Lower Preprandial Insulin and Altered Fuel Use in HIV/Antiretroviral-Exposed Infants in Cameroon

Abstract

Context: Intrauterine HIV/antiretroviral (ARV) and postnatal ARVs are known to perturb energy metabolism and could have permanent effects on future metabolic health. Such maladaptive effects could be mediated by changes in mitochondrial function and intermediary metabolism due to fetal and early-life ARV exposure in HIV/ARV-exposed uninfected (HEU) infants.

Objective: The objective of the study was to understand the relationship(s) between mitochondrial fuel use (assessed via acylcarnitines and branched chain amino acids) and preprandial insulin in infants exposed to in utero HIV/ARV plus postnatal zidovudine or nevirapine compared with HIV/ARV-unexposed uninfected (HUU) infants.

Design: This was a prospective cohort study with the following three groups: 1) intrauterine HIV/ARV/postnatal zidovudine-exposed (HEU-A), 2) intrauterine HIV/ARV/postnatal nevirapine-exposed (HEU-N), and 3) HUU infants. Principal component analysis and linear regression modeling were performed to assess the association between in utero HIV/ARV exposure and infant insulin.

Setting: The study was conducted at Cameroonian urban antenatal centers.

Participants: HIV-infected and -uninfected pregnant woman/infant dyads participated in the study.

Main outcome: Preprandial insulin was the main outcome measured.

Results: Of 366 infants, 38 were HEU-A, 118 HEU-N. Forty intermediary metabolites were consolidated into seven principal components. In a multivariate analysis, both HEU-A (β = -.116, P= .012) and HEU-N (β = -.070, P= .022) demonstrated lower insulin compared with HUU infants. However, at high levels of plasma metabolites, HEU-A (β = .027, P= .050) exhibited higher insulin levels than HEU-N or HUU infants. A unique array of short-chain acylcarnitines (β = .044, P= .001) and branched-chain amino acids (β = .033, P= .012) was associated with insulin.

Conclusion: HEU-A and HEU-N infants have lower preprandial insulin levels at 6 weeks of age and appear to use metabolic fuel substrates differently than HUU infants. Future studies are warranted to determine whether observed differences have lasting metabolic implications, such as later insulin resistance.

Figure 1.

Relationship between principal components. +, PC1, long-chain fatty acylcarnitines, C0, and C5; *, PC3, short-chain fatty acyl carnitines and BCAAs; ^, PC5, BCAAs (isoleucine, leucine, valine). OXPHOS, oxidative phosphorylation.

Figure 2.

Relationships between C2 and insulin and C3 and insulin. A, Spearman's correlation between C2 and preprandial insulin in entire cohort. B, Regression analysis comparing relationship between C2 and preprandial insulin by exposure group. C, Spearman's correlation between C3 and preprandial insulin in entire cohort. D, Regression analysis comparing relationship between C3 and preprandial insulin by exposure group. *, Insulin is quarter-root-transformed for all analyses shown.

Figure 3.

Relationship between BCAAs and C5. A, Spearman's correlation between BCAAs and C5 in entire cohort. B, Regression analysis comparing relationship between BCAAs and C5 by exposure group. *, BCAAs are represented as PC5 in all analyses shown.