232 - Sex-Specific Metabolomic and Lipidomic Alterations in HIV-Positive Mothers on Antiretroviral Therapy and Their Exposed Infants in Kenya and Zambia

Publication Number: 1220.232

Kiana Asgari, Stanford University School of Medicine, Stanford, CA, United States; Steven Hawken, Ottawa Hospital Research Institute and University of Ottawa, Ottawa, ON, Canada; Kumanan Wilson, University of Ottawa Faculty of Medicine, Ottawa, ON, Canada; Monica Lamoureux, Children's Hospital of Eastern Ontario Research Institute, Ottawa, ON, Canada; Robin Ducharme, Ottawa Hospital Research Institute, Ottawa, ON, Canada; Nancy A.. Otieno, Kenya Medical Research Institute - Centre for Global Health Research, Kisumu, Nyanza, Kenya; Gary Darmstadt, Department of Pediatrics, Stanford University School of Medicine, Palo Alto, CA, United States; Ivana Maric, Stanford University School of Medicine, Stanford, CA, United States; Victoria Ward, Stanford University School of Medicine, San Francisco, CA, United States; David K. Stevenson, Stanford University School of Medicine, Stanford, CA, United States; Pranesh Chakraborty, University of Ottawa Faculty of Medicine, Ottawa, ON, Canada

Background: HIV infection and antiretroviral therapy (ART) during pregnancy may disrupt maternal metabolism and impact fetal development. Despite improved maternal health and safer breastfeeding with ART, HIV-exposed uninfected children remain at increased risk for infectious morbidity, mortality, and impaired growth compared to HIV-unexposed uninfected children. Identifying metabolomic and lipidomic disruptions related to HIV and ART exposure is important for understanding underlying biological changes and their impact on infant health, and for further improving outcomes of HIV-exposed children.

Objective: Our goal was to identify such biomarkers in the maternal metabolome and lipidome, and also in the neonatal metabolome, that distinguish HIV-positive mothers on ART from controls, and to characterize sex-specific alterations among HIV-exposed neonates.

Design/Methods: Clinical and demographic data, plus plasma samples (22-24 weeks of gestation), were collected from 529 Kenyan women. Using targeted and untargeted mass spectrometry, 641 lipids and 639 metabolites were quantified. Neonatal metabolites were measured from heel-prick blood at birth in 1,211 infants from Kenya and Zambia. Univariate and machine learning analyses identified HIV/ART-associated biomarkers, with Elastic-Net stability analysis used for model training and performance comparison beyond clinical and demographic factors.

Results: Demographic variations were observed across cohorts (Tab.1). Maternal lipidomic and metabolomic profiles were significantly altered in HIV-positive participants receiving ART (Fig.2a-e). Integrating omics data with clinical and demographic variables improved AUROC from 0.667±0.047 to 0.901±0.032, indicating a strong metabolic impact of HIV/ART during pregnancy. Neonatal biomarkers were also affected by HIV/ART exposure, with stronger alterations in males (Fig.1h; 0.77±0.03 to 0.82±0.03 with p-value 0.015) than in females (Fig.1i; 0.85±0.02 to 0.87±0.02 with p-value 0.13, only increasing marginally beyond the effect of clinical/demographic covariats).

Conclusion(s): We identified specific biomarkers in both mothers and neonates affected by maternal HIV exposure and ART. These findings may help improve infant health through targeted clinical follow-up. These findings may help improve infant health through targeted clinical follow-up. The results also reveal a sex-dependent impact, with male infants showing stronger biomarker alterations and potentially
greater vulnerability to HIV and ART exposure.

Figure 1: Maternal Results.
Maternal biomarkers altered by HIV/ART. Top row: Elastic-Net coefficients across
100 resampling trials. Middle row: Volcano plots of adjusted p-values. Bottom row: ROC curves
comparing models trained on clinical/demographic data (orange) versus adjusted biomarkers (blue),
highlighting HIV-specific effects beyond clinical differences.

Figure 2: Neonatal Results
Maternal biomarkers altered by HIV/ART. Top row: Elastic-Net coefficients across
100 resampling trials. Middle row: Volcano plots of adjusted p-values. Bottom row: ROC curves
comparing models trained on clinical/demographic data (orange) versus adjusted biomarkers (blue),
highlighting HIV exposure effects beyond clinical differences in Male infants.

Table 1
Characteristics of mothers in Kenya and infants in the Kenya and Zambia cohorts.