Cholesterol lowering drugs and risk of age related maculopathy: prospective cohort study with cumulative exposure measurement

Participants, methods, and results

This investigation was part of the Rotterdam study, a population based cohort study of people aged 55 years and more. After the baseline phase from 1990 to 1993, two follow up examinations were performed at mean intervals of 2 and 6.5 years. Of all the subjects at risk of age related maculopathy, 4822 (83%) participated at follow up. A diagnosis of age related maculopathy was based on stereoscopic colour fundus transparencies graded according to the international classification system.4 The incidence of the disorder was defined as the development of soft distinct drusen with pigmentary irregularities, indistinct drusen, or the end stages of atrophic or neovascular age related macular degeneration.

A register of prescriptions filled by local pharmacies provided continuous data on use of cholesterol lowering drugs. These data were available for 99% of the cohort from 1 January 1991 onwards. We used Cox proportional hazards regression analysis to calculate hazard ratios, with age in days as the time axis to ensure optimal controlling for age. Cumulative exposure to drugs was represented as a time dependent covariate and was analysed both as a dichotomous and a categorical variable. The model compared each incident case of age related maculopathy with all subjects in the cohort who were alive and free of the disorder at the age when the case of maculopathy was diagnosed.5

During 26 781 person years of follow up, 457 patients used cholesterol lowering drugs for one or more days, and 419 cases of incident age related maculopathy were observed. Use of cholesterol lowering drugs at any time, defined as a binary variable, was not associated with the incidence of age related maculopathy (hazard ratio 1.0 (95% confidence interval 0.7 to 1.5)). Compared with patients who had never used cholesterol lowering drugs, cumulative exposure for less than one month, for one month to a year, or for more than a year did not have a protective effect on the risk of maculopathy (see table). Additional adjustment for body mass index, hypertension, smoking, and peripheral arterial disease (ankle:arm index <0.9) did not change the association. When we performed the same analysis with progression of age related maculopathy as the outcome variable, we obtained the same results.

Comment

Exposure to cholesterol lowering drugs did not change patients' risk of age related maculopathy. In contrast with the studies that reported a protective effect, we used a prospective design and assessed drug use by means of data registered by pharmacies. This minimised potential selection and information bias, and our data provided quantitative information for each patient's cumulative exposure to drugs. This prevented misclassification of the duration of drug use. Even though the total number of participants was high, the number of subjects using cholesterol lowering drugs who developed age related maculopathy was low, possibly leading to a type II error. With a two sided α of 0.05, we had a power of 80% to show a relative risk of 0.7 or lower. The fact that we did not find an association between cholesterol lowering drugs and age related maculopathy makes a protective effect of statins unlikely.