HIV, Antiretroviral Therapy, Pregnancy, Lactation and Bone Health in Uganda

Change log
Nabwire, Florence 

Globally, ~17 million women and ~2.1 million children are living with HIV. Sub-Saharan Africa accounts for 70% of HIV-infected (HIV+) persons. Mother-To-Child Transmission of HIV (MTCT) during pregnancy, delivery and breastfeeding, is the main route of HIV infection in children. The World Health Organisation recommends lifelong antiretroviral therapy (ART) for all HIV+ pregnant and breastfeeding mothers to prevent MTCT, and breastfeeding for ≥24 months for optimal child health in resource limited settings (Option B+ strategy).

Initiation of ART in HIV+ adults is associated with a 2-6% decrease in areal bone mineral density (aBMD) regardless of ART regimen, but data are limited in pregnant and lactating women. Tenofovir, a preferred first-line drug in Option B+ ART regimen, is associated with 1-2% greater decreases in aBMD. Pregnancy and lactation are associated with physiological changes in maternal bone mineral density, but most evidence shows that this is recovered after cessation of breastfeeding. The hypothesis of this thesis is that ART may accentuate the normal process of bone mobilisation during pregnancy and lactation, leading to bone loss that is not recovered in the mother and/or compromised infant growth and bone mineral accretion. The primary objective of this research was to investigate if HIV+ women experience greater reductions in bone mineral compared to HIV-uninfected (HIV-) counterparts.

Two groups of pregnant women, 95 HIV+ on ART (Tenofovir-Lamivudine-Efavirenz, previously ART naïve) and 96 HIV- were followed prospectively in Kampala, Uganda. Data were collected at 36 wks gestation (PG36), 2 (PP2) and 14 wks postpartum (PP14). Dual-energy x-ray absorptiometry was used to measure bone phenotype (aBMD, bone mineral content (BMC), bone area (BA), and size-adjusted BMC (SA-BMC, adjusted for height or length, weight and BA) of the whole body (WB) and lumbar spine (LS) in mother-baby pairs, and total hip (TH) in mothers. The primary outcome was the difference between groups in % change (± SE) in maternal LS aBMD between PP2 and PP14. Secondary outcomes included changes in maternal markers of bone formation (P1NP and BAP) and resorption (CTX), serum 25-hydroxy vitamin D (25(OH)D), parathyroid hormone (PTH), plasma and urine concentrations of creatinine (Cr), calcium (Ca), phosphate (PO4) and magnesium (Mg), urine mineral:creatinine ratios, TmCa/GFR and TMP/GFR, respectively), breastmilk mineral composition (Ca, P, Na, K and Na/K ratio); and infant growth Z-scores and bone mineral. Statistical models were adjusted for potential confounders.

Median maternal age was 24.5 (IQR 21.1, 26.9) yrs. Mean gestation was 40.9±1.8 wks and not significantly different between groups. All women were breastfeeding at PP2 and PP14. More HIV+ women reported exclusive breastfeeding (PP2: 82.9% v 58.7%, p=0.0008; PP14: 86.7% v 66.2%, p=0.002). Body weight was 4-5% lower in HIV+ women. By PP14, mean duration of ART was 29.3±5.1 wks, adherence was >95%, and the median CD4 count was 403 (IQR 290-528) cells/mm3.

Maternal aBMD decreased between PP2 and PP14 at all skeletal sites in both groups as expected in lactation. Reductions in LS aBMD were not significantly different between groups (-1.8±0.4% vs -2.5±0.4%, p=0.3). However, HIV+ women had a significantly greater reduction in TH aBMD which persisted after adjustment for body size (-3.7±0.3% vs -2.7±0.3%, p=0.04). Median serum 25(OH)D was 67.4 nmol/L (IQR 54.8, 83.7) at PG36 and 57.6 nmol/L (48.7, 70.1) at PP14 with no significant difference between groups. Changes in 25(OH)D and PTH from PG36 to PP14 were not significantly different between groups (25(OH)D: -13.9±4.1% vs -11.1±3.1%; PTH: +60.0±6.4% vs +57.6±6.4%; both p>0.05). However, HIV+ women had 33-35% greater plasma PTH concentrations at both PG36 and PP14.

Bone formation and resorption markers increased in both groups between PG36 and PP14. HIV+ women had greater increases (CTX: +74.6±5.9% vs +56.2±5.9%; P1NP: +100.3±5.0% vs +72.6±5.0%; BAP: +67.2±3.6% vs +57.1±3.6%, all p<0.05). They also had a greater decrease in plasma Ca (-6.6±0.5% vs-3.8±0.5%, p≤0.0001) and greater increase in plasma phosphate (+14.4±2.0% vs +7.7±2.0%, p=0.02). Changes in plasma Cr and Mg, TmP/GFR and urine mineral:creatinine ratios were not significantly different between the groups. However, at both PG36 and PP14, HIV+ had significantly lower mean plasma Ca (PG36: -1.0±0.5%; PP14: -4.1±0.6%) and TmP/GFR (PG36: -11.4±3.1%; PP14: -7.2±3.0%) but higher PTH (PG36: +33.0±7.0%; PP14: +35.3±7.6%) compared to HIV- women (all p<0.05).

Mean breastmilk Ca decreased between PP2 and PP14, and the changes were not different between the groups (-19.9±3.0% vs -24.2±3.1%, p=0.3). There were no significant changes in breastmilk phosphorus (P) in both groups, but HIV+ women had significantly higher concentrations (PP2: +9.7±3.8%, p=0.01; PP14:+9.6±3.5 %, p=0.007). Breastmilk P was significantly correlated with maternal plasma [CTX] in a separate ANCOVA model (β = +0.13±0.04% per 1% increase in CTX, p=0.0003). Mean breastmilk Na, K concentrations and Na/K decreased between PP2 and PP14 in both groups. However, HIV+ women had a smaller decrease in breastmilk Na (-44.3±8.9% vs -72.6±9.0%, p=0.03). They also had a trend towards smaller reduction in Na/K ratio (-22.2±9.3% vs -46.6.6±9.5%, p=0.07).

Babies born to HIV+ mothers (HIV-exposed infants, HEI) had significantly lower gains in weight +53.0±1.4% vs +57.5±1.4%, p=0.02) compared to HIV-unexposed infants (HUI), and also lower weight-for-age (-0.47±0.16, p=0.003) and length-for-age (-0.53±0.18, p=0.005) Z-scores at PP14. HEI had a slower gain in WB BMC (+51.2±1.9% vs +57.3±1.9%, p=0.02), but the difference was not significant after adjustment for body size (-6.0±3.5% vs -7.6±3.8%, p=0.2); showing that the bone mineral accretion was appropriate for achieved infant size. In contrast, HEI had a greater increase in LS BMC (+29.5±1.7% vs +24.4±1.7%, p=0.03), a difference which remained after size-adjustment (+9.4±5.8% vs +4.3±6.2%, p=0.02).

This is the first study to compare changes in maternal aBMD and bone metabolism between HIV+ mothers on Option B+ ART and HIV- counterparts. The results show a greater reduction in TH aBMD in Ugandan HIV+ women on Option-B+ ART compared to HIV- in the first three months of lactation, consistent with their greater increases in bone turnover markers, lower TmP/GFR and plasma phosphate, and higher breastmilk phosphorus concentration. Also, HEI have slower growth and whole body bone mineral accretion compared to HUI. It is important to determine if these changes are temporary or have long-term consequences for the bone health of the mother and child.

Goldberg, Gail
HIV, Antiretroviral therapy (ART), Bone Mineral Density, Bone Biomarkers, Pregnancy, Lactation, Uganda, Growth, HIV exposed infants, Option B-plus ART, Breastfeeding
Doctor of Philosophy (PhD)
Awarding Institution
University of Cambridge
PhD studentship was funded by the Gates Cambridge Trust The research was jointly funded by the Medical Research Council (MRC) (programme code U105960371) and the Department for International Development (DFID) under the MRC/DFID Concordat agreement.