What tryptophan depletion tells us

Serotonin (5-hydroxytryptamine) is made from the essential amino acid tryptophan. Researchers cannot measure serotonin directly in a living human brain, so they often lower it temporarily with an experimental method: acute tryptophan depletion. In the laboratory, participants drink a mixture containing large amounts of other amino acids but no tryptophan; within hours blood tryptophan falls sharply and the brain has less raw material for serotonin. A control group drinks a balanced mixture containing tryptophan.

These experiments show that serotonin is involved in responses to unfavourable signals: with lowered serotonin, people respond more readily to unfairness with punishment. They do not show that more serotonin is always better, nor that everyday low mood is caused by too little serotonin.

There is one direct human measurement linking light and serotonin.

Another kind of light-sensitive cell in the eye

Besides the rods and cones used for vision, the retina contains a small population of intrinsically photosensitive retinal ganglion cells (ipRGCs). They contain melanopsin, are most sensitive to blue-green light, respond slowly and persistently, and project mainly to the suprachiasmatic nucleus of the hypothalamus, the master circadian clock.

The intrinsic period of the human circadian clock averages about 24.2 hours, slightly longer than a day (Czeisler et al. 1999). Without correction by light, the clock would drift a little later each day. Light acts differently at different times:

A phase model

# A simplified phase model of the human circadian clock under daily light (cross-disciplinary
# simplification; real models such as Kronauer's are limit-cycle oscillators).
# State: clock time of the core-body-temperature minimum (CBTmin). Intrinsic period 24.2 h
# (Czeisler et al. 1999), so without light the clock drifts 0.2 h later each day.
# Light shortly before CBTmin delays the clock, light shortly after advances it (Khalsa et al. 2003);
# the response saturates with illuminance (half-maximal near 100 lux, Zeitzer et al. 2000).
import math

TAU = 24.2
def prc(hours_after_cbtmin):
    x = (hours_after_cbtmin + 12) % 24 - 12          # wrap to -12..+12 h
    return 0.6 * math.sin(math.pi * x / 12)        # h of advance (+) per hour of saturating light

def drive(lux):
    return lux / (lux + 100.0)

def settle(schedule, days=90):
    """schedule: list of (start hour, end hour, lux) of light at the eye, repeated daily."""
    cbt = 4.0                                        # start: CBTmin at 04:00
    for _ in range(days):
        shift = 0.0
        for start, end, lux in schedule:
            h = start
            while h < end:
                shift += 0.25 * drive(lux) * prc(h + 0.125 - cbt)
                h += 0.25
        cbt = (cbt + (TAU - 24) - shift) % 24        # advance moves CBTmin earlier
    return cbt

def clock(h):
    h %= 24
    return f"{int(h):02d}:{int(round(60 * (h % 1))) % 60:02d}"

indoor = [(7, 23, 150)]
plans = {
    "indoor light only, 07:00-23:00 at 150 lux": indoor,
    "plus 30 min outdoors after waking (10,000 lux)": [(7, 7.5, 10000), (7.5, 23, 150)],
    "plus bright room 21:00-24:00 (500 lux)": [(7, 21, 150), (21, 24, 500)],
    "both: morning outdoors, dim evening (20 lux)": [(7, 7.5, 10000), (7.5, 21, 150), (21, 23, 20)],
}
print("light plan                                     CBTmin   estimated DLMO (CBTmin - 7 h)")
for name, sched in plans.items():
    c = settle(sched)
    print(f"{name:46s} {clock(c)}    {clock(c - 7)}")
# light plan                                     CBTmin   estimated DLMO (CBTmin - 7 h)
# indoor light only, 07:00-23:00 at 150 lux      03:19    20:19
# plus 30 min outdoors after waking (10,000 lux) 03:09    20:09
# plus bright room 21:00-24:00 (500 lux)         04:35    21:35
# both: morning outdoors, dim evening (20 lux)   02:23    19:23

In this model, adding 30 minutes of outdoor morning light alone advances the clock by only about 10 minutes, because a full day of indoor light already supplies much of the “daytime” signal; three hours of bright indoor light in the evening delays the clock by more than an hour; and outdoor morning light together with dark evenings advances it by about an hour. This agrees in direction with Wright’s camping study: what matters is not only more light in the morning but also less light at night. The model’s numbers should not be taken as predictions: the real phase response curve is not a sine wave, and people differ widely.

Summary

Serotonin helps regulate responses to negative signals, but “serotonin = calm” and “low serotonin = depression” are both too simple. Light is one of the few inputs a person can change at will: through ipRGCs it corrects the phase of the circadian clock, and in one human measurement brain serotonin turnover was higher on sunny days. The next chapter turns to a system that can be moved even more directly by intention: breathing and heart rate.