The Mayo Clinic cohort study of personality and aging: design and sampling, reliability and validity of instruments,...

We established a historical cohort of 7,216 subjects who completed the Minnesota Multiphasic Personality Inventory (MMPI) at the Mayo Clinic from 1962 through 1965 for research (not clinical indication), and who resided within a 120-mile radius centered in Rochester, Minnesota. We describe here the overall cohort design and sampling, we report results concerning reliability and validity, and we describe age and sex patterns at baseline for four MMPI scores of primary interest (depression, anxiety, social introversion, and negativity). Subjects excluded from the cohort because of missing data had MMPI scores similar to subjects included (after appropriate rescaling). A cut-off specific for age and sex at the 75th percentile of the distribution of raw scores was valid compared with the traditional clinical cut-off (T scores plus one standard deviation). Baseline scores for all four scales were higher in women than in men at all ages (all p < 0.0001). Depression and social introversion scores showed an increasing trend with age in both sexes (Spearman rank correlation, rho = 0.05 and 0.08, respectively, p < 0.0001 for both). Baseline scores on the anxiety scale showed a decreasing trend with age in both sexes (rho = -0.06, p < 0.0001). Negativity scores remained relatively stable with age in both sexes (rho = 0.03, p = 0.01). We found a high correlation between the anxiety score and the negativity score (rho = 0.90, p < 0.0001) even after the exclusion of overlapping items (rho = 0.68, p < 0.0001). This newly established historical cohort study provides opportunities to test hypotheses regarding the link between personality and aging, aging-related diseases, and overall mortality.