Rate of Dyslipidemia Higher Among Women Living With HIV

A Comparison of Metabolic and Cardiovascular Health in a Cohort to Study Aging in HIV

EAB Russell; AYK Albert; HCF Côté; AYY Hsieh; A Nesbitt; AR Campbell; EJ Maan; J Brophy; N Pick; MCM Murray


HIV Medicine. 2020;21(7):418-428. 

In This Article


Two hundred and eighty-nine women, including 156 WLWH and 133 HIV-negative controls, were enrolled in the CARMA-ENDO study (Table 2). Within this sample, 49% identified as white/Caucasian, 15% as ACB, and 19% as Indigenous. The median age and body mass index (BMI) of participants were similar in the WLWH and HIV-negative groups (median age 46.2 and 43.5 years in HIV-negative women and WLWH, respectively). However, the distribution of age in WLWH and HIV-negative women differed (Figures S1 and S2). Current or past HCV infection and current or past smoking were more common in WLWH. High school completion, annual income > 15 000 Canadian dollars/year, menopause, and current or past alcohol use were more common in the HIV-negative group. There was no significant difference in mean unadjusted blood mtDNA content between groups, but LTL was significantly shorter in WLWH compared to HIV-negative women (Table 2).

Among WLWH, median time since HIV diagnosis was 15.7 [interquartile range (IQR) 10.9–19.9] years. Median CD4 nadir was 190 (IQR 88–289) cells/μL, and median current CD4 count was 570 (IQR 335–725) cells/μL. Among WLWH, 52.6% had a peak HIV viral load > 100 000 copies/mL. Current antiretroviral regimens included NNRTIs (23.7%), integrase inhibitors (43.6%), PIs (39.7%) and the nucleoside reverse transcriptase inhibitor (NRTI) tenofovir (48.7%). Of WLWH, 144 of 156 were on current cART.

Overall, the prevalences of diabetes, moderate/high-risk FRS, hypertension and metabolic syndrome were similar in WLWH and HIV-negative women. However, WLWH were significantly more likely to have dyslipidaemia (Figure 1).

Figure 1.

Comorbidity prevalence in women living with HIV and negative controls. A Framingham risk score (FRS) ≥ 10% indicates moderate to high 10-year cardiovascular risk.


Similar but very low numbers of WLWH (10 of 156; 6.4%) and HIV-negative women (four of 133; 3%) in this cohort had been diagnosed with diabetes. The prevalence of diabetes in the cohort was too low for further regression analysis (Figure 1).

FRS for Cardiovascular Disease

Among women > 30 years of age with FRS data (n = 126 WLWH, and n = 97 HIV-negative women), 8.7% (11 of 126) of WLWH had FRS in the moderate to high range (i.e. ≥ 10% risk of myocardial infarction in the next 10 years), which was not statistically different from the 5.1% (five of 97) observed among HIV-negative women. Most women in both groups were low risk and very few showed very high risk, with a median FRS of 3.45 (IQR 2.23–6.75; range 0.69–50.18) and 4.12 (IQR 2.08–6.0; range 0.46–27.93) in WLWH and HIV-negative women, respectively (Figure S3).


Among all women, 13% (37 of 289) had hypertension; 15% among WLWH and 10% among HIV-negative women (Figure 1). There was a significant relationship between ethnicity and the odds of hypertension (P = 0.02). Although none of the pairwise comparisons was significant in post hoc Tukey tests, Indigenous and women of other ethnicity tended to have lower odds of hypertension compared to ACB women in the multivariable model (Figure 2, Table 3). Univariately, there was a nonsignificant trend towards higher odds of hypertension among WLWH [OR 1.98; 95% confidence interval (CI) 0.94–4.35; P = 0.07; Figure 2]; the trend remained in the multivariable model [adjusted odds ratio (AOR) 1.95; 95% CI 0.90–4.40; P = 0.09].

Figure 2.

Factors associated with risk of comorbidities among all Children and Women: Antiretrovirals and Markers of Aging (CARMA)-endocrine (ENDO) participants. All regressions were adjusted for age. Shown are the crude odds ratios (ORs) and 95% confidence intervals (CIs) for variables with P-values ≤ 0.10. *P < 0.05, **P < 0.001 in the age-adjusted analysis. All covariates included in the multivariable model are present in this figure. The mtDNA ratio was divided by 10 so that ORs could be interpreted for every increase in mtDNAcn of 10 units. ACB, African/Caribbean/black; HCV, hepatitis C virus; LTL, relative leukocyte telomere length; WLWH, women living with HIV.

Among WLWH, higher current CD4 count was associated with increased risk of hypertension (Figure 3). For every increase in CD4 count of 100 cells/μL, the odds of hypertension increased by approximately 18%.

Figure 3.

HIV-specific factors associated with comorbidities among women living with HIV (WLWH). All regressions were adjusted for age. Shown are the crude odds ratios (ORs) and 95% confidence intervals (CIs) in logistic regressions for variables with P-values ≤ 0.10. *P < 0.05, **P < 0.0001 in the age-adjusted analysis. All covariates included in the multivariable model are present in this figure. ORs for current CD4 count are for increments of 100 cells/μL. PI, protease inhibitor; pVL, plasma viral load; TDF, tenofovir disoproxil fumarate.

Metabolic Syndrome

Of the 289 women in the study, 71 (24.6%) met the criteria for metabolic syndrome, and the proportions affected were similar in WLWH (43 of 156; 27.6%) and HIV-negative women (28 of 133; 21.2%) (OR 1.6; 95% CI 0.90–2.89; P = 0.11; Figures 2 and 3). The odds of metabolic syndrome were increased among all women with a current (OR 2.21; 95% CI 1.14–4.32) or past smoking history (OR 2.22; 95% CI 1.01–4.82), or with lower mtDNA content (OR 0.91; 95% CI 0.83–0.99) (Figure 2).

In WLWH, current PI use and higher CD4 count were associated with a diagnosis of metabolic syndrome (Figure 3). Both remained significant in the multivariate model (CD4 count: AOR 1.41; 95% CI 1.21–1.68; P < 0.0001; PI use: AOR 2.74; 95% CI 1.18–6.61; P = 0.02).


Over half of the women in the study (151 of 289; 52.2%) met the criteria for dyslipidaemia (Table 1). Low HDL cholesterol was the predominant dyslipidaemia criterion met by the women, followed by high triglycerides and diagnosis by a physician (Table S1). A significantly higher proportion of WLWH (62.8%; 98 of 156) than HIV-negative women (39.8%; 53 of 133) had dyslipidaemia (OR 2.57; 95% CI 1.60–4.16; P < 0.0001; Figure 1). Other factors significantly positively associated with dyslipidaemia included smoking (overall P = 0.01) and HCV status (P = 0.01). Alcohol (P = 0.08), LTL (P = 0.07) and mtDNA content (P = 0.09) showed weak (P < 0.10) associations in the univariable analysis (Figure 2). In the multivariable model, HIV status (AOR 2.89; 95% CI 1.69–5.01; P = 0.0001) and smoking (current vs. never: AOR 1.39; 95% CI 0.75–2.62; past vs. never: AOR 2.84; 95% CI 1.33–6.35; P = 0.02) were significantly associated with dyslipidaemia. Alcohol use did not remain significantly associated with dyslipidaemia when HIV status was also included in the model (P = 0.39), suggesting that the marginally significant association univariably may have been attributable to a greater number of alcohol users among the HIV-negative women who were less likely to have dyslipidaemia.