do depression and anxiety travel with migraine, in both directions?
“person's mood depression anxiety are comorbid with migraine right they coexist at a higher than expected by chance and there's birectional relationship between mood disorders and migraine.”
a guest, a headache neurologist said on The Peter Attia Drive, 2026-08-31, at 58:00. hear it in context →
first, the part that holds
the neurologist is right on both counts as an epidemiologist would mean them: people with migraine have roughly double the odds of depression and anxiety, and in a community cohort followed over time, each condition predicted new onset of the other.
what we found when we went and looked
we went and looked. a 2019 meta-analysis of 16 studies found a pooled odds ratio of 1.95 for migraine and depression, and a 2021 systematic review found an average odds ratio of 2.33 for migraine and anxiety in cross-sectional studies. a 2003 community cohort found that baseline major depression predicted first-onset migraine over two years (odds ratio 3.4) and baseline migraine predicted first-onset depression (odds ratio 5.8). so "higher than expected by chance" and "bidirectional" both hold as statements about association and timing. what is still open is why. an 8-year canadian cohort found both directions largely shrank once stress was accounted for, and a genetic study found evidence that depression raises migraine risk but none for the reverse.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| depression and anxiety occur together with migraine more often than expected by chance | confirmed | a meta-analysis of 16 studies in adults gives a pooled odds ratio of 1.95 (95% ci 1.61-2.35) for migraine and depression. a systematic review of 8 studies gives an average odds ratio of 2.33 (95% ci 2.20-2.47) for migraine and anxiety in prevalence and cross-sectional studies, and a relative risk of 1.63 in cohort studies. limitation: the review's own authors flag small samples and poor adjustment for confounders in many of the included studies. [1][2] |
| the relationship between mood disorders and migraine runs in both directions: each predicts later onset of the other | confirmed | in a community cohort (496 with migraine, 151 with other severe headaches, 539 controls, ages 25-55, interviewed in 1997 and 1999), baseline major depression predicted first-onset migraine (or 3.4, 95% ci 1.4-8.7) and baseline migraine predicted first-onset major depression (or 5.8, 95% ci 2.7-12.3). this pattern did not hold for other severe headaches. an 8-year canadian cohort of 9,288 people also found both directions after adjusting for age and sex (depression to migraine hr 1.62, migraine to depression hr 1.55). limitation: the confidence intervals are wide, and the evidence covers major depression, not anxiety. [3][4] |
| popular reading: bidirectional means each condition causes the other | overstated | in the canadian cohort, both directions shrank and stopped being statistically significant once stressors were adjusted for (hr 1.30, 95% ci 0.80-2.10, and hr 1.19, 95% ci 0.86-1.66). the authors concluded that much of the link may come from stress. a two-sample mendelian randomization study found that genetic liability to major depression raised migraine risk (or 1.61) but found no effect of migraine liability on depression (or 1.00, p=0.90). causation in both directions has not been shown. [4][5] |
| side detail: depression worsens the course of migraine | unverifiable | the speaker did not say this. it is listed because people often assume it. in the 2003 cohort, comorbid major depression did not change migraine attack frequency, persistence or the progression of migraine-related disability over two years. one cohort is not enough to settle the question either way. [3] |
both directions
for the speaker: every source we read finds migraine linked to depression and anxiety at about twice the expected rate. the best-known prospective cohort found that each condition predicted new onset of the other, and that this was specific to migraine, not severe headache in general. a second, larger cohort found the same two-way pattern after adjusting for age and sex. the genetic study supports at least one causal direction, from depression to migraine.
against: the two-way finding rests on observational data with wide confidence intervals (the 3.4 odds ratio has a lower bound of 1.4). in the canadian cohort, the link mostly went away once chronic stress, childhood trauma, financial and work stress were accounted for, which points to shared causes rather than one condition driving the other. the genetic study found no effect from migraine to depression. on anxiety, the review found the comorbidity but could not establish which comes first. also, the speaker grouped anxiety under "mood disorders", but the two-way evidence we found is for major depression only.
what this cannot say
comorbidity estimates depend on how migraine and depression are measured (diagnostic interview, questionnaire or self-report) and on whether people came from clinics or the general population, so the "about double" figure is a range, not one number. mendelian randomization depends on assumptions about the genetic instruments, and its null result for migraine to depression may reflect weak instruments rather than a true absence of effect. we read abstracts and summaries, not every full text. pubmed pages were blocked by a cookie wall, so we read those abstracts through ncbi's own e-utilities service. we did not look for a separate prospective test of anxiety and migraine in both directions.
sources, each one fetched and read
- Amiri S, Behnezhad S, Azad E. Migraine headache and depression in adults: a systematic review and meta-analysis. Neuropsychiatrie. 2019;33(3):131-140. doi:10.1007/s40211-018-0299-5. PMID 30666527 What we read there: 16 eligible articles; pooled odds ratio 1.95 (95% CI 1.61-2.35) for migraine and depression. Fetched 2026-10-18 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/30666527/
- Karimi L, Wijeratne T, Crewther SG, et al. The migraine-anxiety comorbidity among migraineurs: a systematic review. Frontiers in Neurology. 2021. doi:10.3389/fneur.2020.613372 What we read there: 8 studies; average OR 2.33 (95% CI 2.20-2.47) in prevalence and cross-sectional studies, average RR 1.63 (95% CI 1.37-1.93) in cohort studies; limited by small samples and lack of adjustment for confounders. Fetched 2026-10-18 (HTTP 200). https://www.frontiersin.org/articles/10.3389/fneur.2020.613372/full
- Breslau N, Lipton RB, Stewart WF, Schultz LR, Welch KM. Comorbidity of migraine and depression: investigating potential etiology and prognosis. Neurology. 2003;60(8):1308-12. doi:10.1212/01.wnl.0000058907.41080.54. PMID 12707434 What we read there: Major depression at baseline predicted first-onset migraine during the 2-year follow-up (OR 3.4; 95% CI 1.4-8.7); migraine at baseline predicted first-onset major depression (OR 5.8; 95% CI 2.7-12.3); comorbid depression did not influence migraine frequency, persistence or disability progression. Fetched 2026-10-18 (HTTP 203). https://pubmed.ncbi.nlm.nih.gov/12707434/
- Swanson SA, Zeng Y, Weeks M, Colman I. The contribution of stress to the comorbidity of migraine and major depression: results from a prospective cohort study. BMJ Open. 2013;3(3):e002057 What we read there: 9,288 Canadians aged 18-64 followed for 8 years; depression to incident migraine HR 1.62 (1.03-2.53), migraine to incident depression HR 1.55 (1.15-2.08); fully stress-adjusted HRs 1.30 (0.80-2.10) and 1.19 (0.86-1.66); 'Much of the apparent association between migraine and depression may be explained by stress.' Fetched 2026-10-18 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC3612807
- Lv X, Xu B, Tang X, et al. The relationship between major depression and migraine: a bidirectional two-sample Mendelian randomization study. Frontiers in Neurology. 2023. doi:10.3389/fneur.2023.1143060 What we read there: Major depression to migraine OR 1.606 (95% CI 1.324-1.949); migraine to major depression OR 1.002 (95% CI 0.975-1.029, p=0.903); only MDD showed a causal effect on migraine risk. Fetched 2026-10-18 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC10140565/
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