Prevalence of dyslipidemia among HIV-infected patients using first-line highly active antiretroviral therapy in Southern Ethiopia: a cross-sectional comparative group study

Background Data on lipid profile abnormalities among patients receiving highly active antiretroviral treatment in Ethiopia are very limited. The aim of this study was to determine the prevalence of dyslipidemia and characteristics of lipid profiles among patients living with human immunodeficiency virus (HIV) using first-line highly active antiretroviral therapy (HAART) in Southern Ethiopia. Methods This cross sectional comparative group study was conducted between March and May 2012, and included 113 HIV infected patients treated for a minimum of one year with first-line HAART regimens that included Efavirenz and Nevirapine (HAART group) and others 113 who had never received HAART (pre-HAART group). Serum lipid profiles were determined after overnight fasting and dyslipidemia was assessed according to the United State National Cholesterol Education program-III guideline. For statistical analysis Chi-square, student’s t-test, and logistic regression were used using Statistical Package for Social Sciences (SPSS) Version 20. Result Ninety-three (82.3%) of HAART and 87 (76.9%) pre-HAART patients had at least one laboratory abnormality, which is compatible with a diagnosis of dyslipidemia. Total cholesterol ≥ 200 mg/dl occurred in 43.4% of HAART and 15.9% pre-HAART patients (p=<0.0001), whereas HDL-cholesterol below 40 mg/dl occurred in 43.4% and in 63.7% respectively, (p=0.002). The LDL-cholesterol ≥ 130 mg/dl occurred in 33.6% of HAART and 15% pre-HAART patients (p=0.001), while triglycerides ≥ 150 mg/dl occurred in 55.8% and 31.0% respectively, (p=0.001). Receiving of HAART was significantly and positively associated with raised total cholesterol, LDL-cholesterol, and triglycerides. The adjusted odds ratio (95% CI) of HAART-treated vs. pre-HAART was 3.80 (1.34-6.55) for total cholesterol ≥ 200 mg/dl; 2.64 (1.31-5.32) for LDL- cholesterol ≥ 130 mg/dl and 2.50 (1.41-4.42) for triglycerides ≥150 mg/dl. Conclusion Use of first-line antiretroviral therapy regimens that contain Efavirenz and Nevirapine were associated with raised total cholesterol, LDL-cholesterol, and triglycerides, an established atherogenic lipid profiles. Lipid profiles should be performed at baseline before commencement of antiretroviral therapy and then periodically through treatment follow-up to monitor any rising trends.


Introduction
In 2011, an estimated 34 million people were living with human immunodeficiency virus /acquired immunodeficiency syndrome (HIV/AIDS) worldwide; of them 22.9 million were living in Sub-Saharan Africa. About 1.2 million people were estimated to be living with HIV in Ethiopia [1].The introduction of highly active antiretroviral therapy (HAART) has led to a marked reduction in AIDSrelated morbidity and mortality [2]. Since its introduction patients have started to live longer, however co-morbid problems have been emerged. Dyslipidemia, insulin resistance, and diabetes are some of metabolic complications of long-term use of HAART [3]. The characteristics of dyslipidemia in HIV-infected patients receiving HAART includes, elevated level of total cholesterol (TC), LDLcholesterol (LDL-c), triglycerides (TG), and decreased HDL-cholesterol (HDL-c), include with severe hypertriglyceridemia in some patients [4]. Some antiretroviral drugs, such as stavudine (d4T) [5], and protease inhibitors (PIs) [6], increase the blood levels of TC, LDL-c, and TGs with variable effects on levels of HDL-c. Nevirapine (NVP) use is associated with increases in LDL-c [7], whereas increases in TC and TG are observed with use of efavirenz (EFV), particularly with longer duration of therapy [8]. Therefore, the use of HAART raises concerning metabolic disorders and cardiovascular risk in HIV infected patients who now present an extended life expectancy [9]. The prevalence of dyslipidemia in resource-limited settings has not been well characterized and current World Health Organization (WHO) antiretroviral therapy (ART) guidelines do not include a recommendation that lipid monitoring should be conducted in patients receiving first-line HAART [10]. In addition, evidences in support of dyslipidemia associated with first-line HAART that include EFV and NVP in Sub-Sahara African countries are scarce [11,12].The aim of the present study was to determine the prevalence of dyslipidemia and characteristics of lipid profiles among people living with HIV infection receiving first-line HAART in Southern Ethiopia.

Study setting and study population
This was a cross-sectional comparative group study. Subjects were recruited between March 2012 and May 2012 at ART clinic of Hawassa University Referral Hospital. Two groups of participants were selected for this study. The first group included HIV-infected individuals who had been receiving WHO recommended first-line HAART for a minimum of one year (HAART group). Participants used first-line HAART regimens that included nucleoside reverse transcriptase inhibitors (NRTIs): lamivudine (3TC), ZDV, or d4T, and non-nucleoside reverse transcriptase inhibitors (NNRTIs): NVP or EFV. Patients who had had their therapy regimens changed during follow-up were not included. The second group was pretested and confirmed HIV-positive individuals who were not yet getting HAART. All participants included were ≥18 years of age and HAART treated group to have a good ART adherence (adherence rate ≥ 95%). A good adherence is defined by missing < 2 dose of 30 doses or < 3 dose of 60 doses; and it was adopted from Ethiopian Federal Ministry of Health, HIV Care/ART follow-up form. Participants receiving lipid altering therapies, pregnant women, known diabetes mellitus patients and renal failures were excluded.

Assessments and measurements
For all participants, data were collected on the sociodemographic informations along with body mass index, medical history including diabetes mellitus, renal failures, use of drugs that alter lipid profiles, and current use of anti-TB drugs. CD4 + lymphocyte count was done by using flow cytometry instrument (Becton Dickinson, CA, USA). Blood sample was collected from each participant after a 12 hour overnight fast and centrifuged at 3000 cycles/ minute, and then serum was obtained for lipid profiles. Fasting serum TC and TG were assayed by enzymatic method (Linear chemicals, Montgat, Spain). Serum HDL-c determined by enzymatic method after a selective precipitation of apolipoprotein containing lipoproteins (very low density lipoprotein, LDL-c, and apolipoprotein a (Lpa) by phosphotungstic acid/MgCl 2 (Linear chemicals, Montgat, Spain). LDL-c was assessed by enzymatic method after precipitation of LDL by polyvinyl sulfate (Linear chemicals, Montgat, Spain) and then LDL-c was calculated by subtracting the supernatant cholesterol fractions from the TC of the sample. And also the TC/HDL-c ratio was calculated.

Statistical analysis
The estimation of the sample size was based on the difference between proportions and the following parameters were considered: alpha (α) = 5%, beta (β) = 10 % and power = 90%. Concerning dyslipidemia, we considered frequencies of LDL-c ≥130 mg/dl, which was 21% for the group of individuals with HIV/AIDS without HAART treatment [14]; and 40.8% for patients receiving HAART treatment [15]. With these parameters, the sample size calculated was 230 participants (115 pre-HAART and 115 HAART-treated patients).Data entry and Database management was completed using EPI-INFO 2002. Statistical analyses were done using Statistical Package for Social Sciences (SPSS) Version 20. Chi-square test was used to evaluate differences in frequency distribution. Student's t-test was applied to assess differences between two means. Logistic regression was also used to determine the association of independent factors (age, gender, and treatment status, type of treatment, current CD4 counts, and history of smoking, body mass index and current use of anti-TB drugs) with abnormal level of each lipid profile. P-value less than 0.05 considered as statistically significant at 95% confidence interval (CI).

Ethical consideration
The study was approved by the ethical review committee of School of Biomedical and Laboratory Sciences, University of Gondar. All individuals participated in the study provided their informed consent.

Dyslipidemia and characteristics of lipid profiles
Ninety-three (82.3%) of HAART treated and 87 (76.9%) pre-HAART patients had at least one laboratory abnormality, which is compatible with a diagnosis of dyslipidemia ( Figure 1). The prevalence of TC ≥ 200 mg/dl, LDL-c ≥130 mg/dl and TG ≥150 mg/dl were significantly higher in HAART group when compared to pre-HAART group. The prevalence of HDL-c below 40 mg/dl was significantly higher in pre-HAART group when compared to those on HAART. However, there was no significant difference in TC/HDL-c ratio between HAART treated and pre-HAART patients ( Table 2).

Dyslipidemia and first-line HAART
No significant difference observed in lipid profile derangements between patients receiving ZDV when compared to those on d4T; and patients treated with EFV when compared to those treated with NVP (Table 3 and Table 4). In addition, no significant difference observed in lipid derangements between HAART treated males and females; however, the proportion of raised LDL-c is slightly higher in females when compared to males (Figure 2).

Dyslipidemia and risk factors
Multivariate analysis was adjusted for potential confounding factors such as age, sex, current history of smoking, past history of smoking, BMI, CD4+cells, current use of anti-TB drugs, and HAART treatment. Receiving of HAART was significantly and positively associated with raised TC, LDL-c, and TG. The adjusted odds ratio (95% CI) of HAART-treated vs. pre-HAART was 3.80 (1.34-6.55) for TC ≥200 mg/dl; 2.64 (1.31-5.32)  (Table 5).
Univariate and multivariate analysis were also applied to assess independent risk factors for each abnormal lipid profile in the subgroup of 113 HAART treated patients. In these both models, age was the risk factor of raised TC and decreased HDL-c; and BMI was the risk factor of raised TC in HAART treated patients (Table 6).

Discussion
The aim of this cross-sectional study carried out in a resource limited East African setting was to assess the prevalence of dyslipidemia (lipid profile derangements) among HIV-infected patients receiving first-line HAART that include NNRTIs. We found that CD4 cell count in the HAART group was significantly different from the pre-HAART group, thus showing the effects of HAART in improving the immunological properties of the subjects.
In the present study, the majority of HAART patients (82.3%) had at least one laboratory abnormality, which is compatible with a diagnosis of dyslipidemia   according to NCEP-ATP III criteria. We found that the proportions of raised TC, LDL-c and TG were significantly higher in HAART group when compared to pre-HAART group. The proportion of HDL-c below 40 mg/dl was significantly higher in the pre-HAART group when compared to those on HAART, thus indicating the effects of HAART treatment in raising the lipid profiles, in addition to the restoration of health. However, there was no significant difference between treated group and non-treated group regarding TC/HDL-c ratio, this because increases in TC were counterbalanced by HDL-c improvement. The described raised lipid profiles (TC, LDL-c and TG), are atherogenic [13,16,17], and suggest a potential risk for the development of cardiovascular diseases in a significant proportion of HIV-infected patients in the near future.
The association between lipid derangements and HAART has been mainly depicted for regimens that contain PIs [6,18]. In addition, NNRTIs derange lipid profiles during therapy [7,8]. However, supportive evidences are very scarce in Sub-Sahara African countries concerning lipid derangements in patients receiving NNRTs including regimens [11,12]. We found that the prevalence of raised TC in HAART group was (43.4%). This prevalence rate is higher than the study reported from a resource poor West African setting [19]. The prevalence rate in this study was 4.5%. However, in this study, the cutoff limit was >210 mg/dl; and this value is higher than that of NCEP-ATP III guidelines, as used in our study. Based on HDL-c cut off value, the prevalence of HDL-c in our treatment group was 43.4%. This is comparable with the prevalence rate reported from Spain [20]. However, it is higher than the prevalence reported from rural South India [21]. The prevalence rate in this study was 24.8%. The prevalence of raised LDL-c in our HAART group was 33.6%. It is similar with the prevalence reported from India [21]. We found that the prevalence of raised TG in HAART group was 55.8%. This is not in line with the prevalence reported from Cameroon [14], Kenya [15], and India [21]. The prevalence rates in these three studies were respectively, 43.5%, 22.5%, and 42.1%. However, there are suggestions that the magnitude of lipid profile derangements induced by first-line HAART could show variation with duration of treatment, across populations and setting.
The reports from rural South India and Brazil indicated that the proportion of lipid profile derangements  were higher among patients received d4T when compared to other NRTIs [21,22]. We found no difference in lipid derangements when patients received d4T compared to those received ZDV. This finding is similar with the reports from Uganda, and Cameroon [11,14]. The randomized trial report of 2 non-nucleosides (2NN) indicated that patients on NVP group had significantly improved HDL-c concentration and had relatively low lipid profile derangements when compared to those on EFV [23]. In line with the reports from rural Uganda [11], Cameroon [14] and India [24], we found no significant difference in lipid profile of patients on EFV compared to those on NVP. In the present study, the raised TC and LDL-c were significantly and positively associated with the use of HAART treatment, and the findings are in line with the study conducted in Cameroon [14].

Limitations of the study
Comprehensive cardiovascular risk stratifications were not assessed in this study. However, the increased risk of cardiovascular diseases associated with described lipid derangements is well known [25][26][27], and long term use of first-line HAART may have an impact on cardiovascular system. The other limitations were the crosssectional nature of the study, small number of male participants, and lack of HIV negative controls.

Conclusion
Our study indicates HIV-infected patients receiving WHO-recommended first-line HAART that include Efavirenz and Nevirapine have a high prevalence of lipid profile derangements when compared to those nontreated HIV-infected patients. Uses of first-line HAART regimens are significantly associated with atherogenic lipid profiles. Therefore, the findings indicate that the need to assess lipid profiles at baseline before initiation of HAART treatment and lipid profile monitoring during therapy to monitor any rising trends. Additionally, the results also recommend implementation of wellcontrolled cohort studies for the evaluation of long-term effects of HAART treatment on lipid profiles.

Competing interests
We announce that we have no any competing interests.
Authors' contributions AT generated and designed the study, performed analysis and interpretation of data including with manuscript drafting, ZA, HA and SB assisted with the design, interpretation of data and the critical appraisal of the manuscript. All authors read and agreed the final manuscript.  HAART, highly active antiretroviral therapy therapy; UOR, un adjusted odds ratio; AOR, adjusted odds ratio; CI, confidence interval; TC, total cholesterol; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol; TG, triglyceride; d4T Stavudine; BMI, body mass index; EFV, Efavirenz; * p-value < 0.05; ** p-value < 0.01. Reference category: BMI ≤ 25 Kg/m 2 , Females, Age ≤ 40 years, no smoking, Nevirapine group, Zidovudine group; HAART treatment >2years.