Abstract:
|
Objective: The objective of this study is to investigate the association between statin use and all-cause mortality.
Methods and Results: We developed propensity scores (PS) models based on the greedy matching and inverse probability of treatment weighing (IPTW) algorithms. We performed covariate balance using standardized mean differences (SMD) for all the variables used in our PS models. Of the 56 variables considered in the PS modeling, 42 were significantly associated (p < 0.05) with statin use at baseline. The prevalence of statin-use and all-cause mortality in the JHS were 13.9% (95% CI: 12.9% -15.0%) and 10.8% (95% CI: 9.9%-11.6%), respectively. In the crude analysis, the odds of dying was 80% higher (crude odds ratio 1.8 (95% CI: 1.4-2.3 if an individual is a statin-user as opposed to a non-user. After adjustment for PS using the greedy matching and IPTW algorithms, the statin use-mortality odds ratios were 0.77 (95% CI=0.53-1.12) and 0.80 (95% CI=0.68-0.95), respectively. There was no statin-use by age interaction. Covariate balance was obtained for all the variables used in the development of the PS models (SMD < = 0.2).
|