Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Antidepressants improve depression in adults with physical illnesses

ACP J Club. 1999 July-Aug;131:8. doi:10.7326/ACPJC-1999-131-1-008

Source Citation

Gill D, Hatcher S. A systematic review of the treatment of depression with antidepressant drugs in patients who also have a physical illness. Cochrane Review, latest version 11 Nov 1998. In: The Cochrane Library. Oxford: Update Software.



Are antidepressants effective for depression in adults with physical illnesses?

Data sources

Studies were identified by searching MEDLINE (to June 1998), the Cochrane Clinical Trials Register (to June 1998), the Cochrane Depression Anxiety and Neurosis Trial Register, and the British National Formulary (to March 1998). 2 journals and a specialist textbook were scanned, bibliographies were reviewed, and experts were contacted.

Study selection

Randomized controlled trials were selected if the patients had a physical illness plus documented depression; antidepressants were compared with placebo or no treatment; and outcomes of changes in depression, changes in physical illness, and treatment acceptability (percentage of patients who completed the trial) were measured. Exclusion criteria were heterogeneous study groups; patients with dementia, psychosis, or addictive behaviors; use of antidepressants for pain relief; studies of euphoriants, adjuvants, or combination therapy; and data from the second arm of crossover studies.

Data extraction

Data were extracted on study quality and characteristics, depression, drug category, physical disease, disability, and outcomes.

Main results

18 trials (838 patients) were included. Diseases studied were HIV or AIDS; stroke; cancer; mixed studies; and multiple sclerosis, heart disease, renal failure, diabetes mellitus, and head injury. Atypical antidepressants were used in 3 studies, tricyclic antidepressants (TCAs) were used in 11 studies, and selective serotonin-re-up-take inhibitors (SSRIs) were used in 6 studies. Reduction in depression was significant for all antidepressants (P < 0.001) and for TCAs (P = 0.002) (Table) but not for individual categories of antidepressants. 2 studies measured function and quality of life; neither showed improvements with antidepressants.7 studies evaluated change in physical illness; of these, 1 showed worse control of disease in diabetes (nortriptyline).


Antidepressants improve depression in adults with physical illnesses.

Sources of funding: Medical Research Council and Waitemata Health.

For correspondence: Dr. D.D. Gill, Department of Liaison Psychiatry, Whipps Cross Hospital, London E11 1NR, England, UK. FAX 44-181-535-6521.

Table. Antidepressants vs placebo for improvement in depression at the end of treatment for adults with physical illnesses (mean duration 6.5 wk)*

Type of antidepressant Weighted event rate RBI (95% CI) NNT (CI)
Antidepressants Placebo
All 47.1% 25.9% 29% (16 to 40) 5 (4 to 7)
Tricyclics 51.1% 24.5% 30% 9 to 47) 4 (2 to 10)

*Abbreviations defined in Glossary; RBI, NNT, and CI calculated from data in article.


Physical and psychiatric conditions frequently occur together in general hospital settings, and psychiatric symptoms in the presence of physical illness are more common than psychiatric symptoms alone in general practice (1). Many persons are depressed, and evidence suggests that patients with physical illnesses do not recover from depression as well as those without. Gill and Hatcher confirm that depression in the presence of physical illness responds to antidepressants.

With such a wide topic, the authors have left several questions unanswered, although they promise to cover some of these in the future. They correctly emphasize the importance of randomization and the use of standardized psychiatric instruments in assessing outcome. More attention could have been given to quantifying the presence of physical illness for either severity or duration. The authors could also have highlighted the wide range of antidepressant doses used (double the therapeutic dose in the case of desipramine), given the small but significant dose-effect relation (2). Studies of different types of antidepressants were combined for analysis because these agents have been shown to be more alike than different in terms of treatment response. It remains uncertain, therefore, if patients who are physically ill and depressed tolerate SSRIs better than TCAs.

It is also unclear to what extent the findings of trials of depressed patients without physical illnesses can be extrapolated to trials of depressed patients with physical illnesses. The authors suggest that such extrapolation might be possible, although their evidence comes from a nonrandomized comparison of trials rather than from trials that randomized patients to either TCAs or SSRIs.

While we wait for answers to some of these questions, evidence highlights the importance of recognizing and treating depression in patients who are physically ill as well as understanding and using the wide range of effective antidepressants available.

Steve Kisely, MBChB, MD, MSc
University of Western AustraliaFremantle, Western Australia, Australia


1. Bridges KW, Goldberg DP. Somatic presentation of DSM III psychiatric disorders in primary care. J Psychosom Res. 1985; 29:563-9.

2. Bollini P, Pampallona S, Tibaldi G, et al. Effectiveness of antidepressants. Meta-analysis of dose-effect relationships in randomized clinical trials. Br J Psychiatry. 1999;174:297-303.