HTB

Haemoglobin A1c (HbA1c) may underestimate glycaemia in patients with diabetes on HIV therapy

Satyajit Das, HIV i-Base

The HbA1c gives an estimate of blood sugar control in last 6-12 weeks. However, previous retrospective studies suggested that HbA1c may not be accurate in HIV-positive patients.

Kim and colleagues from Washington Hospital investigated the relationship between HbA1c and glycaemia in HIV-positive patients with diabetes or hyperglycaemia.

This was a prospective cohort study of over 100 HIV-positive patients (70% male) with type-2 diabetes (77%) or fasting hyperglycaemia without anaemia or known kidney disease. In addition to a detailed medical history, evaluation included fasting and non-fasting glucose, haptoglobin, viral load, HbA1c, and fructosamine levels. Results were compared to 200 HIV-negative, sex-race-age matched type-2 diabetic controls.

Compared to HIV-negative, non-diabetic controls, HbA1c underestimated mean (+SE) plasma glucose in HIV patients by 29 (± 4) mg/dL. Univariate analyses found that current NRTI use, mean cell volume (MCV), haemoglobin, low HIV viral load and low haptoglobin were significantly associated with HbA1c-glucose discordance, but current PI and NNRTI use were not.

In a multivariate analysis of these factors, MCV was the only significant independent predictor (p=.002) of HbA1c-glucose discordance. When individual NRTIs were assessed, only ABC use (p=0.0004) and MCV (p=0.001) were independent predictors in a multivariate model. Fructosamine more closely reflected glycaemia than HbA1c.

The authors concluded that HbA1c may underestimate glucose levels by as much as 16% in patients with HIV on HAART and that “this appeared to be a direct effect of NRTI use, in particular abacavir” and that other markers may be more appropriate in this setting.

Comment

Fructosamine gives a picture of blood sugar control from the previous 2 to 3 weeks and may be a better marker of blood glucose control in HIV patients. However, falsely low fructosamine results are seen with decreased serum total protein and/or albumin.

Reference:

Kim P et al Hemoglobin A1c Does Not Accurately Reflect Glycemia in HIV Patients on HAART. 48th ICAAC, 25-28 October 2008. Washington. Abstract H2304.

Links to other websites are current at date of posting but not maintained.