Abstract
BACKGROUND: The association with cardiovascular disease (CVD) is stronger for mean systolic blood pressure (SBP) estimated using ambulatory blood pressure monitoring (ABPM) vs. office measurements. Determining whether this is due to ABPM providing more measurement reliability or greater ecological validity can inform its use. METHODS: We estimated the association of mean SBP based on 2 office measurements and 2, 5, 10, and 20 measurements on ABPM with incident CVD in the Jackson Heart Study (n = 773). Hazard ratios (HRs) for CVD were estimated per standard deviation higher mean SBP. CVD events were defined by incident fatal or non-fatal stroke, non-fatal myocardial infarction, or fatal coronary heart disease. RESULTS: There were 80 CVD events over a median of 15 years. The adjusted HRs for incident CVD were 1.03 (95% CI: 0.90-1.19) for mean office SBP and 1.30 (95% CI: 1.12-1.50), 1.34 (95% CI: 1.15-1.56), 1.36 (95% CI: 1.17-1.59), and 1.38 (95% CI: 1.17-1.63) for mean SBP using the first 2, 5, 10, and 20 ABPM readings. The difference in the HRs for incident CVD ranged from 0.26 (95% CI: 0.07-0.46) to 0.35 (95% CI: 0.15-0.54) when comparing mean office SBP vs. 2, 5, 10, or 20 sequential ABPM readings. The association with incident CVD was also stronger for mean SBP based on 2, 5, 10, and 20 randomly selected ABPM readings vs. 2 office readings. CONCLUSIONS: Mean SBP based on 2 ABPM readings vs. 2 office measurements had a stronger association with CVD events. The increase in the strength of the association with more ABPM readings was small.