Psychological predictors of hypertension in the Framingham Study. Is there tension in hypertension?

• Published in: JAMA : the journal of the American Medical Association Vol. 270; no. 20; p. 2439
• Main Authors: Markovitz, J H, Matthews, K A, Kannel, W B, Cobb, J L, D'Agostino, R B
• Format: Journal Article
• Language: English
• Published: United States 24.11.1993
• Subjects: Aged; Analysis of Variance; Anger - physiology; Anxiety - physiopathology; Blood Pressure - physiology; Cohort Studies; Female; Humans; Hypertension - epidemiology; Longitudinal Studies; Male; Middle Aged; Predictive Value of Tests; Psychological Tests; Regression Analysis; Risk Factors
• ISSN: 0098-7484
• DOI: 10.1001/jama.1993.03510200045030

To test the hypothesis that heightened anxiety, heightened anger intensity, and suppressed expression of anger increase the risk of hypertension, using the Framingham Heart Study. A cohort of men and women without evidence of hypertension at baseline were followed up for 18 to 20 years. Baseline measures of anxiety (tension), anger symptoms, and expression of anger (anger-in and anger-out) were taken, along with biological and behavioral predictors of hypertension (initial systolic blood pressure, heart rate, relative weight, age, hematocrit, alcohol intake, smoking, education, and glucose intolerance). A total of 1123 initially normotensive persons (497 men, 626 women) were included. Analyses were stratified by age (45 to 59 or > or = 60 years) and gender. Hypertension was defined as either taking medication for hypertension or blood pressures higher than 160/95 mm Hg at a biennial examination. In univariate analyses, middle-aged men who went on to develop hypertension had greater baseline anxiety levels than men who remained normotensive (P = .04). Older hypertensive men had fewer anger symptoms at baseline (P = .04) and were less likely to hold their anger in (P = .01) than normotensives. In multivariate Cox regression analysis including biological predictors, anxiety remained an independent predictor of hypertension in middle-aged men (P = .02). Among older men, anger symptoms and anger-in did not remain significant predictors in the multivariate analysis. Further analysis showed that only middle-aged men with very high levels of anxiety were at increased risk (relative risk, 2.19; 95% confidence interval, 1.22 to 3.94). No psychological variable predicted hypertension in middle-aged or older women in either univariate or multivariate analyses. The results indicate that among middle-aged men, but not women, anxiety levels are predictive of later incidence of hypertension.