To the Editor: Tanja Laukkala, M.D., Ph.D., and colleagues (1) reported underuse of antidepressant medications in the Finnish population. Only 13% of the respondents with an episode of major depression in the preceding year, according to the Composite International Diagnostic Interview Short Form, were taking antidepressant medications at the time of the interview.

Unfortunately, Dr. Laukkala et al. did not report the total proportion of subjects using antidepressants. According to a recent Canadian health survey, the National Population Health Survey (2), which also employs a national probability sample, this proportion was 4.0%. Of note, of those with an apparent episode of depression on the Composite International Diagnostic Interview Short Form in the previous 12 months in the Canadian survey, 27.1% (weighted) reported antidepressant use in the preceding month (2), which is reasonably consistent with the results reported by Dr. Laukkala et al. The estimated 30-day point prevalence of major depression in the National Comorbidity Survey was 4.9% (3), so a 4.0% rate of antidepressant use does not suggest drastic underuse. It seems strange that antidepressants should be so frequently used yet paradoxically appear to be underused to such a great extent. In view of this, an interpretive point is worth raising. The 13% rate of antidepressant use reported by Dr. Laukkala et al., like the 27.1% rate reported elsewhere, represents the proportion of individuals with a recent episode of major depression who were currently taking antidepressants. This should not be confused with the proportion who need treatment and are receiving it. Many who have had a successful outcome of antidepressant treatment more than 12 months previously are removed from estimates of this type.

In the National Population Health Survey, of 668 subjects with an episode of major depression (in the preceding year), 194 reported antidepressant use, compared to 482 of 14,108 subjects without a major depressive episode. As expected, the rate of use was much higher in those with a depressive episode, but the fact remains that over 70% of those taking antidepressants had no episodes of major depression in the preceding year. A proportion of these subjects may derive from the 15%–20% of the population with lifetime major depression (3), some of whom have successfully achieved control of their depressive disorders by using antidepressant medications. Such success is not reflected in the 27.1% use rate reported. To illustrate this point, suppose that 10 of 50 persons with active major depression are found to be taking antidepressants in a survey with a group size of 1,000. The estimation approach of Dr. Laukkala et al. would put the use rate at 20%. However, if another 50 of the remaining 950 members of the population (who were not depressed in the last year) had recurrent major depression that is successfully controlled by medications, then an alternative way of depicting use would be to estimate it at 60 of 100, or 60%.

Psychiatric epidemiologists will need to develop methods to estimate the adequacy of antidepressant use that can account for the various ways in which antidepressants can improve mood status, including inducing and sustaining remission.