Do combined oral contraceptives have long-term effects? Little evidence of an enduring effect on cognitive function in former users

The use of combined oral contraceptives (COCs) is associated with subtle differences in brain activity and in certain cognitive, perceptual, or affective processes among current COC users compared with non-users. Whether any differences persist after the COCs are discontinued is essentially unknown. In a retrospective analysis of a cognitive dataset from healthy adults (N = 221, M_age = 23.78 yrs), we asked whether any residual effects of COC use could be identified in former users (who had discontinued their COCs ∼30 mo prior) on standardized tests of verbal fluency and visuospatial function. A detailed reproductive history was available, including specifics of any past COC use. Cognition was evaluated during the early follicular phase of the menstrual cycle when the ovaries are quiescent and hormone production is lowest. Based on their reproductive history, participants were classified as former users of COCs (N = 86, mean duration of COC use = 29.89 mo), never-users (N = 106, no past use of hormonal contraceptives), or current users (N = 29, currently using a COC, mean duration of COC use = 29.72 mo). All COCs contained either levonorgestrel or a progestin from the norethindrone family, and 30–35 μg/day of ethinyl estradiol. Analysis of covariance (ANCOVA) contrasting cognitive performance in the 3 groups showed that despite the use of androgenic progestins and an estradiol dose that was relatively high, there was no evidence of a difference between former-users and never-users on any of the cognitive tests. Current users scored lower than former or never users on a conventional visuospatial task (Space Relations). Partial correlations controlling for age, estimated IQ, and pregnancy history revealed that duration of past COC use in former users was significantly associated with verbal fluency on one of two phonemic fluency tasks administered. A longer duration of COC use was associated with poorer word generation ability. This dataset produced little evidence of any enduring effects of COC use, however a duration-dependent association between former COC use and verbal fluency reinforces a similar trend reported previously among current users.