BackgroundAir pollution is known to induce systemic inflammation and contribute to cardiovascular disease. The effects of ambient air pollution on adolescents living with perinatally acquired HIV (PHIV) in Africa has been understudied. In this study, we aim to investigate the relationship between particulate matter, a common proxy indicator for air pollution, and markers of inflammation, monocyte activation, and cardiovascular disease risk. MethodsWe recruited a cohort of PHIV-positive and HIV-negative adolescents who were from around Kampala, aged 10–18 years, had no known active infections, and who lived within 40 km of an PM2·5 monitoring site. Adolescents with PHIV were on ART with HIV-1 RNA concentrations ≤400 copies/mL. Daily ambient concentrations of particulate matter (PM2·5), a proxy for air pollution, were measured with continuous central site monitoring using a Beta Attenuation Monitor or E-Samplers from the GeoHealth Hub. We measured carotid intima media thickness (IMT) using ultrasound, plasma concentrations of soluble CD14 and CD163 as markers of monocyte activation, plasma markers of systemic inflammation (hsCRP, IL6, sTNFRI), oxidised lipids, and intestinal permeability (zonulin). FindingsOne hundred and nineteen participants were included (69 with PHIV, 50 were HIV-negative). Median age was 12·7 years (IQR 11·4–14·2) years, 55% were female and 63% lived below the extreme poverty line (living with <US$1·90 a day). Median CD4+ cell count was 962 cells/μL (643–1313), 82% had viral load <20 copies/mL and median ART duration was 12·1 years (9·7–13·3) years. Overall median daily PM2·5 exposure was 29·08 μg/m3 (23·40–41·70), considered a moderate health risk (WHO definition). There was no significant difference in exposure of PM2·5 between groups (p=0·073), nor between sexes (p=0·099). PM2·5 concentrations significantly correlated with intestinal permeability (zonulin; r=0·43, p<0·001), monocyte activation (soluble CD163: r=0·25, p=0·053), and IMT (r=0·35, p=0·004) in participants with PHIV but not in those who were HIV-negative (p≥0·05 for all three correlations in HIV-negative children). In multivariable quantile regression models, after adjusting for age, sex, and, separately, soluble CD163 and zonulin, daily PM2·5 concentrations remained associated with IMT (β=0·005, 95%CI [0·001–0·009], p=0·007 and β=0·004, 95%CI [0·001–0·008], p=0·014, respectively) in children with PHIV. InterpretationPM2·5 exposure was associated with elevated subclinical vascular disease in adolescents with PHIV but not HIV-negative adolescents, despite there being no significant difference in exposure between these two groups. Our findings suggest the combined effects of HIV infection and air pollution may amplify and contribute to early development of cardiovascular disease. FundingNICHD K23HD088295 (SDF); University Hospitals Cleveland Medical Center; The Clinical and Translational Science Collaborative of Cleveland; and UL1TR002548 from the National Center for Advancing Translational Sciences.