Sleep
It Is Not Insomnia If You Fall Asleep Fine at Two in the Morning
The Short Answer
Two different problems look identical at two in the morning. Insomnia is being unable to sleep.
Delayed sleep phase is sleeping perfectly well, several hours later than your alarm allows. They have different treatments, and the second one is badly served by melatonin taken the way the bottle suggests.
There is a specific conversation that goes wrong in surgeries every week. Someone says they cannot sleep.
They are asked about sleep hygiene, given a leaflet about screens, and possibly prescribed something. What nobody asks is the question that separates the two conditions: if you had no alarm at all, and nowhere to be, when would you fall asleep and when would you wake up?
Because if the answer is “about two, and about ten, and I feel fine” — that is not insomnia. That is a body clock running several hours behind the society it lives in, and it responds to completely different treatment.
What is actually measured
Dim-light melatonin onset is the point in the evening at which your brain starts producing the hormone that signals sleep. It is the cleanest available marker of where your internal clock sits. In adults with ADHD it arrives, on average, about ninety minutes later than in controls, and up to 78% show delayed sleep–wake timing.
A systematic review pulling together 62 studies covering 4,462 patients found the pattern holding consistently across the literature: more eveningness, later phase, and a strong association between ADHD and circadian disruption that does not disappear when you control for the obvious confounds.
What That Means In Practice
If your brain does not begin signalling sleep until half past midnight and your alarm is set for half past six, you are running a permanent transatlantic jet lag — while also managing a disorder of executive function. Every morning is a bad-sleep morning, and the reason is not discipline.
SURVIVING ADHD · NINETY MINUTES BEHIND 1
The trial that is more useful than any sleep-hygiene list
In 2022 a Dutch group ran the study this field needed. Forty-nine adults aged 18 to 55, all with ADHD and a diagnosed delayed sleep phase, randomised across three weeks to one of three arms: melatonin 0.5 mg, placebo, or melatonin
0.5 mg plus bright light therapy at 10,000 lux for thirty minutes.
The clock moved. Melatonin alone advanced dim-light melatonin onset by roughly an hour and a half; melatonin plus light by roughly two hours. Placebo barely moved at all.
And then the finding that almost never gets quoted:
The clock advanced by up to two hours. Actual sleep times did not follow.
Participants’ real sleep onset did not significantly shift in any group. The authors’ conclusion is the sentence to build a plan around: biological phase-shifting alone is not enough, and additional behavioural work — the sort used in CBT for insomnia — is necessary rather than optional.
That is a genuinely inconvenient result and it is why you will not find it in a listicle. It is also the most practically useful thing in the whole literature, because it tells you in advance that the pill is half the intervention and that the other half is the part you have to build.
Three things about melatonin that contradict the bottle
All three of those make this a conversation with a clinician rather than a decision made in a pharmacy aisle — and they are the reason so many people conclude melatonin “does not work for them” when what actually happened is that they took twenty times the trial dose at the wrong hour.
The number you can measure tonight
You do not need a laboratory to get a usable reading. Take the midpoint between falling asleep and waking on a work night, and the same midpoint on a free night with no alarm. The distance between the two is your social jetlag — the gap between when your body wants to sleep and when your week makes you.
Under an hour is well aligned. One to two hours is a meaningful weekly displacement. Two hours or more means that functionally, every Monday morning, you are landing from a transatlantic flight.
It is a clock signal, not a sedative. Taken at bedtime as a sleeping aid it does approximately nothing. Its job is to tell your system that evening has arrived, and to do that it has to be taken before your own onset, not at the point where you want to be unconscious.
The dose in the trial was 0.5 mg. Not the 5 mg or 10 mg sold on the shelf. Higher doses are not more phase- shifting; past a point they simply spill over into the following day.
Timing beats dose, and the timing is relative to your onset. Which is not something you can work out from a forum post, because it depends on where your own clock currently sits.
SURVIVING ADHD · NINETY MINUTES BEHIND 2
What to do with this
Why This One Is Worth Doing First
It is the cheapest large intervention available in adult ADHD, it is rarely asked about, and it sits underneath everything else. Executive function measured on five hours of misaligned sleep is not a measurement of your executive function.
Sources van Andel E, Bijlenga D, Vogel SWN, Beekman ATF, Kooij JJS. Effects of chronotherapy on circadian rhythm and ADHD symptoms in adults with ADHD and delayed sleep phase syndrome: a randomized clinical trial. Journal of Biological Rhythms 2022.
Coogan AN, McGowan NM. A systematic review of circadian function, chronotype and chronotherapy in attention deficit hyperactivity disorder. ADHD Attention Deficit and Hyperactivity Disorders 2017;9:129–147.
van Veen MM, Kooij JJS, Boonstra AM, Gordijn MCM, Van Someren EJW. Delayed circadian rhythm in adults with ADHD and chronic sleep- onset insomnia. Biological Psychiatry 2010;67:1091–1096.
Roenneberg T, Wirz-Justice A, Merrow M. Life between clocks: daily temporal patterns of human chronotypes. Journal of Biological Rhythms
2003;18:80–90.
This article is educational material. It is not medical advice, it cannot diagnose anything, and it does not replace evaluation by a clinician who has met you. Evidence current to September 2026.
Change the words you use. Ask about “delayed sleep–wake phase”, not insomnia. It points at a different diagnosis and a different treatment pathway, and it stops the conversation ending at sleep hygiene.
Bring two weeks of times. Actual sleep and wake times, work nights and free nights. Ten minutes of data beats an hour of trying to remember.
Ask about timing before dose. Both are a clinician’s decision, but timing is the one that determines whether anything happens at all.
Plan the behavioural half in advance. The trial is unambiguous: the clock moving does not move your sleep by itself. Decide now what holds the earlier bedtime once the biology permits it.
Rule out the other thing. Heavy snoring, witnessed pauses in breathing, or daytime sleepiness that persists after adequate sleep points at sleep apnoea, which is a different referral and a treatable one. Do not assume the clock explains everything.
SURVIVING ADHD · NINETY MINUTES BEHIND 3
Where these came from
Each of these articles is a short version of material that appears at greater length in Surviving ADHD: A Comprehensive
Guide — a workbook for adults, including the very large number of adults who were never identified as children, who were treated for anxiety or depression for years before anyone asked about attention, and who arrived at a diagnosis somewhere between relief and grief.
The workbook differs from most in one specific way: it prints the strength of the evidence beside every claim and every strategy. A four-bar rating shows at a glance whether you are holding a finding replicated across millions of people or a clinician’s useful hunch. Both belong in a workbook. They do not belong in the same sentence without a label.
Worksheets, each finishable in one sitting
START ANYWHERE
Original diagrams of the mechanism
NOT DECORATION
Sections, from diagnosis to support
NOTHING DEPENDS ON ANYTHING
EARLIER