Adherence to antihypertensive therapy and cardiovascular outcomes: an individual participant data meta-analysis of randomised controlled trials
Q Yang, Z Bidel, D Canoy, G Zeng, W Cushman, A Rodgers, K Teo, B Davis, J Chalmers, C Pepine, J Sundstrom, M Woodward, M Nazarzadeh, K RahimiAbstract
Background
Suboptimal adherence to antihypertensive therapy is common and may attenuate clinical benefit, but randomised evidence quantifying the impact of reduced adherence on clinical outcomes remains limited.
Purpose
To determine whether and to what extent medication adherence modifies the blood pressure and cardiovascular effects of antihypertensive therapy in randomised trials.
Methods
We performed an individual participant data meta-analysis of randomised controlled trials with available adherence data. Adherence measures were harmonised into a binary variable (≥80% vs <80% of medication taken as assigned) for each participant, defining higher- and lower-adherence strata, updated at each follow-up visit. Primary outcomes were longitudinal systolic blood pressure (SBP) and incidence of clinical events, including major cardiovascular disease (stroke, myocardial infarction/ischaemic heart disease, and heart failure causing death or hospitalisation) and all-cause death. Treatment effects were estimated within adherence strata, while preserving randomisation, using linear mixed-effects models for SBP and trial-stratified Cox models for clinical events, with treatment-by-adherence interaction terms to test effect modification.
Results
The analysis included 115,389 participants from 12 trials; 91,339 participants from nine trials had longitudinal adherence data. Adherence declined over time in both intervention and comparator groups. At one year, mean adherence was 88% in the intervention group and 87% in the comparator group; after five years, these figures had decreased to 80% and 77%, respectively. Over a median follow-up of 4.7 years, 15,460 major cardiovascular events and 9,591 deaths were recorded. Compared with the comparator group, the intervention reduced SBP by −3.0 mm Hg (95% CI −3.3 to −2.7) in the lower-adherence stratum and by −5.2 mm Hg (95% CI −5.4 to −5.0) in the higher-adherence stratum (P for interaction < 0.001). The treatment effect on major cardiovascular disease differed by adherence stratum (P for interaction = 0.04), with an HR of 0.89 (95% CI 0.86–0.92) in the higher-adherence stratum versus 1.01 (95% CI 0.92–1.10) in the lower-adherence stratum (Figure 1).
Conclusion
The benefits of antihypertensive therapy were strongly dependent on medication adherence. When comparing randomised treatment effects between adherence strata, we observed substantially greater reductions in SBP and major cardiovascular disease events with higher adherence, with attenuation of benefit at lower adherence levels. These findings underscore the importance of effective long-term adherence strategies to realise the full clinical benefits of BP-lowering therapy.