Abstract
BACKGROUND: Switching to an integrase strand transfer inhibitor (INSTI) during the menopausal transition has been associated with accelerated increases in visceral obesity, a risk factor for insulin resistance. Whether switching to an INSTI modifies the association of HIV and menopause with insulin resistance is unknown. METHODS: From 2006 to 2019, 389 nonpregnant women with HIV (WWH) [133 who switched to an INSTI (INSTI+); 256 who did not switch (INSTI-)] and 334 women without HIV from the Women's Interagency HIV study without diabetes or hepatitis C virus were included in the analysis. Mixed effect models evaluated the change in insulin resistance estimated through log HOMA-IR by HIV status by menopausal phase. We then compared trajectories by INSTI group. Menopausal phase was determined by anti-Müllerian hormone, a biomarker of ovarian reserve. RESULTS: Compared to women without HIV, INSTI+ WWH in premenopause had nonstatistically significant faster annual increases in HOMA-IR [difference in slope: 7.03%; 95% confidence interval (CI): -4.99 to 20.58] whereas INSTI- WWH had nonstatistically significantly faster annual decreases [-1.01% (95% CI: -7.34 to 5.75)]. In late perimenopause, INSTI+ and INSTI- had 4.87% (95% CI: -3.59 to 14.06) and 4.38% (95%CI: -3.10 to 12.44) nonstatistically significantly faster annual increases in HOMA-IR, respectively. In menopause, INSTI+ and INSTI- WWH had 9.18% (95% CI: 1.60 to 17.33) and 11.28% (95% CI: 3.27 to 19.91) statistically significant faster annual increases in HOMA-IR than women without HIV. There was no statistically significant difference between INSTI+ and INSTI- in any menopausal phase. CONCLUSION: Regardless of switching to an INSTI or not, WWH in late perimenopause and menopause have faster increases in insulin resistance when compared to women without HIV. Diabetes screening and prevention in midlife WWH is imperative.