Abstract
Abstract
Day-to-day home blood pressure (BP) variability predicts cardiovascular and cognitive risk, but whether it shows seasonal variation within the same individual remains uncertain. We examined its seasonal patterns using repeated personal health record measurements. We analyzed a Japanese health-insurance database (DeSC database) linked to home BP from a mobile application. Monthly trajectories and winter–summer differences in mean home morning BP, its within-individual standard deviation (SD), and variability independent of the mean (VIM) were estimated using linear mixed-effects models stratified by antihypertensive treatment and sex, with age category as a fixed effect. Winter–summer differences were standardized by the model-based total SD for comparison across indices. Among the 7,908 individuals (mean age, 55 years), 13.6% were untreated females, 10.2% were treated females, 31.1% were untreated males, and 45.1% were treated males. Home systolic BP peaked in December or January and reached its nadir in July–August, regardless of antihypertensive treatment status, sex, or age category, with winter–summer differences of approximately 3–4 mmHg (standardized differences: 0.28–0.41). The SD of home systolic BP followed a similar seasonal pattern (standardized differences: 0.07–0.18). Although winter–summer differences in VIM were also significant in several strata, particularly among males, their magnitudes were small (standardized differences ≤0.09). Home BP levels showed clear seasonal variation, whereas seasonal differences in day-to-day BP variability were minor when indices less dependent on average BP were used. The apparent winter increase in day-to-day BP variability unadjusted for BP level is largely attributable to the seasonal BP elevation.