Preventive statin therapy is often recommended as lifelong treatment.
The aim of this study was to analyze persistence with statin therapy over a decade of use and to identify factors associated with its discontinuation.
Persistence with therapy among new users of statins in 1995 was followed up until December 31, 2005, in Finland using the nationwide drug reimbursement register. Cumulative persistence was analyzed using Kaplan-Meier analysis. A Cox regression model was applied to analyze associations of various baseline covariates with discontinuation. We further modeled the association of time-specific covariates by stratifying the duration of therapy in years and using a logistic regression in which those continuing therapy until the end of follow-up (persistent users) formed the reference group. Adherence, defined as the proportion of days covered by statins, stratified by the timing of discontinuation, was computed for the respective groups.
Of the 18,072 new statin users, 73.3% (n =13,254) were aged >54 years and 54.8% (n =9908) were men. Of this cohort, 43.9% (n = 7926) were using statins throughout and at the end of the tenth year. Sex was not associated with persistence at any point. In the Cox model, persons aged 45 to 74 years at initiation were more likely to continue statin use than younger or older age groups. Among those who still used statins after the fifth year of observation, the age difference was not observed in the logistic regression model. The use of 1, 2, 3, or > or =4 cardiovascular drugs before the initiation predicted continuation relative to no cardiovascular drug use (hazard ratio for discontinuation significantly