Doxepin for sleep

Doxepin is a tricyclic antidepressant from the 1960s that turned out, at very low doses, to be one of the best sleep medications available. The 3mg and 6mg versions (sold as Silenor) are FDA-approved for insomnia. The 25mg, 50mg, 100mg versions are the old-school antidepressant doses, which nobody really uses for depression anymore because we have better, cleaner options. The sleep version of this drug is one of those quiet, half-hidden good ideas that almost no general practitioner mentions, and that some of the better sleep specialists swear by.

The way it works for sleep is different from the way it works as an antidepressant, which is the part most general prescribers don’t know about. At low doses, it’s a selective histamine antagonist. Histamine keeps you awake. Antihistamines make you sleepy. Doxepin at 3mg or 6mg is essentially a precision-targeted antihistamine that gets you sleep without much else attached. At higher doses, it starts hitting serotonin and norepinephrine receptors and a bunch of other targets, which is what makes high-dose tricyclics (the older antidepressant class, things like amitriptyline and nortriptyline) the complicated drugs they are. The sleep-dose version dodges all of that by staying under the threshold where the other receptors get involved.

Why low-dose doxepin is unusually good

Most sleep medications have problems. Ambien works but a meaningful number of patients get weird side effects, sleep eating, sleepwalking, next-day grogginess that lingers into the afternoon. Trazodone works fine for some people but a lot of guys feel hungover the next morning, like they got hit by a truck before breakfast. Benzodiazepines (Xanax, Klonopin, Restoril, the addictive sedative class) work but they’re benzos and the right answer for most patients is to not use them as a sleep solution because the long-term math doesn’t pencil out. Melatonin works if your problem is circadian (your body clock is off, you’re a shift worker, you flew across six time zones), which isn’t most adult insomnia.

Low-dose doxepin keeps you asleep through the night, doesn’t seem to cause much next-day grogginess in most patients, isn’t addictive, doesn’t have the sleepwalking thing. The catch is it’s not great at getting you to sleep. It’s good at keeping you asleep. So if your problem is sleep onset (can’t fall asleep at all, lie there for two hours staring at the ceiling), doxepin alone isn’t ideal. If your problem is sleep maintenance (waking up at 3am and not getting back to sleep until 5am if at all), this is exactly the drug.

What gets prescribed and how

3mg or 6mg, taken about 30 minutes before bed. You should not eat in the three hours before. Food slows absorption and reduces the effect, which catches a lot of patients off guard because nobody warned them. Most guys end up on 6mg. Some respond fine to 3mg. Going higher than 6mg gets you into territory where you start hitting other receptors, which can mean dry mouth, constipation, some next-day sedation. Not worth it for sleep alone, and at that point you’re better off picking a different drug.

It takes about an hour to peak. Half-life is around 15 hours, which sounds long but at the tiny doses being used, it doesn’t cause daytime sedation for most patients. The math of half-life versus dose is one of the things that confuses prescribers who haven’t actually used this drug at the sleep dose. A 15-hour half-life on 100mg is a problem. A 15-hour half-life on 3mg is just a slow taper of a tiny amount.

What it doesn’t do

It’s not a strong hypnotic. You will not feel knocked out. If you have severe sleep onset insomnia, you might still lie there for a while staring at the ceiling. Some patients add a small dose of trazodone or use it in combination with CBT-I (cognitive behavioral therapy for insomnia, the structured worksheet-and-sleep-log version of CBT that’s the gold standard for chronic insomnia, and which most patients have never heard of either).

It doesn’t treat the underlying reason you can’t sleep. If your sleep problem is anxiety, the anxiety still needs treatment. If it’s sleep apnea, doxepin won’t fix that and may make it slightly worse, and the right move is the sleep study, not the pill. If it’s chronic pain, the pain needs work. Doxepin treats the symptom of broken sleep maintenance. It does not treat why your sleep maintenance is broken in the first place, which is honest but worth saying out loud.

What’s nice to hear about this one

For the right kind of patient, doxepin is one of the most satisfying drugs to land on. Picture a guy who comes in having tried trazodone (gave him a hangover so bad he couldn’t drive in the morning), melatonin (didn’t do much), Ambien (had a sleepwalking episode where he sent his daughter a text he didn’t remember and was done with that one), is finally ready to try the boring old generic. Start at 3mg. Works okay, some nights solid, some nights not. Go up to 6mg. From there on it’s 80 to 90 percent solid sleep. Two years in, still on it, asking about coming off, no real medical reason to push him off. The drug is cheap. The quality of life is dramatically better. He’s not going to put it on social media. He just sleeps. That’s the result.

It’s good at keeping you asleep. So if your problem is waking up at 3am and not getting back to sleep until 5am, this is exactly the drug.

Who shouldn’t take it

Patients with narrow-angle glaucoma. Patients with significant urinary retention from BPH, the prostate-enlargement thing that older guys deal with where the bladder doesn’t empty cleanly… doxepin can make it worse. Patients on MAOIs (monoamine oxidase inhibitors, an older antidepressant class that’s mostly off the menu now, but a few patients still take them) because the combination is dangerous. Patients who are very sensitive to anticholinergic effects, meaning the dry-mouth, constipation, slight-cognitive-fog cluster… if Benadryl knocks you sideways, doxepin is going to do similar things.

The other group worth naming: anyone expecting it to work like Ambien. It doesn’t. It’s gentler, slower, more subtle. Some patients try it for two nights, don’t feel anything dramatic, and quit. Give it a couple of weeks. Most of the benefit shows up between night seven and night fourteen, and patients who bail at night three are bailing before the medication has done its job.

Why it’s underused

Two reasons. The first is the brand confusion. Doxepin is generic and dirt cheap (the 25mg and 50mg capsules cost almost nothing). The 3mg and 6mg sleep-specific dose is Silenor, which is branded and more expensive. A lot of prescribers don’t realize that a compounding pharmacy can make 3mg or 6mg from the generic for a fraction of the Silenor cost, which is the move for patients without insurance coverage on the brand. The second is the tricyclic-antidepressant association. A lot of prescribers, hearing “tricyclic,” mentally file the whole class as old, dirty, and complicated. At sleep doses none of that applies, but the perception is hard to shake.

On the autonomy piece

The sleep medication conversation is full of patients who want a stronger drug than they need, because the bad nights felt awful and they want a sledgehammer for next time. The honest answer is the sledgehammer drugs come with sledgehammer trade-offs, and a gentler drug that works on the specific problem you actually have is almost always the better trade. If you want Ambien, you can have Ambien… I’m not refusing. The take is that doxepin is probably the better fit if your problem is the middle-of-the-night wake-up. The choice is yours.

What the field gets wrong on sleep meds generally

The bigger sleep-medication problem in this country is that most patients with chronic insomnia don’t get offered CBT-I, which is the actual evidence-based first-line treatment, because there aren’t enough trained therapists and it takes six to eight sessions and nobody has time. So instead the patient gets handed a sleep pill on the first visit and ends up on it for years. Doxepin isn’t the wrong answer there, it’s just not the only answer, and the honest version of the conversation includes naming what the gold-standard treatment actually is and where to get it. There are app-based versions of CBT-I now (Somryst, SHUTi, the free ones) that work pretty well for motivated patients without an in-person therapist, which is worth mentioning before reaching for any pill.

Bottom line

Low-dose doxepin is one of the cleanest sleep medications available for the right kind of insomnia, which is waking up in the middle of the night and not getting back to sleep. It’s cheap, it’s generic, it’s been around for sixty years. It doesn’t have the abuse profile or the weird side effects of newer agents. If you’ve been on trazodone or Ambien or Lunesta and you’re not happy, this is worth asking about. Most general practitioners have never heard of using it this way, which is a function of the marketing landscape, not a function of the drug.

Sources

1. Krystal AD, Lankford A, Durrence HH, et al. Efficacy and safety of doxepin 3 and 6 mg in a 35-day sleep laboratory trial in adults with chronic primary insomnia. Sleep. 2011;34(10):1433-1442. PMID 21966075.
2. Roth T, Rogowski R, Hull S, et al. Efficacy and safety of doxepin 1 mg, 3 mg, and 6 mg in adults with primary insomnia. Sleep. 2007;30(11):1555-1561. PMID 18041488.
3. Yeung WF, Chung KF, Yung KP, Ng TH. Doxepin for insomnia: a systematic review of randomized placebo-controlled trials. Sleep Med Rev. 2015;19:75-83. PMID 25047681.