Differential Effects of Antihypertensive Drug Classes on Home and Office Blood Pressure Variability

Abstract

Objective Blood pressure (BP) variability provides prognostic information beyond mean BP. However, the class‐specific associations of antihypertensive medications with BP variability remain uncertain, and few studies have directly contrasted home and office BP variability within the same patients while accounting for the white‐coat effect. Methods In a multicenter prospective registry, patients with hypertension and adequate measurements (≥ 10 home and ≥ 5 office readings) were analyzed. The primary outcomes were systolic and diastolic average real variability (ARV) of home and office BP. Associations of renin–angiotensin system inhibitors, beta‐blockers, dihydropyridine calcium channel blockers (DHP‐CCBs), and diuretics with BP variability were estimated using inverse probability of treatment weighting (IPTW) and doubly robust models. The white‐coat effect was further adjusted using both a continuous office‐home mean difference and a guideline‐based white‐coat definition. A true‐monotherapy subcohort sensitivity analysis and class‐by‐class interaction tests were additionally performed. Results Among 495 participants, home ARV differed minimally by antihypertensive class; a small beta‐blocker‐associated increase in home systolic BP (SBP)‐ARV was attenuated with additional adjustment. In contrast, office BP ARV was higher with DHP‐CCBs and diuretics, although the beta‐blocker association was weakened after further adjustment. Following white‐coat effect adjustment, home BP ARV remained neutral across drug classes, whereas DHP‐CCB remained associated with higher office SBP‐ARV and diuretics with higher office diastolic BP ARV. DHP‐CCBs were also associated with a lower probability of office BP control. Home BP control rates did not differ significantly by antihypertensive class. Conclusions In this cross‐sectional observational cohort, antihypertensive medication classes showed context‐dependent associations with BP variability, which were minimal for home BP but more pronounced for office BP, partially independent of the white‐coat effect. Given potential residual confounding by indication, these hypothesis‐generating findings warrant prospective confirmation.