Abstract
The prevalence of low HDL-C levels in an HIV population and its related factors was investigated. We undertook a multicenter, cross-sectional study of all HIV patients on regular follow-up in five hospitals (Southern Spain). A physical examination and fasting laboratory analysis were performed and a questionnaire about cardiovascular risk factors was provided. One thousand and seventy-two patients were included, 43.8% of whom had low HDL-C levels. The prevalence of low HDL-C was higher among patients diagnosed with AIDS, those not on antiretroviral therapy, those with a detectable HIV viral load, those with CD4 cell counts ≤350 cells/μl, smokers, and those with hypertriglyceridemia. For patients on antiretroviral therapy, the prevalence of low HDL-C was higher for those on protease inhibitors than those taking nonnucleoside reverse transcriptase inhibitors. In the multivariate analysis, low HDL-C levels were associated with tobacco use (OR 1.37, 95% CI 1.04–1.8; p = 0.04), hypertriglyceridemia (OR 2.94, 95% CI 2.2–3.8; p < 0.00001), CD4 cells count ≤350 cells/μl (OR 1.74, 95% CI 1.2–2.3; p < 0.0001), and a detectable HIV viral load (OR 1.85, 95% CI 1.3–2.5; p < 0.0001). The immunological and virological conditions, in addition to traditional cardiovascular risk factors such as tobacco use and hypertriglyceridemia, affect HDL-C levels in HIV-infected patients. For patients on antiretroviral therapy, the use of protease inhibitors is associated with a higher probability of low levels of HDL-C. Although it is not clear if the higher HDL-C levels associated with antiretroviral use are surrogates for decreased cardiovascular disease risk, this may be another reason to start antiretroviral therapy earlier.
A
To investigate the prevalence of low HDL-C levels in the HIV population and factors associated with this dyslipidemia we undertook a multicenter, cross-sectional study of all adult HIV patients on regular follow-up in five Andalusian hospitals (Southern Spain) who attended the consult in March–August 2007. Patients suffering an AIDS-defining disease in the past 3 months, patients with ascites, pregnant or breastfeeding women, and those with no available data were excluded. A physical examination, a questionnaire about CVR factors, 9 fasting plasma lipid and glucose profiles, and CD4 cell count and HIV viral load (VL) were performed. LDL-C was calculated by the Friedewald equation. 10 Nadir CD4 cell count and baseline HIV VL were collected from clinical reports. Low HDL-C was defined as <40 mg/dl in men and <50 mg/dl in women. The chi-square and Fisher's exact tests were used to analyze categorical variables, the Student's t-test for continuous variables with a normal distribution, and the Mann–Whitney U test for variables without a normal distribution. Multivariate analysis was performed by multiple conditional logistic regression. Statistical analyses were performed using SPSS 11.0
Among 1166 patients attending clinical visits during the period of the study, 1072 were included for the analysis. Levels of HDL-C were low in 469 (43.7%) subjects. Mean HDL-C was correlated positively with CD4 cells (r = 0.159; p < 0.001) and negatively with HIV VL (r = –0.234; p < 0.001) and triglycerides (r = –0.288; p < 0.001). The epidemiological and clinical features of the 1072 patients and factors related to low HDL-C are shown in Table 1. Tobacco use (OR 1.37, 95% CI 1.04–1.8; p = 0.04), hypertriglyceridemia (OR 2.94, 95% CI 2.2–3.8; p < 0.001), CD4 cell count ≤350 cells/μl (OR 1.74, 95% CI 1.2–2.3; p < 0.001), and HIV VL >50 copies/ml (OR 1.85, 95% CI 1.3–2.5; p < 0.001) maintained the association with low HDL-C in the multivariate analysis.
HMX, homosexual; HTX, heterosexual; IDU, intravenous drug use; HAART, highly active antiretroviral therapy; VL, viral load; PI, protease inhibitor; NNRTI, nonnucleoside reverse transcriptase inhibitor; CVD, cardiovascular disease. Quantitative parameters, mean (IQR). Qualitative parameters, absolute number (%).
The prevalence of low HDL-C in our cohort was high; it was similar to that from other series of HIV patients 11 –13 and was higher than in the general population. 14 –16 A worse immunovirological condition, tobacco use, and hypertrigliceridemia were associated with low HDL-C.
Most available data about the prevalence of low HDL-C in the HIV population are from studies evaluating the metabolic syndrome, with ranges from 33.9% to 54%. 11,12 Only one prior study has specifically analyzed low HDL-C in HIV patients, finding a prevalence of 44.7%, similar as ours. 13 The prevalence of low HDL-C in the general population varies depending on demographic and geographic factors, habits, and comorbidities, and ranges from 7–9% in France to 17–38% in the United States. 14,15 In a large Spanish study that included working people, the prevalence of low HDL-C was 25.6%, nearly half that in our series. 16
We observed a direct relationship between HDL-C and CD4 cell count and an inverse relationship with HIV VL, which in addition to higher levels of HDL-C in patients with suppressed VL and in those on HAART suggests that HIV infection itself plays a role in the development of this dyslipidemia. This fact has been observed not only in naive subjects but also in those with HAART interruption, whose HDL-C levels decrease significantly. 4,6,7,17,18 Bernal et al. 13 also found an association between HDL-C and HIV VL, but not with CD4 cells, that we found and that reinforces the hypothesis of the role of HIV in this dyslipidemia.
On the other hand, HAART, mainly nonnucleoside reverse transcriptase inhibitors (NNRTIs), could be considered as a protective factor against low HDL-C. Total cholesterol and its subfractions increase after starting HAART, probably as a normalization of the lipid profile associated with clinical and immunovirological improvement. 6,19 –21 NNRTIs, particularly nevirapine, have been associated with a significant increase of HDL-C in naive and in protease inhibitor (PI)-exposed patients. 22,23 Although the most common mechanism responsible for increasing HDL-C is attributable to decreased HDL-C catabolism, it has been recently observed that nevirapine also stimulates apolipoprotein A1 production. 24 As previously described, 13,21 in our cohort, among patients on HAART, the prevalence of low HDL-C was higher for those on PIs than those taking NNRTIs.
There is a well-established inverse association between HDL-C and fasting triglycerides 25 as well as with tobacco use. 26 Hypertriglyceridemia is the most common dyslipidemia in HIV patients, due to the infection itself and also as an adverse event of some antiretrovirals. 1 –3,12,17 –21 On the other hand, cigarette smoking is more prevalent in the HIV population than in the general population. 1,2,12 . Therefore both factors may also contribute to the high prevalence of low HDL-C concentrations in these subjects.
In summary, the immunovirological condition in addition to traditional CVR factors, such as tobacco use and hypertriglyceridemia, affect HDL-C levels in HIV patients. Although it is not clear if the higher HDL-C levels associated with HAART are surrogates for decreased CVR, this may be another reason to start HAART earlier.
Footnotes
Acknowledgments
Members of the study group: S. Puerta, M. Gallego, J. Ruiz, E. Niño, M. Márquez, R. Palacios and J. Santos, Hosp. Virgen de la Victoria. F. Orihuela, M. Castaño, F. Jiménez-Ońate, M. Delgado and JD. Colmenero, Hosp. Carlos Haya. J. Olalla, A. del Arco and J. de la Torre, Hosp. Costa del Sol. J. Roldan, Hosp. de Antequera. M. Grana, Hosp. de la Serranía.
Author Disclosure Statement
No competing financial interests exist.