Respiratory Tract Infections and Household Fuel Use in HIV-Infected Kenyan Infants

Abstract

Background: Acute respiratory infections (ARIs) are the largest contributor to child morbidity and mortality worldwide. HIV-infected children are particularly susceptible to ARIs. Little data exists on the incidence and cofactors of ARIs in HIV-infected infants receiving early antiretroviral therapy (ART). Household air pollution from wood (high polluting), charcoal, or kerosene (both medium polluting) fuel may also exacerbate risk of ARIs in HIV-infected children compared to exposure to petroleum gas (low polluting). We examined the burden of fuel use and HIV infection on ARI risk in HIV-infected infants receiving ART. Methods: HIV-infected infants initiated ART at ≤12 months of age and were followed monthly for 2 years in Nairobi, Kenya. AT 4-6 years post-ART initiation, fuel exposure was collected on the survivors within this cohort. ARI rates and cofactors were determined using Andersen-Gill models. Results: Among 111 HIV-infected infants (49% male) initiating ART, the median age at ART initiation was 4.5 months. Pre-ART, the median CD4% was 19% and HIV-1 RNA was 6.6 log10 copies/ml. Most infants (29%) were wasted (weight-for-height Z-score <-2). At 2-years post-ART, upper respiratory infection (URI) and pneumonia rates were 118.9 and 32.7 per 100 person-years, respectively. In univariate analyses, lower monthly rent (<1000 Kenya Shillings), higher plasma HIV-RNA level (per 1 log10copies/ml increase) and wasting were associated with pneumonia (hazard ratio (HR) =3.62; p=0.0001; HR=1.91; p = 0.01 and HR=2.82; p=0.001, respectively), and these associations remained in adjusted models (HIV RNA, p=0.1; wasting, p=0.07; rent, p=0.01). At 2-years post-ART, URI and pneumonia rates in those infants for whom fuel exposure data were collected (N=45) were 122.2 and 15.6 per 100 person-years respectively. Of these infants, 4.1% had wood, 64.6% had charcoal and/or kerosene, and 30.4% had liquid petroleum gas exposure, which was often used in combination with higher polluting fuels. In univariate analysis, lower monthly rent was associated with pneumonia (HR=7; p=0.02). Wood exposure was associated with higher risk of URI (HR=1.74; p = 0.004). Conclusions: In early treated HIV-infected infants, lower monthly rent, and wasting prior to ART were independent risk factors for pneumonia, and infants with wood exposure had higher URI. Future effective strategies need to be implemented to decrease this ARI risk in HIV-infected infants.