are amitriptyline and nortriptyline rarely used for depression, at much lower migraine doses?
“if we're using a medication like amatrippleene and norripane often the doses we're using are much lower than what would be needed to treat depression and rarely are these medications used to treat depression anymore.”
a guest, a headache neurologist said on The Peter Attia Drive, 2026-08-31, at 19:45. hear it in context →
first, the part that holds
the first half holds up where we could measure it: in a large canadian primary-care sample, tricyclics made up 2.7% of antidepressant prescriptions written for depression, and the main use of amitriptyline was pain, not mood. low doses such as 10–25 mg a day do show up in migraine trials, and those sit well below the 75–150 mg a day the us label gives for depression.
what we found when we went and looked
we read the us labels for both drugs. both are approved only for depression. for adult outpatients, amitriptyline starts at 75 mg a day and can go up to 150 mg, with 50–100 mg for maintenance. nortriptyline is usually 25 mg three or four times a day, up to 150 mg. in a quebec primary-care study (101,759 prescriptions, 2006–2015), tricyclics were 1,502 of 56,154 depression prescriptions (2.7%), which supports "rarely." the dose clause is where the claim gets stretched: in the placebo-controlled migraine trials that the european headache federation's 2023 meta-analysis pooled, two of the three trials titrated amitriptyline to 100 mg a day, and one migraine guideline it cites gives a 30–150 mg range, which overlaps the antidepressant range. low-dose prevention is common, but the evidence that amitriptyline works for migraine rests partly on doses in the antidepressant range.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| amitriptyline and nortriptyline are rarely used to treat depression today | confirmed | in quebec primary care from 2006 to 2015, tricyclics were 1,502 of 56,154 antidepressant prescriptions for depressive disorders (2.7%). they were 10,936 prescriptions in total, so roughly 14% of tricyclic prescriptions were for depression. amitriptyline was the most-prescribed antidepressant for pain. limits: the data come from one canadian province, end in 2015, and report the tricyclic class as a whole, not nortriptyline by itself. |
| the doses used for migraine prevention are much lower than antidepressant doses | overstated | the label's adult outpatient dose for depression is 75–150 mg a day for amitriptyline (maintenance 50–100 mg) and 75–100 mg for nortriptyline (25 mg three or four times a day, max 150 mg). the migraine trials in the 2023 meta-analysis used 25 mg a day in one trial, but 100 mg a day or the maximum tolerated dose in the other two. another trial in the review used 10–60 mg, and one guideline it cites gives 30–150 mg. the low end of migraine dosing is far below depression dosing, but the ranges overlap, and nothing we read measures what doses are 'often' used in practice. [1][2][3] |
| depression is the only fda-labeled use of both drugs; migraine prevention is off-label | confirmed | both us labels list only 'relief of symptoms of depression' as the indication. the quebec study counted migraine among the uses that were off-label for every antidepressant. [1][2] |
both directions
for the speaker: tricyclics were a small share (2.7%) of what primary-care doctors prescribed for depression in the quebec dataset, and amitriptyline was mostly prescribed for pain. in that dataset, migraine prescriptions went almost entirely to tricyclics. a 25 mg a day amitriptyline arm and a 10–60 mg arm in the migraine trials are clearly below the 75–150 mg the label gives for depression.
against: the trials behind amitriptyline's 'probably effective' (level b) migraine rating mostly titrated to 100 mg a day or the maximum tolerated dose, which is inside the labeled antidepressant range. a migraine guideline cited in the 2023 meta-analysis gives 30–150 mg, the same top dose as outpatient depression treatment. so 'much lower' fits common low-dose prescribing but not the whole evidence base. the meta-analysis also found moderate-certainty evidence that amitriptyline increases dropouts due to side effects compared with placebo, which matters when doses climb.
what this cannot say
the usage data come from one canadian province, stop in 2015, and roll tricyclics up as a class. we found no current us dataset that separates amitriptyline or nortriptyline prescriptions by indication. we read no source that measures the doses actually prescribed for migraine in practice, so 'often' can't be checked. we found no nortriptyline migraine trial doses in what we read. the label summaries came through openfda's structured label api, not the printed label.
sources, each one fetched and read
- us food and drug administration, openfda drug label: amitriptyline hydrochloride tablets, set id 0067b698-ee1d-4524-b1cc-af1028d13736 What we read there: indicated 'for the relief of symptoms of depression.' adult outpatients: 75 mg a day in divided doses initially, may be increased to 150 mg a day; usual maintenance 50–100 mg a day; lower doses for adolescents and older adults. Fetched 2026-10-18 (HTTP 200). https://api.fda.gov/drug/label.json?search=openfda.generic_name:%22amitriptyline%22&limit=1
- us food and drug administration, openfda drug label: nortriptyline hydrochloride capsules, set id 017f7717-e160-4142-9a5a-1a2bd9776707 What we read there: 'indicated for the relief of symptoms of depression.' usual adult dose 25 mg three or four times a day; total daily dose should not exceed 150 mg. Fetched 2026-10-18 (HTTP 200). https://api.fda.gov/drug/label.json?search=openfda.generic_name:%22nortriptyline%22&limit=1
- european headache federation (ehf) critical re-appraisal and meta-analysis of oral drugs in migraine prevention—part 1: amitriptyline. the journal of headache and pain. 2023. doi:10.1186/s10194-023-01573-6 What we read there: three placebo-controlled trials (n = 622): two used 'amitriptyline 100 mg/day or mtd, titrated over 4 weeks,' one used 25 mg a day. one guideline gives 'a dose range between 30 and 150 mg.' moderate-certainty evidence of more dropouts due to adverse events than with placebo. Fetched 2026-10-18 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC10088191/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam