Circadian Biology of Hormones: The Master Rhythm

Almost nothing in human physiology is a constant. Cortisol, testosterone, growth hormone, melatonin, core temperature, blood pressure and insulin sensitivity all follow daily curves, and a value measured at 7am can differ enormously from the same value at 7pm. That is not noise around a true figure. The timing is the physiology. Which means it is possible to have a hormonal problem that is not a hormone problem at all — it is a scheduling one.

The clock, and the clocks

The central pacemaker is the suprachiasmatic nucleus, a paired structure of a few thousand neurons in the hypothalamus directly above the optic chiasm. It receives a dedicated retinal projection, and it is the only structure that keeps time for everything else (Moore, Cell Tissue Res 2002).

Timekeeping at the cellular level is a transcription-translation feedback loop: clock genes drive proteins that accumulate through the day, re-enter the nucleus and switch off their own transcription, after which the cycle restarts. The loop takes approximately, but not exactly, 24 hours — and that "approximately" is why the system needs a daily correction.

The part that surprises people is that the loop is not confined to the brain. Liver, muscle, fat, pancreas, gut and immune cells all run their own copies. The SCN's job is not to generate that rhythm but to keep the peripheral clocks aligned with each other and with the outside world.

Which reframes the topic. Circadian disruption is not fundamentally about being tired. It is internal misalignment — organ systems running on different schedules from each other.

Light is the correction signal

Because the intrinsic period is not exactly 24 hours, the clock drifts without a daily reset. Light is the dominant reset signal, via a small population of retinal ganglion cells containing melanopsin that project straight to the SCN. They are most sensitive to short-wavelength light and respond to intensity rather than images — they work perfectly well in people blind from photoreceptor loss.

The consequence of timing is the part worth internalising. Light in the morning advances the clock, pulling everything earlier. Light in the late evening and early night delays it, pushing everything later. There is a window in the middle of the day where light has little phase effect at all. This response has been mapped experimentally, and it is why identical light exposure can help or harm depending only on when it lands (Khalsa et al., J Physiol 2003).

Food sets the peripheral clocks — and that is where conflict arises

Light entrains the SCN and has surprisingly little direct effect on the liver. Peripheral clocks take their strongest cue from feeding — the arrival of nutrients and the insulin response to them.

Shifting when people eat, while holding sleep and light constant, shifts the phase of peripheral rhythms independently of the central clock (Wehrens et al., Curr Biol 2017). This is the mechanism behind the single most common form of modern misalignment: light says one thing, food says another. A person who sees no morning light, works indoors, and eats their largest meal at 9pm has a brain clock anchored late and a liver clock anchored later still.

It also explains why the same meal has different consequences at different hours. Insulin sensitivity is highest in the morning, so an identical carbohydrate load produces a larger glucose excursion at night — not through willpower or portion size, but because the tissue receiving it is on a different part of its own cycle. Time-restricted eating is largely an attempt to exploit this, with more modest results than the marketing suggests.

The hormone timetable

Each of these is downstream of the clock, and each is measured against it.

• Cortisol bottoms out around midnight and rises sharply before waking, peaking roughly 30-45 minutes after you get up. A flat morning rise is a meaningful finding; a random cortisol with no time attached is close to uninterpretable.
• Testosterone in men peaks in the early morning and declines across the day — the entire reason morning draws are specified rather than preferred.
• Growth hormone pulses with slow-wave sleep, concentrated in the first half of the night, so a late bedtime with a fixed alarm costs GH output specifically.
• Melatonin rises in darkness; its onset is the cleanest marker of internal clock phase, and evening light suppresses it directly.
• Insulin sensitivity is highest in the morning, lowest at night.

Diurnal hormone rhythms covers the individual curves, and sleep architecture and hormones the sleep-stage dependency.

What misalignment actually costs

The strongest human evidence comes from forced desynchrony protocols, where participants live on a schedule the clock cannot entrain to. Under controlled misalignment — same food, same sleep duration, wrong times — glucose tolerance deteriorates, insulin sensitivity falls, blood pressure rises and leptin falls, in some participants to a prediabetic degree within days (Scheer et al., PNAS 2009). Nothing about the diet changed. Only the timing did.

Long-term shift work points the same direction across metabolic, cardiovascular and oncological endpoints, with night work and breast cancer risk the most examined of them (Carcinogenesis 2021). That observational evidence is not uniform and should not be oversold, but it is consistent enough to be treated as a real exposure.

The version most people actually have

Very few readers work night shifts. Almost all of them have social jet lag: a weekday schedule set by an alarm and a weekend schedule set by preference, producing a two-hour phase shift twice a week, every week. The mismatch between biological and social timing tracks with body weight and metabolic markers in large population samples (Roenneberg et al., Curr Biol 2012). The Sunday-night difficulty falling asleep and the Monday fog are not a mood — they are a two-hour westward flight taken every weekend, and the fix is a wake time that does not move much.

What to change, in order of leverage

1. A fixed wake time, weekends included. The single highest-leverage change available. Bedtime follows wake time, not the other way round.
2. Outdoor light within the first hour of waking — ten minutes or more, no sunglasses, overcast days too, since a dull sky still vastly outperforms indoor lighting (sunlight and hormones).
3. Dim the evening. Overhead lighting is more of a problem than screens; a bright bathroom at 11pm delivers more melanopic stimulus than a phone at arm's length.
4. Move the eating window earlier, finishing at least three hours before bed, which aligns the peripheral clocks and removes the late glucose excursion.
5. Caffeine curfew and less alcohol. Caffeine's half-life leaves an afternoon coffee meaningfully present at bedtime; alcohol shortens sleep latency then fragments the second half of the night (why you wake at 3am).

What to expect

Sleep and wake timing usually respond within one to two weeks of consistent morning light and a fixed wake time — the achievable phase shift per day is limited, so this is a gradual correction. Daytime alertness follows. Metabolic markers move over months. A hormone panel, if disruption was a genuine contributor, is worth repeating after six to eight weeks of a stable schedule rather than immediately.

Two honest caveats. Chronotype is substantially genetic: a real night owl can shift their schedule but cannot become a lark, and forcing it produces sleep restriction rather than alignment. And for genuine rotating shift work the realistic goal is damage limitation — consistent anchor sleep, controlled light at both ends of the shift, and an eating window as close to daytime as the job allows.

Bottom line

The circadian system is a network of cellular clocks kept in step by a central pacemaker, entrained mainly by light at the brain and by food timing at the organs. When those two signals disagree the result is internal misalignment — measurably worse glucose handling, blood pressure and appetite regulation, independent of what is eaten. Hormone results are only interpretable against this timetable, which is why draw time and sleep schedule matter as much as the marker. The interventions are behavioural and free, they take weeks rather than days, and they change the value of everything layered on top. If a panel still looks wrong after that, a proper evaluation is the next step rather than the first.

Educational content, not medical advice. Laboratory interpretation and any treatment decision are made by a licensed physician after individual evaluation. Individual results vary.