Association between antihypertensive treatment and hospitalisation or death due to falls according to sex, ethnicity, and social deprivation status: an observational cohort study in English primary care electronic health-care records

SummaryBackground Antihypertensive therapy is associated with an increased risk of adverse events, such as falls, in older patients and those with increasing levels of frailty. This study aimed to explore the association between antihypertensive treatment and serious adverse events in other patient subgroups in primary care according to sex, ethnicity, and socioeconomic status. Methods This retrospective observational cohort study used routinely collected primary care data from England, UK, in the Clinical Practice Research Datalink Aurum database between Jan 1, 1998, and Dec 31, 2018. Patients aged 40 years or older, without previous antihypertensive prescriptions, and with a systolic blood pressure reading of 130–179 mm Hg were included. The main outcome was first hospitalisation or death from a fall within 10 years of follow-up. Exposure was defined as the first antihypertensive prescription within 12 months after inclusion. Cox regression models adjusted for propensity scores were developed in sex, ethnicity, and socioeconomic status subgroups, and absolute risk differences were calculated. Findings 2 614 330 patients were included in the analyses, of whom 337 742 (12·9%) were in the exposed group and 2 276 588 were in the non-exposed group. Median age in the non-exposed group was 53 years (IQR 46–63); 1 141 885 (50·2%) were female and 1 134 703 (49·8%) were male; 1 120 659 (49·2%) were White, 561 427 (24·7%) were of mixed or other ethnicity, 64 377 (2·8%) were Black, and 53 338 (2·3%) were south Asian. In the exposed group, median age was 61 years (51–71); 166 639 (49·3%) were female and 171 103 (50·7%) were male; 196 432 (58·2%) were White, 66 460 (19·7%) were of mixed or other ethnicity, 13 668 (4·0%) were Black, and 10 948 (3·2%) were south Asian. The median follow-up time was 7·0 years (IQR 2·9–10·0). The main outcome of falls was observed in 110 874 patients (4·2% of the total population; 23 925 7·1% in the exposed group and 86 949 3·8% in the non-exposed group). The association between antihypertensive medication and hospitalisation or death due to falls was similar for both sexes (adjusted hazard ratio aHR 1·09 95% CI 1·06–1·12 for male patients and 1·10 1·07–1·12 for female patients). The associations (aHRs) in the different ethnicity groups were 1·32 (95% CI 1·13–1·53) in Black patients, 1·22 (1·05–1·41) in south Asian patients, 1·47 (1·29–1·66) in patients of mixed or other ethnicity, and 1·05 (1·03–1·07) in White patients. The most deprived group had an aHR of 1·02 (95% CI 0·99–1·06) and the least deprived group had an aHR of 1·16 (1·12–1·21). The absolute risk difference was low for all subgroups (≤12 events per 10 000 patients per year). Interpretation All subgroups of patients with antihypertensive prescriptions had an increased risk of hospitalisation or death due to falls versus those who were unexposed. No significant difference in risk of falls was observed based on sex, ethnicity, or social deprivation; therefore, no distinction should be made among these groups when considering the harms of antihypertensives in individual treatment decisions in primary care. Funding Wellcome Trust, Royal Society, and National Institute for Health Research School for Primary Care Research.