Background
Africa has a high prevalence of both Human Immunodeficiency Virus and Non Communicable Diseases (NCDs) but in Zambia there are few data on co-morbid NCDs like Diabetes Mellitus (DM) among HIV-infected individuals. We aimed to identify risk factors for impaired fasting glucose or diabetes among HIV-infected Zambians on long-term Combined Antiretroviral Treatment (cART).

Methods
This was a cross sectional study of adult HIV patients in five health facilities of Copperbelt Province in Zambia. HIV/AIDS patients aged 18 years and above, enrolled in care at those health facilities and had been on cART for more than 2 years were included. All patients known to have Diabetes mellitus were excluded from the study. Participants underwent assessment of random blood sugar levels at enrolment and returned the following morning for fasting glucose measured by glucometers. The primary outcome was proportion with impaired fasting glucose or DM. Multivariable logistic regression was used to examine if demographics, time on ART, type of ART regimen, body mass index and baseline CD4 count were predictors of impaired fasting glucose.

Results
Overall (n = 270) there were 186 females (69%) and 84 males (31%). The prevalence of impaired fasting blood sugar or diabetes after 8 h of fasting was 15% (95%CI: 11.1, 20.0). Ten percent (26/270) had impaired fasting glucose and 5 % (14/270) had diabetes. Impaired fasting glucose was higher in males than females [AOR = 3.26, (95% CI: 1.15–9.25; p-value = 0.03)]; as well as among patients on second line treatment than those on first line [AOR = 3.87 (95% CI 1.16–12.9); p-value = 0.03]. In contrast those with less likelihood of impaired fasting glucose included patients with a normal BMI (18.5–24.9) than overweight or obese patients [AOR = 0.09 (95% CI 0.03–0.31; p-value < 0.001)]; and participants who had less than 4 diabetes symptoms than those with more than 4 diabetes symptoms [AOR = 0.04 (95% CI 0.02–0.12); p-value < 0.001].

Conclusion
We have found high levels of impaired fasting glucose or diabetes among ART patients compared to what is reported in the general population suggesting missed care and support opportunities associated with metabolic imbalance management. There is thus a need to re-package HIV programming to include integration of diabetes screening as part of the overall care and support strategy.