Defining the optimum HOMA-IR cut-off value for insulin resistance among ARV-treated HIV patients in Indonesia

Research output: Contribution to journalArticlepeer-review

Abstract

Background and aims: HIV-infected patients who are treated with anti-retroviral (ARV) drugs are prone to develop insulin resistance. This study aims to determine the cut-off value of HOMA-IR score in ARV-treated HIV patients in Indonesia. Methods: A cross-sectional study was conducted among 234 adults with HIV who received ARV therapy in HIV Integrated Care Unit of Dr. Cipto Mangunkusumo National General Hospital, Jakarta, Indonesia. The duration of HIV diagnosis, duration of ARV therapy, metabolic syndrome status, and calculated HOMA-IR were obtained in this study. HOMA-IR cut-off point was calculated using ROC curve analysis, along with the specificity, sensitivity and likelihood ratio (LR). Results: Among 234 subjects, 58% of subjects were on second-line ARV therapy. The prevalence of metabolic syndrome was 23.9%. The obtained HOMA-IR cut-off value was 2.705 with sensitivity and specificity approaching 70%, PPV 40.9%, NPV 87.6%, with positive LR of 2.15 and negative LR of 0.48. The insulin resistance prediction from the obtained HOMA-IR cut-off value was at moderate strength. Based on this cut-off value, 39.7% of the subjects experienced insulin resistance. Conclusions: Using the new proposed HOMA-IR cut-off point for HIV patient in Indonesia, the prevalence of insulin resistance among HIV-infected patients treated with ARV in Indonesia using optimum HOMA-IR cut-off value of 2.705 was 39.7%.

Original languageEnglish
Article number102719
JournalDiabetes and Metabolic Syndrome: Clinical Research and Reviews
Volume17
Issue number2
DOIs
Publication statusPublished - Feb 2023

Keywords

  • Anti-retroviral therapy
  • HIV
  • HOMA-IR
  • Insulin resistance

Fingerprint

Dive into the research topics of 'Defining the optimum HOMA-IR cut-off value for insulin resistance among ARV-treated HIV patients in Indonesia'. Together they form a unique fingerprint.

Cite this