Background
Equipoise exists regarding the effect of statins on cognitive function, with some observational studies suggesting harm and others suggesting benefit. Limited data among people with HIV (PWH) may be biased by indication for statin prescription.
Methods
To assess statin effects on neurocognitive function among PWH, we leveraged data from participants co-enrolled in REPRIEVE (RCT of statin therapy vs. placebo) and HAILO (observational study including serial assessments of neurocognitive function). All co-enrolled participants with ≥1 measure of neurocognitive function before (median 7) and after enrollment (median 3) into REPRIEVE were included. Neurological function was determined by NPZ-4, the average of the Z scores from: Hopkins Verbal Learning Test Revised (HVLT-R), Trailmaking A and B (TrA, TrB), and Digit Symbol Test (DST) every 48 weeks. NPZ-4 and its component tests were analyzed with general estimating equation models using an exchangeable working correlation structure, with trajectories before and after REPRIEVE randomization estimated separately.
Results
Of 181 co-enrolled participants (pitavastatin 88, placebo 93), the mean (SD) age was 50.3 (5.0) years, 19% female at birth, 31% Black, 23% Hispanic, and 24% had current cigarette use. Characteristics were balanced between randomized groups. Changes in overall and individual neurocognitive scores were small, not meeting threshold for clinically relevant (<0.5), and similar in the placebo and pitavastatin arms (Table). In bivariate analyses adjusting for demographics, comorbidities, and cardiovascular risk factors, the effect sizes were minimally changed and remained non-significant. Although subgroup analyses were limited by a small sample size, we observed trends towards improved TrA in participants with baseline impairment randomized to pitavastatin (Z score [95% CI]: 0.075 [0.024, 0.127]) vs placebo (-0.05 [-0.123, 0.023]), p=0.018 and towards worsened NPZ-4 in females randomized to pitavastatin (-0.061 [-0.133,0.011]) vs placebo (0.033 [-0.023, 0.089]), p=0.068 that similarly did not reach threshold for clinical relevance. Other subgroup effects were minimal and not statistically or clinically significant.
Conclusions
We found no evidence to suggest a detrimental effect of pitavastatin on a limited battery of neurocognitive assessments among people with HIV at low to moderate cardiovascular risk, even among those with baseline impairment.
Table or Figure
