Statins and Cognitive Impairment

Elevated low-density lipoprotein cholesterol levels increase the risk of heart disease and ischemic stroke (1,2). Statins remain the mainstay of lipid-lowering strategies. However, the risks of adverse events are commonly cited by individuals declining and discontinuing statins (3). Whereas statins increase the risk of myopathy, new-onset diabetes mellitus, and hemorrhagic stroke (4), there are 2 important caveats. First, an elegant analysis of the ASCOT-LLA (Anglo-Scandinavian Cardiac Outcomes Trial–Lipid-Lowering Arm) demonstrated that muscle-related adverse effects were increased in statin users only during the unblinded period; perceived risks of muscle symptoms likely led patients and physicians to overestimate these risks (5). Second, the increased risks of diabetes and hemorrhagic stroke are likely outweighed by the significant absolute risk reduction in ischemic events, even in locales such as China where the risk of hypertension-related hemorrhagic stroke is high (2,6). The cognitive effects of statins have been addressed in numerous studies, including several randomized controlled trials (7,8). Though these studies did not find an increased risk of cognitive impairment with statins, concerns about the lack of generalizability to older individuals–who are most at risk for cognitive impairment–persisted (9).

In this issue of the Journal, Samaras et al. (10) report on the association between statin use and cognitive outcomes in older individuals in the MAS (Sydney Memory and Aging Study). The investigators performed longitudinal analyses of prospective observational data that were collected starting in 2005. After excluding those with prevalent dementia, 1,037 noninstitutionalized individuals age 70 to 90 years were included. Participants underwent baseline medical assessments and neuropsychological testing, which was repeated every 2 years over a 6-year period. Neuropsychological assessments were remarkably rigorous; 12 standardized cognitive tests covered 5 cognitive domains. A subsample of 529 participants had brain magnetic resonance imaging for volumetric assessments, including brain regions relevant to memory and cognition, such as hippocampal and parahippocampal regions. Participants were categorized as statin never users, ever users, and continuous users–participants on a statin for the entire study period. The investigators also separately categorized participants who were not taking a statin at baseline, but who initiated statin use while under observation.

The study sample was homogenous with 98% of participants being Caucasian. Among the total sample, 395 were never users and 642 were ever users, of whom 68% had continuous use. Statin therapy was initiated in 99 participants during observation. Statin ever users were slightly younger and had more vascular disease and risk factors, but baseline cognitive performance was similar among statin user groups. Complete 6-year follow-up data were available for only 55% of participants. Using linear mixed models, the investigators attempted to account for bias related to attrition while adjusting for age, education, prevalent cardiovascular and cerebrovascular disease, hypertension, diabetes, and apolipoprotein E (APOE) ε4 allele status. They demonstrated that the overall rate of cognitive decline was not different in statin users compared with in never users. They also showed that statin initiation was associated with a reduction in the rate of global cognitive decline, although this was not significant after correcting for multiple comparisons. However, in exploratory analyses, statin initiation significantly attenuated decline in a memory-domain test. The investigators also performed tests of interaction to determine whether statin use moderated trajectories of cognitive decline associated with recognized risk factors. Statin use, especially continuous use, was associated with a reduced rate of cognitive decline among normotensive individuals. Conversely, APOE ε4 carriers at higher risk of cognitive decline had their rate of cognitive decline slowed by statin use. Important effects on brain imaging parameters did not emerge, which is not surprising given that only a subsample underwent imaging and that imaging was repeated after only 2 years.

Therefore, contrary to popular concern, statin use was not associated with cognitive decline in this observational study. Rather, and provocatively, statin initiation was associated with a favorably altered cognitive trajectory, including among high-risk APOE ε4 carriers. The major strengths of this study are the rigorous, longitudinally repeated neuropsychological assessments and the well-characterized cohort. Although their results are consistent with randomized trials (7,8) and a meta-analysis (11) that demonstrated the cognitive safety of statins, there are several important limitations. First, with complete data available for only 55% of participants, the risk of bias is high. Second, the overall sample size was small, especially for new statin users. Third, there is likely unmeasured confounding related to statin use, and it is unclear that linear mixed modeling is enough to account for unmeasured differences between individuals never taking and continuously taking statins, for example.

Whereas these limitations may temper inferences about the neuroprotective effect of statins, the investigators’ finding that statin use blunted the rate of cognitive decline among high-risk APOE ε4 carriers merits further consideration (10). Two recent meta-analyses also demonstrated a dementia risk reduction with statins, also with evidence of a dose-response relationship (12,13). Interestingly, the risk reduction was seen for Alzheimer type dementia but not vascular dementia (12). Carriers of APOE ε4 have a high burden of amyloid-β and are at high risk of Alzheimer disease, and, although there is no consensus (14,15), some studies have found statins to improve amyloid clearance and reduce its accumulation (16-18). Conversely, the possibility of toxicity to astrocytes has also been suggested (18). Taken together, these data support the view (4) that worries about cognitive impairment should not limit statin use and raise the possibility that statins may favorably alter cognitive trajectories in a group of elderly subjects at high risk of Alzheimer disease. The latter finding is of great interest and requires additional mechanistic clarification and rigorous clinical validation.