A cohort study of incident hypertension in relation to changes in vigorous physical activity in men and women

To assess the dose-response relationship between changes in vigorous exercise (running distance, Deltakm per week) and physician-diagnosed hypertension.

Twenty-four thousand, five hundred and fifty men and 10 113 women were followed prospectively for (mean +/- SD) 7.8 +/- 1.8 and 7.5 +/- 2.0 years, respectively.

Among those who maintained their running distance within +/-5 km per week (5841 men), logistic regression showed that the log odds for hypertension was significantly lower for those who ran longer distances (coefficient+/-SE:-0.019 +/- 0.003 per km per week; P < 0.0001) even when adjusted for body mass index (-0.010 +/- 0.003 per km per week; P = 0.002). Analyses of all 24 550 male and 10,113 female runners showed that the log odds for hypertension declined significantly in relation to Deltakm per week in men (-0.009 +/- 0.001; P < 0.0001) and women (-0.006 +/- 0.003; P = 0.03), which remained significant when adjusted for body mass index in men (-0.005 +/- 0.001; P < 0.0001) but not in women (-0.004 +/- 0.003; P = 0.13). In both sexes, the decline was related to the distance run at the end of follow-up but not at baseline. Compared with men who ran less than 8 km per week, the age-specific rate for incident hypertension in those who ran more than 40 km per week at the end of follow-up was 80% lower in those aged between 35 and 44 years, 66% lower in those between 45 and 54 years, 69% lower in those aged between 55 and 64 years (all P < 0.0001), and 57% lower in those older than 65 years (P = 0.08).

The odds of developing hypertension are reduced in those who remain vigorously active and increased in those whose vigorous activity declined. These effects are dependent on the exercise dose and are due in part to metabolic processes associated with body weight.