Can A1C Values Be Distorted in Patients Taking Protease Inhibitors?
Yes, A1C values can be falsely lowered in HIV-infected patients taking protease inhibitors, potentially leading to under-diagnosis and under-treatment of diabetes.
The Evidence for A1C Distortion with Protease Inhibitors
The most recent and authoritative guidelines explicitly address this issue. The 2022 ADA Standards of Medical Care state that A1C is less reliable than blood glucose measurement in HIV patients treated with certain protease inhibitors (PIs) and nucleoside reverse transcriptase inhibitors (NRTIs) 1. The 2019 ADA guidelines similarly recommend that in conditions associated with altered A1C-glycemia relationships, including HIV treated with certain drugs, only plasma blood glucose criteria should be used to diagnose diabetes 2.
Magnitude of the Problem
Research data quantify this distortion: In a large cohort study of 1,357 HIV-infected men followed over 13 years, at a fasting glucose of 125 mg/dL, the median A1C among HIV-infected men was 0.21% lower than among HIV-uninfected men, with this effect increasing at higher glucose levels 3. Critically, 63% of HIV-infected men had at least one visit with clinically significant A1C discordance (defined as ≥0.5% lower than expected) 3.
Independent Risk Factors
The discordance was independently associated with:
- Regimens containing protease inhibitors (directly answering your question)
- Regimens containing NNRTIs or zidovudine
- Low CD4 cell count (<500 cells/mm³)
- High mean corpuscular volume
- Abnormal corpuscular hemoglobin 3
Clinical Implications and Recommendations
For HIV-infected patients on protease inhibitors, use plasma glucose measurements (fasting glucose or oral glucose tolerance test) rather than A1C for diabetes screening and diagnosis 2, 1. This is a Grade B recommendation from the ADA.
Practical Algorithm:
For diabetes screening in patients on PIs: Use fasting plasma glucose (≥126 mg/dL diagnostic) or 2-hour OGTT (≥200 mg/dL diagnostic)
If A1C has already been measured and is borderline (6.5-6.9%): Confirm with fasting glucose ≥126 mg/dL before diagnosing diabetes 4
For monitoring established diabetes in patients on PIs: Rely primarily on self-monitoring blood glucose and fasting glucose measurements rather than A1C alone
If A1C seems discordant with glucose readings: Strongly suspect assay interference or altered erythrocyte dynamics from the antiretroviral therapy
Mechanism of Distortion
The falsely low A1C occurs because protease inhibitors (and certain other antiretrovirals) affect red blood cell turnover and hemoglobin parameters 3. Since A1C depends on the duration of erythrocyte exposure to glucose, any condition that shortens RBC lifespan or alters hemoglobin characteristics will distort the A1C-glucose relationship 4, 5.
Common Pitfall to Avoid
The major clinical danger is under-diagnosis of diabetes or under-treatment of established diabetes when relying solely on A1C in this population 3. A "normal" A1C of 6.0% in a patient on protease inhibitors might actually reflect poor glycemic control that would show an A1C of 6.5-7.0% in someone without HIV or antiretroviral therapy.
Note on Recent Guideline Evolution
Interestingly, the 2025 HIV treatment guidelines have moved away from recommending A1C for diabetes screening in HIV patients entirely, specifically recommending against its use 6. This represents recognition of the accumulating evidence of A1C unreliability in this population, though the 2024 version notes that earlier concerns may have been partially related to older agents like zidovudine 6.