|dc.description.abstract||A double burden of non-communicable -and communicable diseases exists in South Africa which include high prevalence of cardiovascular disease (CVD) and human immunodeficiency virus (HIV) infection. Coronary heart disease (CHD), a sub-category of CVD, affects more African-American individuals than white individuals. The high prevalence of CVD and HIV decreases the quality of life for those living with HIV. However, being treated with antiretroviral drugs are known to increase the life expectancy for the HIV-infected population. Those living with HIV presented with higher prevalence of arterial stiffness, atherosclerosis, kidney disease and left ventricular hypertrophy (LVH). Despite the advantages of antiretroviral therapy (ART) use, it was shown to have worsen CVD risk for the HIV-infected population. The need for risk assessment in the African HIV-infected population burdened by both coronary heart disease (CHD) and CVD are important.
The aim of this study was to determine the risk scores of 10-year CHD and CVD by making use of the Framingham -and Reynolds risk score models, respectively, and to determine associations of the risk score models with measures of end-organ damage.
This study is embedded in the South African arm of the Prospective Urban and Rural Epidemiological (PURE) study and included African participants from the North-West Province of South Africa who were infected with HIV for at least 10-years. The PURE-South African study consisted of baseline (2005) and follow-up (2015) data. A number of 2010 individuals participated in the PURE study during baseline, where 322 participants were newly identified as being infected with HIV. Ten years later 100 of the 322 HIV-infected participants took part in the follow-up study of which 29 participants were excluded due to incomplete data sets. A number of 71 HIV-infected participants remained and they were matched to 71 HIV-free controls according to age, sex and locality.
Anthropometric measurements included height, weight and waist circumference (WC), followed by the calculation of body mass index (BMI). Regarding cardiovascular measurements, systolic and diastolic blood pressure (SBP and DBP), carotid-femoral pulse wave velocity (cfPWV), only for follow-up, and intima-media thickness (IMT) were determined. Biochemical variables included total cholesterol (TC), high and low-density lipoprotein cholesterol (HDL-C and LDL-C), triglycerides (TG), glycosylated haemoglobin A1c (HbA1c), C-reactive protein (CRP) and HIV status. Creatinine clearance (CrCl) was calculated with the Cockcroft-Gault formula. The CD4 counts were determined, in whole blood, at baseline by the National Health Laboratory using flow cytometric analysis and at follow-up with finger-prick blood and a point-of-care device PIMATM CD4 (Alere, Jena, Germany).
The statistical analyses were performed by using Statistica® 13 (StatSoft, Inc., Tulsa, OK, USA). The Framingham and Reynolds risk score were determined with excel spreadsheets, separately, during baseline and follow-up. Basic descriptive statistics were used to determine normal distribution of the data and logarithmic transformation was applied and presented as geometric mean with 5th and 95th percentiles if skewed. Groups were compared using independent t-tests and Chi-square tests as appropriate. Co-morbidity prevalence was defined by using cut-off values. Associations of measures of end-organ damage with the risk scores were determined by making use of Pearson and partial correlations. Odds ratios with 95% confidence intervals (CI) were calculated. Median values of risk scores and measures of end-organ damage were used as cut-off values.
Results and conclusion
The HIV-infected group presented with lower HDL-C (p<0.01) and CrCl levels (p=0.02) at baseline, compared to the HIV-free group. At follow-up the HIV-infected group had significant lower BMI (p<0.01), WC (p<0.01) and HbA1c (p=0.01) compared to the HIV-free control group.
The CD4 counts of the HIV-infected group was higher (p=0.03) at follow-up compared to baseline.
More HIV-free participants were overweight (p=0.02), had diabetes (p=0.04), had lower HDL-C (men) (p=0.013) and had a higher prevalence of microalbuminuria (p=0.04), compared to the HIV-infected group at follow-up.
No differences were seen between the HIV-infected and HIV-free group with either the Framingham or Reynolds risk score at baseline or follow-up. The Framingham risk score was higher in the HIV-infected group at follow-up when compared to the HIV-free controls, however both the HIV-free controls and HIV-infected group had a higher Reynolds risk score at follow-up than at baseline. No differences were seen between the ART and no-ART group.
A borderline negative correlation (p=0.053) was seen between CHD and CrCl, while CVD risk correlated negatively with cfPWV in the HIV-infected group at follow-up, however after adjusting for age, sex, WC, SBP, CRP, CD4 and ART use no correlations were seen. The CVD risk correlated negatively with Cornell product in the HIV-free group, however after adjusting for age, sex, WC, SBP, CRP, tobacco use and alcohol use no correlations were seen. No significant odds ratios were found for having a higher than median CHD or CVD risk with higher than median PWV, IMT and Cornell product or lower than median CrCl.
To conclude, despite their HIV-status for at least 10-years and 80% of participants receiving ART, we found that the HIV-infected participants did not have higher CHD or CVD risk when compared to the HIV-free participants. Those infected with HIV did also not show associations of risk scores with measures of end-organ damage||en_US