Positive Attitudes and Health

Causality

One way to solve the problem of causality is to use longitudinal studies in which one can establish which comes first—the good health or the happiness. In the Duke Longitudinal Studies, for example, it was found that each predicts the other; good health was one of the strongest predictors of happiness at later rounds of examination, and happiness was one of the strongest predictors of good health at later rounds. This indicates a kind of “chicken and egg” relationship, good health causes happiness, and happiness causes good health.

Another way to solve the problem of causality is to use longevity as the outcome variable to see whether happiness predicts greater longevity. By definition in this type of analysis, the happiness comes before the longevity and therefore must be the cause of the longevity rather than vice versa. Again in the Duke Longitudinal Studies, in multiple regression analysis in which other variables (e.g., age, sex) were held constant, it was found that the happiness rating was the second strongest predictor of greater than expected longevity. Happiness was even more important for [Page 471]women, among whom it was the single strongest predictor of longevity.

Similarly, Morton Lieberman did extensive psychological testing of 180 aged men and women and correlated these tests with longevity (intact survival over 2 years). He found that three of the strongest predictors of longevity were self-esteem and the absence of anxiety and depression. A more recent study of the life trajectories of a group of nuns indicated that those who had expressed the most positive emotions during early adulthood lived up to 10 years longer than did those who expressed the least positive emotions.

Another recent study found that people with more positive attitudes toward their aging lived an average of 7.5 years longer than did those with less positive attitudes. This advantage remained after age, gender, socioeconomic status, loneliness, and functional health were statistically controlled. The study was based on 660 individuals age 50 years and older who participated in a community-based survey in Ohio.

...