Statin Use and Arterial Calcification in the General Population: A Longitudinal Analysis and a Negative Control...

Statins can increase arterial calcification by directly affecting the vessel wall. Insufficient evidence on long-term effects may reflect short follow-up and confounding by indication. We used long-term population-based cohort data to examine the effect of statin use, cumulative duration, and duration-response association with calcification volume across multiple key arteries. The robustness of the findings was evaluated using a negative control population.

Rotterdam Study participants with computed tomography (CT)-assessed calcification volume in coronary arteries (CAC), aortic arch (AAC), and extra/intracranial carotid arteries (ECAC, ICAC) were linked to pharmacy records (n = 2,399) to identify statin prescriptions; those with repeat CT (after 13.6 years) were included as follow-up (n= 815). Associations were estimated using multivariable linear and mixed regressions with varying exposure and covariate specifications, with calcification volume as the outcome. The negative control compared associations between statin transporter function and calcification in statin users vs. never-users.

Cumulative duration until second CT (n = 588, mean 8 years) was statistically significantly associated with increased CAC (β = 0.02; 95% CI 0.01-0.03), AAC (0.01; 0.00-0.02), and ECAC (0.02; 0.01-0.03), but not ICAC. Effect sizes increased across quartiles of duration, reaching statistical significance in the highest quartile. In the negative control analysis, decreasing statin transporter function among statin users was generally associated with higher effect estimates for calcification volume, with statistically significant trends in CAC and ECAC; no such trend was observed in never-users.

Prolonged statin use is associated with increased arterial calcification, most notably in CAC and ECAC, and least in ICAC, with the negative control supporting the robustness of these findings. Future studies using causal inference approaches, such as target trial emulation in large longitudinal datasets with detailed information on time-varying statin exposure and cardiovascular risk factors and longitudinal assessment of arterial calcification, are needed to strengthen causal interpretation.