PDF: Amitriptyline Couch
(1976)Objective
Amitriptyline 50-175 mg/day — to evaluate its efficacy as a prophylactic antimigraine agent and determine whether any benefit is independent of its antidepressant effect.
Study Summary
• Frequency and duration of disabling and severe headaches each fell by roughly 55-60%.
• Nausea decreased 67%, vomiting 83%, and migraine-associated neurologic symptoms fell ∼50%.
• Benefit emerged within 1-42 days (mean 8.8 days) and was sustained out to 30 weeks of follow-up.
• Antimigraine effect correlated only weakly with change in Zung Depression Scale — action appears independent of antidepressant effect.
• Established amitriptyline as an effective migraine prophylactic, now a first-line preventive agent for decades.
Intervention
Amitriptyline started at 50 mg/day, titrated to 75-175 mg/day based on response and tolerability, with follow-up at 4-12 weeks. Open-label, uncontrolled design.
Inclusion Criteria
Adults with at least one severe or disabling migraine per month attending the Kansas University Headache Clinic, without cardiac disease, uncontrolled hypertension, epilepsy, urinary retention, glaucoma, or amitriptyline allergy.
Study Design
Arms: Amitriptyline 50-175 mg/day (single-arm, no placebo control)
Patients per Arm: 110 patients (89 female, 21 male) included in analysis
Outcome
• 72% (79/110) of patients achieved ≥50% reduction in migraine score; 57% (63/110) achieved ≥80% reduction; response distribution was bimodal
• Tension score reduced 56% (288 → 124); GI symptoms and neurologic aura symptoms reduced 50-83%
• Weak correlation between migraine improvement and SDS change (r=0.25, p<0.01) — antidepressant effect insufficient to explain benefit
• Tolerability: 20% discontinued in full cohort for minor AEs (drowsiness, dry mouth, constipation, weight gain); sustained response at 30 weeks
Bottom Line
In 110 patients with severe migraine, amitriptyline (50-175 mg/day, mostly 50-75 mg) reduced migraine score by ≥50% in 72% of patients and by ≥80% in 57%, with only weak correlation to antidepressant effect. Established amitriptyline as an effective migraine prophylactic with action independent of mood modulation — remains a first-line preventive 50 years later.
Major Points
- Single-center open-label prospective study at Kansas University Medical Center Headache Clinic (Couch & Hassanein, 1976)
- 148 patients started; 110 with regular adherence and 4-12 week follow-up analyzed
- Amitriptyline started at 50 mg/day and titrated to 75-175 mg/day; >90% maintained on 50-75 mg
- Primary endpoint: Migraine (M) score (frequency x duration x severity) change from baseline
- Overall M score fell 69.2% (461 → 142); p<0.01
- ≥50% reduction in 72% of patients; ≥80% reduction in 57% — bimodal response distribution
- Nausea reduced 67%, vomiting 83%, neurologic aura symptoms 50%
- Benefit emerged within 1-42 days (mean 8.8 d) and was sustained to 30 weeks in a 55-patient subset
- Antimigraine effect only weakly correlated with Zung depression improvement (r=0.25, p<0.01)
- Tolerable AE profile: drowsiness, dry mouth, constipation, weight gain; 8% early discontinuation
- No comparator arm and no placebo — placebo effect cannot be excluded
- Established amitriptyline as an effective migraine prophylactic independent of antidepressant effect — first-line for ≥50 years
Study Design
- Study Type
- Single-center open-label prospective uncontrolled study (no placebo)
- Randomization
- No
- Blinding
- Unblinded
- Sample Size
- 148
- Follow-up
- 4-30 weeks (mean ~12 weeks)
Primary Outcome
Definition: Migraine (M) score change from baseline (frequency × duration × severity)
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| N/A (no control arm) | 461 → 142 (69.2% reduction) | - (Not reported in 1976 format) | p<0.01 paired t-test |
Limitations & Criticisms
- Open-label uncontrolled design — placebo effect cannot be excluded; authors acknowledged but argued large population and bimodal response distribution made chance unlikely
- Retrospective baseline headache frequency from 8-week recall
- Variable dosing and follow-up intervals (4-12 weeks); not a standardized protocol
- Zung SDS is a screening tool, not a rigorous measure of depression severity — correlation analysis with antidepressant effect may be underpowered
- 25 patients excluded from analysis — selection bias possible (though telephone follow-up suggested excluded patients had similar response rates)
- No comparison with standard prophylactics (propranolol, methysergide) — relative efficacy inferred from historical comparisons only