Abstract 08: Blood Pressure Percentile Charts Better Identify High Risk Young Adults: The Coronary Artery Risk Development in Young Adults (CARDIA) Study

Abstract

Introduction: Current guidelines set a static threshold for diagnosis of hypertension based on absolute blood pressure (BP) levels for all ages. However, BP generally increases with age and relative levels tend to track within individuals from young adulthood on. We aimed to determine whether BP percentiles , compared to the current AHA/ACC clinical thresholds for hypertension, are better able to identify young adults at higher risk for future CVD events. Methods: Among 5,115 CARDIA participants (ppts; mean age 25 yrs; 54% female; 51% Black), we used baseline BP level and over 35 years of follow-up for fatal and non-fatal CVD events. We calculated normative BP percentiles for adults aged 18 and older using NHANES by estimating underlying BP among treated individuals combined with observed BP for untreated individuals (Figure). Using these BP Percentile Charts, we calculated the baseline BP percentile for each ppt. Adjusted survival analyses examined the association between BP percentile (in deciles) and incident CVD events over 30 years. C statistics for models using BP percentiles vs those using current guideline-recommended thresholds were compared to determine relative accuracy. Results: At baseline, mean S/DBP was 112/70 mmHg and <1% were on antihypertensive medications. Adjusted HRs for BP percentile decile demonstrated a dose-response effect with increasing risk associated with higher deciles of SBP. Compared to the 5 th decile young adults with SBPs in the highest decile 90-100% (SBP ≥127 for men and ≥120 for women) had a HR of 1.76 (95% CI 1.10-2.83). C statistics for models using BP percentiles were higher than for models using guideline-recommended thresholds (0.724 vs 0.713). No significant association was seen with DBP deciles. Conclusion: BP percentiles are better able to predict future CVD risk in young adults as compared to current age-invariant thresholds. These methods may help to risk stratify young adults and more accurately identify higher risk individuals for early prevention efforts.