Clin Drug Invest 2009; 29 (2): 101-109 ORIGINAL RESEARCH ARTICLE 1173-2563/09/0002-0101/$49.95/0  2009 Adis Data Information BV. All rights reserved. Long-Term Evaluation of Glucose Homeostasis in a Cohort of HAART-Treated HIV-Infected Children A Longitudinal, Observational Cohort Study Alessandra Vigan` o, 1 Paolo Brambilla, 2 Giulia Pattarino, 1 Sara Stucchi, 1 Silvia Fasan, 1 Chiara Raimondi, 1 Chiara Cerini, 1 Vania Giacomet, 1 Gian V. Zuccotti 1 and Giorgio Bedogni 3 1 Pediatric Clinic, ‘L. Sacco’ Hospital, University of Milan, Milan, Italy 2 Azienda Sanitaria Locale Milano 2, Milan, Italy 3 Clinical Epidemiology Unit, Liver Research Centre, Basovizza, Trieste, Italy Background and objectives: Few and mainly cross-sectional studies of glucose Abstract homeostasis are available in HIV-infected children treated with highly active antiretroviral therapy (HAART). The aim of the present study was to describe a 4-year course of glucose homeostasis in a cohort of HAART-treated children and adolescents, using glucose and insulin levels during an oral glucose tolerance test (OGTT) as outcome measures. In addition, we investigated possible risk factors, both related and unrelated to antiretroviral therapy, associated with insulin resis- tance. Methods: We assessed glucose metabolism yearly for 4 consecutive years in 37 HIV-infected children receiving a protease inhibitor (PI)-based HAART regimen containing lamivudine/stavudine plus indinavir or ritonavir or nelfinavir or a non- nucleoside reverse transcriptase inhibitor (NNRTI)-based HAART regimen con- taining lamivudine/tenofovir/efavirenz. Generalized estimating equations were used to evaluate the relationship between the loge-transformed area under the serum concentration-time curve (AUC) of insulin during OGTT and antiretroviral therapy, controlling for time, sex, baseline age, puberty, body mass index and CD4+ T cells percentage. Results: Ritonavir-unboosted PI-based HAART regimens were administered to most children at baseline; however, their use decreased during follow-up in favour of an NNRTI-based regimen. The nelfinavir/lamivudine/stavudine (regression coefficient = –0.69, p < 0.05) and efavirenz/lamivudine/tenofovir (regression coefficient = –0.93, p < 0.05) regimens, but not the ritonavir/lamivudine/stavudine