It's the first of October.
Which means somewhere out there, somebody is already feeling the floor drop out.
Right. The days are getting shorter, the light is getting thinner, and for people with seasonal affective disorder, this is the week it starts.
Not the calendar week. The light week. It just happens to land around now.
Which is where Daniel's question comes in. He wrote in about a friend of his who struggles with this, and he says it sounds debilitating. And what he wants to unpack isn't really SAD itself. He says SAD he can sort of understand. We associate light with vitality. That part tracks. What he can't get his head around is why depression has such a hold on energy at all. He put it as: for him, that connection has never been obvious.
Because it isn't obvious.
Right. And he's careful about it. He says most people with depression probably don't have SAD, and the connection there is a lot more complicated than "I don't get enough natural light so I feel sad and tired." So the ask is really two things. First, the relationship between mood, light, and energy. Second, and this is the harder one, why non-seasonal depression drains you the way it does.
Two different mechanisms wearing the same symptom.
That's the episode.
Start with what SAD actually is, because people use the term loosely. It's a type of depression with a recurrent seasonal pattern. Symptoms typically run four to five months, starting in late fall or early winter and lifting in spring and summer. And the diagnosis is stricter than people assume. You need depressive episodes in a specific season for at least two consecutive years, occurring more frequently at that time of year than at any other.
So it's not "I feel a bit flat in January."
One in five people in the UK report winter blues. About two percent have actual SAD. That's a big gap.
And the timing. Why October, if the diagnostic window is late fall?
Because it tracks daylight, not the calendar. The depression is related to changes in daylight hours. Norman Rosenthal, who named the condition, makes the point that it isn't strictly a winter thing. Symptoms can show up during cloudy stretches, in dark offices, in dark apartments. Whenever there's insufficient sunlight.
So a person in a windowless office in September could be in it already.
If they're vulnerable, yes.
What does it actually feel like? Because the symptom profile is strange.
It's atypical. That's the clinical term, and it's the opposite of what you'd expect. Classic non-seasonal depression tends toward insomnia and appetite loss. Winter-pattern SAD goes the other way. People oversleep. They overeat, specifically carbohydrates, and they gain weight. And they withdraw socially. The clinical word for it is hibernating.
Hibernating is the word they chose. That's not a metaphor anyone had to reach for.
It's almost embarrassingly on the nose.
So you've got a condition that looks like a mammal winding down for winter. Which raises the question of whether that's literally what it is.
Hold that. That's actually one of the live hypotheses and we'll get to it.
Lay out the light pathway first. Because Daniel's intuition, that light equals vitality, turns out to be more mechanically real than most intuitions people have about mood.
It's real, and the reason it's real is that the eye does more than see. Everyone learns about rods and cones. Those are the image-forming photoreceptors. But there's a third class, the intrinsically photosensitive retinal ganglion cells. They express a pigment called melanopsin.
And they don't build pictures.
They don't build pictures at all. They're not there for vision. They innervate dozens of brain areas to influence physiology, behavior, perception, and mood. That's the phrase from a 2019 review by Do in Neuron. Dozens of brain areas.
For cells that don't help you see anything.
They're doing something else entirely. The term is non-image-forming functions. Circadian photoentrainment, which is how your internal clock gets set to the actual day. The pupillary light reflex. And modulation of sleep, alertness, and mood.
So there's a dedicated channel from the retina to the mood and alertness systems that bypasses vision entirely.
A separate wire. And here's where it gets contested. The standard story is that light affects mood by shifting your circadian clock. You get out of sync, you feel bad. That's the phase-shift hypothesis for SAD.
Which is what?
Later winter sunrises delay your internal rhythms relative to your sleep and wake schedule. So your mood lows land during the day instead of at night, and you wake up in what amounts to night mode. Tired, sluggish, flat.
That's a satisfying story. Light sets the clock, clock drifts, mood follows.
It is satisfying. And Lazzerini Ospri, Prusky and Hattar, in the Annual Review of Neuroscience in 2017, put forward an alternative. They argue light's effect on mood may run through an ipRGC-dependent mechanism independent of circadian disruption. Meaning light might affect mood directly, not only by moving the clock.
So two models. Light moves the clock, mood follows. Or light hits mood circuitry in parallel and the clock is a separate thing.
And the review explicitly presents a model that dispenses with assumptions of causality between circadian disruption and mood. That's a real fork in the field, not a settled question.
Which matters for treatment, presumably.
It matters for what you'd even try to fix. If it's the clock, you fix the clock. If light is acting on mood directly, then timing might matter less than we think, and it's more about total light exposure.
Rosenthal's own suggestion was genetic, right?
He suggested some people with SAD may have abnormalities in genes for serotonin transmission or in retinal light sensitivity. And he specifically flagged the retinal pigment melanopsin as a possible site.
So the receptor itself might be tuned differently in different people.
Which would explain why one person in a dim office is fine and the person at the next desk is not.
Give me the history. Because this condition has an origin date.
1984. Rosenthal and colleagues described it in Archives of General Psychiatry, based on twenty-nine patients in Maryland. And here's the detail that surprises people. Twenty-seven of those twenty-nine had bipolar disorder.
So the first cohort was almost entirely bipolar.
Almost entirely. Which is not how SAD is popularly imagined at all.
Anna Wirz-Justice, at Basel, said of that paper that it was as though this something had been there all the time, but it didn't have a diagnosis or label until then.
Which is the story of a lot of diagnoses. The thing existed. The name didn't.
And Rosenthal wasn't theorizing from a distance. He came to the US from South Africa in 1976 and described his own experience of it. A draining of energy, difficulty getting through his tasks once the days drew short and dark.
He lived the thing he named.
Numbers. How common is this?
By the standard questionnaire criteria, up to three percent of Europeans, ten percent of North Americans, one percent of Asians. And there's a latitude gradient in the US data. New Hampshire around nine point four percent. New York four point seven. Maryland six point three. Florida four.
That's not a clean line.
It isn't, and I'd flag that. Florida at four percent is not trivial, and Maryland comes in above New York. Latitude is doing something, but it isn't doing everything. Cloud cover, indoor life, who responds to the survey, all of it muddies the picture.
Women?
About four times as likely as men.
And the evolutionary hypothesis. The one I told you to hold.
This is the part that makes the hibernating language look less like a metaphor. The idea is that SAD may be a retained mammalian mechanism. Sheep track seasons by the duration of melatonin release. Something analogous may be sitting in us. And becoming more lethargic and depressed during colder months made sense when food was scarce. You conserve energy. You wait it out.
So the low-energy state isn't a malfunction. It's an old program running in an environment that no longer needs it.
That's the argument. And there's a counterexample that complicates the whole framing. The Sami, in northern Scandinavia, embrace the seasonal slowdown rather than fighting it. Which challenges the assumption that year-round high energy is the healthy norm and everything else is pathology.
That's the part I want to sit with, because it reframes the question. Daniel asked why depression drains energy. But if there's a version of the low-energy state that's culturally normal and not distressing, then the drain isn't automatically the disease.
The distress is the disease. The slowdown might just be October.
Alright. That's the easy half. Light, melanopsin, circadian phase, an old mammalian program. All of it intuitive, because light and vitality feel connected.
And now the harder half.
Most depression isn't seasonal. What's the actual split?
In both SAD and general depression, somewhere between ten and twenty percent of patients report a seasonal variation in their symptoms. So there's overlap in both directions. But the trigger differs. SAD onset is usually triggered by shortened daylight exposure, as opposed to negative life events.
Which is a clean distinction on paper and probably messier in a person.
Almost certainly messier. But it's the right frame. One is triggered by the sky. The other is triggered by life, or by biology, or by both.
So take the seasonal piece out. A person with major depression, in July, with plenty of light. Why are they exhausted?
Start with how common it is, because this is the part that gets underestimated. Fatigue occurs in over ninety percent of patients with major depressive disorder.
Ninety.
Over ninety. This isn't a side symptom. For most people with the condition, it's one of the defining features.
And the folk explanation is laziness. Or lack of willpower.
Which is wrong, and it's wrong in a specific, mechanistic way. There are at least two distinct systems involved. The first is reward processing.
Meaning what, exactly?
Depression involves impaired reward processing and a reduced willingness to exert effort for reward. And that's linked to dopamine function. There's a 2024 framework paper in Translational Psychiatry on exactly this. And separately, anhedonia severity predicts lower motivation for both physical and cognitive effort. Both kinds. Not just getting off the couch. Thinking hard, too.
So it's not that the reward isn't there. It's that the system that converts reward into effort has stopped paying out.
The person can still want things. The machinery that turns wanting into doing is what's degraded.
And the second system?
Inflammation. This is the one I find fascinating, and it's the one that connects back to the hibernation thread. Proinflammatory cytokines, interleukin-one, interleukin-six, TNF-alpha, produce a specific syndrome. The list is lethargy, depression, anorexia, energy conservation, fever, anhedonia, cognitive impairment.
Read that list again and tell me it isn't a description of being ill.
It's a description of being ill. That's the point. It's called sickness behavior, and it's not a malfunction. It's what the body does when it's fighting something. You lose your appetite, you lose interest, you want to lie down in the dark. It's an energy-conserving response.
So the body has a program for "shut down and recover," and it runs it whether or not there's an infection.
And Maes and colleagues, in BMC Medicine in 2012, described depression and sickness behavior as Janus-faced responses to shared inflammatory pathways. Same pathways, two faces. And they framed it as having a good, acute side and a bad, chronic side.
The acute side being: you get sick, you shut down, you recover, you get up.
And the chronic side being: the shutdown never lifts. The program that was supposed to last a week is still running in month four.
That reframes the whole thing. Low energy in depression isn't a bug. It's a conserved adaptive response that's gone chronic.
That's the thesis. And it's testable, which is what makes it more than a nice story. If you activate the innate immune system in healthy people, you induce fatigue, worsened mood, and pain sensitivity. That's from Lasselin and colleagues in Frontiers in Neuroendocrinology in 2018. You give someone an inflammatory challenge, and they start reporting exactly the symptoms we're describing.
So the causation runs in the direction the model predicts.
In healthy volunteers, yes. Which is about as close to a controlled experiment as you can ethically get.
And the two systems, reward and inflammation, they're not separate, are they?
They appear to converge. There's a scoping review in Translational Psychiatry from 2020 arguing that fatigue and anhedonia may share a common mechanism, running through dopamine and inflammatory cytokines. So the motivational account and the immune account may be two descriptions of one underlying process.
Which would explain why fatigue and anhedonia so often arrive together.
They cluster. And if they share a mechanism, that's not a coincidence.
So now pull the two halves together. SAD and non-seasonal depression.
They may share a final common pathway, which is energy conservation, but they arrive there differently. SAD gets there through seasonal light and circadian signals. Non-seasonal depression gets there through inflammation and reward-circuitry failure. Same destination. Different roads.
The hibernation program and the sickness program, both ending in the same place.
And both of them, in their original context, being sensible. Conserve energy when food is scarce. Conserve energy when you're infected. The problem is a system designed to run for a season, or a week, running indefinitely.
Does the treatment split follow the mechanism split?
Partly, and this is where it gets interesting. Light therapy works for SAD. The standard dose is ten thousand lux for thirty to forty-five minutes daily, usually in the morning, from fall through spring. The box is about twenty times brighter than ordinary indoor light.
Twenty times.
That's why a bright kitchen doesn't do it. You need the intensity, and you need it at the right time of day.
And for non-seasonal depression?
The evidence for bright light in non-seasonal major depression is newer and more mixed. It's not the same slam dunk. Which is what you'd expect if the mechanisms are different. If the problem isn't light, adding light is a weaker bet.
There's a trial from last year worth mentioning.
Volf and colleagues, in Chronobiology International. They used dynamic LED lighting that mimics a daylight cycle, rather than static lighting, in female inpatients with major depression. Daytime melanopic illuminance peaked at five hundred and seventy-six lux, versus sixty-six in the static condition, and dimmed to zero point three lux in the evening.
And the result?
A significantly greater antidepressant effect at week three in the female inpatients. Seventy-one percent response. But not in the whole group. And I'd flag that honestly, because that's a partial result, not a clean one.
Seventy-one percent in a subgroup is real, but it's a subgroup.
It's a signal, not a verdict. And there's a bigger point sitting under it. The reason this matters is that it suggests the shape of the light, the dynamic rise and fall, might matter more than the raw dose. Which loops back to the mechanistic fork. If timing and rhythm matter, the clock is involved. If raw intensity matters, it's more direct.
And then there's the finding that cuts against the pure-photon story entirely.
Cognitive behavioral therapy for SAD outperformed light therapy at long-term follow-up. Kelly Rohan, at the University of Vermont, who does that work, put it this way. She doesn't argue there's no strong physiological component, and she agrees it's tied to the light-dark cycle. But she argues the person has some control over how they respond to and cope with that.
So the meaning we assign to winter matters, not just the photons.
Which is a awkward finding if you're a strict mechanism person. The light is doing something. But so is what the person believes about the light, and what they do when it goes.
And the timeline on treatment is worth saying out loud, because people quit early. Antidepressants take four to eight weeks to work.
Four to eight weeks. And SAD symptoms last an average of forty percent of the year. So you're talking about a condition that eats two fifths of the calendar, and a treatment that takes two months to show up.
Which is a brutal mismatch.
It's a real one.
Before we go further, I want to flag something about Daniel's framing. He said SAD is easier to understand because we associate light with vitality. And that intuition is correct, but it's correct for a reason most people don't know. It's not a metaphor. There's a dedicated photoreceptor system for it.
The intuition is right and the mechanism is stranger than the intuition.
People think "light makes me feel good" the way they think "sunshine is nice." It's actually a separate neural channel that doesn't involve seeing at all.
Hilbert: It's the flicker.
Sorry?
Hilbert: The flicker matters more than the brightness. I maintained the lighting in a greenhouse for two years, late seventies. Twelve hundred tubes, and the ones that were failing didn't go dark. They flickered at a rate you couldn't consciously see. The plants didn't care. The people did.
Because flicker at certain frequencies is known to affect...
Hilbert: The women on the potting line would move. Not complain, just move. They'd drift down the bench to a section where the tubes were newer. Same lux, same color, but they'd end up in the same six feet of bench by the end of a shift. Took me a while to work out why.
And the plants didn't care.
Hilbert: Plants don't have melanopsin.
You had a natural experiment. Identical measured light, different subjective effect, and the people self-corrected without being able to explain it.
Hilbert: I had four of those ballasts in my garage for years. Never got around to them.
Four.
Hilbert: Four. Anyway, the level on your second mic is running about three decibels hot. I'll fix it in the edit.
The point he's making is real, though. It connects to the individual-sensitivity question. If the response depends on the quality of the light and not just the quantity, then two people under identical measured conditions can have completely different outcomes. Which is exactly what the melanopsin-variant hypothesis would predict.
It also explains why the latitude data is messy. Florida at four percent. Maryland above New York. If it were purely about photon count, those numbers would line up in a straight line.
They don't. Because the photon count isn't the whole variable.
Where does that leave Daniel's question? Why does depression have such a strong impact on energy?
The honest answer is that there isn't one mechanism. There are at least three, and they overlap. Light and circadian timing. Reward and effort circuitry. Inflammation and sickness behavior. And the reason the connection isn't obvious is that the folk model only has one of them, which is "you feel sad, so you don't do things."
The folk model has the causality backwards.
It does. The energy drain isn't a consequence of the sadness. In a lot of cases it's upstream of it. The sickness behavior program produces the lethargy and the anhedonia together, as one package. The low mood is part of the shutdown, not the cause of it.
Which is why telling a depressed person to push through it doesn't work. You're asking them to override a program that their immune system is actively running.
It's why the "just get some sunlight" advice fails for non-seasonal depression. You're applying the seasonal fix to a non-seasonal problem.
What's still unresolved?
Most of it. There's no single definitive cause of SAD established. The serotonin, melatonin, vitamin D, retinal-sensitivity, and phase-shift hypotheses are all live, and they may all be partly true at once. The vitamin D evidence specifically is mixed. Some studies find it as effective as light therapy, others find no effect at all.
And summer-pattern SAD?
Under-researched. Less is known about it, and more research is needed. Which is odd, because it's the natural control group. If the mechanism is "not enough light," summer-pattern SAD shouldn't exist.
The mechanistic fork, clock versus direct?
Unresolved. The Hattar review explicitly presents the alternative model that dispenses with the causality assumption. Nobody has closed that question.
Here's the thing I keep circling, and it's what Hilbert's flicker story actually points at. If individual sensitivity varies that much, and the quality of light matters as much as the quantity, then the line between "this person has a disorder" and "this person is responding normally to a bad environment" gets very thin.
The Sami point again.
If a culture can embrace the seasonal slowdown and not call it illness, then some fraction of what we're measuring as pathology might be a mismatch between an old program and a modern expectation.
Year-round high energy as the healthy norm. Which is a recent assumption, historically.
Very recent. And possibly wrong.
The forward-looking version of that is interesting, though. If the shared pathway really is energy conservation, and inflammation is one of the roads into it, then the next generation of treatments might not be about light at all. They might target inflammation directly, or target the reward circuitry, or both.
Treating the shutdown program instead of the mood on top of it.
Treating the mechanism rather than the symptom. And the reason that's plausible is that we now have a candidate mechanism. Twenty years ago, "depression drains energy" was just an observation. Now it's a set of pathways you can point at.
Which is progress, even if none of the questions are closed.
Especially because none of the questions are closed.
That's the episode, then. No single cause, several live theories, and a mechanism that looks less like a malfunction and more like an old program that won't switch off.
Thanks to Hilbert Flumingtop for producing.
This has been My Weird Prompts. If you enjoyed it, leave us a review on whatever platform you're listening on. It helps.
We'll be back soon.
See you tomorrow.
