Blood pressure responses to exercise as predictors of blood pressure level after 5 years

There is considerable disagreement in the literature on the clinical usefulness of exercise testing as a tool for prediction of future hypertension. Much of the discrepancy between various reports is attributable to the difficulties of blood pressure (BP) measurement during exercise. Therefore, we investigated whether accurate intraarterial BP measurement will increase the predictive power of exercise testing. The BP responses to dynamic and isometric exercise were evaluated in 97 healthy, unmedicated men, of whom 34 were normotensive (NT), 29 borderline (BHT), and 34 mildly hypertensive (HT) using three criteria: 1) achieved BP during the test, 2) the change of the BP from baseline to exercise, and 3) the group was divided into high responders (HIGH, n = 19, systolic BP > or = 220 and diastolic BP > or = 105 mm Hg) and normal responders (n = 60). Five years later the BP was reassessed by casual measurements and noninvasive ambulatory 24-h monitoring (NAMB) in 79 (81%; 27 NT, 24 BHT, and 28 HT) subjects. The achieved isometric BP correlated well with the follow-up BP (casual systolic BP r = 0.43, diastolic BP r = 0.45, and NAMB systolic BP r = 0.44, diastolic BP r = 0.58, P < .001). However, achieved dynamic BP showed a poorer relationship to future BP (r range, 0.09 to 31, P = NS to P < .01). Because the intraarterial preexercise sitting BP also correlated well with follow-up BP (r range, 0.33 to 0.48, P < .01 to P < .001), and the r values were close to those of achieved isometric BP we used multiple regressions (including all resting and exercise BP values as independent variables) to evaluate the contributions of the baseline and exercise values for prediction of the follow-up BP. The baseline value explained 12% to 23% (from casual diastolic BP to NAMB diastolic BP, systolic BP values) of future BP variance, whereas achieved isometric BP ranged an additional 1% to 11% (from casual systolic BP to NAMB diastolic BP) of variance. In general, BP change from baseline with stressors did not correlate with follow-up measurements. In the high responder group the achieved dynamic BP did not correlate significantly with the follow-up BP, whereas the achieved isometric diastolic BP did correlate (casual diastolic BP r = 0.56, P < .05, NAMB systolic BP and diastolic BP r = 0.52, P < .05). Both groups had similar future BP levels. In conclusion, even with very accurate BP readings the reactivity to dynamic exercise is a weak predictor of future BP, and does not improve the prediction compared to resting BP values. Intraarterial BP response to isometric exercise marginally improves the prediction of future BP levels.