | tell us about the paper you selected.

01:39:58.160 | I'm really eager to learn more.

01:40:00.300 | Well, I wanted to pick a paper

01:40:02.560 | that was kind of interesting as a paper.

01:40:05.340 | And this paper, I think, is interesting

01:40:08.580 | in that it is kind of the landmark study

01:40:12.540 | of a class of drugs.

01:40:14.620 | But in the same way that you kind of picked a paper

01:40:17.400 | that I think has a much broader overarching importance,

01:40:22.320 | the reason I picked this paper,

01:40:24.200 | which is from the New England Journal of Medicine,

01:40:26.520 | it's about 10 years old, no correction, 13 years old,

01:40:30.100 | is because it is kind of the landmark study

01:40:34.300 | in a class of drugs that I believe

01:40:36.980 | are the most relevant class of drugs

01:40:39.460 | we've seen so far in cancer therapy.

01:40:42.500 | And even though the net effect of these drugs

01:40:46.140 | has only served to reduce mortality

01:40:49.300 | by maybe eight to 10%, which is not a huge amount,

01:40:53.240 | it's the manner in which they've done it

01:40:56.380 | that gives me great hope for the future,

01:40:59.060 | even if it's through other means.

01:41:00.960 | So I'll take a step back before we go into the paper

01:41:04.300 | for, again, just the context and background.

01:41:06.540 | So the human immune system is kind of a remarkable thing.

01:41:14.140 | It's hard when you're sort of trying to imagine

01:41:18.080 | what's the most amazing part of the human system.

01:41:20.860 | And maybe it's my bias as well,

01:41:22.440 | because just as you spent your time in the light system

01:41:25.680 | and the photosensing system,

01:41:26.900 | I spent my time in the immunology world,

01:41:29.920 | but it is remarkable to me how our immune systems evolved.

01:41:33.380 | And they have this really brutal task,

01:41:37.260 | which is how can they be tuned

01:41:41.120 | to detect any foreign pathogen that is harmful

01:41:46.120 | without knowing a priori what that could be,

01:41:51.260 | while at the same, so in other words,

01:41:52.600 | how can you tune a system to be so aggressive

01:41:56.680 | that it can eradicate any virus or bacteria

01:41:59.760 | billions of years into the future

01:42:01.940 | without knowing what it's going to be,

01:42:04.000 | but at the same time, it has to be so forgiving of the self

01:42:09.600 | that it doesn't turn around and attack the self.

01:42:12.620 | It's remarkable.

01:42:14.080 | And of course, we can always think of the exceptions.

01:42:16.120 | There are things called autoimmune conditions.

01:42:18.040 | So clearly the system fails,

01:42:20.600 | and the immune system turns around and attacks the self.

01:42:23.840 | If you see a person with vitiligo,

01:42:25.440 | I have a little bit of vitiligo on my back,

01:42:27.680 | a couple of spots,

01:42:28.760 | clearly the immune system is attacking something there

01:42:32.320 | and destroying some of the pigment.

01:42:33.960 | - I didn't realize vitiligo was autoimmune.

01:42:35.920 | - Yeah, so there are lots of

01:42:39.320 | more serious autoimmune conditions.

01:42:41.120 | Of course, somebody that has lupus

01:42:42.740 | or where the immune system can be attacking the kidney,

01:42:45.600 | the immune system can be attacking anywhere.

01:42:47.040 | Autoimmune conditions can be deadly,

01:42:49.460 | but fortunately they are very rare.

01:42:51.040 | And for the most part,

01:42:52.160 | this immune system works remarkably well.

01:42:55.720 | So how does it work?

01:42:58.020 | And why is it that cancer seems to evade it

01:43:02.820 | virtually all of the time?

01:43:04.140 | This is the question.

01:43:05.640 | Now let's first of all talk about how it works.

01:43:08.660 | And then when I tell you how it works,

01:43:11.060 | you'll say, that sounds amazing.

01:43:12.920 | Clearly it should be able to destroy cancer.

01:43:17.800 | I'm gonna simplify it by only talking about one system,

01:43:20.680 | which is how T cells recognize and get activated,

01:43:25.280 | how T cells recognize antigens.

01:43:27.160 | So we have something called an antigen.

01:43:31.060 | So an antigen is an antibody generating peptide.

01:43:35.880 | So it's a protein, almost always a protein,

01:43:39.320 | they can be carbohydrates,

01:43:40.600 | but they're almost always proteins.

01:43:41.900 | And they're very, very small peptides.

01:43:43.700 | Like we're talking as little as nine amino acids,

01:43:47.060 | maybe up to 20 amino acids.

01:43:48.780 | So teeny tiny little peptides.

01:43:51.780 | But it's amazing that in such a short peptide,

01:43:55.340 | the body can recognize if that's Andrew or not Andrew.

01:43:59.940 | I mean, when again, think about like,

01:44:01.660 | we talk about proteins in kilodaltons, right?

01:44:04.400 | We're talking about proteins in terms of thousands

01:44:07.460 | of amino acids that make up every protein in your body.

01:44:10.540 | And yet if it samples a protein and sees that,

01:44:13.860 | hey, this little nine, 10, 15 peptide amino acid

01:44:17.860 | is not part of you, I know it's bad.

01:44:21.300 | And therefore I'm going to generate

01:44:23.360 | an immune response to it.

01:44:25.180 | So we have what are called antigen presenting cells.

01:44:29.020 | You have cells that go around sampling peptides,

01:44:33.700 | and they will on these things called MHC class receptors,

01:44:38.300 | bring the peptide up to the surface

01:44:41.100 | and serve it up to the T cell.

01:44:43.700 | There are two types of these.

01:44:44.860 | There's MHC class one and MHC class two.

01:44:48.360 | I only bring--

01:44:49.200 | - So this is major histocompatibility complex.

01:44:51.460 | - That's correct.

01:44:52.300 | And we refer to them that way because of the context

01:44:58.260 | in which they were discovered,

01:44:59.800 | which was for organ rejection.

01:45:02.060 | So not surprisingly, when you need to put a kidney

01:45:05.620 | into another person, if that kidney is deemed foreign,

01:45:09.660 | it will not last long.

01:45:10.860 | In the early days of organ transplantation,

01:45:13.180 | we're rife with immediate rejections.

01:45:15.580 | And by not, I mean,

01:45:17.260 | the immediates are the ABO incompatibilities,

01:45:19.920 | but you know, the sort of next layer of incompatibility

01:45:23.220 | was MHC incompatibility, which would lead to, you know,

01:45:26.140 | within weeks the organ is gone as opposed to within hours.

01:45:30.080 | So you have these two classes of MHC.

01:45:34.220 | You have class one and class two.

01:45:35.420 | Class one is what we call endogenous.

01:45:38.860 | So this is basically what happens when a protein

01:45:43.300 | or an antigen is coming from inside the cell.

01:45:46.260 | So let's consider the flu.

01:45:48.240 | So if you get the flu,

01:45:50.020 | the influenza virus infects the respiratory epithelium

01:45:53.340 | of your, you know, your larynx.

01:45:55.780 | And that virus, as you know, folks listening might remember

01:46:00.300 | from our days of talking about COVID,

01:46:02.400 | viruses can't replicate on their own.

01:46:04.260 | What they do is they hijack the replication machinery

01:46:08.740 | of the host and they use that either to insert their RNA

01:46:12.740 | or DNA to replicate.

01:46:14.220 | And in the process, proteins are being made.

01:46:16.260 | Well, those proteins are the proteins of the virus,

01:46:18.540 | not of us.

01:46:19.880 | So some of those peptides get launched

01:46:23.020 | onto these MHC class one gloves.

01:46:27.180 | Basically the glove comes up to the surface

01:46:29.760 | and a T cell comes along.

01:46:31.420 | And in the case of MHC class one, it's a CD8 T cell.

01:46:35.660 | These are what are called the killer T cells, right?

01:46:38.620 | And so this cell comes along and with its T cell receptor,

01:46:42.660 | the T cell receptor meets the MHC class one receptor

01:46:47.120 | with the antigen in it.

01:46:48.720 | And if that's a lock, it realizes that's my target.

01:46:52.540 | And it begins to replicate and proliferate and target those.

01:46:56.860 | And that creates the immune response.

01:46:58.580 | And by the way, that's how it works

01:47:00.060 | when you vaccinate somebody.

01:47:01.140 | You're basically pre-building that thing up.

01:47:03.820 | - So would this fall under the adaptive immune response

01:47:06.760 | or the innate immune response?

01:47:08.500 | - No, this is adaptive, yep.

01:47:10.580 | Innate is just these pure antibody response

01:47:13.180 | on the B cell side.

01:47:14.020 | I won't get into that for the purpose of this discussion.

01:47:17.420 | The other example is MHC class two.

01:47:21.260 | And that's also part of the adaptive system

01:47:24.700 | or the innate system, which is more what we call

01:47:27.580 | the exogenous form.

01:47:29.780 | So these are peptides that are usually coming

01:47:31.440 | from outside the cell.

01:47:32.900 | So we're gonna focus more on the MHC class one

01:47:36.020 | because this is peptides that come from inside the cell.

01:47:39.900 | Okay, so just keep in the back of your mind,

01:47:41.920 | if a foreign protein gets presented from inside a cell

01:47:45.280 | to outside a cell, the T cells recognize that

01:47:48.140 | and they will mount a foreign response.

01:47:50.020 | And by the way, that's why we basically can beat any virus.

01:47:53.740 | Like if you consider how many viruses are around us,

01:47:57.120 | the fact that we almost never die from a viral infection

01:48:00.520 | is a remarkable achievement

01:48:02.700 | of how well this immune system works.

01:48:04.340 | We're constantly combating these viruses.

01:48:06.100 | Constantly, and by the way,

01:48:07.480 | we don't really have very effective antiviral agents.

01:48:10.900 | It's not like antibiotics.

01:48:11.980 | Like we have antibiotics up the wazoo.

01:48:14.260 | I mean, we're way better at fighting viruses than bacteria.

01:48:17.600 | Can I just ask one question?

01:48:18.860 | I've always wondered about this.

01:48:19.900 | To what extent is our ability to ward off viruses

01:48:23.080 | on a day-to-day basis as an adult,

01:48:26.380 | reliant on us having been exposed

01:48:28.900 | to that virus during development?

01:48:30.980 | Like as I walk around today,

01:48:32.420 | maybe I'll be exposed to 100,000 different viruses.

01:48:35.740 | Would you say that half of those,

01:48:37.520 | I've already got antibodies too,

01:48:38.900 | because I was exposed to them

01:48:40.480 | at some prior portion of my life?

01:48:42.060 | Yeah, harder to quantify.

01:48:43.600 | And the other ones, I'm just building up antibodies.

01:48:45.900 | Like I was on a plane last night,

01:48:47.760 | someone was coughing, so I was hiding.

01:48:49.820 | And I had COVID a little while ago,

01:48:51.020 | so I wasn't too worried about that.

01:48:52.380 | And I feel great today.

01:48:53.620 | But I just assume that on that plane,

01:48:59.560 | I'm in a swamp of viruses, no matter what,

01:49:03.360 | and that most of them I've already been exposed to

01:49:05.940 | since I was a little kid.

01:49:07.240 | So I've got all the antibodies

01:49:08.400 | and they're just fighting it back,

01:49:09.440 | binding up those viruses and destroying them.

01:49:12.600 | Yeah, I think it's part that,

01:49:13.740 | and I also think it's part of them

01:49:15.100 | that our body can destroy

01:49:16.560 | without mounting much of an immune response.

01:49:18.480 | So therefore your immune system is doing the work,

01:49:20.560 | and yet it's not mounting a systemic inflammatory response

01:49:23.640 | that you're not sensing.

01:49:24.560 | So is it also a physical trapping in my nasal epithelium,

01:49:28.620 | it's a virus issue?

01:49:29.460 | Yeah, so yeah, you have huge barriers, right?

01:49:31.940 | So the skin, the hairs in your nose,

01:49:34.600 | all of these things are huge barriers.

01:49:37.240 | But assuming that still a bunch of them are getting in,

01:49:39.920 | at least the respiratory ones,

01:49:41.100 | that's the other thing to keep in mind, right?

01:49:42.420 | There are certain viruses that are totally useless

01:49:44.460 | floating around the air, right?

01:49:46.260 | There are certain viruses,

01:49:47.780 | the viruses that most people are really afraid of,

01:49:49.920 | Hep C, Hep B, HIV, well,

01:49:52.860 | if they're sitting on a table or floating around the air,

01:49:56.000 | they're of no threat to you.

01:49:57.220 | They have to be sort of transmitted through the barrier.

01:50:00.680 | But again, some of these viruses you're gonna defeat

01:50:04.080 | without an enormous response.

01:50:06.300 | And then some of them,

01:50:07.600 | why is influenza quote unquote, such a bad virus?

01:50:10.760 | Whereas the common respiratory cold

01:50:13.240 | kind of sidelines you for a day.

01:50:15.040 | It's the immune response that you're feeling.

01:50:17.200 | The worse, the bigger the immune response to the virus,

01:50:20.760 | the more you're feeling that.

01:50:21.880 | You feel your immune system going crazy, right?

01:50:25.920 | The interleukins that are spiking,

01:50:28.120 | the third spacing that occurs

01:50:29.780 | to get more and more of the immune cells there,

01:50:32.000 | the spike of your temperature

01:50:33.600 | as your body basically tries to cook the virus,

01:50:35.940 | all of that stuff is your body doing-

01:50:37.300 | - The fatigue.

01:50:38.240 | - Yeah, yeah, you're being drained and all this happening.

01:50:40.540 | So one more point I'll mention,

01:50:43.840 | just to close the loop on the autoimmunity,

01:50:46.440 | how is it that we learn not to attack ourselves?

01:50:50.280 | That's something called thymic selection

01:50:52.120 | that occurs in infancy.

01:50:54.420 | So you and I have a no good for nothing,

01:50:57.980 | tiny little thymus that would be,

01:50:59.580 | it's almost impossible to see these things.

01:51:01.740 | You know, when we used to operate on people,

01:51:03.980 | you know, the thymus is barely visible in an adult,

01:51:07.860 | in a healthy adult outside of thymic tumors.

01:51:10.520 | But in a child, the thymus is quite large.

01:51:12.760 | And the purpose of the thymus is to educate T cells

01:51:16.680 | and basically show the T cells what self is.

01:51:20.580 | And any T cell that doesn't immediately recognize it

01:51:24.700 | gets killed.

01:51:25.840 | So it's a really clever system

01:51:27.760 | where we basically teach you to recognize self

01:51:31.500 | at a very early age.

01:51:32.780 | And if you can't do that, you're weeded out.

01:51:35.060 | And then the thymus involutes thereafter

01:51:37.600 | because it's sort of served its purpose.

01:51:39.580 | Okay, now let's talk about cancer.

01:51:42.720 | So what do we know about cancer?

01:51:44.300 | So we know that, again, you know,

01:51:46.100 | cancer is a genetic disease in the sense

01:51:48.740 | that every cancer has genetic mutations.

01:51:52.480 | Most of those mutations are somatic,

01:51:55.540 | which means most of those mutations are mutations

01:51:59.380 | that occur during the course of our life.

01:52:01.140 | They're not germline mutations.

01:52:02.820 | The germline being the eggs and sperm, right?

01:52:07.060 | So it's all other cells.

01:52:08.300 | And I love that you pointed out that, you know,

01:52:11.060 | there can be, that cancer can be genetic

01:52:14.180 | but isn't necessarily inherited, right?

01:52:16.580 | In fact, it's rarely inherited.

01:52:17.400 | They're genetic and they think inherited, right?

01:52:19.380 | Yes.

01:52:20.220 | Inherited is always genetic to some extent,

01:52:22.420 | but genetic isn't always inherited.

01:52:25.300 | Yeah, so there are a handful of cancers

01:52:27.780 | that are derived from inherited mutations.

01:52:32.220 | So Lynch syndrome is an example of that.

01:52:35.620 | Hereditary polyposis is an example of that,

01:52:39.500 | where you have a gene that gets passed through the germline

01:52:43.540 | and that gene codes for a protein like all genes do.

01:52:47.400 | And it's either you have too much of a gene

01:52:49.300 | or too little of a gene.

01:52:50.400 | So it's either a gene that promotes cancer

01:52:52.860 | and you have too much of that,

01:52:54.080 | or it's a gene that prevents cancer

01:52:55.660 | and you have too little of it

01:52:57.380 | or a dysfunctional version of it, right?

01:52:59.220 | So BRCA is an example of that.

01:53:00.680 | BRCA is hereditary.

01:53:02.360 | BRCA codes for a protein and the women and men,

01:53:07.360 | but mostly the women that we think about

01:53:09.200 | who have a BRCA mutation that all, in some cases,

01:53:12.760 | almost guarantees breast cancer,

01:53:14.520 | it's because of a defective copy.

01:53:17.620 | So it's like they don't get the protein

01:53:19.420 | that they need to protect them from breast cancer.

01:53:21.740 | So what do we know?

01:53:23.580 | Well, we know that,

01:53:25.620 | and this is probably one of the most remarkable things

01:53:28.980 | I've ever learned,

01:53:31.020 | and it still blows my mind every time.

01:53:33.100 | Well, actually, before I get to that point,

01:53:35.600 | I want to make another point.

01:53:36.820 | Okay, so you might think,

01:53:38.980 | so cancer, you know, our cells become cancerous,

01:53:43.360 | but they're clearly hijacked

01:53:45.060 | because they have these mutations,

01:53:46.640 | and as a result of these mutations,

01:53:48.920 | they make proteins that allow cancers to behave differently.

01:53:53.280 | And cancers behave differently from non-cancers

01:53:55.880 | in two very critical ways.

01:53:58.600 | The first way is that they do not respond

01:54:00.740 | to cell cycle signaling.

01:54:02.340 | So if you cut your skin, it heals,

01:54:07.740 | but how does it know to heal just right

01:54:10.140 | and not to keep growing and growing and growing

01:54:12.300 | and growing and growing?

01:54:13.340 | Well, it knows that because there are cell cycle signals

01:54:16.120 | that tell it time to grow, time to stop.

01:54:18.120 | If, believe it or not, this is an extreme example,

01:54:21.800 | if you donated to me half of your liver,

01:54:24.640 | which I know you would.

01:54:25.860 | - Absolutely.

01:54:26.700 | I'd give you more than half of my liver.

01:54:28.580 | - Well, you'd only need to give me half.

01:54:29.420 | - If it meant that we could keep doing

01:54:30.580 | these journal clubs, right? - Yeah, yeah, yeah.

01:54:32.520 | Within months, you would regenerate a full liver.

01:54:36.340 | - Yeah, that's so--

01:54:37.180 | - That's amazing.

01:54:38.020 | - That's so wild.

01:54:38.840 | It's like a salamander.

01:54:39.680 | You cut off a salamander limb,

01:54:40.520 | and please don't do that experiment

01:54:41.340 | 'cause other people are doing it anyway,

01:54:43.580 | and it grows back.

01:54:44.660 | - Yeah, and it knows how much to grow back.

01:54:47.340 | - So wild.

01:54:48.180 | - So when the cell is perfectly functioning,

01:54:50.580 | it knows how much to grow, and it's,

01:54:52.700 | well, cancer loses that ability.

01:54:54.140 | That is one of the hallmarks of cancer.

01:54:56.500 | It just keeps growing.

01:54:57.820 | It doesn't grow faster, by the way.

01:54:59.060 | That's a misnomer.

01:54:59.900 | People think cancers grow faster than non-cancers.

01:55:02.320 | There's no real evidence that that's the case.

01:55:04.020 | They just don't stop growing.

01:55:05.700 | The second property of cancer is the capacity

01:55:09.420 | to leave the site of origin,

01:55:12.620 | go someplace else, and take up residence.

01:55:14.700 | - So that's metastases.

01:55:15.540 | - That's the metastases component.

01:55:17.580 | So if you think about it for a minute,

01:55:19.840 | a cell that never stops replicating

01:55:22.820 | and has the capacity to up and leave and move

01:55:25.660 | and take up residence is clearly different

01:55:27.700 | from the cell itself, right?

01:55:29.080 | So if I have a cell of colonic epithelium,

01:55:31.940 | the cell that lines the inside of my colon,

01:55:34.220 | it's clearly got a set of proteins in it.

01:55:36.300 | But if all of a sudden that thing can grow, grow, grow,

01:55:39.860 | grow, grow, not stop, not stop, not listen to the signal,

01:55:43.100 | and then somehow wind its way into the liver

01:55:45.240 | and just keep growing and growing and growing,

01:55:47.020 | it must have different proteins.

01:55:49.080 | So the question then becomes, why does cancer even exist?

01:55:53.140 | How has our immune system not figured out a way

01:55:56.780 | to just silence this and eradicate it the way it does

01:56:00.220 | to virtually every virus you encounter?

01:56:03.460 | And to me, this is one of the most interesting questions

01:56:05.740 | in all of biology, and it really comes down

01:56:07.660 | to how clever cancer is, unfortunately,

01:56:11.220 | how evolutionarily clever it is.

01:56:14.420 | It basically does a lot of things

01:56:17.820 | to trick the immune system.

01:56:19.700 | So it has its own secretory factors

01:56:22.560 | that tamp down the immune system.

01:56:25.060 | It grows in an environment because of its nature.

01:56:28.940 | So one of the things that's long understood about cancer

01:56:31.820 | is it's heavily glycolytic.

01:56:34.060 | And when something is heavily glycolytic,

01:56:36.140 | it's going glucose to pyruvate to lactate nonstop.

01:56:39.900 | There are lots of reasons for that.

01:56:41.660 | I think there's more than one.

01:56:45.180 | - What does that afford it?

01:56:46.420 | Does that, is that afforded a migratory potential?

01:56:49.700 | - No, so it's super interesting.

01:56:50.780 | So that's the effect that what I just described

01:56:52.420 | is called the Warburg effect.

01:56:54.100 | And when Warburg proposed this,

01:56:57.260 | which God was probably in the 1920s,

01:57:00.740 | it was before World War II,

01:57:03.120 | he proposed it because he thought the mitochondria

01:57:06.820 | of cancer cells were defective.

01:57:09.140 | So he proposed that the cancer cells' mitochondria

01:57:12.260 | don't work, hence they have to undergo glycolysis.

01:57:15.540 | They can't undergo aerobic metabolism.

01:57:19.840 | We now know that that's not the case.

01:57:23.700 | So we now know that the Warburg effect

01:57:26.280 | or the Warburg effect, if I'll refer to him correctly

01:57:28.620 | by his name, almost assuredly does not have to do

01:57:31.900 | with defective mitochondria.

01:57:33.420 | Others have proposed several mechanisms.

01:57:35.900 | I think there's probably more than one thing going on.

01:57:37.620 | So a paper that came out in 2009, very influential paper

01:57:41.260 | by a guy named Matt Vander Heiden and Craig Thompson

01:57:46.220 | and Lou Cantley proposed that the reason

01:57:50.360 | that cancer cells do the Warburg effect

01:57:52.560 | is that they're not optimizing for energy,

01:57:54.620 | they're optimizing for cellular building blocks.

01:57:56.660 | And if you do the mass balance,

01:57:58.120 | it completely makes sense.

01:57:59.500 | Like dividing cells need building blocks more than energy.

01:58:03.960 | And glycolysis, while very inefficient for generating ATP,

01:58:07.840 | is much more efficient at generating substrate

01:58:09.760 | to make more cells.

01:58:11.200 | But another proposed mechanism is exactly at this one.

01:58:14.020 | Glycolysis lowers the surrounding pH because of lactate.

01:58:17.860 | Lactate attracts hydrogen, pH goes down.

01:58:19.960 | And guess what that does to the immune system?

01:58:22.260 | It attracts the immune system.

01:58:23.840 | So it's also a way to hide from the immune system.

01:58:26.680 | - So there's like a pH cloaking.

01:58:29.700 | Leveraging pH to cloak the signal

01:58:32.520 | that the immune system would otherwise see.

01:58:34.320 | - Yep.

01:58:35.160 | And then when you layer on top of that,

01:58:36.980 | that it knows how to secrete things like IL-10, TGF beta,

01:58:40.280 | all of these other secretory factors

01:58:42.040 | that also inhibit the immune system,

01:58:44.000 | basically it's figured out a way

01:58:46.420 | to kind of hide itself from the immune system.

01:58:48.300 | - The way you describe it, cancer sounds like a virus.

01:58:51.400 | - Yes.

01:58:52.240 | - I mean, it sounds a lot like a virus.

01:58:53.060 | And that leads me to ask,

01:58:55.460 | are there any examples of contagious cancers?

01:58:57.940 | I recall seeing some studies about these little critters

01:59:00.800 | down in Australia, Tasmanian devils,

01:59:04.180 | that like they would,

01:59:05.780 | they scratch each other and fight as Tasmanian devils do.

01:59:09.260 | They're actually quite cute.

01:59:11.240 | And they would get cancers and tumors growing on their faces.

01:59:14.360 | - Yes, so it-

01:59:16.180 | - And so it was like a literal physical interaction

01:59:20.160 | that could transmit cancer from one animal to the next.

01:59:22.840 | - So it's less that there are viruses

01:59:25.020 | that cause cancer.

01:59:26.740 | So in that sense, you could argue, yes,

01:59:28.980 | there are contagious cancers.

01:59:30.740 | - Well, HPV.

01:59:31.840 | - Sure, yeah, HPV, Hep B, Hep C,

01:59:36.060 | but there are even cancers like cutaneous cancers

01:59:38.500 | that arise from viruses.

01:59:40.220 | But I don't know if that's quite the same

01:59:43.580 | as what you're saying, like-

01:59:44.620 | - No, no, they're both,

01:59:46.060 | what you're saying is an important point.

01:59:47.620 | I mean, we don't wanna go down the rabbit hole of HPV,

01:59:50.700 | but right, that's increasing susceptibility

01:59:53.220 | to cervical cancer.

01:59:55.300 | Now there's a vaccine against HPV, right?

01:59:57.180 | There wasn't when we were in college, as we all knew.

02:00:00.020 | There was no vaccine, but the, okay, so, but yeah,

02:00:03.620 | direct transmission of cancers from one organism

02:00:06.820 | to the next, more rare.

02:00:08.380 | - Yes, okay, so now, a moment ago I said

02:00:11.860 | there's this really incredible thing about cancer

02:00:14.620 | that blows my mind and about our immune system,

02:00:16.320 | which is that at least 80% of solid organ tumors,

02:00:21.660 | and we're gonna mostly talk about solid organ tumors,

02:00:23.820 | 'cause that's where the field of oncology

02:00:26.500 | has made very little progress.

02:00:28.300 | So if you go back 50 years,

02:00:30.020 | where has oncology made huge progress?

02:00:31.940 | It's made great progress in blood tumors,

02:00:36.020 | leukemias, and some kinds of lymphomas.

02:00:38.860 | In fact, there's two kinds of lymphomas

02:00:40.380 | where the progress has been remarkable.

02:00:41.920 | One has been in Hodgkin's lymphoma,

02:00:44.980 | and the other has also been in immunotherapy,

02:00:47.900 | has been in a type of B-cell lymphoma,

02:00:50.660 | where that B-cell demonstrates

02:00:52.900 | or presents something called a CD19 receptor.

02:00:55.660 | So in B-cell lymphomas with CD19,

02:00:59.300 | there's a very unique niche immunotherapy,

02:01:01.960 | we won't talk about that today,

02:01:03.480 | called CAR-T therapy that has got rid of those guys,

02:01:05.920 | and then leukemias have also been pretty good.

02:01:08.540 | But in solid organ tumors,

02:01:11.380 | there've been only two real breakthroughs

02:01:14.140 | in the last 50 years.

02:01:15.220 | One has been the therapy for a certain type

02:01:18.920 | of testicular cancer, and it's really just

02:01:22.300 | a chemotherapy cocktail that has been found

02:01:24.160 | to work really well, and the other has been

02:01:26.120 | in this really rare kind of gastric cancer

02:01:28.380 | called the GI stromal tumor, which happens to result

02:01:31.420 | from one mutation in a kinase pathway,

02:01:33.860 | and there's one drug that can now target that,

02:01:36.620 | and it works, it's kind of amazing.

02:01:38.180 | Cures that cancer.

02:01:39.180 | Cures that.

02:01:40.020 | What I'm talking about are the cancers

02:01:42.540 | that kill virtually everybody else.

02:01:45.120 | This is what, when you sort of line up,

02:01:47.180 | what are the big causes of cancer death?

02:01:49.720 | Let's start at the top.

02:01:50.680 | It's lung, it's then breast and prostate in men and women,

02:01:55.320 | it's colorectal, it's pancreas.

02:01:57.080 | Those are the big five.

02:01:58.200 | They kill more than 50% of Americans.

02:02:00.980 | Cancer-wise, not, sorry, let me restate that.

02:02:04.100 | More than 50% of cancer deaths in Americans

02:02:06.400 | come from those five.

02:02:08.260 | These are what we call the solid epithelial tumors,

02:02:10.700 | and you can march down the list,

02:02:12.080 | and most cancers that most people are thinking of

02:02:13.960 | are those cancers.

02:02:14.800 | Good thing, more than 80% of those cancers

02:02:17.400 | have antigens that are recognized

02:02:20.920 | by the host's immune system.

02:02:23.240 | I will state it again because it is so profound.

02:02:26.300 | 80% at least of those cancers actually generate an antigen,

02:02:31.300 | meaning a little peptide in that cell

02:02:35.280 | gets presented to the T cell, and it is recognizable.

02:02:41.300 | And now the question is, why is that not sufficient

02:02:45.000 | to induce remission?

02:02:47.520 | And the short answer is there are not enough T cells

02:02:52.520 | that are able to act,

02:02:54.160 | and/or they are being sufficiently inhibited from acting,

02:02:58.620 | which gets me to the point of this paper.

02:03:01.020 | One of the ways in which the body inhibits

02:03:05.000 | the immune system, which we should remind ourselves

02:03:07.920 | is an important thing, right,

02:03:10.340 | is something called a checkpoint inhibitor.

02:03:13.320 | Okay, so go back to that idea that I talked about before.

02:03:17.060 | You have an antigen-presenting cell.

02:03:19.320 | It brings up an MHC receptor with a peptide on it,

02:03:24.320 | and there is a T cell that is coming,

02:03:26.860 | and I actually brought a diagram,

02:03:28.200 | which I'm going to link to this

02:03:29.680 | 'cause I don't want to make this too complicated,

02:03:31.960 | but I really think that this figure

02:03:34.480 | is helpful to understand how these drugs work.

02:03:38.000 | So the MHC receptor with the peptide is sitting there,

02:03:41.380 | and it binds to the T cell receptor on the T cell,

02:03:45.580 | but there is another receptor on the T cell,

02:03:48.560 | a CTLA-4 receptor,

02:03:51.960 | and that binds to a receptor that I won't bother naming now

02:03:56.240 | 'cause the names don't matter,

02:03:58.100 | but there's another receptor on the antigen-presenting cell

02:04:01.300 | that binds to that,

02:04:03.040 | and that acts as the breaks in the reaction.

02:04:07.440 | So CTLA-4, which is on the T cell,

02:04:11.320 | binds to another CD receptor on the antigen-presenting cell,

02:04:15.320 | and it says, "Tamp down the response."

02:04:18.680 | And the reason for that is

02:04:21.600 | we want to keep our immune system in check.

02:04:23.440 | This basically is a way of asking the immune system,

02:04:26.680 | 'cause remember, when the immune system sees that antigen,

02:04:29.900 | it wants to go nuts.

02:04:31.340 | It wants to start replicating and killing.

02:04:33.480 | This is a CD8 T cell.

02:04:34.800 | It is a targeted killer T cell.

02:04:37.540 | The checkpoint says, "Let's double-check that.

02:04:41.160 | "Let's be sure, let's tamp down the response."

02:04:43.940 | And as a result of that, a thought experiment emerged,

02:04:47.960 | which was, "What if we block CTLA-4?

02:04:51.800 | "What if we block the checkpoint?

02:04:54.320 | "Could we unleash the immune system a little bit more?"

02:04:58.780 | And I will say this.

02:05:01.580 | At the time it was proposed,

02:05:04.240 | it seemed a bit far-fetched.

02:05:06.960 | Because of the complexity of the immune system,

02:05:10.100 | it seemed a little far-fetched

02:05:11.920 | that simply blocking the checkpoint would have any effect.

02:05:15.560 | It's also worth noting that prior to this,

02:05:20.980 | one immunotherapy had found some efficacy,

02:05:24.620 | which was trying the exact opposite strategy.

02:05:27.200 | Rather than blocking the inhibitor,

02:05:30.080 | it was throwing more accelerant at the fire,

02:05:33.320 | which was giving something called interleukin-2.

02:05:35.600 | So interleukin-2 is, for lack of a better word,

02:05:39.280 | candy and fuel for T cells.

02:05:42.440 | So the idea was, if we have T cells

02:05:45.040 | that innately recognize a cancer antigen,

02:05:48.520 | can we just give high doses of interleukin-2

02:05:52.000 | and have them undergo proliferation and response?

02:05:56.160 | And the answer turned out to be yes,

02:05:57.960 | but only in two cancers, melanoma and kidney cancer,

02:06:01.640 | and only at very small levels.

02:06:03.360 | About 10% of the population would respond to these things.

02:06:06.880 | Now, look, that's 10% of people

02:06:08.280 | who were gonna be dead within six months,

02:06:10.040 | 'cause these are devastating cancers,

02:06:12.320 | and once they spread, there are no treatments

02:06:14.880 | that have any efficacy whatsoever.

02:06:16.800 | In fact, I think median survival for metastatic melanoma

02:06:19.240 | at the time was probably four months.

02:06:21.080 | So this was a very grim death sentence.

02:06:23.400 | But the idea now was,

02:06:27.360 | what about doing the exact opposite approach?

02:06:29.260 | Instead of trying to throw more fire at the T cell,

02:06:34.100 | what if we can take its breaks down?

02:06:35.740 | Less gas, pardon me, instead of giving more gas,

02:06:38.540 | let's give less breaks.

02:06:39.840 | And there were some phase two,

02:06:43.580 | some phase one studies that demonstrated efficacy.

02:06:45.460 | Phase two, and the paper I'm gonna talk about today

02:06:47.820 | is the phase three study

02:06:51.100 | that compared the first version of these.

02:06:55.820 | So the drug we're gonna talk about today

02:06:58.660 | is an anti-CTLA-4 drug called ipilimumab.

02:07:03.120 | There is another drug out there

02:07:07.000 | that came along shortly thereafter

02:07:09.200 | that is an anti-PD-1 drug.

02:07:12.780 | So PD-1 turns out to be another one of these checkpoints

02:07:17.780 | on T cells.

02:07:20.200 | And the Nobel Prize, by the way,

02:07:22.180 | I think it was 2018 or 2019 in medicine or physiology,

02:07:25.680 | was actually awarded to the two scientists

02:07:28.700 | who discovered CTLA-4 and PD-1.

02:07:31.900 | So I believe this is the only Nobel Prize in medicine

02:07:35.820 | for immunotherapy.

02:07:37.180 | It's a very big deal.

02:07:38.220 | So this study sought to compare the effect of anti-CTLA-4

02:07:44.160 | to a placebo.

02:07:50.480 | And the placebo in this case was not a real placebo.

02:07:53.020 | It was a peptide vaccine called GP-100

02:07:57.360 | to ask the question, in patients with metastatic melanoma,

02:08:02.380 | what would be the impact on median survival

02:08:06.180 | and overall survival?

02:08:08.400 | So let's talk a little bit about the paper.

02:08:12.380 | So again, one of the funny things about this is,

02:08:15.980 | I used to read these papers a lot, Andrew.

02:08:17.740 | These used to be my bread and butter papers.

02:08:20.580 | So I mean, reading these, it's my hobby.

02:08:25.580 | And I don't read them that much anymore.

02:08:29.700 | So it was kind of amazing how long it took me

02:08:32.720 | to remind myself of stuff I used to remember.

02:08:35.720 | But you do have to kind of go back and read the methods

02:08:38.080 | and figure out who were the patients in this?

02:08:40.220 | What was the eligibility criteria?

02:08:42.380 | Why did they do it this way?

02:08:43.860 | And of course it all kind of came back to me,

02:08:46.140 | but it took a minute.

02:08:48.220 | So the first thing is, these are all patients

02:08:51.820 | who had progressed through every standard therapy.

02:08:56.260 | So these are patients for whom there were no other options.

02:08:59.380 | These patients either had very advanced stage three melanoma,

02:09:05.300 | which means it was local regional melanoma,

02:09:08.520 | but it couldn't be resected.

02:09:10.140 | So an example of that would be a cancer

02:09:15.020 | that was completely engulfing,

02:09:19.060 | let's say the primary site was the cheek,

02:09:20.980 | and it had completely grown

02:09:23.140 | into all of the surrounding soft tissue.

02:09:25.900 | It hadn't spread anywhere,

02:09:27.400 | but it was all the lymph nodes of the neck.

02:09:30.380 | And I've seen patients like this

02:09:32.340 | and it's just completely disfiguring.

02:09:34.580 | And they'd already been through the standard chemotherapy

02:09:37.980 | and nothing was working and the thing was growing.

02:09:40.200 | And then it was mostly made up of patients

02:09:42.820 | with stage four cancer.

02:09:44.060 | Now melanoma has a very funny staging system.

02:09:47.120 | So in cancer, we typically talk about something

02:09:49.700 | called the TNM staging system.

02:09:52.300 | It is the standard way that cancers are staged.

02:09:54.740 | T refers to the tumor size,

02:09:56.780 | N refers to the lymph node status,

02:09:59.340 | and M refers to the presence or absence of metastases.

02:10:03.260 | And for most cancers, it is a very simple system.

02:10:07.300 | It is, you know, T is typically a number one, two,

02:10:11.300 | sometimes up to three and four,

02:10:12.960 | N is typically zero, one or two,

02:10:15.140 | and M is zero or one.

02:10:16.660 | Either there's no METs or there are METs.

02:10:18.580 | So for example, in colorectal cancer,

02:10:21.200 | the T staging determines the depth

02:10:24.120 | in the colon wall that it went.

02:10:26.460 | N is did it go to METs?

02:10:28.220 | And I think in colon, I'm a little rusty on this,

02:10:30.400 | I think colon has N zero, one or two,

02:10:32.600 | depending on how many lymph nodes.

02:10:34.300 | And then M zero, did it go to anything beyond that,

02:10:37.540 | like to the liver, lung, et cetera, or not.

02:10:40.220 | Melanoma is a bit more complicated.

02:10:41.720 | It has M zero, meaning no METs,

02:10:44.820 | but it also has M1A, M1B, M1C, and M1D.

02:10:49.820 | And within each of those,

02:10:52.260 | it has a threshold for high and low lactate dehydrogenase,

02:10:56.560 | or LDH.

02:10:57.860 | So it's both a staging based on imaging and biochemical.

02:11:01.460 | And the reason for that is LDH level

02:11:03.380 | is such a strong prognostic indicator of survival,

02:11:06.620 | in addition to M staging.

02:11:08.660 | Higher LDH levels tend to reflect more acidity,

02:11:12.200 | which we talked about why that's problematic,

02:11:14.160 | tends to reflect faster growing tumors,

02:11:16.200 | higher turnover, higher metabolic activity.

02:11:19.000 | M1A, let me see if I can remember this,

02:11:22.960 | M1As are cancers that have metastasized

02:11:27.960 | to surrounding soft tissue,

02:11:32.740 | or soft tissue anywhere in the body,

02:11:34.080 | so anywhere else on the skin.

02:11:35.680 | And you might think, well, that's kind of crazy,

02:11:37.360 | like, how does that happen?

02:11:38.520 | And it's really bizarre.

02:11:39.480 | You can have a patient who had a melanoma

02:11:41.600 | that showed up in one part of their body,

02:11:43.240 | and then they have metastases on other parts of their skin.

02:11:46.920 | M1B is, and I always get B and C confused,

02:11:51.920 | I think B is the lung.

02:11:54.520 | So M1B is to the lung, M1C is to any internal organs,

02:11:59.440 | so liver, et cetera, and M1D is to the CNS.

02:12:03.080 | And as those numbers increase,

02:12:04.520 | as those letters increase,

02:12:05.340 | the prognosis gets lower and lower and lower.

02:12:07.800 | So one of the first things I always look at

02:12:09.320 | when I look at a paper like this is,

02:12:10.800 | tell me about the patient population.

02:12:12.880 | Like, what was the breakdown of patients?

02:12:17.380 | And in table one, so that's, again,

02:12:19.960 | in clinical papers like this,

02:12:21.680 | table one is always, always, always baseline characteristics.

02:12:26.200 | Oh, I should mention one other thing, Andrew.

02:12:28.320 | This was done as a three to one to one randomization.

02:12:32.940 | So again, in the simplest form,

02:12:35.860 | a study would have two groups, right?

02:12:38.300 | You would have,

02:12:39.320 | we're gonna just have a treatment group and a placebo group.

02:12:43.560 | But in this arm, you had three groups,

02:12:47.060 | with one of them being the placebo.

02:12:49.080 | The placebo got just GP100, which is just a cancer vaccine.

02:12:53.980 | By the way, this is a cancer vaccine

02:12:56.020 | that never showed any efficacy.

02:12:58.500 | So it was a cancer vaccine that had been tested

02:13:00.940 | both with interleukin-2 directly,

02:13:03.860 | and as an adjuvant for patients

02:13:06.760 | who had metastatic melanoma,

02:13:08.840 | or had, sorry, not metastatic melanoma,

02:13:10.320 | who had melanoma resected, who were tumor-free,

02:13:13.200 | and then given the vaccine as adjuvants to see,

02:13:16.680 | did that have an effect on outcomes, and it didn't.

02:13:19.160 | So it's kind of a known placebo.

02:13:21.400 | So you had that group, then you had the anti-CTLA-4 group,

02:13:25.960 | and then you had anti-CTLA-4 plus GP100.

02:13:30.080 | Yeah, what's the rationale?

02:13:31.440 | For the three to one to one.

02:13:33.280 | It's basically, it increases statistical power, right?

02:13:37.400 | So this total study was a little under 700 people.

02:13:42.360 | They put 400 in the anti-CTLA-4 plus GP100 group,

02:13:47.140 | and then a little over 130 in each of the other two groups.

02:13:51.780 | So you're always gonna be able

02:13:53.000 | to make these two comparisons, right?

02:13:54.740 | What you can check by doing this is,

02:13:57.580 | is there any effect of GP100 in this setting,

02:14:00.820 | which had never been done before.

02:14:02.060 | So again, GP100 is a known protein expressed by melanoma,

02:14:07.060 | and all of these people were haplotyped

02:14:11.220 | to make sure that their immune system would recognize it.

02:14:14.560 | And the question was, would giving people anti-CTLA-4,

02:14:19.560 | i.e. taking the breaks off their immune system,

02:14:22.080 | with or without GP100 make a difference?

02:14:24.840 | So kind of going through this,

02:14:26.240 | you can see it sort of skews about 60% to 40% male to female.

02:14:30.860 | They talk about something called the ECOG performance status.

02:14:33.560 | That refers to how healthy a patient is coming in.

02:14:36.320 | So ECOG0 is no limitations whatsoever,

02:14:39.960 | which is kind of amazing

02:14:40.920 | when you really consider something.

02:14:42.180 | I think this speaks to just how devastating this disease is.

02:14:45.120 | These are patients who all have like six months to live,

02:14:49.240 | right, you know, a year max, and yet look at this,

02:14:52.800 | 58 to 60% of them have no limitation

02:14:56.480 | on their quality of life at this very moment.

02:14:58.480 | That's going to change dramatically, you know,

02:15:02.320 | absent a cure here.

02:15:03.460 | And then ECOG1 has some limitation,

02:15:05.940 | and you can see that ECOG1 plus ECOG0

02:15:08.500 | is basically 98% of the population.

02:15:11.600 | You can see the staging there.

02:15:13.380 | So again, very, very few of these patients

02:15:15.960 | are in the M0 category.

02:15:17.740 | M0s are people who have stage three disease

02:15:20.560 | that is so aggressive it can't be resected.

02:15:22.740 | That's about 1%,

02:15:24.260 | but the majority of these people are the M1As,

02:15:26.900 | M1Bs, M1Cs.

02:15:29.100 | So these are people with very aggressive cancers.

02:15:32.800 | You can also see that about 10 to 15% of these people

02:15:36.140 | also have CNS metastases.

02:15:38.760 | Again, the poorest prognosis of the poor.

02:15:41.300 | And then you can see the,

02:15:43.560 | about 40% of them have the LDH level above cutoff.

02:15:48.560 | All of this is to say,

02:15:50.460 | we're talking about a group of patients who have,

02:15:55.220 | you know, a very high likelihood of not surviving

02:15:59.260 | more than, you know, a year.

02:16:02.100 | It would be very, you know,

02:16:03.260 | unlikely that many of these patients

02:16:04.820 | would survive more than a year.

02:16:05.920 | So basically more than 70% of these people

02:16:08.720 | have visceral metastases.

02:16:10.620 | A third have high LDH and 10,

02:16:13.700 | more than 10% have brain mets.

02:16:15.420 | They've also all progressed through standard therapy.

02:16:19.920 | So-

02:16:21.480 | - Radiation, chemo.

02:16:23.020 | - Yeah, and the chemo for melanoma can be, you know,

02:16:27.080 | kind of a toxic chemo that really just

02:16:30.940 | doesn't really do anything.

02:16:33.140 | - So is it commonplace to use a treatment

02:16:36.780 | that failed in clinical trials as a placebo

02:16:40.900 | in these sorts of studies?

02:16:42.220 | - Yeah, it's interesting.

02:16:43.340 | I think you're referring obviously to the GP100.

02:16:45.860 | And I think the thinking was,

02:16:49.940 | okay, it hasn't been effective in other treatments,

02:16:54.000 | for example, when combined with IL-2 or as an adjuvant,

02:16:57.500 | but never before has it been tried

02:17:00.340 | with a checkpoint inhibitor,

02:17:01.900 | which is the technical term for this type of drug.

02:17:04.940 | I think there was also some belief

02:17:09.580 | that it would be easier to enroll patients.

02:17:12.120 | I don't think they stated this, but that's often the case.

02:17:14.460 | It would be easier to enroll patients

02:17:16.020 | if they would know that even in the placebo arm,

02:17:18.700 | they're still getting an active agent.

02:17:20.860 | - Got it.

02:17:21.700 | And I suppose there's always the possibility

02:17:23.340 | that the combination of the failed drug

02:17:26.720 | with a new drug would work.

02:17:28.320 | And then, so you're increasing the probability

02:17:30.240 | for novel discovery.

02:17:31.640 | - For sure.

02:17:32.480 | And again, if you go back to the randomization

02:17:34.380 | of three to one to one,

02:17:37.180 | it's really only one fifth or 20% of the participants

02:17:42.160 | that would get just the GP100.

02:17:44.440 | So in other words, you're basically telling people

02:17:46.500 | when they come into this study,

02:17:48.020 | there's an 80% chance you're going to get anti-CTLA-4.

02:17:52.660 | That's a much better set of odds than your typical study

02:17:56.260 | where you're going to be 50% likely

02:17:57.940 | to get the agent of interest.

02:18:00.500 | - Right, and people who are literally dying of cancer,

02:18:04.820 | they don't want to be in the control group.

02:18:07.240 | - Right, that's right.

02:18:09.320 | So the primary outcome for this study

02:18:11.980 | actually changed in the study.

02:18:14.380 | Now, they have to get permission to do that.

02:18:17.340 | So the original primary endpoint

02:18:21.940 | was the best overall response rate.

02:18:24.820 | So I have to explain how response rates are measured.

02:18:27.040 | This is a bit complicated.

02:18:29.640 | Remember, all of these patients by definition

02:18:31.560 | have measurable visible cancer.

02:18:34.280 | By visible, either on the surface of their body,

02:18:35.940 | but more likely on an MRI or CT scan.

02:18:39.240 | So all of these patients had to be scanned head to toe

02:18:41.740 | within 12 weeks of enrollment.

02:18:44.160 | Again, there's another thing I should point out here,

02:18:45.680 | which I know you understand,

02:18:46.700 | but it's always worth reminding people.

02:18:48.900 | When a study like this takes place,

02:18:50.860 | it usually takes place over many years.

02:18:53.220 | And so it's not the case that all 700 of these patients

02:18:56.700 | were enrolled on the same day and finished, you know,

02:18:59.280 | we finished observing them on the same day.

02:19:00.860 | No, no, no, this took place for a very long period of time.

02:19:03.860 | This took place across tens of centers.

02:19:07.540 | I can't remember if this was just globally

02:19:09.060 | or across the world, it might've been across the world.

02:19:12.540 | And so every center really needs to adhere

02:19:14.700 | to a very strict protocol.

02:19:16.260 | And you have a central organization that is running this.

02:19:20.220 | So you have a drug company.

02:19:21.500 | I think this is Bristol-Myers Squibb that makes the drug.

02:19:24.300 | They provide the drug.

02:19:25.940 | And then you have a CRO, a clinical research organization

02:19:28.780 | that is basically managing the trial.

02:19:32.620 | And the trial is being done at cancer centers

02:19:35.540 | all over the world or all over the country.

02:19:37.800 | And, you know, enrollment I think began in 2008 for this.

02:19:41.760 | No, no, I think it completed in 2008.

02:19:43.780 | It probably started in about 2004, 2005.

02:19:47.280 | And therefore you had to kind of have

02:19:49.140 | real clear protocols around this.

02:19:50.380 | So a complete response is the easier of these to understand.

02:19:53.620 | A complete response is everything vanishes completely.

02:19:58.620 | That's very rare in cancer therapy.

02:20:03.280 | So instead, what we kind of look for is a partial response.

02:20:06.420 | A partial response,

02:20:07.580 | and there are really different ways to define this.

02:20:10.140 | There are different criteria,

02:20:11.260 | but this is the most common way

02:20:12.760 | you define a partial response.

02:20:14.480 | A partial response is at least a 50% reduction by diameter.

02:20:19.480 | 'Cause remember in this type of imaging,

02:20:23.900 | you're looking at 2D versus 3D.

02:20:26.700 | So if you're looking at a lung lesion and it's this big,

02:20:29.420 | it has, you know, if it's two centimeters long,

02:20:31.420 | it has to go to at least one centimeter in diameter.

02:20:34.100 | So it's a 50% reduction at least of every single lesion

02:20:38.820 | with no new lesions appearing and no lesions growing.

02:20:42.740 | So it's very strict criteria, right?

02:20:45.440 | Again, CR means everything vanishes.

02:20:48.380 | PR means at least a 50% by diameter,

02:20:52.300 | which by the way is a much bigger diameter,

02:20:54.420 | much bigger reduction in terms of tumor volume.

02:20:56.980 | When you consider the linear versus the third power

02:20:59.520 | relationship of length and volume of every single lesion

02:21:03.500 | with nothing new appearing regardless of how small

02:21:05.940 | and no lesion growing.

02:21:08.640 | So that's a PR.

02:21:09.740 | So you basically have no response, progression,

02:21:14.100 | you know, we talk about those together,

02:21:15.940 | and then partial response and complete response.

02:21:18.960 | So initially the authors of this study were going to,

02:21:23.140 | the primary end point of this was going to be

02:21:24.940 | the best overall response rate.

02:21:26.700 | So what was the proportion of patients that hit PR?

02:21:29.940 | What was the proportion that hit a CR?

02:21:32.560 | That's very common in this type of paper

02:21:36.020 | where the outcomes are typically so dire.

02:21:39.220 | However, oh, I think I said,

02:21:41.300 | I think I said that the study was,

02:21:43.160 | I don't remember when the study ended,

02:21:45.820 | but the amendment was made to change the primary end point

02:21:50.540 | to overall survival at some point during the study.

02:21:55.540 | So, and by the way, that tends to be the metric

02:21:59.740 | everybody cares most about.

02:22:01.540 | So the overall survival for metastatic melanoma is zero,

02:22:07.100 | with the exception of people who respond to interleukin-2,

02:22:11.060 | high dose interleukin-2,

02:22:12.580 | and that will boost the overall survival rate

02:22:16.300 | to somewhere between eight and 10%, very, very low.

02:22:21.300 | These patients, many of whom had already taken

02:22:25.900 | and progressed through interleukin-2.

02:22:28.920 | Let me refresh my memory on what percentage

02:22:33.140 | of those patients.

02:22:35.140 | About a quarter of these patients

02:22:37.060 | had already taken high dose interleukin-2,

02:22:40.780 | and by definition, the fact that they're in this study

02:22:43.420 | means they had already progressed through that.

02:22:45.460 | That treatment had failed.

02:22:47.420 | Just reiterate just kind of the state these patients are in.

02:22:51.660 | So now let's look at figure one.

02:22:55.560 | So again, I'll describe it

02:22:57.060 | because I realized many people are just listening to us.

02:22:59.460 | All of this will be available both in the video,

02:23:01.780 | and then we'll link to the paper.

02:23:03.760 | So figure one is a figure that probably looks

02:23:06.840 | really familiar to people who look at any data

02:23:10.460 | that deal with survival.

02:23:11.440 | It's called the Kaplan-Meier survival curve.

02:23:14.200 | So on the x-axis for this curve is time,

02:23:17.720 | and time here is shown in months,

02:23:20.340 | and on the y-axis is the overall survival.

02:23:23.820 | At the very top, 100%, at the bottom, 0%.

02:23:27.720 | And it has three graphs or three curves

02:23:31.660 | that are superimposed on one another.

02:23:33.540 | For each of the three groups.

02:23:34.880 | Again, the control group, which is the GP100,

02:23:38.100 | the anti-CTLA-4 group by itself,

02:23:41.020 | and the anti-CTLA-4 plus GP100.

02:23:45.040 | And one of the characteristics of a Kaplan-Meier curve

02:23:48.100 | is by definition, they have to be decreasing

02:23:50.640 | in a monotonic fashion

02:23:52.320 | because it's cumulative overall survival.

02:23:54.320 | That just means it can't like come down and go back up.

02:23:57.740 | Nobody comes back to life.

02:23:59.020 | So once a person dies, they are censored from the study,

02:24:03.320 | and the curve drops and drops and drops.

02:24:06.000 | And you can see that they kind of highlight,

02:24:07.680 | and I actually think it makes the graph

02:24:09.160 | a little harder to read

02:24:10.760 | when they put some of those marks on there.

02:24:13.240 | But what really becomes clear when you look at this

02:24:16.880 | is that there's a key,

02:24:18.900 | there's a clear distinction between the curve

02:24:23.320 | for the placebo group, the GP100 group,

02:24:25.840 | and the other two, the two treatment groups.

02:24:29.680 | Now, you'll note at the very end

02:24:32.840 | that the two treatment groups

02:24:34.040 | appear to separate a little bit.

02:24:35.440 | I'll talk about that in a second.

02:24:37.760 | So when I look at these, Andrew,

02:24:39.620 | the first thing I always turn my attention to,

02:24:41.640 | I can't resist,

02:24:42.600 | I have to look at the right-hand side of the graph.

02:24:45.120 | 'Cause what is that really telling me, right?

02:24:46.940 | The tail of this is showing me the true overall survival.

02:24:50.380 | And I want to sort of figure out what is going on.

02:24:53.680 | So in the GP100 group, which is the placebo group,

02:24:57.800 | it is kind of amazing to think

02:24:59.500 | that there is still one person who is alive at 44 months.

02:25:03.960 | It's amazing.

02:25:05.840 | I mean, it's both sobering and amazing

02:25:07.780 | that like one person made it to 44 months.

02:25:10.520 | The next thing I ask myself is,

02:25:13.360 | well, how long did half of the people make it?

02:25:15.860 | That's called median survival.

02:25:17.720 | And to do that, you go up to the Y axis

02:25:21.560 | and you draw a little line from the 50 over,

02:25:24.720 | and then you bring that down.

02:25:27.040 | And that's awfully low.

02:25:30.100 | It's about, yeah.

02:25:31.360 | In fact, the table will tell us exactly what that is.

02:25:34.480 | 'Cause I think it's really hard to eyeball that stuff.

02:25:36.920 | So let's go to,

02:25:38.620 | so there's always a table that will accompany these things.

02:25:41.600 | And let's pull up that table.

02:25:44.220 | I've got this paper spread out over so many things.

02:25:46.800 | That's adverse events.

02:25:48.840 | Where's our survival table here?

02:25:50.640 | Your two subgroup analysis of overall survival.

02:25:53.960 | (silence)

02:25:56.120 | It would probably be helpful

02:25:59.760 | if I stapled these things together,

02:26:01.240 | 'cause it would be easier.

02:26:02.080 | Well, this is always a trade-off.

02:26:03.020 | Actually, since this is a Journal Club episode,

02:26:05.960 | I will say that stapling helps,

02:26:08.000 | but it also prevents one from separating things out,

02:26:11.020 | writing in the margins.

02:26:11.880 | I like these little mini clips.

02:26:14.000 | No financial relationship to the mini clips either.

02:26:17.400 | Just have to state that.

02:26:18.880 | 'Cause I always get, if you don't say that,

02:26:20.760 | people will go, oh, you must have a stake

02:26:22.120 | in these mini clips.

02:26:23.240 | I like these little mini clips.

02:26:24.480 | In fact, I'm such a nerd.

02:26:25.880 | I always have one of these Pilot V5, V7s on my pocket,

02:26:30.680 | or my hip.

02:26:31.520 | And then my pockets are always filled

02:26:33.180 | with these little mini clips.

02:26:36.160 | And, but then again, I have a friend who's a musician,

02:26:38.640 | and he's always raining guitar picks.

02:26:40.120 | So, you know, as far as occupational hazards go

02:26:42.860 | of being a nerd, these mini clips are-

02:26:44.960 | I'm a big fan of the mini clip as well,

02:26:46.880 | but I went without it today.

02:26:48.800 | All right, so thank you.

02:26:49.640 | Yes, table two.

02:26:51.000 | All right, so let's look at table two

02:26:53.400 | while looking at the Kaplan-Meier curve,

02:26:55.920 | 'cause now this allows us to see a couple of things.

02:26:57.520 | By the way, remember how I said there's like that one person

02:27:00.480 | who kind of is still alive in the treatment group?

02:27:03.960 | Well, you can tell that he's not a complete responder.

02:27:08.000 | He or she is not a complete responder,

02:27:10.040 | because under evaluation of therapy in table two,

02:27:14.720 | it says best overall response,

02:27:16.840 | and it says complete responders, zero.

02:27:19.640 | So there was zero complete responders in the placebo.

02:27:22.360 | There were two partial responders.

02:27:24.960 | Again, a partial responder is some lesions got smaller,

02:27:29.720 | some got bigger.

02:27:30.780 | Stable disease is, it didn't really change that much.

02:27:35.720 | And progressive disease is obviously, it went beyond.

02:27:39.960 | Not-

02:27:40.800 | When you say partial response, like lesions got smaller,

02:27:43.320 | are they literally just tracing the circumference

02:27:47.560 | of one of these skin lesions and saying,

02:27:49.760 | okay, it got bigger, smaller?

02:27:51.320 | Literally, we'd had rulers in the clinic.

02:27:53.440 | Yep, yep.

02:27:54.280 | Gosh, this feels so crude in terms of like,

02:27:57.360 | I mean, it makes total sense,

02:27:58.360 | but like in terms of like modern medicine,

02:28:00.240 | oh, like your lesion grew from like three millimeters

02:28:02.440 | to six millimeters, and we're literally

02:28:04.960 | like drawing little boundaries

02:28:06.140 | around little blotches on the skin.

02:28:08.320 | Yeah, you're putting a little measuring tape on them.

02:28:10.780 | Now, again, most of these are happening

02:28:12.480 | in the radiology suite,

02:28:13.720 | 'cause most of the disease for these patients

02:28:15.960 | is inside the body.

02:28:17.200 | Remember, more than 70% of these patients

02:28:19.400 | had visceral metastases.

02:28:21.280 | So liver, soft tissue, lung, brain, you know, these are,

02:28:26.280 | in fact, if you include lung, liver, brain, and viscera,

02:28:29.980 | it's pretty much all the patients.

02:28:33.280 | So most of this is looking at a CT scan

02:28:35.080 | or an MRI for the brain.

02:28:36.280 | Got it.

02:28:37.120 | Okay, so that's kind of the first thing that comes up.

02:28:40.500 | The median response rate should be shown

02:28:44.440 | pretty prominently here.

02:28:45.720 | So I'm looking through this, and where is median response?

02:28:50.720 | Maybe it's shown in a different table.

02:28:56.040 | Let's see.

02:28:58.500 | Not disease control rate, time to progression.

02:29:04.060 | I remember it's about 10 months,

02:29:08.120 | but maybe that's just in the text.

02:29:10.940 | Yeah, here it is.

02:29:12.160 | So I thought this would be in a table,

02:29:13.980 | but it's on page 715 of the paper.

02:29:17.420 | It just reports it.

02:29:18.800 | So, and I'm sorry, I misspoke.

02:29:21.200 | The 10 months was for the anti-CTLA-4 plus GP-100,

02:29:27.040 | and 6.4 months for the GP-100 alone.

02:29:31.600 | That's the control.

02:29:32.760 | And then 10.1 for the anti-CTLA-4 alone.

02:29:37.760 | Okay, so again, I always, and again,

02:29:40.080 | I'm just always doing this.

02:29:41.060 | I'm kind of going back to the paper to be like,

02:29:42.720 | does that make sense?

02:29:43.600 | And yeah, you kind of called it, right?

02:29:45.300 | You said median survival was about eight.

02:29:47.420 | Well, it turns out it's actually like six and change,

02:29:50.400 | 'cause it has that little ding in it,

02:29:52.700 | and it's out to a little past 10 on the two others.

02:29:55.140 | So the net takeaway here is,

02:29:57.140 | again, just to put that in English,

02:29:59.020 | 'cause it's so profound,

02:30:01.380 | 50% of the patients in the control group

02:30:04.640 | were dead in six months.

02:30:06.380 | 50% of the patients in the treatment group,

02:30:10.240 | both treatment groups were dead in 10 months.

02:30:13.680 | So what that means in cancer speak is,

02:30:16.920 | these drugs extended median survival by four months.

02:30:21.120 | Now, that's an important concept.

02:30:25.600 | You know, when we think about how has cancer therapy changed

02:30:30.080 | over the past 50 years,

02:30:31.880 | median survival for metastatic cancer

02:30:34.640 | has increased across the board.

02:30:36.840 | So a person today with metastatic colorectal cancer,

02:30:40.460 | or a woman today with metastatic breast cancer,

02:30:43.480 | or a person with metastatic lung cancer,

02:30:45.700 | these people will live longer with those diseases today,

02:30:50.740 | thanks mostly to treatments.

02:30:53.360 | This is not an early detection lead time bias issue.

02:30:56.820 | This is treatments allowing people to live longer.

02:31:01.120 | And that's an important part of the story,

02:31:04.180 | but it's only half of the story,

02:31:06.040 | yet it often gets touted as the story.

02:31:08.960 | The other half of the story,

02:31:10.180 | and frankly the story that I think is more important,

02:31:12.220 | is what is overall survival doing?

02:31:15.160 | And if you go back to those cancers, the answer is zero.

02:31:18.920 | So overall survival hasn't changed

02:31:23.280 | for solid epithelial tumors.

02:31:25.080 | It was 0% in 1970, and it's 0% today.

02:31:30.080 | - Everyone dies.

02:31:32.360 | Everyone dies from metastatic solid organ tumors.

02:31:36.040 | Now again, there's those niche examples I gave you.

02:31:38.240 | Testicular cancer is now an exception.

02:31:40.720 | GI stromal tumors would be an exception,

02:31:43.960 | and I'm not including leukemias and lymphomas,

02:31:46.720 | where now there are exceptions.

02:31:48.280 | - Okay, within, not to try and be overly optimistic,

02:31:53.000 | but if I look at the graph in figure one,

02:31:57.240 | and I look out at the tail of the graph.

02:31:59.320 | - That's right.

02:32:00.480 | - And for those that are just listening,

02:32:01.760 | what I see, and I'm far less, far less familiar

02:32:06.440 | with this type of work and this,

02:32:08.960 | analyzing these type of data.

02:32:10.160 | But what I see is that people in the placebo group,

02:32:12.560 | they're all dead, except that one.

02:32:15.140 | They're basically all dead at 44 months.

02:32:18.000 | But when I look at the number,

02:32:20.440 | how long it takes for everyone to be dead

02:32:23.080 | in the true treatment groups,

02:32:25.440 | it's like 50, looks like 53, 54 months or so.

02:32:29.960 | - And they're not dead.

02:32:31.040 | - That's the point.

02:32:31.880 | - They're hanging in there, right?

02:32:32.700 | So, because an extra,

02:32:35.340 | as somebody who lost both of my scientific advisors,

02:32:40.340 | two of the three, the other one, the suicide,

02:32:42.640 | we've talked about this before,

02:32:44.020 | but the other two to different cancers,

02:32:46.400 | both had the BRCA2 mutation, by the way,

02:32:48.500 | an extra eight to 10 months with your kids

02:32:53.720 | or with your spouse, or to quote unquote,

02:32:56.760 | get your affairs in order is a big deal.

02:32:59.480 | I mean, it's still depressing in the sense

02:33:00.840 | that nobody survives long-term,

02:33:03.280 | but an extra 10 months,

02:33:05.880 | as long as one is not miserable in that time,

02:33:08.480 | completely miserable.

02:33:10.320 | I mean, that's an extra 10 months of living, right?

02:33:13.120 | - Well, and what's interesting here is,

02:33:15.240 | the observation period stops

02:33:17.880 | and some of these patients are still going.

02:33:20.160 | So, what you're highlighting is kind of the point

02:33:22.720 | I wanna make, which is overall survival

02:33:26.060 | is the most important metric.

02:33:28.480 | And it's the highest bar, make no mistake about it,

02:33:31.480 | and it's certainly not the bar any drug company

02:33:35.040 | is ever going to wanna talk about for a cancer drug.

02:33:38.000 | - But why not?

02:33:39.200 | - Because they don't, none of them work, right?

02:33:40.760 | Like we don't have, you know, like drugs-

02:33:42.840 | - They only wanna talk about cures.

02:33:44.100 | They don't wanna talk about-

02:33:44.940 | - No, they only wanna talk about median survival.

02:33:46.160 | They only wanna talk about extending median survival.

02:33:49.360 | And there are lots of people out there

02:33:52.480 | that are on this platform, I don't need to get onto it,

02:33:55.320 | but who will say like, look, it's a real racket in oncology

02:33:59.040 | today, where drugs that are extending median survival

02:34:02.360 | by four weeks are being put on the market

02:34:06.000 | at a tune of, you know, 50 to $100,000 per treatment.

02:34:09.640 | That's not uncommon in oncology.

02:34:11.760 | There was one drug that was approved for pancreatic cancer.

02:34:14.280 | I believe it extended median survival by nine days

02:34:17.520 | and it costs $40,000.

02:34:18.960 | - And it's being advertised as significantly-

02:34:21.600 | - Yes, 'cause that was a statistical significant improvement

02:34:24.720 | in median survival.

02:34:25.600 | - So I'm just, yeah.

02:34:27.280 | - So it's, look, it's really understandable

02:34:29.160 | why people are very skeptical of the pharma industry.

02:34:31.820 | And I think, you know, a much more nuanced view is necessary.

02:34:35.400 | Clearly, I don't think pharma is all bad,

02:34:37.920 | but I really understand why people lose faith in pharma

02:34:41.080 | when, you know, these types of products

02:34:43.840 | somehow make regulatory approval.

02:34:46.640 | - Does insurance cover these kinds of drugs?

02:34:49.100 | - It can, in fact, it often does.

02:34:51.260 | It depends on the FDA approval, of course,

02:34:53.040 | and the indication, but a lot of times they do, right?

02:34:57.000 | So yeah, there's a societal cost to these things,

02:35:00.440 | but there's also a patient cost, right?

02:35:02.760 | So a lot of times insurance doesn't fully cover it

02:35:05.120 | and a patient has to bear the cost difference.

02:35:07.720 | And on top of that, you alluded to this a second ago,

02:35:09.840 | which is what if your quality of life

02:35:11.620 | is dramatically compromised as a result of this treatment?

02:35:14.580 | And yes, statistically, you're gonna live nine days longer

02:35:17.640 | or three weeks longer, but at what cost to your health

02:35:22.080 | in those final remaining days?

02:35:23.580 | And by the way, you're potentially straddling

02:35:26.640 | your loved ones with enormous debt in your absence.

02:35:30.220 | So it's a super complicated topic.

02:35:32.320 | - Yeah, there's a dignity component too.

02:35:34.660 | I mean, I've seen this in people dying.

02:35:36.480 | You know, at some point they become

02:35:38.640 | such a diminished version of their former selves

02:35:41.760 | that they don't wanna be seen by people that way.

02:35:45.320 | - So what is exciting about this drug,

02:35:50.680 | although this paper is not the one that shows it,

02:35:53.080 | the reason I chose this paper, Andrew,

02:35:54.980 | is because it was the first approval.

02:35:58.500 | A second drug came along that is an anti-PD-1 drug.

02:36:03.000 | That drug is called Keytruda.

02:36:04.720 | That drug turned out to be even better

02:36:06.960 | and has even a greater response rate,

02:36:10.140 | both in terms of median survival and overall survival.

02:36:13.360 | But this was the landmark paper.

02:36:15.460 | I also have a slight bias here,

02:36:18.600 | and I'll disclose in a moment why,

02:36:21.160 | but I think it just talks about very interesting biology.

02:36:24.040 | So let's talk about a couple of things

02:36:25.720 | that stuck out to me in this paper.

02:36:27.480 | The first thing that stuck out to me,

02:36:30.320 | and the authors didn't comment on it,

02:36:32.480 | unless they did and I missed it, is look at figure two.

02:36:37.160 | So figure two is the subgroup analyses

02:36:42.160 | where you're sort of showing a similar graph

02:36:47.100 | to the one you showed earlier,

02:36:49.380 | where you show the response rate

02:36:51.900 | or the change in response between the groups,

02:36:54.920 | and then you put the error bars on it.

02:36:57.140 | And this is where we talk about how,

02:36:59.260 | well, it's a 95% confidence interval,

02:37:01.160 | so does it touch the unity line?

02:37:03.080 | So these are called like tornado plots typically.

02:37:05.540 | And what you'll notice is that in the top,

02:37:10.840 | you're looking at sort of,

02:37:16.480 | it's comparing the anti-CTLA-4 with GP-100 versus the GP-100,

02:37:21.340 | and in the bottom,

02:37:22.180 | you're looking at the anti-CTLA-4 versus the GP-100.

02:37:25.000 | So at a glance, you can see GP-100 is not doing anything.

02:37:29.120 | I mean, that's the first takeaway of comparing A to B.

02:37:31.720 | What I find most interesting

02:37:34.760 | is look at the subgroup analysis of females.

02:37:37.640 | Notice that in females,

02:37:40.000 | while there's a trend towards risk reduction,

02:37:43.760 | and this is risk reduction for overall mortality.

02:37:46.720 | So again, I just want to restate that.

02:37:48.100 | The primary outcome of this trial

02:37:49.740 | was changed to overall survival,

02:37:51.540 | which I think is the better outcome, by the way.

02:37:53.640 | And overall, for all patients,

02:37:57.480 | when you compare anti-CTLA-4 plus placebo versus placebo,

02:38:02.480 | there was a 31% risk reduction in overall mortality.

02:38:08.120 | That's the mathematical interpretation

02:38:12.440 | of what you're seeing at the tail end

02:38:14.120 | of that Kaplan-Meier curve.

02:38:15.520 | - Living longer.

02:38:16.720 | - Living longer.

02:38:17.560 | - And it sounds like a big difference.

02:38:19.200 | And in some sense it is a big difference.

02:38:21.240 | - Well, it is for those people

02:38:22.800 | because you're really looking at basically

02:38:25.040 | 0% surviving in the placebo group

02:38:29.440 | versus 20% of people are still alive

02:38:33.760 | at 56 months in the treatment group.

02:38:36.560 | But look, that means 80% have died, right?

02:38:40.800 | But notice that, and sorry,

02:38:43.960 | when you just look at the anti-CTLA-4 plus GP-100

02:38:47.340 | in the subgroup B,

02:38:48.720 | that hazard ratio's even showing more compression.

02:38:51.640 | It's a 36% reduction in risk of death.

02:38:56.480 | But notice that the females did not reach significance.

02:39:01.360 | So in the first group, they barely do.

02:39:05.120 | And you can see that because the confidence interval

02:39:07.280 | runs from 0.55 to 0.92.

02:39:09.760 | And notice the error bar almost touches the line.

02:39:13.160 | And in the second one,

02:39:14.000 | it does not reach significance at all.

02:39:15.760 | So I actually went and kind of did a little reading

02:39:19.120 | on this after, and I said,

02:39:20.340 | "Hey, you know, how much did this study,

02:39:22.880 | was this an outlier study?"

02:39:24.680 | And it turned out it wasn't.

02:39:27.060 | And that about half the studies of anti-CTLA-4

02:39:31.040 | did indeed find that the drug was less effective

02:39:34.320 | in women than men.

02:39:35.880 | Which I found interesting.

02:39:36.720 | Now, I couldn't find any great explanation for it,

02:39:40.800 | but the most plausible explanations fit into two categories.

02:39:44.600 | The first are, maybe there are differences

02:39:46.840 | in the immune response to the drug,

02:39:49.300 | if you're a man or a woman.

02:39:51.040 | The second comes down to dosing.

02:39:54.120 | I should have said this at the outset,

02:39:55.260 | but of course these drugs are not like a pill,

02:39:57.400 | where it's like everybody gets, you know,

02:39:58.840 | 50 milligrams of this.

02:40:00.080 | They're all dosed based on weight.

02:40:02.160 | So this study is dosed, I believe,

02:40:04.400 | at three milligrams per kilogram.

02:40:06.400 | And because most men are heavier than women,

02:40:10.080 | men are getting a higher dose than women.

02:40:12.460 | And weight and body surface area and immune system,

02:40:17.520 | like these things are not all perfectly linear.

02:40:20.320 | So I kind of wonder if this difference

02:40:24.520 | is simply explained by men, on average,

02:40:27.220 | getting a higher dose than women.

02:40:29.140 | - Interesting.

02:40:29.980 | - Last thing I want to talk about here

02:40:33.960 | is in table three.

02:40:38.080 | So table three, always an important table

02:40:41.640 | to look at in any paper,

02:40:42.800 | is what are the adverse outcomes, right?

02:40:45.760 | What are the adverse effects of the drug?

02:40:47.320 | - Yeah, I spent a little bit of time with this

02:40:49.660 | and I confess it.

02:40:51.280 | You know, I definitely don't want cancer

02:40:53.280 | to the extent that I can avoid it,

02:40:54.920 | but this table made me wonder whether or not

02:40:57.800 | I would also want to just avoid cancer treatment,

02:41:00.280 | given the life extension provided.

02:41:03.160 | I mean, these adverse events are pretty uncomfortable.

02:41:07.280 | They sound like it.

02:41:08.120 | - Yeah, so just to put in perspective,

02:41:10.480 | and you always have to kind of be mindful

02:41:12.920 | of how many of these adverse events are occurring in people

02:41:16.280 | just because their disease is progressing.

02:41:18.280 | So the first thing I always want to look at

02:41:19.520 | is total adverse events in all three groups,

02:41:23.760 | not just grade.

02:41:24.640 | So grade three and grade four are real toxicities, right?

02:41:28.320 | Grade four toxicity is life-threatening toxicity, by the way.

02:41:31.580 | Grade three is pretty significant toxicity.

02:41:33.700 | Grade one and two, we typically just, you know,

02:41:36.120 | that's not that severe, right?

02:41:37.400 | A little rash to put some corticosteroids on it,

02:41:39.200 | it went away kind of thing.

02:41:40.560 | Okay, so in the treatment plus GP 100 group,

02:41:44.780 | 98.4% of people reported some event.

02:41:48.880 | So all but 1.6%.

02:41:51.420 | In the anti-CTLA-4 group alone, it was 96.7%,

02:41:55.900 | so only 3.1% did not.

02:41:58.640 | But in the placebo group, it's 97%.

02:42:00.880 | So it's important to keep in mind, like, you know,

02:42:03.580 | everybody's having some adverse effect.

02:42:05.300 | Okay, well, what if you say,

02:42:06.140 | well, let's just limit it to the most severe events?

02:42:10.520 | Well, let's just talk about grade four toxicities.

02:42:13.580 | There were 6.1% of those in the placebo group,

02:42:16.900 | 8.4% in the anti-CTLA-4 group,

02:42:20.120 | and 6.8% in the combined group.

02:42:23.180 | So not a huge difference in grade four toxicity.

02:42:26.840 | - Meaning that whatever adverse events are occurring

02:42:30.280 | may not be related to the treatment.

02:42:31.960 | - They may not be related to the treatment.

02:42:33.720 | Again, these are, if you think about it,

02:42:35.760 | and it's a very awful, sad, morbid thought to imagine,

02:42:38.880 | you're looking at the adverse responses of people,

02:42:42.920 | more than 80% of whom died

02:42:44.600 | during the course of a very, very short study.

02:42:47.520 | And so, you know, it's very difficult to disentangle

02:42:51.360 | what effects or what side effects a person is having

02:42:54.080 | just from that process as they are from the actual treatment.

02:42:58.400 | But if there is an area where there's a really clear

02:43:01.420 | difference, it's down in the autoimmune category.

02:43:05.360 | So if you look at any immune-related events,

02:43:08.840 | you can see that in the anti-CTLA-4 plus GP-100 group,

02:43:13.840 | it's about 60% in both of those treatment groups

02:43:17.560 | versus 30%.

02:43:19.520 | And if you look at the grade three and four toxicities,

02:43:23.080 | it's 10% in the anti-CTLA-4, 15%,

02:43:28.720 | in the anti-CTLA-4 alone group,

02:43:32.040 | and only 3% in the treatment.

02:43:34.240 | So that's a real difference.

02:43:36.740 | - Well, it makes sense that people getting this drug

02:43:39.520 | plus placebo or just the drug would have autoimmune issues

02:43:44.000 | because this is an immunotherapy.

02:43:45.920 | - It's an immunomodulator.

02:43:47.320 | In fact, what is it doing?

02:43:48.340 | It is taking the brakes off the immune system.

02:43:51.680 | - But then again, the things that they list out,

02:43:54.760 | pruritis, is that a irritation of the skin?

02:43:57.640 | - Yeah, irritation of the skin.

02:43:59.120 | - I'm not a physician,

02:44:00.040 | but I know that any itis is gonna be like an inflammation.

02:44:02.920 | And OMA, unfortunately, likely a cancer or cell replication.

02:44:07.080 | - Look at the difference in vitiligo.

02:44:09.400 | - I mean, wow.

02:44:12.640 | Yeah, so very little.

02:44:14.040 | - Sorry, sorry, look at the gastrointestinal differences.

02:44:16.640 | Yeah, and the vitiligo, right?

02:44:18.200 | So 3.7%, 2.3%, 0.8%.

02:44:21.680 | The GI stuff is the most common stuff

02:44:23.880 | you're gonna see there.

02:44:24.960 | Those are the really big ones.

02:44:26.420 | - And of course there's diarrhea and there's diarrhea.

02:44:28.760 | - Oh, yeah, yeah.

02:44:29.600 | - Like there's traveler's diarrhea,

02:44:31.000 | there's, you know, ate an overly spicy large meal

02:44:34.280 | the night before diarrhea.

02:44:35.280 | And then there's like, can't really do anything

02:44:38.040 | besides make trips back and forth to the bathroom.

02:44:40.080 | - Well, put it this way.

02:44:41.500 | There's, you know, colitis here is diarrhea so significant.

02:44:44.440 | These patients require IV fluids.

02:44:46.380 | Now, what you don't see here is how many of these patients

02:44:49.160 | actually required corticosteroids

02:44:50.800 | to reverse the autoimmunity.

02:44:52.840 | So a lot of times what'll happen here in these studies

02:44:54.960 | or with these drugs is the autoimmunity becomes

02:44:57.320 | so significant that you have to stop the drug

02:45:00.300 | and give corticosteroids.

02:45:02.040 | Do the exact opposite.

02:45:02.980 | You now have to shut the immune system down.

02:45:05.220 | So you just took the brakes off it with the drug

02:45:07.820 | and now you need to shut it down with corticosteroids.

02:45:10.520 | When I was, was I in med school?

02:45:14.420 | No, when I was in my fellowship,

02:45:17.240 | I wrote a paper about autoimmunity correlating

02:45:24.800 | with response rate in anti-CTLA-4 early on.

02:45:29.160 | This was during the phase two work.

02:45:31.280 | So, you know, so the NCI was a very early

02:45:35.480 | adopter of participating in these trials.

02:45:40.700 | And, you know, it was observed that,

02:45:45.280 | or at least hypothesized,

02:45:46.380 | this is what the paper basically wrote about,

02:45:48.380 | which was, is there any correlation

02:45:50.520 | between autoimmunity and response?

02:45:53.800 | And it turned out the answer was yes.

02:45:55.440 | There was a very strong correlation.

02:45:56.860 | So there was no difference in autoimmunity

02:46:01.840 | between the doses.

02:46:03.240 | And so the paper we wrote was two dosing schedules.

02:46:07.140 | So it was basically the full dose,

02:46:08.560 | the three milligrams per kilogram versus a low dose,

02:46:10.800 | one milligram per kilogram.

02:46:11.740 | This is a phase two trial.

02:46:12.960 | Those are your two arms.

02:46:14.560 | There turned out to be no difference

02:46:15.940 | in autoimmunity between them,

02:46:17.820 | but there was a big difference between

02:46:22.880 | the response rate that tied to autoimmunity.

02:46:26.320 | In other words, autoimmunity predicted response.

02:46:30.220 | Now, I think over time, these investigators,

02:46:34.660 | the doctors who administer these treatments,

02:46:36.000 | are getting better and better

02:46:37.040 | at catching these things earlier,

02:46:38.760 | because these autoimmune conditions

02:46:40.880 | can actually be devastating.

02:46:42.200 | So on a very personal note,

02:46:44.440 | when Kytruda came out,

02:46:48.280 | I want to say it was around 2011,

02:46:52.100 | no, no, no, gosh, it must've been 2013, 2014, thereabouts.

02:46:55.880 | Again, it was for treatment of metastatic melanoma.

02:47:00.600 | I want to come back and explain why melanoma

02:47:02.640 | gets all of the attention in autoimmune condition,

02:47:04.840 | in immunotherapy conditions, I'll state that.

02:47:07.700 | But anyway, a friend of mine got pancreatic cancer,

02:47:12.700 | and he got the bad type of pancreatic cancer.

02:47:17.420 | So this is like the adenocarcinoma of the pancreas.

02:47:22.420 | So this is a non-survivable type of cancer.

02:47:26.000 | Furthermore, his was unresectable, so--

02:47:29.040 | - Can you explain what that is?

02:47:30.400 | - Yeah, so about 20% of people who have pancreatic cancer

02:47:35.400 | technically have it in a way

02:47:38.040 | where you could still take out the head of the pancreas.

02:47:40.280 | - The Whipple procedure.

02:47:41.320 | - The Whipple procedure, right.

02:47:43.380 | Now, tragically, most of those patients will still recur.

02:47:46.180 | My understanding is that pancreatic cancer

02:47:49.840 | progresses from anterior to posterior in the pancreas,

02:47:53.680 | and that the Whipple is a removal

02:47:55.240 | of the front and the anterior.

02:47:56.960 | That's the Whipple procedure.

02:47:58.320 | So if the cancer has progressed far enough caudal

02:48:02.100 | into the posterior pancreas,

02:48:04.400 | then there's nothing left to cut out, basically.

02:48:07.120 | Can we survive without a pancreas for any amount of time?

02:48:10.700 | - Oh yeah, absolutely.

02:48:11.540 | - So why don't they just remove the whole pancreas?

02:48:13.160 | - Oh, that's my point, it's already micrometastisized.

02:48:16.340 | So it's not, the surgical procedure

02:48:18.660 | is not the challenge anymore, it used to be.

02:48:20.940 | So, you know, at Johns Hopkins,

02:48:23.080 | which is one of the hospitals where this was pioneered,

02:48:24.940 | like, the 30-day mortality for a Whipple procedure

02:48:29.060 | was, I don't know, 80%.

02:48:33.100 | And the reason was to figure out

02:48:36.020 | how to suture a pancreas to the bowel without the,

02:48:40.660 | so the pancreas is such an awful organ to operate on,

02:48:43.840 | because its enzymes are designed

02:48:46.000 | to digest anything and everything.

02:48:47.960 | So imagine now you have to cut the pancreas in half,

02:48:52.000 | take out the head of the pancreas with the duodenum,

02:48:54.760 | and then, somehow, sew that open half of a raw pancreas

02:48:59.320 | to the end of the jejunum, and not let it digest itself.

02:49:03.820 | - Someone at Hopkins figured this out?

02:49:05.960 | - No, the first one was actually done by A.O. Whipple,

02:49:09.320 | but yes, at Hopkins is where they figured out

02:49:12.980 | the way to put drains in, the surgical technique,

02:49:16.820 | how to do it in two layers, what type of stitches to use,

02:49:19.840 | like, all of the nuances of this were worked out

02:49:23.080 | in a few places, but I would say Hopkins

02:49:24.900 | more than any place else.

02:49:26.240 | - And are there physicians who, like,

02:49:28.600 | try this on non-human primates or something,

02:49:30.740 | or is this always just done on patients, and you keep trying?

02:49:32.760 | - Well, nowadays, I mean, put it this way,

02:49:34.280 | even 25 years ago, at a major center like Hopkins,

02:49:39.560 | the mortality of that procedure was less than 1%.

02:49:42.800 | - Amazing. - Yeah, totally amazing.

02:49:43.640 | - So there have been some victories.

02:49:45.300 | - Well, yes, but here's my point.

02:49:47.120 | That's no longer the bottleneck, right?

02:49:49.300 | Taking out the pancreas safely,

02:49:50.840 | as complicated and challenging as that is,

02:49:52.800 | and if you need a Whipple procedure,

02:49:54.360 | you only want to have it done by someone

02:49:56.080 | who just does that night and day,

02:49:57.820 | 'cause you don't want weekend warriors doing it.

02:50:00.320 | That's not why people are living or dying.

02:50:03.880 | They're dying because the cancer just comes back.

02:50:06.640 | It was already spread to the liver

02:50:08.640 | by the time you did it.

02:50:09.600 | You just didn't realize it yet.

02:50:11.680 | So whether you took out the whole pancreas,

02:50:13.380 | or the head of the pancreas, or the tail of the pancreas,

02:50:15.560 | the location of the tumor is predictive of survival,

02:50:20.560 | only in the extent that it basically is a window

02:50:25.240 | into how soon did symptoms occur.

02:50:27.220 | So pancreatic cancers in the tail tend to be more fatal,

02:50:32.160 | even though they're way easier surgically to take out,

02:50:36.180 | because by the time you develop symptoms

02:50:39.000 | of a tail pancreas cancer, it's a big cancer.

02:50:43.680 | - I was gonna ask this question later,

02:50:44.860 | but I'll just ask it now.

02:50:46.780 | Given the link between the immune system and these cancers,

02:50:50.520 | is there an idea in mind that people who are,

02:50:54.800 | let's say, 40 and older, or 50 and older,

02:50:59.140 | who don't yet, they're not diagnosed with any cancer,

02:51:02.700 | would periodically just stimulate their immune system

02:51:06.240 | to wipe out whatever early cancers might be cropping up?

02:51:10.840 | Just take a drug to just ramp up the immune system,

02:51:13.680 | even to the point where you start having a little diarrhea,

02:51:15.760 | maybe a few skin rashes, and then come off the drug,

02:51:19.920 | just basically to fight back whatever little cell growths

02:51:24.040 | are starting to take place in skin or liver,

02:51:26.440 | maybe for three weeks out of each year.

02:51:30.680 | I mean, why not?

02:51:32.480 | - Yeah, it's an interesting question.

02:51:34.740 | I've never thought of it through that lens.

02:51:36.540 | I suppose the question is, what can we do

02:51:39.980 | to keep our immune systems as healthy as possible

02:51:42.940 | as we age, because--

02:51:43.780 | - Stay on a normal circadian schedule.

02:51:45.900 | There's evidence for that.

02:51:47.040 | - Sure, no, there's evidence that,

02:51:49.000 | certainly if it promotes sleep,

02:51:50.420 | anything that promotes better rest

02:51:52.220 | is going to promote immune health.

02:51:54.000 | Because if you ask the macro question,

02:51:57.380 | which is like, why does the prevalence of cancer

02:51:59.480 | increase so dramatically with age?

02:52:02.440 | There are certain diseases where it's really obvious

02:52:06.480 | why the prevalence of the disease increases with age.

02:52:08.840 | - Yeah, like age-related macular degeneration.

02:52:11.680 | - Sure, or cardiovascular disease is by far the most obvious

02:52:15.240 | because it's an area under the curve exposure problem.

02:52:17.760 | The more exposure to lipoproteins

02:52:19.240 | and the more the endothelium gets damaged,

02:52:21.680 | the more likely you are to accumulate plaque.

02:52:23.400 | And again, it totally makes sense

02:52:25.120 | why 10-year-olds don't have heart attacks

02:52:27.520 | and 80-year-olds do.

02:52:29.780 | But when you sort of acknowledge that,

02:52:33.060 | well, hey, anybody's accumulating genetic mutations.

02:52:37.880 | We're always surrounded and being bombarded by things

02:52:40.820 | that are altering the genome of ourselves.

02:52:43.020 | Is it simply a stochastic process

02:52:45.500 | where the longer you live,

02:52:46.860 | the more of these mutations you're going to occur

02:52:48.600 | until at some point one of them just wins?

02:52:51.360 | I think that's got to be a big part of it.

02:52:53.980 | But I think another part of it,

02:52:55.660 | and clearly I'm not alone in thinking this,

02:52:57.940 | is that our immune system's getting weaker and weaker

02:52:59.980 | as we age, right?

02:53:01.200 | I mean, people become more susceptible to infections

02:53:05.100 | as they get older.

02:53:06.540 | And I think that that's equally playing a role

02:53:09.660 | in our susceptibility to cancer.

02:53:11.260 | So yeah, I think the question is

02:53:12.780 | how do you modulate immunity as you age?

02:53:17.100 | And to me, that's one of the most interesting things

02:53:19.020 | about rapamycin potentially is that

02:53:21.420 | when taken the right way,

02:53:22.680 | it seems to enhance cellular immunity,

02:53:26.860 | which again, that's potentially a really big deal.

02:53:29.180 | Again, at least in short-term human experiments

02:53:32.180 | in response to vaccination, it's enhancing vaccine response.

02:53:35.700 | So the question is, would that translate into cancer?

02:53:38.580 | Nobody knows.

02:53:39.860 | Could that be one of the reasons

02:53:41.440 | why animals treated with rapamycin live longer

02:53:44.580 | and get less cancer?

02:53:45.860 | Don't know.

02:53:46.700 | You know, it could also be that it's at a fundamental level

02:53:49.120 | that's targeting nutrient sensing.

02:53:50.820 | Where I was going with that story was that,

02:53:56.420 | maybe I'll back up for a moment.

02:53:57.940 | Why melanoma?

02:53:59.140 | So we didn't really know this like 30, 40 years ago

02:54:05.060 | in the early days of immunotherapy.

02:54:07.860 | But what we know now is that most cancers

02:54:11.820 | probably have about 40 mutations in them.

02:54:14.860 | That's like ballpark.

02:54:15.760 | 40, 50 mutations is standard fare for a cancer.

02:54:20.140 | But melanoma happens to be one of the cancers

02:54:23.020 | that has many, many more mutations.

02:54:25.740 | And the more mutations a cancer has,

02:54:28.340 | the more likelihood that it will produce an antigen

02:54:31.380 | that's recognized as non-self.

02:54:33.860 | And that's why in the early days of immunotherapy,

02:54:37.460 | the only things that worked were IL-2

02:54:40.140 | against metastatic melanoma and kidney cancer,

02:54:42.100 | because kidney cancer turned out to also be

02:54:44.260 | one of those cancers that, for reasons that are not clear,

02:54:47.840 | produced hundreds of mutations.

02:54:50.900 | And so it's no surprise that the early studies

02:54:53.700 | of checkpoint inhibitors were also done in

02:54:55.940 | metastatic melanoma,

02:54:57.280 | where you basically have more shots on goal.

02:54:59.900 | Again, if I'm going to take the breaks off my immune system,

02:55:03.020 | I might as well do it in an environment

02:55:05.360 | where there are more chances for my T cells

02:55:08.140 | to find something to go nuts against.

02:55:10.640 | So it's 2013, 2014, and this friend of mine

02:55:16.660 | who has something called Lynch syndrome,

02:55:19.700 | which is one of those few hereditary or germline mutations

02:55:24.640 | that results in a huge increase in the risk of cancer.

02:55:27.060 | He had already had colon cancer at about the age of 40

02:55:30.460 | and had survived that.

02:55:33.120 | It was a stage three cancer, but he had survived it.

02:55:35.620 | Well, now, five years later,

02:55:37.820 | had developed pancreatic cancer.

02:55:41.220 | And when he went to see the surgeon, they said,

02:55:44.300 | "Yeah, there's nothing we can do.

02:55:45.980 | "Like, it's too advanced."

02:55:48.300 | So that's, you know, to put that in perspective,

02:55:50.580 | that is a death sentence.

02:55:52.000 | And that's not, that's a six month survival.

02:55:56.300 | And at around that time,

02:55:59.020 | there was a study that had come out

02:56:00.380 | in the New England Journal of Medicine

02:56:01.780 | that had talked about how patients with Lynch syndrome

02:56:04.820 | had lots of mutations.

02:56:07.260 | And so we, you know, talked with his doctors

02:56:11.340 | about the possibility of enrolling him

02:56:13.540 | in one of the Kytruda trials.

02:56:15.740 | There was one going on, I think, at Stanford.

02:56:17.840 | And, you know, the thinking being,

02:56:20.440 | well, you know, you would want to target

02:56:22.500 | a checkpoint inhibitor against somebody

02:56:24.560 | who has a lot of mutations.

02:56:25.700 | And even though typically we don't see that

02:56:27.180 | in pancreatic cancer, his is a unique variant of it

02:56:30.760 | because it's based on this.

02:56:32.360 | And so sure enough, he was tested

02:56:34.400 | for these mismatch repair genes.

02:56:36.380 | He had them enrolled in the trial and amazingly,

02:56:40.020 | had not only a complete regression of his cancer,

02:56:42.900 | and he's still alive and cancer free today, 10 years later.

02:56:46.420 | But the treatment worked so well

02:56:50.440 | at activating his immune system

02:56:52.120 | that his immune system completely destroyed his pancreas.

02:56:55.080 | So now he has effectively had a pancreatectomy

02:57:00.120 | based on his immune system.

02:57:01.200 | So now he actually has type one diabetes.

02:57:03.880 | He has no pancreas.

02:57:05.180 | - He injects insulin to deal with it?

02:57:06.980 | - Yeah, no, he has to use insulin

02:57:10.160 | just like someone with type one diabetes.

02:57:11.000 | - Okay, but he had to pick being alive

02:57:12.620 | with type one diabetes.

02:57:14.140 | - Yeah, of course, and no comparison.

02:57:15.600 | But it's just an interesting example

02:57:17.540 | of how remarkable this treatment was able to work

02:57:22.040 | when you could completely unleash

02:57:24.460 | the immune system of a person

02:57:26.560 | and you eradicate the cancer

02:57:29.320 | and the rest of the cells around it.

02:57:31.440 | And there are many organs we could live without.

02:57:35.400 | There are certain organs you can't live without.

02:57:38.460 | You can't live without your heart, lungs, liver, kidneys.

02:57:41.660 | But many things that kill people arise from organs,

02:57:45.820 | the breast, you could live without all breast tissue.

02:57:48.140 | - Prostate.

02:57:48.980 | - Prostate, you can live without all prostate tissue.

02:57:49.820 | - No one would choose to live without these, but--

02:57:52.940 | - Right, but I'm saying if you had metastatic cancer

02:57:55.560 | and you had a bullet that could selectively target a tissue,

02:58:00.180 | you would take it.

02:58:01.020 | And right now, the only tissue we can do that against

02:58:04.340 | is a CD19 B cell, and that's what those CAR T cells are.

02:58:08.060 | So right now, these are not tissue-specific treatments,

02:58:10.580 | but they're mutation-specific.

02:58:13.020 | The last thing I'll say about this paper

02:58:14.100 | that I found interesting, and I was looking for it

02:58:16.620 | and I was surprised, they didn't at all comment

02:58:19.500 | on if there was any correlation

02:58:20.900 | between autoimmunity and response.

02:58:22.700 | So they obviously acknowledge the autoimmunity in table three

02:58:28.140 | but I would have loved to have seen a statistical analysis

02:58:31.060 | that said, "Hey, is there any correlation

02:58:33.540 | "between response rate and autoimmunity?"

02:58:35.660 | But they didn't comment to that effect.

02:58:38.460 | So we're left kind of wondering

02:58:40.560 | what the current state of that is.

02:58:43.300 | And I guess in summary, I'll say that the reason

02:58:46.400 | I thought this was an interesting paper to present

02:58:49.000 | is that I still believe that immunotherapy

02:58:52.460 | is probably the most important hope we have

02:58:56.820 | for treating cancer.

02:58:59.460 | And while I think we're still only scratching

02:59:01.340 | the surface of it, so collectively,

02:59:04.260 | the overall survival increase for patients

02:59:07.060 | with metastatic solid organ tumors

02:59:09.460 | is about 8% better than it was 50 years ago.

02:59:13.340 | And virtually all of that has come from

02:59:15.420 | some form of immunotherapy, I think is promising.

02:59:19.860 | And I think the holy grail, meaning the next step,

02:59:24.220 | if you go back to where we started the discussion,

02:59:27.060 | is coming up with ways to engineer T cells

02:59:32.060 | to be even better recognizers of antigens.

02:59:36.900 | And there's many ways to do that.

02:59:38.340 | One is to directly engineer them.

02:59:40.440 | Another is to find T cells

02:59:42.880 | that have already migrated into tumors.

02:59:45.020 | Those are called tumor infiltrating lymphocytes, or TIL.

02:59:48.620 | And expanding those and engineering them

02:59:52.300 | to be better and younger.

02:59:53.860 | - Is it possible to engineer our own T cells

02:59:56.820 | to be more pH variant tolerant?

03:00:00.100 | Meaning, since this cloaking of a local area

03:00:05.140 | by changing the pH, could we pull some T cells?

03:00:09.380 | I'm always thinking about the inoculation stuff,

03:00:11.380 | like pull some T cells as part of our standard exam

03:00:14.920 | when we're 30 and grow some up in an environment

03:00:19.220 | that the pH is slightly more acidic than normal,

03:00:24.180 | and then reintroduce them to the body.

03:00:26.300 | I mean, after all, they are our T cells.

03:00:28.340 | In other words, give them a little opportunity to evolve

03:00:33.260 | so that the conditions they can thrive in, right?

03:00:36.180 | Or even just keep them in the freezer in case we need them.

03:00:38.500 | - Yes, so the interesting thing is,

03:00:40.260 | I don't know that if you just got them to be comfortable

03:00:43.260 | in a lower pH, it would be sufficient,

03:00:45.360 | because there are still so many other things

03:00:49.620 | that the cancer is doing

03:00:51.660 | as far as using other secreting factors.

03:00:54.980 | It seems that by far the most potent thing

03:00:59.340 | comes down to expanding the number of T cells

03:01:03.680 | that recognize the antigen,

03:01:05.880 | and making sure that you can get that number big enough

03:01:09.580 | without aging them too much.

03:01:11.180 | So in some senses, it has become

03:01:13.340 | a longevity problem of T cells.

03:01:15.980 | The way to think about it is you want an army of soldiers

03:01:20.580 | who are wise enough to recognize the bad guys,

03:01:24.420 | which comes with age, but young enough to go and kill.

03:01:29.180 | And right now, both extremes seem to be unhelpful, right?

03:01:33.880 | When you go and find tumor-infiltrating lymphocytes

03:01:36.380 | in a tumor, they're very wise.

03:01:38.220 | They know which one.

03:01:39.380 | They've demonstrated that they can do everything.

03:01:41.700 | They can outmaneuver the cancer,

03:01:43.500 | but they're too old to do anything about it.

03:01:46.100 | And when you take them out to try to expand them

03:01:47.940 | by three logs, which is typically what you need to do,

03:01:50.340 | expand them by a thousand fold, they can't do anything.

03:01:53.300 | - Got it.

03:01:54.140 | And what about avoiding melanoma altogether?

03:01:55.740 | I mean, obviously avoiding sunburn.

03:01:58.440 | Somehow I got couched as anti-sunscreen

03:02:00.880 | and that is absolutely not true.

03:02:02.820 | I said some sunscreens contain things

03:02:05.660 | that are clearly immune disrupt, endocrine, excuse me,

03:02:08.700 | disruptors, and we're gonna do a whole episode on sunscreen.

03:02:11.580 | Maybe we could do some journal clubs on them.

03:02:12.980 | - It's funny, I think I'm actually planning something

03:02:14.860 | on that as well.

03:02:15.680 | I wanna do a deep dive on that.

03:02:16.520 | - Yeah, I mean, and some dermatologists reached out,

03:02:18.420 | some very, very skilled dermatologists reached out

03:02:21.220 | and said that indeed some sunscreens are downright dangerous,

03:02:24.660 | but of course melanoma is super dangerous.

03:02:27.140 | Physical barrier, no one disputes physical barriers

03:02:29.680 | for sunscreen, like everyone agrees that that is unlikely

03:02:33.020 | to have endocrine disruption.

03:02:34.660 | So physical barriers are undisputed,

03:02:37.540 | but aside from limiting sunlight exposure to the skin,

03:02:42.540 | what are some other risks for melanoma?

03:02:46.080 | - I mean, I think that's the biggest one.

03:02:47.280 | I do not believe that smoking poses a risk for melanoma

03:02:50.560 | and if it does, it's gonna be very small.

03:02:53.140 | There are hereditary cases, so one needs to be pretty

03:02:56.300 | mindful when taking a family history.

03:02:58.060 | And by the way, there are really weird genetic conditions

03:03:02.500 | that link melanoma to other cancers,

03:03:04.780 | such as pancreatic cancer, by the way.

03:03:06.520 | So whenever I'm taking somebody's family history

03:03:08.820 | and I hear about somebody that had melanoma

03:03:11.040 | and someone that had pancreatic cancer,

03:03:13.600 | there's a couple of genetic tests we'll look at

03:03:16.820 | to see if that's a person that's particularly sensitive

03:03:19.740 | just from a genetic predisposition.

03:03:23.180 | But I do think that first and foremost it's,

03:03:25.580 | and by the way, I think with melanoma,

03:03:28.340 | although it's not completely agreed upon,

03:03:31.460 | I think it's less about sun exposure

03:03:34.180 | and more about sunburn, right?

03:03:36.640 | So, and again, I'm sure there's somebody listening to this

03:03:39.100 | who will chime in and apply a more nuanced response to that.

03:03:43.860 | But I think there's a fundamental difference

03:03:45.900 | between I'm out in the sun getting sun,

03:03:48.380 | making some vitamin D versus I'm getting scorched

03:03:52.140 | and undergoing significant UV damage.

03:03:55.700 | There might also be something to be said

03:03:57.460 | for the time in one's life.

03:03:58.960 | And I've certainly seen things that suggest

03:04:01.240 | that early repeated sunburns would be more of a risk.

03:04:05.620 | So look, I think that's not a controversial point

03:04:10.640 | in the sense that like who wants to be sunburned, right?

03:04:13.280 | So it's like whatever one needs to do to be sunburned,

03:04:15.800 | whether it's being mindful of what the UV index is,

03:04:20.100 | wearing the appropriate sunscreen.

03:04:23.380 | I also find the whole kind of anti-sunscreen establishment

03:04:28.020 | to be a little bit odd.

03:04:29.620 | - Well, the anti-sunscreen establishment is odd.

03:04:32.420 | I'm trying to open the door for a nuanced discussion

03:04:34.980 | about the fact that some sunscreens

03:04:36.660 | really do contain things like oxybenzines

03:04:39.780 | and things that are real.

03:04:40.620 | - Yeah, but when you're spraying them on kids.

03:04:42.860 | - Yeah, but when you just look at the straight,

03:04:45.140 | the good old fashioned mineral sunscreens.

03:04:47.740 | - Perfectly safe.

03:04:49.060 | - As far as we know.

03:04:50.180 | I mean, I also dare we cross the seed oil debate into this.

03:04:55.180 | Some of the folks who are really anti-seed oil

03:04:57.860 | also claim that seed oils increase risk for sunscreen.

03:05:01.380 | Peter and I are smiling because we have teed up

03:05:04.740 | a debate soon with some anti-seed oil

03:05:10.220 | and less anti-seed oil experts.

03:05:13.180 | So that's forthcoming.

03:05:14.620 | That's gonna be a fun one.

03:05:15.740 | I will be doing all of that with our shirts on.

03:05:18.840 | (laughing)

03:05:19.680 | - I really appreciate you walking us through this paper,

03:05:22.560 | Peter, I've never looked at a paper on cancer

03:05:25.560 | and certainly not one like this.

03:05:28.120 | I learned a lot and it's such an interesting field,

03:05:33.120 | obviously because of the importance of getting people

03:05:34.960 | with cancer to survive longer and lead better lives,

03:05:37.720 | but also because of the interaction with the immune system.

03:05:41.440 | So we learned some really important immunology.

03:05:43.760 | - Yeah, and this was great.

03:05:47.440 | I feel much more confident now in the belief

03:05:50.720 | that the exposure to light early and late in the day

03:05:54.840 | can actually have benefits.

03:05:56.160 | And as I said, I think that there's some causality here

03:06:00.720 | and I think it shouldn't be ignored.

03:06:02.560 | - Cool, well, this was our second journal club.

03:06:05.280 | I look forward to our third.

03:06:07.160 | Next time you'll go first, we'll just keep alternating.

03:06:09.760 | And we've also switched venues,

03:06:11.200 | but we both wore the correct shirt.

03:06:14.840 | And I hope people are learning

03:06:18.440 | and not just learning the information,

03:06:21.240 | but learning how to parse and think about papers.

03:06:23.240 | And I certainly learned from you, Peter.

03:06:24.920 | Thank you so much.

03:06:25.880 | - Yeah, thanks, Endor, this was great.

03:06:27.280 | - Thank you for joining me

03:06:28.160 | for today's journal club discussion with Dr. Peter Attia.

03:06:31.600 | If you're learning from Endor and enjoying this podcast,

03:06:33.980 | please subscribe to our YouTube channel.

03:06:35.920 | That's a terrific zero cost way to support us.

03:06:38.320 | In addition, please subscribe to the podcast

03:06:40.560 | on both Spotify and Apple.

03:06:42.320 | And on both Spotify and Apple,

03:06:43.760 | you can leave us up to a five star review.

03:06:45.780 | Please also check out the sponsors mentioned

03:06:47.480 | at the beginning and throughout today's episode.

03:06:49.640 | That's the best way to support this podcast.

03:06:52.100 | If you have questions or comments for me

03:06:53.980 | or guests or topics that you'd like me to include

03:06:56.140 | on the Huberman Lab Podcast,

03:06:57.660 | please put those in the comments on YouTube.

03:06:59.700 | I do read all the comments.

03:07:01.360 | Not on today's episode,

03:07:02.560 | but on many previous episodes of the Huberman Lab Podcast,

03:07:05.120 | we discuss supplements.

03:07:06.640 | While supplements aren't necessary for everybody,

03:07:08.840 | many people derive tremendous benefit from them

03:07:10.960 | for things like improving sleep, for hormone support,

03:07:13.800 | and for focus.

03:07:14.860 | To learn more about the supplements

03:07:16.120 | on the Huberman Lab Podcast,

03:07:17.440 | you can go to Live Momentus, spelled O-U-S.

03:07:20.080 | So that's livemomentus.com/huberman.

03:07:23.280 | If you're not already following me on social media,

03:07:25.260 | I am @hubermanlab on all social media platforms.

03:07:28.320 | So that's Instagram, Twitter,

03:07:29.800 | now called X, Threads, LinkedIn, and Facebook.

03:07:32.800 | And on all of those platforms,

03:07:34.280 | I discuss science and science-related tools,

03:07:36.600 | some of which overlaps with the content

03:07:38.040 | of the Huberman Lab Podcast,

03:07:39.280 | but much of which is distinct from the content

03:07:41.280 | on the Huberman Lab Podcast.

03:07:42.740 | Again, that's Huberman Lab on all social media platforms.

03:07:45.800 | If you haven't already subscribed

03:07:46.940 | to our neural network newsletter,

03:07:48.840 | it is a zero-cost newsletter

03:07:50.720 | that includes summaries of podcast episodes,

03:07:53.280 | as well as protocols in the form of brief

03:07:55.380 | one to three page PDFs

03:07:57.080 | for things like how to improve sleep,

03:07:59.040 | how to regulate dopamine, deliberate cold exposure,

03:08:01.800 | deliberate heat exposure, exercise protocols, and much more.

03:08:05.720 | To sign up, simply go to HubermanLab.com,

03:08:08.200 | go to the menu tab, scroll down to newsletter,

03:08:10.720 | and provide us with your email.

03:08:12.040 | And I want to point out

03:08:12.900 | that we do not share your email with anybody.

03:08:14.960 | And again, the newsletter is completely zero cost.

03:08:17.580 | Thank you once again for joining me

03:08:19.000 | for today's Journal Club discussion with Dr. Peter Attia.

03:08:22.600 | And last, but certainly not least,

03:08:24.880 | thank you for your interest in science.

03:08:26.800 | (upbeat music)

03:08:29.380 | (upbeat music)

Journal Club with Dr. Peter Attia | Effects of Light & Dark on Mental Health & Treatments for Cancer

Chapters