Background Infection with, and treatment of HIV is associated with effects on glycaemia and renal function. The purpose of this study was therefore to compare glycaemic control and albuminuria in HIV-positive and HIV-negative type 2 diabetic patients. Materials and methods Diabetic patients with and without HIV infection were recruited from a diabetic clinic at Chris Hani Baragwanath Hospital in Soweto, South Africa. Data was collected on weight, height, HbA1c, fasting glucose, urine albumin:creatinine ratio, HIV status, CD4 counts, viral load and concomitant therapies. Multivariable regression analysis was used to isolate the determinants of fasting glucose and HbA1c levels and risk factors for albuminuria. Results Data were collected from 106 HIV-positive and 214 HIV-negative diabetics. All HIV infected subjects were receiving anti-retroviral therapy. The determinants of fasting glucose levels (log) were HIV infection (β = 0.04, p = 0.01) and use of anti-hypertensive agents (β = 0.07, p = 0.0006), whilst for HbA1c levels (log) they were HIV infection (β = -0.03, p = 0.03), BMI (β = 0.004, p = 0.0005), statin use (β = 0.04, p = 0.002) and glucose levels (β = 0.01, p<0.0005). In HIV-positive subjects, CD4 counts were negatively associated with glucose levels (β = -0.0002, p = 0.03). The risk factors for albuminuria were (odds ratio [95% CIs]) dyslipidaemia (1.94 [1.09, 3.44], p = 0.02) and HbA1c levels (1.24 [1.12, 1.38], p<0.0001). Discussion These data suggest that glycaemic control is worse in type 2 diabetic subjects with HIV infection and that HbA1c underestimates glycaemia in these patients. Albuminuria was not associated with HIV-positivity. The negative relationship of CD4 counts with glucose levels may reflect viral removal and easing of the associated inflammatory response. It is possible that the association of statin and anti-hypertensive therapies with high HbA1c and glucose levels, respectively, is due to such therapies being given largely to subjects with poor glycaemic control.
