Thresholds for white-coat and masked blood pressure effects: an ambulatory blood pressure monitoring study

Objective: The values used to define white-coat and masked blood pressure (BP) effects are usually arbitrary. This study aimed at investigating the accuracy of various thresholds based on the differences (DBP) between office BP (OBP) and 24h-ambulatory BP monitoring (ABPM) to identify white-coat (WCH) and masked (MH) hypertension, which are phenotypes assumed to carry adverse prognosis. Design and method: This cross-sectional study included 11,350 [Derivation cohort; 45% men, mean age = 55.1 ± 14.1 years, OBP = 132.1 ± 17.6/83.9 ± 12.5 mmHg, ABPM = 121.6 ± 11.4/76.1 ± 9.6 mmHg, 25% using antihypertensive medications (AH)] and 7,220 [Validation cohort; 46% men, mean age = 58.6 ± 15.1 years, OBP = 136.8 ± 18.7/87.6 ± 13.0 mmHg, ABPM = 125.5 ± 12.6/77.7 ± 10.3 mmHg; 32% using AH] unique individuals who underwent OBP and ABPM measurements. We compared the sensitivity, specificity, positive (PPV) and negative (NPV) predictive values and area under the curve (AUC) of 8 different ↗BP thresholds to detect WCH (↗systolicBP/↗diastolicBP = +28/+17, +20/+15, +20/+10, +16/+11, +15/+9, +14/+9 mmHg and ↗systolicBP = +13 and +10 mmHg) and eight different ↗BP thresholds to detect MH (↗systolicBP/↗diastolicBP = -14/-9, -5/-2, -3/-1, -1/-1, 0/0, +2/+2 mmHg and ↗systolicBP = -5 and -3 mmHg), built from formerly reported criteria in the literature. WCH was defined as OBP> = 140/90mmHg and ABPM < 130/80mmHg, and MH was defined as OBP < 140/90mmHg and ABPM> = 130/80mmHg. Results: The +20/+15 mmHg threshold showed the best AUC (0.804, 95%CI = 0.794 - 0.814) to detect WCH in the Derivation cohort, with sensitivity, specificity, PPV and NPV of 80.6%, 80.2%, 42.3%, and 95.8%, respectively. The +2/+2 mmHg threshold showed the highest AUC (0.741, 95%CI = 0.728 - 0.754) to detect MH, with sensitivity, specificity, PPV and NPV of 78.9%, 69.3%, 22.0% and 96.8%, respectively. Both threshold values also had the best accuracy to detect WCH (0.767, 95%CI = 0.754 - 0.780) and MH (0.767, 95%CI = 0.750 - 0.784) in the Validation cohort. In secondary analyses, these thresholds had the best accuracy to detect WCH and MH in individuals using or not AH, and to detect individuals with higher and lower office-than-home BP stages, respectively, in both cohorts. Conclusions: The +20/+15 and +2/+2 mmHg ↗BP thresholds had the best accuracy to detect hypertensive patients with WCH and MH, respectively, and may be indicators of marked white-coat and masked BP effects.