Title: HbA1c for the diagnosis of diabetes in black South Africans: impact of anaemia, HIV infection, and anti-retroviral treatment (ART)
Authors: Tom R Hird1, Fraser J Pirie2, Tonya M Esterhuizen3, Brian O’Leary4, Elizabeth H Young5, Mark I McCarthy6, Manjinder S Sandhu7, Ayesha A Motala2
1 Department of Public Health and Primary Care, University of Cambridge, Cambridge, UK
2 Nelson R. Mandela School of Medicine, University of KwaZulu-Natal, Durban, South Africa
3 Centre for Evidence-Based Health Care, Faculty of Medicine and Health Sciences, Stellenbosch University, Stellenbosch, South Africa
4Research and Policy Department, Office of Strategy Management, eThekwini Municipality, Durban, South Africa
5 Wellcome Trust Sanger Institute, Hinxton, U.K
6 Oxford Centre for Diabetes, Endocrinology and Metabolism, University of Oxford, UK
7 Department of Medicine, University of Cambridge, Cambridge, UK

Introduction
Glycated haemoglobin (HbA1c) is recommended as an additional tool to glucose-based measures (fasting plasma glucose [FPG], 2-hr plasma glucose [2PG] during OGTT) for the diagnosis of diabetes mellitus (DM). However, some clinical conditions may affect the ability of HbA1c to reflect glycaemia. In SSA, the prevalence of some of these conditions, such as HIV infection and anaemia, is high and could potentially limit the utility of HbA1c for the diagnosis and detection of diabetes in these populations. We assessed the association of anaemia, HIV, and antiretroviral therapy (ART) with HbA1c in an urban black South African population. 

Methods
The Durban Diabetes Study was a population-based cross-sectional survey of participants aged ≥18 years from the city of Durban in South Africa. HbA1c and plasma glucose (PG) were measured using the BioRad Variant II turbo 2.0 analyser and the Abbott Architect 2: ci 8200 analysers, respectively. The 1999 WHO diagnostic criteria for disorders of glycaemia were used for the PG results and the 2011 WHO criteria for diabetes for the HbA1c. This analysis only included participants without known DM. Haemoglobin (Hb) concentration (g/dL) was obtained as part of a full blood count. Anaemia was defined as Hb <12.0 g/dl for women and <13.0 g/dl for men. HIV infection was determined by screening and confirmatory enzyme immunoassay. Current ART was self-reported.

Results
The analysis included 1067 participants (318 men and 749 women) with a mean age of 38.0 ±16.0 years. The prevalence of DM was 3.5%, based on OGTT, and 3.7%, based on HbA1c. The prevalence of anaemia was 28.8% and the prevalence of HIV was 45.6% (untreated HIV=29.6% and ART-treated HIV=16.1%). The age-adjusted mean Hb was 12.9 g/dl (95% CI 12.8-13.0). Mean Hb was lower in those with anaemia (11.0 g/dl [95%CI 10.9-11.1]) than those without anaemia (13.7 g/dl [95% CI 13.6-13.8]), and lower in those with untreated HIV (12.5 g/dl [95%CI 12.3-12.7]) and ART-treated HIV (12.2 g/dl [95% CI 12.0-12.5]) than in those without HIV (13.4 g/dl [95% CI 13.2-13.5]). After adjustment for confounders including PG measures, anaemia was associated with higher HbA1c (β=0.10% [95% CI 0.05-0.17] p<0.001), whereas ART-treated HIV was associated with lower HbA1c (β=-0.11% [95%CI -0.19- -0.03], p<0.01). However, there were no statistically significant differences in the prevalence of DM based on OGTT and HbA1c measurements in those with anaemia (2.9 vs. 3.3%), untreated HIV (1.6 vs. 1.6%) or ART-treated HIV (2.8 vs. 1.3%).
Conclusion
In this urban black South African population, the association of anaemia, HIV infection, and ART with HbA1c was modest and did not appear to affect the utility of HbA1c for the detection of diabetes. However, the use of HbA1c as a diagnostic and screening tool require further evaluation in this and other populations in SSA. 
Disclosure statement
This study was supported in part by Sanofi-Aventis, NovoNordisk (South Africa), MSD Pharmaceuticals (Pty) Ltd (Southern Africa), and the NIHR Cambridge BRC.

HbA1c for the diagnosis of diabetes in black South Africans: impact of anaemia, HIV, and anti-retroviral treatment (ART)