Maternal Anthropometry,Body Composition,and Fat Distribution by HIV Status and Antiretroviral Therapy Class in South African Women

Abstract

Pregnancy affects adiposity, which may be influenced by HIV infection or antiretroviral therapy (ART). The objective of this study was to examine adiposity measures in the perinatal period, by HIV status and ART class. A total of 214 women (113 women with HIV [WWH], 71 initiated ART postconception), enrolled between 24 and 28 weeks of gestation and followed until 6–12 months postpartum, were assessed for longitudinal weight and cross-sectional postpartum anthropometry. A subset of 65 (52 WWH, 42 initiated ART postconception) had cross-sectional adiposity (body composition and fat distribution) measured at 6–12 months postpartum using dual-energy X-ray absorption scan. Multivariable linear and modified Poisson regression, adjusted for maternal age, pre-pregnancy body mass index, socioeconomic status, and postpartum months, examined associations of HIV status and postconception ART (dolutegravir-based [DTG] vs. efavirenz-based [EFV]) with anthropometry and adiposity outcomes. At enrollment, the median age was 30 years (interquartile range, 26–34) and 82% were multiparous. Between pre-pregnancy and postpartum, women gained an average of 2.33 kg (0.90 kg WWH), 30% lost weight (35% WWH), and 48% gained weight (38% WWH). WWH gained weight slower during pregnancy (0.27 vs 0.38 kg/week, p = .03) and were less likely to gain weight postpartum (RR = 0.72 95% CI 0.55, 0.93; p = .01) compared with women without HIV. Postpartum, mean body mass index was 32 kg/m² (standard deviation = 7.33) and 58% (53% WWH) of women had obesity. HIV was not associated with cross-sectional measures of postpartum anthropometry and adiposity. Among WWH, compared with EFV-based ART, DTG-based ART was not associated with weight gain during pregnancy or anthropometry and adiposity postpartum. Despite high rates of postpartum weight gain and obesity, no significant differences were observed in anthropometry and adiposity measures by HIV status and postconception ART. Nonetheless, these findings underscore the need for interventions to support healthy weight gain in pregnancy and postpartum weight loss to minimize pregnancy-associated obesity.

Keywords

obesity HIV body composition fat distribution antiretroviral therapy perinatal

Introduction

Obesity in the perinatal period is an important driver of pregnancy complications and long-term adverse maternal metabolic health.^1,2 It has been reported that over one-third of women with HIV (WWH) experience higher gestational weight gain (GWG) than recommended by the Institute of Medicine^3,4 and retain and/or gain weight postpartum.^5,6 A meta-analysis of 21 countries in sub-Saharan Africa (SSA) suggested that excessive GWG could exacerbate the high burden of noncommunicable diseases, including higher type 2 diabetes (T2D) risk among WWH.⁷ In addition, excessive GWG has long-term deleterious consequences for offspring. Specifically, a longitudinal study of school-aged children found that excessive GWG was associated with childhood obesity among 7- to 11-year-old children.⁸ WWH may, in particular, be more susceptible to weight gain related to the use of dolutegravir (DTG)-based antiretroviral therapy (ART),^9,10 which is first-line HIV treatment in SSA, including in pregnancy. Notably, randomized controlled trials in SSA countries, including Botswana, South Africa, and Uganda, found greater weight gain during the postpartum period in women who initiated DTG-based ART in pregnancy compared with those who initiated efavirenz (EFV)-based ART.^11,12 However, there are limited data on trajectories of weight change from pregnancy to postpartum among WWH overall, and those initiating DTG-based ART in pregnancy.

Although weight and body mass index (BMI) are widely used measures due to their ease of measurement, direct measures of body composition and fat distribution have been shown to be better predictors of future health outcomes.^13,14 For example, in a longitudinal study among SSA women, the accumulation of central body fat, particularly visceral adipose tissue (VAT), predicted the development of T2D 13 years before its onset.¹⁵ Another meta-analysis reported that higher visceral fat mass was associated with a fourfold risk of T2D among women, while fat-free mass reduced the risk.¹⁶ Historically, older nucleoside reverse transcriptase inhibitor (NRTI) ART such as thymidine analogues and protease inhibitors (PIs) has been associated with abdominal fat redistribution linked to metabolic dysregulation and long-term risk of metabolic disorders.¹⁷ Although lipodystrophy has been minimized with non-NRTI ART, the risk of metabolic disorders might persist due to DTG-related weight gain.¹⁸ However, there is a lack of data describing the body fat depot that is most affected by DTG-based ART initiation in pregnancy. To address this evidence gap, we followed WWH (initiating DTG- or EFV-based ART) and without HIV attending an antenatal care clinic in Cape Town from 24 to 28 weeks of gestational age (GA) to 6–12 months postpartum. The objective of this study was to examine trajectories of weight change, anthropometry, and body fat distribution in the perinatal period, comparing these by HIV status and postconception ART class.

Methods

Study design and population

A total of pregnant WWH (n = 200) and women without HIV (n = 200) (≥18 years of age) were recruited between November 2019 and June 2022 and enrolled into a prospective cohort study, called the CArdioMetabolic complications in Pregnancy (CAMP) study.^19,20 Briefly, participants completed three study visits, enrollment at 24–28 weeks of gestation (baseline) and follow-up at 33–38 weeks of gestation and 6–12 months postpartum. Women receiving treatment for diabetes or hypertension were excluded from the study. The recruitment coincided with the transition from EFV-based ART to DTG-based ART in South African antenatal care clinics, which took place from 2019. As a result, WWH in CAMP were either using pre- or postconception EFV-based (tenofovir 300 mg + [emtricitabine 200 mg or lamivudine 300 mg] + efavirenz 600 mg) or DTG-based (tenofovir 300 mg + [emtricitabine 200 mg or lamivudine 300 mg] + dolutegravir 50 mg) ART provided as a fixed-dose combination pill taken once daily. The CAMP study took place at the Gugulethu primary health care facility in Cape Town. Gugulethu is a township situated 18 km southeast of Cape Town in the Western Cape Province, South Africa.

Two study populations were included. First, of the 400 participants, we included 214 women who were followed longitudinally and attended their postpartum visit between 6 and 12 months, and we compared anthropometry measures between WWH (on pre- and post-conception ART) and those without HIV (Fig. 1). Second, of the 214 participants, we included a subset of 65 women who had cross-sectional dual-energy X-ray absorption (DXA) scan measures between 6 and 12 months, and we compared WWH (on pre- and postconception ART) and those without HIV. For both study populations, the ART comparison was restricted to women who initiated ART postconception and we compared postconception DTG- versus EFV-based ART. The study protocols were reviewed and approved by the Faculty of Health Sciences Human Research Ethics Committee of the University of Cape Town. All women provided written informed consent before study participation.

FIG. 1.

Participant enrollment and retention flowchart showing the selection of participants included in the different analyses conducted. ART, antiretroviral therapy; DTG, dolutegravir; DXA, dual-energy X-ray absorptiometry; EFV, efavirenz; GA, gestational age; PI, protease inhibitor.

Measures

Exposure and outcome assessments

The main exposures included HIV status and postconception ART class (EFV vs. DTG based). Both HIV status and ART regimen (including the timing of initiation) were collected through self-report and confirmed using medical records.

Longitudinal weight trajectory outcomes

Longitudinal weight outcomes included the rate of weight gain or loss (measured as average gain or loss in kilograms per week) and total weight change (kilograms) between pre-pregnancy and postpartum. Total weight change was calculated as postpartum weight minus pre-pregnancy weight and categorized as lost weight (<2 kg), stayed same (±2 kg), and gained weight (>2 kg).⁵ Weight was self-reported pre-pregnancy and measured by study staff at 24–28 weeks of gestation, 33–38 weeks of gestation, and 6–12 months postpartum. Although some authors have criticized self-reported pre-pregnancy weights,²¹ others have shown that self-reported versus measured pre-pregnancy weights are similar.²² Weight (kilogram) measurements were conducted with the participants wearing light clothing and no shoes, using a calibrated scale (Charder, Taichung City, Taiwan) accurate to within 0.2 kg.

Postpartum anthropometry outcomes

Postpartum anthropometry outcomes were weight, BMI, waist and hip circumferences, waist–hip ratio, and visceral adiposity index (VAI). BMI was calculated as weight divided by squared height, categorized as underweight (<18.5), normal (18.5–24.9), overweight (25.0–29.9), and obese (≥30.0) in kilogram per square meter. Waist (at the umbilicus) and hip (at the largest protrusion of the buttocks) circumferences were measured using a metal anthropometric tape measure (Seca, Birmingham, UK). VAI, a marker of adipose tissue dysfunction,²³ was calculated as waist circumference (cm)/[36.58 + (BMI × 1.89)]) × [triglycerides (mmol/L)/0.81] × [1.52/high-density lipoprotein (mmol/L)].²³ Triglycerides and high-density lipoprotein used in this formula were measured in serum samples collected after a 10–12-h fast; testing was performed in real time by the ISO 15189-accredited South African National Health Laboratory Services using an enzymatic colorimetric test (Roche Cobas^TM 6000 analyzer, Roche Diagnostics, Basel, Switzerland).

Postpartum body composition and fat distribution outcomes

Whole-body DXA scan [Hologic Discovery-W (S/N 71201), Bedford, MA, USA, software version 12.7.3.7], operated by a trained study radiographer, was used to measure total body composition and regional distribution of fat (kilogram) and fat-free mass (kilogram)²⁴ in a subset of women postpartum. In vivo precision was previously determined in our laboratory for fat-free soft tissue mass (FFSTM) (0.7%) and fat mass (FM) (1.67%) by measuring 30 individuals twice on the same day with repositioning.²⁵ Total body composition included subtotal (minus head) FM, FFSTM, and total body FM presented relative to total body mass and reported as a percentage (%BF). Regional fat distribution included trunk, arm, leg, android, and gynoid FM (expressed in kilograms). In addition, VAT and subcutaneous adipose tissue were estimated using DXA software algorithms developed for South African women.²⁶

Covariates

Maternal sociodemographic and clinical data were collected via trained interviewer-administered questionnaires at baseline, in the participants’ home language. Socioeconomic status (SES) was a composite score based on level of education, employment status, type of housing, and presence of a toilet, running water, electricity, fridge, telephone, and television in the house,²⁷ which were categorized into tertiles corresponding to the lower, middle, and higher SES group. Alcohol use was measured using a 3-item Alcohol Use Disorders Identification Test-Consumption (AUDIT-C; range 0–12). An AUDIT-C score ≥3 indicates hazardous drinking in the previous 12 months for women.²⁸ Perceived household food insecurity was assessed using a measure adapted from the Household Food Insecurity Access Scale and was categorized as “yes” and “no.”²⁹ Gestational age at enrollment was primarily determined by an ultrasound (89%) operated by an experienced research sonographer, the remaining 11% were assessed by either symphysis fundal height measurement or last menstrual period. Self-reported pre-pregnancy weight was used to calculate pre-pregnancy BMI. Postpartum month period (months) since delivery was calculated as the date of postpartum visit attendance minus delivery date, which was confirmed from obstetric records.

Statistical analysis

The analysis was divided into three parts as follows: (1) longitudinal weight trajectory analysis, which included all participants with valid weight measurements at enrollment (24–28 weeks of gestation, 33–38 weeks of gestation, and 6–12 months postpartum (n = 214); (2) cross-sectional postpartum anthropometry analysis, which included all participants with complete data on anthropometry at 6–12 months postpartum (n = 214); and (3) cross-sectional postpartum body composition and fat distribution analysis, which included all participants with complete data for DXA scans at 6–12 months postpartum (n = 65). In all three analyses, we compared measures of anthropometry, body composition, and body fat distribution by HIV status and postconception ART regimen (EFV vs. DTG based). Postconception ART group included women who were ART naive and those who were reinitiating ART. Because there were so few women on DTG before conception (N = 9), no preconception ART initiator analysis was conducted. Linear regression models were fit to estimate the associations between HIV status, postconception ART, and continuous anthropometry measures; and modified Poisson regression was used for categorical total weight change and BMI category. Modified Poisson regression is a robust statistical approach that is primarily used to estimate risk ratios (RRs) directly for binary outcomes in prospective cohort studies, especially when the outcome is not rare.³⁰ Similarly in a subset, linear regression models were used to examine the associations between HIV status, ART, and continuous body composition and fat distribution measures. Models for both HIV status and ART were adjusted for maternal age, pre-pregnancy BMI, SES, and postpartum months since delivery. All data were analyzed using STATA version 15.0 (Stata Corporation, College Station, TX, USA) and R Studio (R Foundation, Vienna, Austria).

Results

Participant characteristics

A total of 214 women (113 WWH) were assessed for longitudinal weight trajectories and cross-sectional postpartum anthropometry, and 65 (52 WWH) for cross-sectional postpartum body composition and fat distribution. Overall, at enrollment, the median age was 30 years (IQR, 26–34), gestational age was 25 weeks (IQR, 24–27), and 82% (175/214) of women were multiparous (Table 1). The median pre-pregnancy BMI was 31 kg/m² [interquartile range (IQR), 26–36], with 26% (55/214) living with overweight and 56% (120/214) living with obesity. Of the 120 women with obesity, 51% (61/120) did not have HIV and 49% (59/120) were WWH. WWH had a higher median age (32 vs. 28 years, p < .01), a higher proportion were multiparous (89 vs. 73%, p < .01), and a lower proportion had higher SES (32 vs. 53%, p < .01) compared with those without HIV. Among WWH, 63% (71/113) initiated ART postconception (DTG 75% [53/71]; EFV 25% [18/71]). Median postpartum months since delivery were 8.1 months (IQR, 6.6–10.1); WWH attended the postpartum visit earlier than women without HIV (7.6 vs. 9.4 months, p = .02). Women assessed in this sub-study (n = 214) had similar demographic characteristics as the overall CAMP cohort (n = 400) (Supplementary Table S1).

Table 1.

Pregnancy (Baseline) Characteristics, Overall and Stratified by Maternal HIV Status

		HIV status
	Total N = 214	Without HIV N = 101	With HIV N = 113
Demographics
Age (years)^a
Median (IQR)	30 (26–34)	28 (23–31)	32 (28–35)
Gestational age (weeks)
Median (IQR)	25 (24–27)	25 (24–27)	26 (24–27)
Pre-pregnancy BMI (kg/m²)
Underweight (<18.5)	0 (0)	0 (0)	0 (0)
Normal (18.5–24.9)	39 (18)	19 (19)	20 (18)
Overweight (25–29.9)	55 (26)	21 (21)	34 (30)
Obese (≥30)	120 (56)	61 (60)	59 (52)
Median (IQR)	31 (26–36)	32 (28–37)	30 (26–35)
Parity^a
Primiparous	39 (18)	27 (27)	12 (11)
Multiparous	175 (82)	74 (73)	101 (89)
Median (IQR)	3 (2–3)	2 (1–3)	3 (2–4)
Socioeconomic status^a
Lower	69 (32)	25 (25)	44 (39)
Middle	55 (26)	22 (22)	33 (29)
Higher	90 (42)	54 (53)	36 (32)
Perceived food insecurity
No	168 (78)	80 (79)	88 (78)
Yes	46 (22)	21 (21)	25 (22)
Hazardous alcohol use
No	196 (92)	93 (92)	103 (91)
Yes	18 (8)	8 (8)	10 (9)
ART initiation timing
Preconception			42 (37)
Postconception			71 (63)
Preconception ART duration (months)
Median (IQR)			57.82 (22.37–105.60)

	WWH on postconception ART (n = 71)	Postconception ART
	WWH on postconception ART (n = 71)	EFV-based (n = 18)	DTG-based (n = 53)
HIV characteristics for women on postconception ART
ART duration (months)^a
Median (IQR)	1.20 (0.63–2.00)	2.58 (1.33–3.73)	0.97 (0.47–1.87)
CD4 count (cells/µL)
≤350	20 (28)	6 (33)	14 (26)
351–500	14 (20)	4 (22)	10 (19)
>500	17 (24)	7 (39)	10 (19)
Median (IQR)	422 (275–578)	481 (293–579)	403 (260–574)

p < .05. Missing data: pre-pregnancy BMI n = 1, CD4 count n = 20. For maternal HIV characteristics, n = 1 on protease inhibitor-based ART.

ART, antiretroviral therapy; BMI, body mass index; DTG, dolutegravir; EFV, efavirenz; IQR, interquartile range; WWH, women with HIV.

Longitudinal weight trajectory analysis

At pre-pregnancy, WWH on pre- and postconception ART had slightly lower weight (WWH 78 kg vs without HIV 82 kg, p = .12) and gained weight more slowly throughout pregnancy [between pre-pregnancy and 24–28 weeks of GA (0.16 kg/week WWH vs. 0.23 kg/week without HIV, p < .01); between 24 and 28 weeks of GA and 33 and 38 weeks of GA (0.27 kg/week WWH vs. 0.38 kg/week without HIV, p = .03)] compared with women without HIV (Fig. 2A). However, the rate of weight loss between 33 and 38 weeks of GA and 6–12 months postpartum did not differ by HIV status (−0.13 kg/week WWH vs. −0.13 kg/week without HIV, p = .87). Among WWH on postconception ART, no differences were observed in the rate of weight gain in pregnancy, or the rate of postpartum weight loss between those who initiated EFV- and DTG-based ART (Fig. 2B).

FIG. 2.

Rate of weight gain/loss between consecutive visits from pre-pregnancy to postpartum period by HIV status (A) and postconception ART (B), the p-values indicate group differences in weekly rate of weight gain/loss between visits. Categorized postpartum total weight change (postpartum weight minus pre-pregnancy weight) by HIV status (C) and postconception ART (D) are also shown, categories were lost weight (<2 kg), stayed same (±2 kg) and gained weight (>2 kg). Risk Ratios (RRs) from modified Poisson regression in panels C and D show associations for HIV status with the risk of (i) losing weight versus staying the same and (ii) gaining weight versus staying the same and for associations of postconception DTG-based ART (vs. EFV-based ART) on the same weight gain/loss outcomes. Models in panels C and D were adjusted for maternal age, pre-pregnancy body mass index (BMI), socioeconomic status (SES), and postpartum months since delivery.

Between pre-pregnancy and postpartum, the overall mean total weight change was 2.33 kg [standard deviation (SD) = 8.45]; and 65/214 (30%) lost weight, 46/214 (22%) stayed the same, and 103/214 (48%) gained weight. A higher proportion of WWH on pre- and postconception ART lost weight [39/113 (35%) WWH vs. 26/101 (26%) without HIV] or stayed the same [31/113 (27%) WWH vs. 15/101 (15%) without HIV], and a lower proportion gained weight [43/113 (38%) WWH vs. 60/101 (59%) without HIV] postpartum compared with women without HIV; p < .01 (Fig. 2C). In models adjusted for age, pre-pregnancy BMI, SES, and postpartum months since delivery, WWH were less likely to gain weight postpartum [RR = 0.72 95% confidence interval (CI) 0.55, 0.93; p = .01] compared with women without HIV. In models adjusted for the same covariates, among WWH on postconception ART, there was no evidence that DTG initiation in pregnancy was associated with higher postpartum weight gain (RR = 1.72 95% CI 0.73, 4.02) compared with those who initiated EFV-based ART (Fig. 2D). Similarly, there were no apparent differences in trajectories of absolute weight over time (Supplementary Fig. S1A, B) and total weight change distribution (Supplementary Fig. S1C, D) by HIV status and ART regimen.

Cross-sectional postpartum anthropometry analysis

Overall, mean BMI at 6–12 months postpartum was 32 kg/m² (SD = 7.33): 2/214 (1%) underweight, 39/214 (18%) normal, 49/214 (23%) with overweight, and 124/214 (58%) with obesity. Obese BMI was prevalent among WWH, as well as those without HIV [60/113 (53%) WWH vs. 64/101 (63%) without HIV, p = .50] (Table 2). Using unadjusted regression models, WWH on pre- and postconception ART had significantly lower weight, BMI, as well as waist and hip circumferences, but all differences were attenuated after adjusting for confounders. In models adjusted for the same covariates, among WWH on postconception ART, although there were modest increases in hip circumference (mean difference = 1.95 95% CI −3.16, 7.05) and overweight (RR = 1.60 95% CI 0.74, 3.46) among women who initiated DTG-based ART compared with EFV-based ART, there was no evidence of significant differences in all postpartum anthropometry measures by ART regimen.

Table 2.

Differences in Postpartum Anthropometry by HIV and Postconception Antiretroviral Therapy Regimen

	HIV status				Postconception ART regimen
	Without HIV N = 101 mean (±SD)	With HIV N = 113 mean (±SD)	With HIV unadjusted mean difference (95% CI)	With HIV adjusted mean difference (95% CI)^a	EFV-based N = 18 mean (±SD)	DTG-based N = 53 mean (±SD)	DTG-based ART unadjusted mean difference (95% CI)^a	DTG-based ART adjusted mean difference (95% CI)^a
Weight (kg)	86.3 (19.9)	79.1 (19.8)	−7.14 (−12.49, −1.78)	−1.66 (−5.14, 1.82)	76.0 (17.6)	80.1 (21.0)	4.09 (−6.89, 15.07)	1.30 (−5.61, 8.20)
BMI (kg/m²)	33.7 (7.4)	31.0 (7.0)	−2.73 (−4.68, −0.79)	−0.66 (−1.65, 0.33)	29.7 (6.5)	31.0 (7.1)	1.35 (−2.44, 5.15)	0.20 (−1.53, 1.93)
Waist circumference (cm)	103.5 (16.3)	97.5 (13.6)	−5.93 (−9.96, −1.90)	−2.23 (−5.26, 0.81)	94.7 (13.1)	97.2 (13.6)	2.48 (−4.84, 9.79)	1.01 (−4.24, 6.25)
Hip circumference (cm)	117.4 (14.5)	112.1 (13.8)	−5.30 (−9.11, −1.49)	−2.41 (−5.35, 0.52)	109.3 (15.5)	113.2 (12.9)	3.92 (−3.46, 11.29)	1.95 (−3.16, 7.05)
Waist–hip ratio	0.88 (0.09)	0.87 (0.07)	−0.01 (−0.03, 0.01)	−0.05 (−0.03, 0.02)	0.87 (0.06)	0.86 (0.07)	−0.01 (−0.05, 0.03)	−0.01 (−0.05, 0.03)
Visceral adiposity index	1.4 (0.8)	1.3 (0.8)	−0.01 (−0.23, 0.20)	0.03 (−0.21, 0.27)	1.29 (0.61)	1.32 (0.89)	0.03 (−0.42, 0.48)	0.01 (−0.47, 0.46)
	N (%)		Risk ratio (95% CI)	Risk ratio (95% CI)	N (%)		Risk ratio (95% CI)	Risk ratio (95% CI)
BMI (kg/m²)
Underweight (<18.5)	1 (1)	1 (1)	0.52 (0.03, 7.94)	1.02 (0.89, 1.16)	1 (5)	0 (0)	Not estimable	Not estimable
Normal (18.5–24.9)	13 (13)	26 (23)	1.00 (ref)	1.00 (ref)	4 (22)	13 (25)	1.00 (ref)	1.00 (ref)
Overweight (25–29.9)	23 (23)	26 (23)	0.78 (0.54, 1.13)	0.90 (0.63, 1.28)	3 (17)	13 (25)	1.17 (0.45, 3.03)	1.60 (0.74, 3.46)
Obese (≥30)	64 (63)	60 (53)	0.84 (0.71, 1.00)	0.96 (0.82, 1.14)	10 (56)	27 (51)	0.05 (0.63, 1.41)	0.80 (0.60, 1.07)

Bold indicates p-value <.05.

Adjusted for age, pre-pregnancy BMI, socioeconomic status, and postpartum time since delivery; reference—women without HIV for HIV models and EFV-based ART for postconception ART models.

CI, confidence interval; SD, standard deviation.

Cross-sectional postpartum DXA-derived body composition and body fat distribution

WWH on pre- and postconception ART had lower body composition and fat distribution compared with those without HIV (Table 3). However, after adjusting for age, pre-pregnancy BMI, SES, and postpartum months since delivery, these differences were no longer significant. Among WWH on postconception ART, initiation of DTG-based ART compared with EFV-based ART in pregnancy did not affect postpartum body composition and fat distribution.

Table 3.

Differences in Postpartum Fat Distribution by HIV and Postconception Antiretroviral Therapy Regimen

	HIV status				Post-conception ART regimen
	Without HIV N = 13 mean (±SD)	With HIV N = 52 mean (±SD)	With HIV unadjusted mean difference (95% CI)	With HIV adjusted mean difference (95% CI)^a	EFV-based N = 13 mean (±SD)	DTG-based N = 29 mean (±SD)	DTG-based ART unadjusted mean difference (95% CI)^a	DTG-based ART adjusted mean difference (95% CI)^a
Fat mass (kg)	45.9 (14.6)	34.2 (13.3)	−11.73 (−20.15, −3.31)	−6.24 (−14.32, 1.85)	30.8 (11.4)	35.7 (14.5)	4.89 (−4.30, 14.08)	−2.28 (−10.49, 5.93)
Fat-free mass (kg)	45.6 (6.5)	39.8 (6.9)	−5.84 (−10.09, −1.59)	−4.12 (−8.49, 0.25)	36.5 (5.01)	41.2 (7.5)	4.71 (0.09, 9.32)	−0.44 (−4.67, 3.79)
Body fat (%)	49.1 (5.9)	45.3 (6.4)	−3.80 (−7.69, 0.95)	−0.43 (−3.59, 4.45)	44.4 (7.6)	45.8 (6.3)	1.35 (−3.18, 5.89)	0.59 (−3.42, 4.61)
Trunk FM (kg)	21.3 (7.6)	15.6 (6.6)	−5.71 (−9.91, −1.51)	−2.88 (−6.76, 0.99)	14.0 (5.7)	16.1 (7.3)	2.05 (−2.56, 6.66)	−1.79 (−5.52, 1.94)
Arm FM (kg)	5.7 (2.6)	4.3 (2.3)	−1.45 (−2.90, −0.01)	−0.02 (−1.84, 1.89)	3.5 (1.4)	4.5 (2.1)	0.99 (−0.28, 2.27)	0.33 (−1.15, 1.82)
Leg FM (kg)	17.9 (6.1)	13.9 (5.1)	−3.96 (−7.27, −0.65)	−1.82 (−4.86, 1.22)	13.3 (4.8)	14.9 (5.4)	2.62 (−0.91, 6.16)	0.20 (−2.78, 3.19)
Android FM (kg)	3.7 (1.5)	2.7 (1.3)	−1.02 (−1.85, −0.19)	−0.36 (−1.09, 0.36)	2.3 (1.1)	2.8 (1.4)	0.51 (−0.38, 1.40)	−0.17 (−0.88, 0.55)
Gynoid FM (kg)	7.9 (2.4)	6.2 (2.2)	−1.70 (−3.10, −0.30)	−0.59 (−1.93, 0.75)	5.6 (2.2)	6.5 (2.3)	0.87 (−0.68, 2.43)	−0.14 (−1.46, 1.17)
Visceral adipose tissue (VAT, cm²)	166.7 (64.7)	117.4 (54.0)	−49.22 (−84.05, −14.38)	−19.38 (−47.03, 8.27)	99.6 (48.7)	121.8 (54.9)	22.27 (−13.58, 58.13)	−1.19 (−27.93, 25.55)
Subcutaneous adipose tissue (SAT, cm²)	551.9 (168.4)	442.2 (158.3)	−109.67 (−208.97, −10.37)	−19.12 (−114.29, 76.01)	411.0 (157.3)	456.8 (168.3)	45.85 (−65.50, 157.20)	−21.92 (−119.24, 75.41)
VAT-SAT ratio	0.29 (0.07)	0.26 (0.06)	−0.04 (−0.07, 0.01)	−0.01 (−0.06, 0.03)	0.23 (0.05)	0.26 (0.07)	0.03 (−0.01, 0.07)	0.02 (−0.02, 0.06)

Adjusted for age, pre-pregnancy BMI, socioeconomic status, and postpartum time since delivery; reference—women without HIV for HIV models and EFV-based ART for postconception ART models.

FFSTM, fat-free soft tissue mass; FM, fat mass; SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue.

Discussion

In this cohort of women from a peri-urban setting in South Africa followed from pre-pregnancy to 6–12 months postpartum, we investigated longitudinal weight trajectories, and cross-sectional postpartum anthropometry, body composition, and fat distribution by HIV status and postconception ART class. HIV infection was associated with slower weight gain in pregnancy and reduced risk of postpartum weight gain. However, there were no significant differences by HIV status and postconception ART class for the other anthropometry and DXA-derived adiposity measures. Despite this, over one-third of both WWH and without HIV gained weight postpartum, and over half had obese BMI postpartum.

Although weight gain in pregnancy is needed for optimal development of the fetus, gaining too much or too little can adversely affect both maternal and child health outcomes.^2,31 WWH were lighter at antenatal care entry and they proceeded to gain weight more slowly in pregnancy, and were less likely to gain weight postpartum, compared with women without HIV. In addition, at the postpartum period, WWH had lower values of other anthropometry and body composition or fat distribution measures, compared with women without HIV. However, all differences were attenuated after accounting for potential confounders, including age, pre-pregnancy BMI, SES, and postpartum months since delivery. These results indicate that HIV-related differences in these characteristics drive the differences in anthropometry and body composition or body fat distribution but not the HIV infection itself. On the contrary, it is worth noting that WWH were compared with women without HIV who have considerably elevated measures of adiposity. Therefore, these findings of lower weight gain among WWH should not be equated to a higher risk of undernutrition, as it was the case in the pre-lifelong ART era.³²

Among WWH, several studies have reported increased weight and adiposity associated with DTG-based ART initiation.^9,10,33,34 However, the risk of metabolic disease depends on the body site or body fat depot in which the body fat is deposited. To our knowledge, there are no data on postconception DTG-based ART and body composition or body fat distribution assessed using a gold standard technique, such as DXA, in SSA postpartum women. Although we observed slightly higher weight gain between pre-pregnancy and postpartum, as well as modest increases in overweight and hip circumference postpartum in DTG- versus EFV-based ART initiators, there was no evidence of significant differences. These findings differ from randomized controlled studies conducted in South Africa, Uganda, and Botswana, which found significantly higher weight gain among postpartum WWH who initiated DTG-based ART in pregnancy compared with other ART classes.^11,12 In our study, we also found that initiation of DTG-based ART in pregnancy was not associated with postpartum trunk, visceral, or android FM, which are body fat depots linked to metabolic disease risk. This is in contrast to findings of greater increases in trunk FM in nonpregnant women on DTG-based ART followed to 144 weeks in the ADVANCE South African trial.¹⁸ A shorter follow-up time in our study might explain these discrepancies, highlighting a need for more studies in pregnant and postpartum women with longer follow-up periods.

For both WWH and without HIV, a notably high proportion entered pregnancy with obesity and gained weight and/or developed obesity postpartum. These rates are higher than those obtained in a study that was conducted in this setting 4 years ago,⁵ suggesting that postpartum obesity might be going up among WWH and without HIV in this population. For women residing in resource-limited settings, contributors to obesity are myriad, in part, involving the consumption of high-energy-dense foods due to their affordability, sedentary lifestyle, as well as pregnancy weight gain and related postpartum weight retention/gain,^4,5,35 in addition to the potential effect of ART regimen for WWH.³ Considering the greater baseline risk of metabolic comorbidities such as T2D and cardiovascular diseases among WWH,³⁶ high levels of postpartum weight gain and obesity, as observed among those without HIV, are likely to increase the burden of these metabolic complications over time. Regardless of HIV or ART effects, escalating levels of obesity among women of reproductive age in low- to middle-income countries are a public health concern that urgently needs attention.

Our study is not without limitations. First, we note that there was a variation in the timing of the postpartum visit attendance, which may have led to increased variability in anthropometry and adiposity measures assessed. However, to account for this, we restricted the analysis to 6–12 months postpartum, and included time since delivery (in months) in the regression models. Second, by restricting to women with a visit between 6 and 12 months postpartum (mostly due to COVID-19-related restrictions), we may have induced selection bias. However, the sample included in this analysis had similar demographic characteristics as the full cohort, suggesting that possible selection bias is unlikely to meaningfully change our results. In addition, we had a small sample size for postconception DTG-based versus EFV-based ART comparisons, which may have limited the detection of significant differences and may help to account for differences with some previous studies, suggesting higher postpartum weight gain for women on DTG.^11,12 The strengths of this study include that it is among the first to explore the effect of DTG-based ART on robust measures of body composition and body fat distribution among SSA women followed across the peripartum period, thereby contributing to addressing evidence gap in this topic in high HIV-burdened settings.

Conclusions

In South African women assessed from pre-pregnancy to postpartum period, we observed no significant differences in anthropometry, body composition, and fat distribution measures, including weight and body fat change by HIV status and postconception ART class. Our findings, particularly those using objective measures of body composition and fat distribution, are in a relatively small cohort. Larger cohort studies are needed to confirm the impact of DTG-based ART on maternal body composition and fat distribution using robust measures such as DXA, particularly in low- to middle-income settings where the majority of WWH reside. Despite this, over one-third of WWH and without HIV gained weight relative to before pregnancy and half had obese BMI by 6–12 months postpartum. These findings underscore the need for interventions to support healthy weight gain in pregnancy and postpartum weight loss to minimize pregnancy-associated obesity, which has the potential to affect long-term maternal and child metabolic health.

Authors’ Contributions

Conceptualization, investigation, methodology, and resources: H.P.M., L.M., J.J., L.D., A.E.M., G.P., S.C.-U., S.T.M., J.H.G., and A.M.B. Funding acquisition: H.P.M., L.M., S.C.-U., and A.M.B. Data curation and project administration: H.P.M., A.F., D.M., and A.E.M. Formal analysis and visualization: H.P.M. and H.G. Supervision: L.M. and A.M.B. Writing of original draft: H.P.M. All authors reviewed, edited, and approved the final article.

Footnotes

Acknowledgments

The authors thank the participants, clinic staff at Gugulethu Community Health Clinic, and the staff members of the “CAMP” study.

Author Disclosure Statement

The authors declare that they have no competing interests.

Funding Information

This work was supported by the Providence/Boston Center for AIDS Research (P30AI042853), the Population Studies and Training Center at Brown University (P2CHD041020), the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK U01-DK-18–018/019), and the Fogarty International Center at the National Institutes of Health (R21TW011678, K43TW012887).

Supplemental Material

References

Endres

, Straub

, McKinney

, et al.; Community Child Health Network of the Eunice Kennedy Shriver National Institute of Child Health and Human Development. Postpartum weight retention risk factors and relationship to obesity at 1 year. Obstet Gynecol, 2015; 125(1):144–152; doi: 10.1097/AOG.0000000000000565

Institute of Medicine. Weight Gain During Pregnancy:. Reexamining the Guidelines: Washington, DC; 2009; doi: 10.17226/12584

Caniglia

, Shapiro

, Diseko

, et al. Weight gain during pregnancy among women initiating dolutegravir in Botswana. EClinMed, 2020; 29–30:100615; doi: 10.1016/j.eclinm.2020.100615

Madlala

, Myer

, Jao

, et al. Gestational weight gain and adverse birth outcomes in South African women with HIV on antiretroviral therapy and without HIV: A prospective cohort study. J Int AIDS Soc, 2024; 27(6):e26313; doi: 10.1002/jia2.26313

Bengtson

, Phillips

, Le Roux

, et al. Postpartum obesity and weight gain among human immunodeficiency virus-infected and human immunodeficiency virus-uninfected women in South Africa. Matern Child Nutr, 2020; 16(3):e12949; doi: 10.1111/mcn.12949

Madlala

, Myer

, Geffen

, et al. Measurement of body composition in postpartum South African women living with and without HIV infection. Front Nutr, 2024; 11:1280425; doi: 10.3389/fnut.2024.1280425

Moyo-Chilufya

, Maluleke

, Kgarosi

, et al. The burden of non-communicable diseases among people living with HIV in Sub-Saharan Africa: A systematic review and meta-analysis. EClinMed, 2023; 65:102255; doi: 10.1016/j.eclinm.2023.102255

Baran

, Weres

, Czenczek-Lewandowska

, et al. Excessive gestational weight gain: Long-term consequences for the child. J Clin Med, 2020; 9(12):3795; doi: 10.3390/jcm9123795

Kouanfack

, Mpoudi-Etame

, Omgba Bassega

, et al.; NAMSAL ANRS 12313 Study Group. Dolutegravir-based or low-dose efavirenz-based regimen for the treatment of HIV-1. N Engl J Med, 2019; 381(9):816–826; doi: 10.1056/NEJMoa1904340

10.

Venter

WDF

, Moorhouse

, Sokhela

, et al. Dolutegravir plus two different prodrugs of tenofovir to treat HIV. N Engl J Med, 2019; 381(9):803–815; doi: 10.1056/NEJMoa1902824

11.

Jao

, Shan

, Legbedze

, et al. Dolutegravir use is associated with higher postpartum weight compared to efavirenz. CROI 2020. Boston MA. 8–11 March; 2020. Poster abstract 772.

12.

Malaba

, Tao

, Kenneth

, et al. Postpartum weight changes in women initiating DTG vs EFV in pregnancy: DolPHIN-2. CROI 2020. Boston MA. 8–11 March; 2020. Poster abstract 771.

13.

, Choi

, Lee

, et al. Changes in body composition are associated with metabolic changes and the risk of metabolic syndrome. J Clin Med, 2021; 10(4):745; doi: 10.3390/jcm10040745

14.

de Morais

, Azevedo

, de Freitas Rocha

, et al. Body fat is superior to body mass index in predicting cardiometabolic risk factors in adolescents. Int J Environ Res Public Health, 2023; 20(3):2074; doi: 10.3390/ijerph20032074

15.

Mtintsilana

, Micklesfield

, Chorell

, et al. Fat redistribution and accumulation of visceral adipose tissue predicts type 2 diabetes risk in middle-aged black South African women: A 13-year longitudinal study. Nutr Diabetes, 2019; 9(1):12; doi: 10.1038/s41387-019-0079-8

16.

Gupta

, Lanca

, Gan

ATL

, et al. The Association between body composition using dual energy x-ray absorptiometry and type-2 diabetes: A systematic review and meta-analysis of observational studies. Sci Rep, 2019; 9(1):12634; doi: 10.1038/s41598-019-49162-5

17.

Barbaro

. Highly active antiretroviral therapy-associated metabolic syndrome: Pathogenesis and cardiovascular risk. Am J Ther, 2006; 13(3):248–260; doi: 10.1097/01.mjt.0000162013.66614.16

18.

Venter

, Moorhouse

, Sokhela

, et al. The ADVANCE trial: Phase 3, randomised comparison of TAF/FTC+DTG, TDF/FTC+DTG or TDF/FTC/EFV for first-line treatment of HIV-1 infection. 23rd International AIDS Conference 2020;July 6-10, 2020. Abstract OAXLB0104.

19.

Bengtson

, Madlala

, Matjila

, et al. Associations of HIV and antiretroviral therapy with gestational diabetes in South Africa. AIDS, 2023; 37(13):2069–2079; doi: 10.1097/QAD.0000000000003678

20.

Bengtson

, Madlala

, Matjila

, et al. Cardiometabolic health in pregnancy and postpartum: Findings from a prospective cohort study in South Africa. Open Forum Infect Dis, 2024; 11(3):ofae093; doi: 10.1093/ofid/ofae093

21.

Han

, Abrams

, Sridhar

, et al. Validity of self-reported pre-pregnancy weight and body mass index classification in an integrated health care delivery system. Paediatr Perinat Epidemiol, 2016; 30(4):314–319; doi: 10.1111/ppe.12286

22.

Natamba

, Sanchez

, Gelaye

, et al. Concordance between self-reported pre-pregnancy Body Mass Index (BMI) and BMI measured at the first prenatal study contact. BMC Pregnancy Childbirth, 2016; 16(1):187; doi: 10.1186/s12884-016-0983-z

23.

Amato

, Giordano

. Visceral adiposity index: An indicator of adipose tissue dysfunction. Int J Endocrinol, 2014; 2014:730827; doi: 10.1155/2014/730827

24.

Chen

, Wang

, Lohman

, et al. Dual-energy X-ray absorptiometry is a valid tool for assessing skeletal muscle mass in older women. J Nutr, 2007; 137(12):2775–2780; doi: 10.1093/jn/137.12.2775

25.

Micklesfield

, Reid

, Bewerunge

, et al. A proposed method to measure body composition in obese individuals using dual-energy X-ray absorptiometry. Int J Body Comp Res, 2007; 5:147.

26.

Micklesfield

, Evans

, Norris

, et al. Dual-energy X-ray absorptiometry and anthropometric estimates of visceral fat in Black and White South African Women. Obes (Silver Spring), 2010; 18(3):619–624; doi: 10.1038/oby.2009.292

27.

Myer

, Stein

, Grimsrud

, et al. Social determinants of psychological distress in a nationally-representative sample of South African adults. Soc Sci Med, 2008; 66(8):1828–1840; doi: 10.1016/j.socscimed.2008.01.025

28.

Bush

, Kivlahan

, McDonell

, et al. The AUDIT alcohol consumption questions (AUDIT-C): An effective brief screening test for problem drinking. Alcohol Use Disorders Identification Test. Arch Intern Med, 1998; 158(16):1789–1795; doi: 10.1001/archinte.158.16.1789

29.

Labadarios

, McHiza

, Steyn

, et al. Food security in South Africa: A review of national surveys. Bull World Health Organ, 2011; 89(12):891–899; doi: 10.2471/BLT.11.089243

30.

Zou

. A Modified poisson regression approach to prospective studies with binary data. Am J Epidemiol, 2004; 159(7):702–706; doi: 10.1093/aje/kwh090

31.

Goldstein

, Abell

, Ranasinha

, et al. Association of gestational weight gain with maternal and infant outcomes: A systematic review and meta-analysis. JAMA, 2017; 317(21):2207–2225; doi: 10.1001/jama.2017.3635

32.

Villamor

, Saathoff

, Msamanga

, et al. Wasting during pregnancy increases the risk of mother-to-child HIV-1 transmission. J Acquir Immune Defic Syndr, 2005; 38(5):622–626; doi: 10.1097/01.qai.0000143601.48986.47

33.

Eckard

, McComsey

. Weight gain and integrase inhibitors. Curr Opin Infect Dis, 2020; 33(1):10–19.

34.

Bengtson

, Myer

, Abrams

, et al. INSTIs and weight gain in pregnancy. Lancet HIV, 2020; 7(10):e663–e665; doi: 10.1016/S2352-3018(20)30251-4

35.

Heart and Stroke Foundation South Africa. Obesity: A ticking time bomb in South Africa. National Obesity Week. 2017. Available from: https://www.heartfoundation.co.za/wp-content/uploads/2017/10/Obesity-%E2%80%93-a-ticking-time-bomb-1.pdf [Last accessed: July 8, 2025].

36.

Marcus

, Chao

, Leyden

, et al. Narrowing the gap in life expectancy between HIV-infected and HIV-uninfected individuals with access to care. J Acquir Immune Defic Syndr, 2016; 73(1):39–46; doi: 10.1097/QAI.0000000000001014

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.32 MB