Skip to main content
. 2009 Jul 8;2009(3):CD007954. doi: 10.1002/14651858.CD007954

Thomson 1982.

Methods Amitriptyline vs placebo vs L‐tryptophan vs combination L‐tryptophan and amitriptyline
 RCT in 5 Scottish group general practices.
Participants N=115. Amitriptyline n=31 placebo n=28 (2 other arms L‐tryptophan n=29; combination Amitriptyline + L‐tryptophan n=27)
 127 women & 34 men. Patients 18‐65 yrs
 HMAD of ?12, depressed for? 2 weeks & considered by their GPs to need antidepressant medication but not need psychiatric referral. 
 Exclusions:
 Tricyclic antidepressants in previous 2 weeks or contraindications to tricyclics
Interventions 1 week run‐in period then 25 mg amitriptyline 3x daily for 1 week then 2 tablets 3 x daily for remaining 11 weeks.
 Could also take diazepam 5mg or nitrazepam as hypnotic but had to continue this throughout study.
 2 other arms to study which involved amitriptyline + tryptophan & tryptophan. Only those on active medication had blood levels for amitriptyline & tryptophan. 12 week duration
Outcomes 37% (43) drop outs.
 At 12 weeks HAMD for amitriptyline group mean = 4.9 (SD 4.9) n=21 vs placebo group mean = 7.93 (SD 4.2) n=15.
 At 4 weeks: Amitriptyline group n=26 mean = 6.85 (SD 7.3) vs placebo group n=19 mean = 11.55 (SD 9.6).
 At 12 weeks: complete remission (i.e. a fall of 4 on HAMD) 67% (14/21) in amitriptyline group vs 33% (5/15) in placebo group (estimates from graphs).
 80% in active group had positive blood levels for amitriptyline.
 Side effects not necessarily leading to withdrawal: amitriptyline 7/31 vs placebo 0/28
 Adverse effects leading to withdrawal: amitriptyline 7/31 vs placebo 0/28
 Withdrawal due to treatment failure: Amitriptyline 0/31 vs placebo 8/28
 Treatment withdrawal for any reason: Amitriptyline 10/31 vs placebo 13/28
Notes Alternating visit assessments done by psychiatrist and patient's own GP 2‐weekly. Good correspondence of assessments. Research psychiatrist ratings used in analysis (Weeks 0,4, 8 and 12).
Risk of bias
Bias Authors' judgement Support for judgement
Allocation concealment (selection bias) Unclear risk B ‐ Unclear