| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||
LETTER TO THE EDITOR |
Departments of Clinical Epidemiology and Community Studies St. Mary's Hospital Department of Epidemiology and Biostatistics McGill University
Department of Psychiatry St. Mary's Hospital Department of Psychiatry McGill University
Douglas Hospital Research Centre Department of Psychiatry McGill University
Department of Epidemiology, Biostatistics and Occupational Health McGill University
Division of Geriatric Medicine Sir Mortimer B. Davis Jewish General Hospital Department of Family Medicine McGill University Montreal, Canada
Address correspondence to Jane McCusker, MD, DrPH, Department of Epidemiology, McGill University, 3830 ave Lacombe, Montreal, QC H3T1M5, Canada. E-mail: jane.mccusker{at}mcgill.ca
To the Editor:
We read with interest the letter from Dr. Rozzini and colleagues, which reports a study of the prevalence, correlates, and 6-month mortality of major and minor depression in a sample of older medical inpatients. We have compared these results with those from our own study (1,2), and from other comparable studies (3–9). Their sample was of patients admitted to an acute geriatric unit whereas ours was from all medical units at two hospitals. Our sample appeared to be more severely ill (mean Acute Physiology Score of 2.9 vs 1.8) and had a higher 6-month mortality rate (21% vs 14.1%) (2). The prevalence of major depression in their sample was similar at 13.3% (vs 14.2% and 44.5% in the two hospitals in our study), whereas the prevalence of minor depression was much higher (41.0% vs 9.4% and 7.9% at our two hospitals) (1).
In our study, a history of prior depression was an important modifier of the effect of depression on mortality (2). Among patients with no prior history, a depression diagnosis was associated with higher mortality, but this association disappeared in multivariate analyses, after adjustment for age and other covariates. However, among patients with a history of depression, major depression at hospital admission was associated with decreased mortality, even after adjustment for covariates. It would therefore be of interest to know whether the association between a diagnosis of major depression and lower mortality in the Rozzini data was found both in patients with and without a history of depression, and whether it persisted after adjustment for age, comorbidity, severity of illness, and other potential confounders.
Other comparable studies of the association between depression and mortality in hospitalized samples have had mixed results (3–9). We have suggested that these studies should be interpreted with caution, as there are differences in hospital catchment populations and admission practices, factors associated with study participation may introduce various types of selection bias, and there may be residual confounding by other risk factors for mortality. Nevertheless, the data from our study and that of Rozzini and colleagues suggest that future research should address possible differences in the threshold for hospitalization of depressed versus nondepressed patients.
References
| ||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
