Mediation by Antihypertensive Drug Adherence in the Impact of Ambulatory Blood Pressure Monitoring on the Cardiovascular Outcomes in the Newly Treated Hypertension Patient

Objective: Non-adherence has been a growing challenge for breakthrough in the improvement of hypertension control. As the standard method for out of office BP measurement, we reported that ambulatory BP monitoring (ABPM) is also associated with improved antihypertensive drug adherence and improved CV outcome. But the magnitude of the contribution of improved adherence by ABPM on the better CV outcomes is uncertain. Design and method: In the database of The Korean National Health Insurance System, 27357 newly prescribed hypertension patients performed ABPM before the prescription were propensity score matched with the randomly selected 118594 newly prescribed patients without ABPM during 2008.1∼2009.12. Among them, 25312 patients without history of cardiovascular diseases were followed. Primary composite end points were any event of ischemic stroke, intracranial hemorrhage, and acute myocardial infarction. Secondary composite end points were any event of the primary composite end points, and admission for heart failure. Proportion mediated (PM) by good adherence was calculated by using the equation for non-linear outcome model, i.e., PM = natural direct effect (NDE) x [natural indirect effect (NIE)-1]/(NDE x NIE-1). Results: Age of the subject was 49.7+/-12.4 years and the proportion of the female was 41.1%. Median follow-up duration was 8.58 years. Adjusting the covariate and the adherence as the mediator, the hazard ratio of ABPM was 0.65 (95% CI:0.58, 0.74), p <0.0001) and 0.72 (95% CI: 0.66, 0.79, p <0.0001) for the primary and the secondary composite outcomes, respectively. In the mediation analysis, PNDE was 0.71 (95% CI: 0.64, 0.79) and TNIE was 0.95 (95% CI: 0.88, 1.02). PM was calculated as 11.1% (p <0.0001). When considering the interaction term of ABPM∗adherence in the log linear model, PNDE was 0.76 (95% CI: 0.64, 0.91) and TNIE was 0.93 (95% CI: 0.93, 0.83). PM was calculated as 19.1% (p < 0.0001). Conclusions: ABPM before antihypertensive drug prescription could have a good impact on both the drug adherence and CV prognosis. The contribution of good drug adherence to the overall impact of ABPM on improved cardiovascular outcomes is minor.