Prenatal adipose tissue estimation using early second trimester anthropometry

Introduction

Prenatal adipose tissue changes influence pregnancy and birth outcomes and reflect energy reservoir shifts that support gestation [1–3]. Visceral adipose tissue (VAT), which surrounds abdominal cavity organs, can influence metabolic health [4]. During pregnancy, ultrasound-assessed VAT and total adipose tissue (TAT) have been linked to adverse outcomes, including a 23% increased preeclampsia risk [5], twofold greater odds for gestational diabetes [6] and a 6–50% odds increase of large-for-gestational age [7, 8]. However, routine prenatal adiposity assessment is limited by access to imaging equipment, underscoring the need for reliable alternatives.

Current methods to assess prenatal adiposity are limited to research settings and are not widely available clinically [9]. Many available methods rely on assumptions that may be violated during pregnancy, like changes in fat-free mass hydration, which can alter tissue density and reduce the accuracy of adiposity estimates [10]. Computerized tomography (CT) for VAT and dual-energy X-ray absorptiometry (DXA) for TAT and VAT are used in non-pregnant individuals; however, they are contraindicated in pregnancy due to radiation exposure [9]. Bioelectrical impedance analysis (BIA) is used clinically but provides indirect adiposity estimates, assumes stable hydration, which is less applicable during pregnancy due to increased total body water and plasma volume, and cannot distinguish between fetal and maternal compartments, potentially leading to underestimated fat mass [9–11]. Air-displacement plethysmography with correction equations estimates body fat [12], but not adiposity distribution. Ultrasonography is commonly used to estimate VAT during pregnancy [5–8, 13–15], but image quality and accuracy decline with gestational progression due to uterine expansion [9, 15–18]. In contrast, MRI can precisely quantify TAT and VAT during pregnancy, but is costly and widely inaccessible [9, 19, 20]. In a trial that included whole-body MRI (n = 210), women from New York City with overweight or class I obesity experienced a 30% VAT increase from 15-weeks gestation to 59-weeks postpartum [21]. A large US healthcare system (Kaiser Permanente) study based in the Pacific Northwest assessed abdominal MRI (n = 50) and found a 56% increase in VAT and 15% increase in fat mass, which were both correlated with gestational weight gain (VAT, r = 0.63 and fat mass, r = 0.83; p < 0.001) in pregnant women with BMIs 17–39 kg/m² [22]. In contrast, large cohorts like the Avon Longitudinal Study of Parents and Children (ALSPAC), the NICHD Fetal Growth Studies (FGS), and Project Viva collect repeated anthropometric measures, but lack adiposity distribution measures and do not have prediction equations to estimate overall or regional fat mass, limiting the ability to distinguish changes in adipose tissue and its distribution [23–25]. Altogether, these findings underscore the need for accessible, validated approaches to assess prenatal adiposity.

Given TAT and VAT contribute to adverse pregnancy outcomes, anthropometric measurements may be a practical alternative to improve gestational adiposity estimation. In this study, we examined associations between anthropometrics and MRI-assessed TAT and VAT in the early second trimester, evaluated for differences by prepregnancy BMI category, and developed prediction equations for TAT and VAT. We hypothesized that anthropometric measures would be positively associated with TAT and VAT, and that these associations would vary by prepregnancy BMI.

Methods

Results

Demographic characteristics and prenatal measurements for the 61 participants are summarized in Table 1. Most had a healthy prepregnancy BMI with 40% classified as overweight or obesity. Participants were 24–45 years old, predominantly non-Hispanic, almost all had >12 years education, and over half were nulliparous (62%).

TAT was positively and more strongly correlated with BMI, WC, HC, MUAC, and TC (r = 0.76–0.92), while moderate correlations were observed between TAT with CC, iliac, triceps and subscapular SF (range 0.59–0.67). VAT demonstrated moderate positive correlations with WC, MUAC, HC, iliac and subscapular SF, and BMI (r = 0.48–0.60), with the strongest correlation observed between VAT and WC (r = 0.60), and weaker correlations with CC, TC, triceps, and thigh skinfolds (r = 0.25–0.42) (Table 2). When stratified by prepregnancy BMI category, TAT demonstrated strong correlations with WC (r = 0.77), and HC (r = 0.92) among participants with healthy BMI (Supplementary Table 1), and stronger correlations with WC (r = 0.84), MUAC (r = 0.83), and HC (r = 0.92) for overweight or obesity (Supplementary Table 2). VAT showed moderate correlations with WC (r = 0.47) among participants with healthy BMI and moderate correlations for WC (r = 0.60) and MUAC (r = 0.52) for overweight or obesity. Observed and predicted TAT and VAT for anthropometric measures across BMI categories are shown in Figs. 1 and 2.

Discussion

This cohort provides insight into associations between anthropometric measurements with MRI-assessed maternal adipose tissue in the early second trimester of pregnancy. Circumferences and skinfolds were differentially associated with TAT and VAT, with HC as the strongest predictor for TAT (b = 0.51, CI: 0.36, 0.66, p < 0.001). Most circumferences (waist, mid-upper arm, hip) and some skinfolds (iliac, subscapular) were highly correlated with TAT while calf, thigh, and triceps measures were weaker. Stratified by prepregnancy BMI, predictive models had a stronger fit for TAT (R² = 0.86–0.93) compared to VAT (R² = 0.60–0.79), indicating anthropometrics may better capture total adiposity in early pregnancy.

The adiposity predictor models differed by prepregnancy BMI category. Among women with healthy BMI (n = 36), MUAC and iliac SF were positively associated with both TAT and VAT, suggesting that peripheral and trunk subcutaneous measures capture variation in both total and visceral adiposity in this group. In contrast, among women with overweight or obesity BMI (n = 25), WC and HC were positively associated with TAT, while HC and subscapular SF were associated with VAT, indicating a greater contribution of central and upper-body adiposity measures. Across the sample, HC emerged as a consistent predictor of TAT, while BMI was the primary predictor of VAT. These patterns are consistent with early pregnancy physiologic changes, where the hip region is a key depot for subcutaneous fat storage and energy reserves, particularly in the first half of pregnancy [2], contributing to TAT but less directly to VAT [28, 29].

Our findings align with prior research in both pregnant and non-pregnant women comparing anthropometry to gold-standard adiposity measures. A cross-sectional study conducted with pregnant women in Brazil (n = 149) found that MUAC and subscapular SF were strongly correlated with ultrasound-measured VAT (r = 0.60 MUAC and r = 0.59 subscapular SF) and TAT (r = 0.79 MUAC and r = 0.74 subscapular SF) than prepregnancy BMI, suggesting that alternative measures may better capture adiposity during pregnancy [16]. This is consistent with our results in women with healthy BMI, where MUAC and iliac crest SF were associated with both TAT and VAT, highlighting the utility of peripheral and trunk subcutaneous measures. Similarly, a longitudinal study of pregnant women in the US (n = 50) reported that gestational weight gain was positively correlated with increases in MRI-measured VAT (r = 0.63), SAT (r = 0.74), and total body fat (r = 0.83), with variation by pre-pregnancy BMI [22], supporting our observation that predictors of adiposity differ by BMI category. Together, these findings support our results that anthropometric measures beyond BMI are informative for estimating adiposity and tracking metabolically relevant fat depots across pregnancy [30].

In a study on non-pregnant women with overweight or obesity in China (n = 561, mean age 31 years), CT-measured VAT was strongly correlated with WC (r = 0.86) [31]. This aligns with our findings among women with overweight or obesity, where WC and HC were associated with TAT, and HC and subscapular SF were associated with VAT, suggesting a greater contribution of central and upper-body adiposity. Similarly, a French study on non-pregnant women with type 1 diabetes or obesity in midlife (n = 71, mean age 46.5 ± 15.42 years) found that TC, WC, and BMI predicted VAT measured by CT or DXA, shown in this predictive model: VAT = 2.15 × WC − 3.63 × Thigh Circumference + 1.46 × Age + 6.22 × BMI − 92.713; R² = 0.836 [32]. This is consistent with our observation that central anthropometrics are more informative of VAT in higher-BMI groups, whereas TC was not predictive in our cohort, likely reflecting its contribution to peripheral subcutaneous fat [33, 34]. Age differences can also influence VAT depots, as our cohort was younger (33 years) than the French cohort (46 years), supporting the evidence of adiposity tracking with age and menopausal changes [2].

Additional MRI-based studies further support our findings. An Austin-based cohort of non-pregnant White and Hispanic women with obesity (n = 55) found that MRI-measured VAT was positively associated with central obesity depth, defined as the mid-point between the umbilicus and widest frontal plane region (b = 11.48 cm) and waist-to-hip ratio (b = 204.74 cm), but was inversely associated (b = −5.09 cm) with central obesity width, defined as the largest umbilical width (R² = 0.71; mean error = −2.00 and 95% CI = −20.36, 16.37), and also found positive associations between MRI-measured TAT with WC (b = 7.10 cm) and sagittal diameter (b = 5.74 cm; R² = 0.89; mean error = −32.04 and 95% CI = −90.70, 26.61) [35]. The aforementioned findings were most similar to ours, as WC was also a strong predictor of TAT in women with overweight or obesity BMI. Further, the European Prospective Investigation into Cancer and Nutrition (ERIC) cohort of non-pregnant women (n = 300), found that BMI, WC, and HC best estimated MRI-measured TAT (predictive equation: 2.3 * 10⁻¹⁵ + 0.905 × BMI (kg/m²) + 0.087 × WC (cm) + 0.318 × HC (cm) [36], which is most similar to our findings of HC as a robust, consistent predictor of TAT across BMI category.

The study’s primary limitation is the small cohort size, which prevented validation and can restrict clinical application. Additionally, the sample was predominantly White, had healthy weight, was of high socioeconomic status, nulliparous, and due to the study design did not include women with underweight (<18.5 kg/m²), Class II (35.0–34.9 kg/m²), or Class III (>40.0 kg/m²) BMI categories, prepregnancy diabetes, or other chronic health conditions, which constrains generalizability to healthy pregnancies. Further, VAT model fit varied by prepregnancy BMI, limiting predictive application. Despite this, the study advances understanding of early second-trimester TAT estimation by providing accessible anthropometric alternatives to gold-standard methods like MRI. While BIA is more cost-effective and used in clinical settings, it provides indirect estimates of adiposity, relies on stable hydration assumptions, and cannot distinguish between fetal and maternal tissues. BIA also does not capture adipose tissue distribution and may underestimate fat mass, whereas MRI offers precise quantification of TAT and VAT [9, 10]. However, both are often infeasible in large epidemiologic cohorts due to cost and logistical constraints. This highlights the need for scalable, non-invasive approaches for estimating maternal adiposity in population-based studies. Our findings highlight future research needs, including tracking TAT and VAT changes throughout gestation and examining how adiposity estimates align with MRI measures to predict maternal cardiometabolic outcomes.

Conclusions

Within our prospective cohort, circumferences and skinfolds were associated with TAT and VAT in early pregnancy, with associations differing by BMI. Although WC is a strong predictor of TAT outside of pregnancy, our findings suggest that other anthropometric measures may also reflect pregnancy adiposity, particularly HC. Further investigation is warranted to evaluate the use of anthropometrics for TAT and VAT in the second trimester to advance current understanding of prenatal adiposity and its relation to maternal health.

Supplementary information

Author contributions

The authors’ responsibilities were as follows – EMW: had oversight over data collection; EMW, SD, and MRA: conceptualized this analysis; MRA: conducted data cleaning and analysis; EMW, SD, JL, IJ, JV, AND, LMH, AGC, and MRA: contributed to the analysis planning and development; and EMW, SD, DAP, JL, IJ, JV, AND, ARN, RRR, SFF, LMH, AGC, DD, and MRA interpreted the data, contributed to the manuscript during drafting and editing, and approved the final manuscript.

Funding

The Mother Infant NuTrition Study (MINT) cohort was supported by the Eunice Kennedy Shriver National Institute of Child Health & Human Development (R00HD086304, PI: Widen). The MINT Study was funded in part by The University of Texas at Austin Texas Biologics, Dell Medical School’s Health Transformation Research Institute, the College of Natural Sciences, and the School of Human Ecology at The University of Texas at Austin. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Data availability

Data described in the manuscript will be made available upon reasonable request pending application and approval by EMW. Analytic code will be available from the first author upon request, pending application and approval.

Competing interests

The authors declare no competing interests.

Ethics approval

The Institutional Review Board at The University of Texas at Austin approved all study procedures (IRB Approval Number: 2018050127), and each participant signed a written, IRB-approved consent form. ClinicalTrials.Gov Registration: NCT04132310.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41430-026-01801-7.