A longitudinal study of students' depression at one medical school

Using a standardized measure of depression at three assessment points, to examine depression in medical students during their training. Students entering the University of Massachusetts Medical School in the fall in 1987, 1988, and 1989 were mailed a recruitment letter and baseline questionnaire four weeks prior to the start of classes. Subsequent assessments took place in the middles of year 2 and year 4 and included only the students who had participated in the baseline assessment. The baseline assessment included the Center for Epidemiological Studies Depression (CES-D) scale, the Bortner Type A Behavior scale, the Spielberger Trait Anger scale, and the Spielberger Anger Expression scale. In addition, the baseline package included a rating of perceived stress, a demographics questionnaire, and a social-life survey. The follow-up assessments included the CES-D scale, the rating of perceived stress level, and the social-life survey. Analytic methods used were univariate descriptive statistics, correlation, and multiple-linear-regression analyses, two-sample t-tests, analysis of variance, and chi-square tests. Of the initial pool of 300 students, 264 responded at the baseline assessment (88% response rate; 53% men); 171 of these participated in the year-2 assessment (65% response rate; 51% men), and 126 participated in the year-4 assessment (48% response rate; 48% men); a total of 99 students participated in all three assessments. CES-D scores > or =80th percentile were obtained for 18% of the entering students. This rose to 39% at year 2 and 31% at year 4 (p = .0001). No gender difference was found at baseline; however, the women experienced higher depression levels than did the men at year 2 (p = .004) and at year 4 (p = .04). Overall, gender and increases in perceived stress (from baseline to year 2) were significant predictors of increased CES-D scores (from baseline to year 2; p = .01 and p = .0001, respectively). For the women, increased perceived stress, angerin, and frequency of social contacts outside work/school were significant predictors of the magnitude of increases in CES-D scores (baseline to year 2; p = .0001, p = .02, and p = .03, respectively). These preliminary data support the view that, upon entering medical school, students' emotional status resembles that of the general population. However, the rise in depression scores and their persistence over time suggest that emotional distress during medical school is chronic and persistent rat