His Melatonin Dose Was Never the Problem. The Clock on the Bottle Was.
He has been taking more and more melatonin, later and later, chasing a bigger dose right before bed. The actual fix required moving the same modest dose to a completely different clock time — several hours earlier, not stronger.
T.J., a 21-year-old man, is a college senior majoring in mechanical engineering who has never, for as long as he can remember, been able to fall asleep before 3 or 4 a.m. The pattern goes back to early adolescence, and through most of high school and his first three years of college he managed it reasonably well simply by arranging his life around it — taking afternoon classes when he could, staying up late finishing problem sets during the hours he’s always felt sharpest, and sleeping until nearly noon on a schedule that was, on its own terms, internally consistent and left him rested.
This semester broke that arrangement. A required 9 a.m. lab, the only section offered, means he is now chronically sleep-deprived on a schedule that was never going to work for him — falling asleep during lecture afterward, relying on increasing amounts of caffeine to get through the day, and, for the first time, genuinely worried about his GPA. He is otherwise healthy: no psychiatric history, no chronic medical conditions, and a family history notable mainly for his father, who has described a similar lifelong pattern of being “a night person” that he never sought treatment for and simply built a career around instead.
Frustrated and increasingly desperate as midterms approach, T.J. has spent the last two months buying over-the-counter melatonin and steadily increasing the dose — starting around 3 mg, now up to 10 mg most nights — taken right before he actually wants to fall asleep, typically between 2 and 3 a.m. He has noticed no meaningful change in how long it takes him to drift off despite the escalating dose, and came in today, tired and a little embarrassed, to ask whether an even higher dose might finally be the thing that works.
In clinic, after months of a rising dose that wasn’t working
His reasoning makes intuitive sense but gets the pharmacology backward. Melatonin’s circadian effect is governed by a phase-response curve, not a dose-response curve — timing relative to a person’s own internal clock determines whether it shifts sleep earlier or has essentially no phase effect at all. Lewy and colleagues first characterized this curve directly in the early 1990s, and taking melatonin right at his desired bedtime, when his own endogenous melatonin hasn’t even begun rising yet, lands squarely in the flat part of that curve.
To shift someone with delayed sleep-wake phase disorder earlier, low-dose melatonin (typically 0.5–1 mg, not his escalating 10 mg) needs to be taken several hours BEFORE his natural dim-light melatonin onset — in his case, likely somewhere in the early evening, not at 2 a.m. He has been increasing the dose while leaving the one variable that actually matters, timing, completely unchanged.
I’d add the practical piece, because the timing fix on its own is a hard sell for a patient who’s been chasing higher doses — he needs to hear why less, earlier, isn’t a downgrade. Formal delayed sleep-wake phase disorder treatment protocols pair low-dose melatonin timed several hours pre-bedtime with morning bright-light exposure and a fixed, earlier wake time, moving the whole circadian system gradually rather than trying to force sleep onset directly at the desired hour.
You’re right that dose isn’t the lever here — but timing alone, without the light and wake-time anchoring, tends to produce a smaller and less durable shift than the combined approach. I’d want all three pieces moving together, not just the melatonin timing corrected in isolation.
Agreed: retimed low-dose melatonin combined with morning bright-light exposure and a fixed, gradually-advanced wake time, with his prior high-dose bedtime regimen discontinued entirely.
No real disagreement remained once the mechanism was explained — the sleep physician’s addition of light exposure and wake-time anchoring built directly on the pharmacologist’s timing correction rather than contesting it, and both voices converged on the same combined plan.