Blood pressure variability predicts the development of hypertension in normotensive individuals

Objective: Hypertension is one of the important risk factors for atherosclerosis leading to cardiovascular diseases. Although several factors have been reported as predictors of new onset of hypertension, the association of blood pressure (BP) variability in normotensive subjects with future development of hypertension is not intensively discussed. We tested the hypothesis that the yearly variability of BP in normotensive subjects predicts new onset of hypertension. Design and Methods: Subjects who participated in our annual physical check-up program for 5 years in a row during 2008 and 2013 were screened (screening period) and those under antihypertensive medication or with the average BP for the 5 visits ≧ 140/90 mmHg were excluded. The rest of 2006 participants (male = 1287) were eligible in the present study. Standard deviation (SD) and coefficient of variation (CV) of their systolic BP (SBP) during the 5 consecutive visits were calculated (screening period) and, then, they were followed up for the next 5 years (follow-up period). The onset of hypertension was defined as average BP during the follow-up period ≧ 140/90 mmHg or the use of antihypertensive medications. Results: The average, SD, and CV of SBP were 118.7 ± 10.0 mmHg, 7.44 ± 2.98 mmHg, and 6.30 ± 2.54%, respectively. During the follow-up, hypertension developed in 259 subjects, with the incidence being more frequent in male (n = 180, 14.0%) than female subjects (n = 79, 11.0%). In retrospective analysis, the average (128.7 ± 7.7 vs. 117.2 ± 9.5 mmHg, p < 0.001) and SD (8.05 ± 3.48 vs. 7.34 ± 2.89 mmHg, p < 0.001) were higher in subjects with than without future hypertension, whereas CV was not different in the both groups (6.28 ± 2.73 vs. 6.30 ± 2.51%). Non-adjusted odds ratio (OR) (95% confidence interval [CI]) of average, SD and CV for the incident hypertension were 1.169 (1.146–1.193), 1.078 (1.034–1.124) and 0.997 (0.947–1.050), respectively. Multivariate logistic regression analysis adjusted for age, gender, family history of hypertension, frequent alcohol consumption (six or seven times per week), current smoking, body mass index, heart rate, serum creatinine, fasting plasma glucose, LDL-cholesterol, and triglyceride at baseline demonstrated that average (OR [95%CI] = 1.193 [1.138–1.187]) and SD (1.058 [1.014–1.105]), but not CV (0.992 [0.941–1.046]), were associated with new onset hypertension. Similar results were obtained using diastolic BP (data not shown). Conclusions: In normotensive subjects, an increase in year-to-year variability as well as average of SBP may be a predictor of the new onset of hypertension. Careful interpretation of year-to-year BP variability is necessary for evaluation of future cardiovascular risk in normotensive individuals.