Using Science to Optimize Sleep, Learning & Metabolism

The third episode of the Huberman Lab podcast is framed as office hours, modeled on students visiting a professor to ask for clarification or to explore a topic in more depth. Questions were collected from the comment sections of the first two episodes on YouTube and from Instagram. From that pool, the episode selected two kinds: questions that were asked and liked very often, and questions that could expand on topics already covered. Unanswered questions were to be tracked, with further episodes on sleep, wakefulness and learning planned for January and possibly into February.

The questions were sorted roughly into light, exercise, supplementation, temperature, learning, plasticity and mood-related disorders. Before any of the material, the host repeats a disclaimer: he is a professor, not a physician, and does not prescribe anything, including behavioral protocols. Listeners are told to filter the information through their own health or illness and to consult a licensed healthcare professional before adding or removing anything from their daily routines.

I'm not responsible for your health, you are, so be smart with this information and be a stringent filter, as we say.

The first question asks whether moonlight, firelight or candlelight at night affects circadian rhythms. To answer it, the episode first defines the unit used throughout earlier discussions. Light intensity is measured in lux, sometimes also described in candelas. One lux equals the illumination of a one-square-meter surface one meter away from a single candle. On that basis, 6,000 or 10,000 lux can be pictured as that many candles shining on one square meter from one meter away.

The answer given is that moonlight, candlelight and even a roaring fireplace do not reset the circadian clock at night or trick the brain into registering morning, even though such sources can seem very bright up close. The explanation rests on the melanopsin ganglion cells in the eye, also called intrinsically photosensitive ganglion cells, discussed in the previous episode. These cells respond best to the blue/yellow contrast of the low-angle rising and setting sun, and they adjust their sensitivity across the day so that moonlight or fire does not trigger the brain's daytime signals.

There is one qualification. The episode acknowledges folklore about full moons changing behavior, mentions that a little quality science exists on this that may be covered later, and allows that a very full, very bright moon could possibly have some waking effect. Otherwise, the advice is to enjoy fireplaces and candles, provided they are used safely, without worrying about circadian disruption.

Many listeners asked about red light, partly because commercial products sell sheets of very bright red lights to be viewed early in the day, with claims about improving mitochondrial function and metabolism. The host declines to name brands or specific studies. He says that, apart from one exception, the studies he has seen supporting positive effects of red light on the visual system were not published in top-tier journals, and that he is unsure of their peer review and stringency.

The exception is a study from Glen Jeffery's lab at University College London, published the previous year. The host says he knows Jeffery and considers the data high-quality. According to the episode, the study showed that viewing red light for a few minutes each morning can benefit mitochondria in photoreceptors, the retinal cells that turn light into electrical signals and that tend to decline in function with age. The episode links this decline to conditions such as age-related macular degeneration and to general worsening of photoreceptor function. The reported effect applied to red light flashes early in the day, not late in the day.

The evidence is explicitly described as preliminary. The study needs support from additional studies, a clinical trial is underway, and the host recalls that the report covered about 12 patients, while hedging on that number.

Most questions, though, were about using red light later in the day. In principle, red light does not stimulate the melanopsin neurons that signal daytime. However, the episode warns that the red lights on commercial panels are far too bright and would wake up the body and brain. Anyone who uses such products, which the host neither recommends nor discourages, should use them early in the day. For evening and nighttime use, red light should be very dim, much dimmer than most commercial products. Red light is not required at night, but dim red light lets people see reasonably well without waking the circadian clock or causing the dopamine disruption discussed in the previous episode.

The red-light discussion leads to blue light. The episode argues that people focused on avoiding blue light are missing the point. Lots of blue light is desirable early in the day and throughout the day, so blue blockers should not be worn then, and perhaps not at all. At night, blue blockers do not help if the lights are bright enough, because the relevant cells and mechanisms will still be activated.

The host traces the confusion to how the research was read. Melanopsin cells do respond best to blue light, which led to the belief that blocking blue light would prevent circadian resetting at night and the harmful effects of screens. But many of the early papers studied these neurons isolated in a dish, cut off from their natural connections in the eye. In a living eye, these cells respond to blue light and also to other wavelengths, because they receive input from photoreceptors as well as detecting light directly. Blue light being the best stimulus does not make it the only one.

a scientific paper can be accurate without being exhaustive, and a lot of claims about products can be accurate but not exhaustive.

The practical rule is to get screen light, overhead light and especially sunlight during the day, and to avoid bright light at night whatever its color. The episode says the papers and data were misunderstood when most of the product recommendations were made.

Some listeners mentioned claims that light shone into the ears, onto the roof of the mouth or up the nose can help set circadian rhythms. The answer given is no, at least not directly. The episode uses the question to draw a distinction it presents as more useful than the usual idea of a placebo effect: the difference between modulation and mediation.

Many things can modulate biology. Putting lights up one's nose, which the host asks people not to do, might shift biology through the stress hormone released by the act itself. The previous episode noted that almost anything different and dramatic enough can phase-shift circadian rhythms. The real question is whether the light is mediating the process, meaning whether it acts through the natural biology of the system being targeted.

The host also rejects the language of biohacking. He defines a hack as using something for a purpose it was not intended for, a kind of cheat, and argues that biology does not work well that way. He has read the papers behind commercial ear-light products and says they were not in top-tier journals and often disclosed commercial conflicts of interest in the fine print. He suspects they lacked controls for heat, which can itself shift circadian rhythms.

He leaves the claim open to revision. If someone sends a peer-reviewed paper from a quality journal showing that light delivered anywhere other than the eyes can mediate human circadian rhythms or wakefulness, he says he will review it and publicly change his position. Until then, his stance is that in humans these light effects work through the eyes. Other animals have extraocular photoreception, but humans do not. The episode adds that it favors behavioral and zero-cost tools first, while deferring to doctors when there is a clinical need for prescribed drugs.

Many people asked whether sunlight through a window works. The episode says that setting the circadian clock with sunlight through a window takes fifty to a hundred times longer, and points to papers discussed in the previous episode and in conversations with Jamie Zeitzer of Stanford.