Insomnia

Insomnia, treated at the driver.

Almost nothing in psychiatry improves while sleep stays broken, and almost no sleeping pill fixes the reason sleep broke. Insomnia care here starts with the driver: the anxiety that will not power down, the ADHD brain that schedules its best thinking for midnight, the bipolar rhythm that treats sleep loss as both symptom and trigger, the apnea nobody has screened for. Treat that, measure what remains with the ISI, and only then talk about medication, non-habit-forming first, with supplement options that hold up to scrutiny.

The driver first

Four engines of a sleepless night

The 60-minute evaluation treats your sleep history like a diagnostic instrument, because it is one. Where the night breaks tells us what is driving it.

Caffeine arithmetic, alcohol's false promise as a sedative, and the medications already on your list that fragment sleep all get the same audit. Sometimes the most effective insomnia prescription is a subtraction.

Medication, in its place

Non-habit-forming first, honestly labeled

When medication belongs in the plan, the shelf here starts with options that do not create the problem they treat. Older medications used at low doses for sleep, trazodone and low-dose doxepin among them, have long track records and no dependence curve. Melatonin, used at the right dose and the right hour, is a timing signal more than a sedative, and it is dosed here like it matters, because at the wrong dose it is just an expensive placebo with good marketing. The supplement shelf gets the same discipline: magnesium glycinate and L-theanine are reviewed, dosed, and interaction-checked alongside everything else you take, through the standards on the functional psychiatry page.

And the honest paragraph about the other shelf: the benzodiazepine family and the related Z-drugs, zolpidem and its cousins, sedate effectively and then keep their own ledger: tolerance, rebound insomnia worse than the original, and a stop that has to be tapered rather than taken. They are generally not started here. If you are already on one, you are not in trouble and you are not judged; you are a candidate for the same slow, supported taper described on the deprescribing page, with the rest of your sleep plan rebuilt underneath first so the night has something to stand on.

Adjustment over time

The goal is sleep that holds without escalation

Sleep medications are reviewed at every follow-up against your ISI line, and the direction of travel matters more than any single night. A dose that creeps is a flag, not a solution. When sleep has been stable and the driver is treated, we talk about stepping the sleep medication down, because the best long-term sleep aid is the one you eventually do not need. When sleep destabilizes, the first question is never "what should we add" but "what changed": the mood, the schedule, the other medications, the season. Insomnia is a symptom with excellent PR. Treating it as a messenger, not a malfunction, is most of the craft.

The audit nobody did

Your current medications, read for sleep

Before anything is added for sleep, everything already on your list is read with one question: what is this doing at 2 a.m.? The offenders are often respectable medications taken at unrespectable hours. A diuretic dosed in the evening is a scheduled awakening. A stimulant whose tail runs long is a bedtime argument. Steroids, decongestants, some blood pressure medications, activating antidepressants taken at night, thyroid medication taken a little too generously: each is a small, findable theft. Caffeine keeps its own ledger, with a half-life measured in hours that surprises people, and alcohol, the most trusted sleep aid in the country, fragments the second half of the night it was meant to protect. Moving a dose is free. It is also, regularly, the whole treatment.

When the flags point that way, the prescription is a sleep study, and this practice will say so plainly and coordinate the referral with your primary care physician. Treating apnea is not psychiatric medication management, and pretending otherwise would waste your year. What happens here is the psychiatric half done properly: the driver treated, the list audited, the measures kept, and the sleep that remains medicated carefully and temporarily, with the dose reviewed at every follow-up and the exit discussed from the first prescription. Sleep medication in this practice is a bridge with a destination, not a residence.

You will keep the same simple log through any change, because the log is how we know the change worked. Nights are noisy data. Weeks are not.

Wearables earn a careful word. A watch can show restlessness and rough sleep windows, and that data is welcome here, but it cannot stage sleep the way a laboratory can, and a bad score after a good night has ruined more mornings than apnea has. Bring the data. We will weigh it like data, not like a verdict, and the ISI and your log stay the measures of record.