Associations between maternal adiposity during pregnancy and neonatal brain cortical development Authors: Xiaoxu Na, Brooke L Yancey-Ward, Yan Xue, Charles M Glasier, Yongming Liu, Xiawei Ou Published: 2026 ABSTRACT AbstractBackground:Previous studies have suggested that obesity during pregnancy is associated with changes in child neurodevelopmental outcomes, indicating in utero effects of maternal obesity on offspring brain development.Purpose:This study aims to evaluate whether there are associations between maternal body fat mass percentage (a direct measure of adiposity) at different stages of pregnancy and neonatal cortical brain development.Materials and methods:169 mother/newborn dyads (79 boys and 90 girls) were included in this longitudinal and prospective study. Maternal fat mass percentage was assessed using air displacement plethysmography methodology at 12 and 36 weeks of pregnancy. At ~2 weeks of postnatal age, the neonates underwent a brain MRI during natural sleep without sedation using high-resolution 3D T1- and T2-weighted imaging. Multiple linear regression models, controlling for appropriate covariates, were used to evaluate associations between maternal fat mass percentage and neonatal cortical features (volume, thickness, and surface area), as well as to test for interaction effects between neonatal sex and maternal fat mass percentage on these outcomes.Results:Maternal fat mass percentage at 12 weeks of pregnancy negatively associated with neonatal cortical volume of left entorhinal cortex (β = −1.36, t = −3.53, P = 0.04), and left parahippocampal gyrus (β = −1.50, t = −3.31, P = 0.04); maternal fat mass percentage at 36 weeks of pregnancy negatively associated with neonatal cortical volume of left entorhinal cortex (β = −1.76, t = −3.25, P = 0.04), left parahippocampal gyrus (β = −2.26, t = −3.53, P = 0.04), and right temporal pole (β = −2.24, t = −3.19, P = 0.04). Sex interaction with maternal fat mass percentage at 12 weeks significantly associated with cortical volume in left pars triangularis (β = −8.28, t = −3.46, P = 0.05).Conclusion:Significant relationships were identified between maternal body fat mass percentage during pregnancy and neonatal cortical volume, suggesting potential in utero and sex-specific impacts of increased maternal adiposity on offspring brain development.Keywords: Maternal obesity, BMI, Fat mass, Pregnancy, Neonate neuroimaging, Cortical volume Introduction The most recent national data indicate that about 40% of adults younger than age 60 in the United States have obesity (body mass index [BMI] ≥30) (Emmerich et al., 2024), and it is more common in women than men. The data also showed a persistent upward trend in obesity over recent decades. Furthermore, obesity during pregnancy poses significant health risks not only for the pregnant woman but also for the offspring, including increased susceptibility to chronic diseases such as obesity, coronary heart disease, stroke, type 2 diabetes, asthma, and adverse neurodevelopmental outcomes (Godfrey et al., 2017). Recent neuroimaging studies have provided initial evidence linking maternal obesity to potential alterations in both structural and functional development of the offspring brain across a wide age range, from fetal stages through adulthood (Parsaei et al., 2024). Among these developmental windows, the neonatal time point is especially critical, as it captures the culmination of fetal programming and establishes the foundation for subsequent neurodevelopment. A few recent small-scale studies have identified negative associations between maternal obesity and newborn brain development, including lower cortical thickness in frontal regions (Na et al., 2021) and functional connectivity alterations in domains involving sensory cue processing, reward processing, cognitive control and motor control (Rajasilta et al., 2021; Salzwedel et al., 2019). Furthermore, several studies have associated maternal obesity with enduring alterations in offspring brain structure and function. For example, boys (but not girls) showed significant negative relationships between maternal pre-pregnancy BMI and hippocampal volume at ages 7–11 years (Alves et al., 2020); maternal overweight/-obesity was linked to altered brain activity and reduced connectivity in regions responsible for cognitive control at ages 4–6 years (Shapiro et al., 2020); and higher maternal BMI was associated with decreased neural reactivity to food cues in the child’s anterior cingulate cortex and dorsolateral prefrontal cortex following glucose intake at ages 7–11 years (Luo et al., 2021). The majority of studies reporting relationships between obesity and brain development have utilized BMI as the primary measure. Although BMI is widely used as a screening tool for obesity, it does not provide an accurate measure of adiposity, as it is limited by its reliance on a simple mathematical construct: weight divided by height squared (National Academies of Sciences, E., and Medicine 2023 2023). BMI-based classification cutoffs operate under the assumption that for a given height, any increase in body mass is primarily attributable to the accumulation of adipose tissue, thereby failing to account for concomitant gains in lean mass, bone mineral density and fluid retention. This can mask differences in body fat content among women with similar BMI values, especially for those who are pregnant. On the other hand, direct measure of body composition offers a more precise understanding of health and fitness by differentiating fat and lean body mass (Holmes and Racette, 2021). Research studies have demonstrated that body fat percentage is a better predictor of cardiovascular risk and mortality risk compared with BMI (Zeng et al., 2012; Anderer, 2025), indicating its superior utility in establishing the link to health outcomes. In addition, BMI can be particularly problematic near category thresholds, as these cutoffs may not accurately reflect an individual’s actual body fat percentage. Consequently, health risks associated with body composition independent of BMI are frequently overlooked. For example, participants within the normal BMI range who exhibit higher body fat percentages demonstrated a greater prevalence of dyslipidaemia, hypertension, and cardiovascular disease compared with those with lower body fat percentage (Romero-Corral et al., 2010). Early pregnancy fat mass percentage and caridometabolic risk factors were also associated with higher incidence of gestational diabetes and offspring with higher proportions of adiposity (Flanagan et al., 2025; Heard-Lipsmeyer et al., 2020). Among women with a normal pre-pregnancy weight, maternal fat mass accounts for about a quarter of total gestational weight gain (Barbour, 2012; Dalfra et al., 2022). On the other hand, for women with obesity, gestational weight gain in the third trimester is predominantly driven by significant increases in fetal weight and fat-free mass, as maternal fat mass decreases (Most et al., 2019). Studies with maternal body composition data will add significant value in understanding the impact of maternal adiposity on offspring neurodevelopment. To date, there is only one study with a small sample size that has identified negative associations between maternal fat mass percentage and neonatal cortical thickness in frontal regions (Na et al., 2021). Furthermore, despite potential differences in neonatal brain development associated with sex (e.g., larger intracranial and white matter volume but smaller cortical volume in boys as reported by one study (Khan et al., 2024)), many previous studies did not check the interaction with sex and further analyze brain development in boys and girls separately when studying relationships with maternal obesity, possibly due to sample size limitations. In this study, we report neuroimaging findings in neonates born to a well-characterized cohort of healthy pregnant women with or without obesity during pregnancy. Maternal body fat mass was measured at both early and late pregnancy, and its relationships with neonatal cortical features (including volume, thickness, and surface area) were reported. Interaction terms were also included in the models, followed by stratified analyses to investigate whether the relationships are sex-specific. Method Study participants This is part of the ongoing Maternal Obesity and Offspring Neuro-development study (the MOON study) which has concluded subject enrollment and completed all neonatal study visits. Pregnant women and their offspring were recruited for this longitudinal and prospective observational study. All procedures were performed in compliance with relevant laws and institutional guidelines and have been approved by the University of Arkansas for Medical Sciences Institutional Review Board (IRB protocol number: 239304; first approved on March 19, 2019). Inclusion criteria for the pregnant women were: singleton pregnancy; 11–13 weeks of gestation; ≥18 years of age, self-reported BMI at the time of enrollment between 18.5–25 (normal-weight: NW) or 30–50 (obese: OB). Exclusion criteria for the pregnant women were: hypertension, diabetes, or other preexisting medical conditions known or suspected to influence fetal growth; family history of psychological or neurogenetic disorders that may increase the risk of adverse neurodevelopmental outcomes in offspring; medications known to influence fetal growth; recreational drugs, nicotine or tobacco, or alcohol use while pregnant; and medical conditions developed during pregnancy (e. g. gestational diabetes, preeclampsia) known or suspected to influence fetal growth. In addition, neonates born preterm (<37 weeks of gestation), with congenital defects, intrauterine growth restriction, small for gestational age, macrosomia, hypoglycemia, low APGAR (appearance, pulse, grimace, activity, respiration) score (<7), or any other medical complications at birth suspected to affect growth and development were excluded. Maternal body fat mass percentage was measured at 12 weeks and 36 weeks of pregnancy using air displacement plethysmography (ADP) methodology (BodPod, Concord, CA) at Arkansas Children’s Nutrition Center. Family socioeconomic status of the neonates (including parental education and household income) was also obtained during the mother’s pregnancy visits. Neonates who were unable to complete an MRI examination of the brain were not excluded from the MOON study but were excluded in this report. In total, there were 169 mother/neonate dyads included in this report. Fig. 1 shows the flow-chart of the participants. Table 1 shows the demographic information of these participants. Flowchart of the study participants. Dash line represents participant dropout. Table 1Demographic information of study participants.Obese(N = 73)Normal-weight(N = 96)P valueMaternal BMI (kg/m2)35.0 ± 4.622.1 ± 2.0<0.001Maternal fat mass % at 12 weeks of pregnancy45.2 ± 5.828.7 ± 6.1<0.001Maternal fat mass % at 36 weeks of pregnancy42.8 ± 5.031.0 ± 5.4<0.001Maternal age at pregnancy (year)29.6 ± 4.729.2 ± 4.10.57Neonate sex (boys/girls)34/3945/510.97Gestational age at birth (day)274.2 ± 6.5274.9 ± 6.90.50Postmenstrual age at MRI (day)293.1 ± 9.5292.8 ± 8.90.83Neonate race (White / African American / Asian or more than one race)57/8/883/7/60.41Mother education (no college / partial college / bachelor degree / graduate training)9/14/36/147/22/42/250.47Father education (no data / no college / partial college / bachelor degree / graduate training)1/12/26/23/111/9/28/45/130.19Household annual income (≤$50,000/≤$100,000/≤$150,000/≤$200,000/>$200,000)17/30/19/6/111/46/26/12/10.32 Demographic information of study participants. MRI data acquisition At ~2 weeks of age, all neonates underwent an MRI examination of the brain during natural sleep without sedation on a 3 Tesla Prisma scanner (SIEMENS Healthineers). The neonates were swaddled with a Medvac infant immobilizer and securely positioned inside the coil. Mini-muffs and headsets were placed over the ears to protect the neonates from the noise. Pulse-oximeters were attached to a foot to monitor heart rate and blood oxygenation. The neonatal brain MRI protocol included 3D T1-weighted and 3D T2-weighted imaging and other sequences. Specifically, the 3D T1-weighted images were acquired with TR=2400 ms, TE=2.24 ms, flip angle=8°, and resolution=0.8 × 0.8 × 0.8 mm3, while the 3D T2-weighted images were acquired with TR=3200 ms, TE=564 ms, and resolution=0.8 × 0.8 × 0.8 mm3. Both were in sagittal orientation and had enough slices to cover the entire brain. Imaging quality control was performed on the scanner by experienced MRI technologists. The scans were repeated in case of severe motion artifacts; those who were unable to stay still and with unusable T1/T2 images were excluded from this study. MRI data processing MR images were exported to the local PACS system to screen for incidental findings needing medical attention by one neuroradiologist, and then exported for further processing using iBEAT V2.0 software (Developing Brain Computing Lab and Baby Brain Mapping Lab) (Wang et al., 2023). Processing steps included inhomogeneity correction, skull stripping, and tissue segmentation (gray matter, white matter and cerebrospinal fluid [CSF]). Left/right hemisphere separation, topology correction, and inner/outer cortical surfaces reconstruction were followed. Segmentation and surface reconstructions were visually inspected for quality control purposes, and the reconstructed surfaces were further parcellated based on UNC neonatal cortical surface atlas (Li et al., 2015). Mean cortical volume, thickness, and surface area of each of the 68 brain regions defined in the UNC atlas were extracted, and a total intracranial volume (ICV) was also computed for each subject. Statistical analysis For the comparison of demographic parameters between the OB and NW groups, Wilcoxon rank sum tests were performed to determine whether there were significant group differences (P < 0.05) for numeric parameters, while Chi-Square tests were used for categorical parameters. Maternal BMI was also tested for association with maternal fat mass percentage using Pearson’s correlation. To evaluate the relationships between maternal body fat mass during pregnancy and neonatal cortical features, multiple linear regression models were used. To control for potential confounders that may also impact the brain developmental trajectory, neonatal sex, postmenstrual age at MRI, race as well as mother’s age, parental education, and household income were included in the analyses (Bethlehem et al., 2022; Hackman et al., 2010; Du et al., 2021). Additionally, neonatal ICV was also controlled to account for the variance of brain size (Wang et al., 2024). The multiple linear regression models were repeated with an additional term of the interactions between sex and fat mass percentage at 12 weeks and 36 weeks respectively. False discovery rate (FDR) correction was used for multiple-comparison correction associated with the 68 brain cortical regions evaluated in total, and relationships with FDR-corrected P ≤ 0.05 were regarded as significant. For regions where the interaction between sex and maternal fat mass percentage was significant, the analysis was stratified by sex to further examine these relationships. Results Maternal BMI strongly correlated with maternal body fat mass percentage at 12 weeks (R = 0.89, P < 0.001) and 36 weeks (R = 0.83, P < 0.001), respectively. Fat mass percentage of women in the OB group was mostly higher than those in the NW group (Table 1); however, there was a considerable overlap of fat mass percentage range between the two groups of women at both 12 weeks (overlaps at 33% - 41%) and 36 weeks (overlaps at 33% - 45%) of pregnancy, respectively, despite that there was a clear gap in BMI between NW and OB (Fig. 2). Specifically, some individuals classified as having a healthy BMI nonetheless exhibited a body fat percentage that falls outside the recommended range (ACSM, 2021). These results suggested that BMI was a good but not optimal measure of maternal obesity/adiposity during pregnancy. Scatterplots of BMI and fat mass percentage at 12 and 36 weeks of pregnancy, respectively. BMI: body mass index, NW: normal weight, OB: obese. Significant associations (FDR-corrected P ≤ 0.05) between neonatal cortical brain features and maternal fat mass percentage at both time points were identified. Specifically, maternal fat mass percentage at 12 weeks of pregnancy was significantly associated with neonatal cortical volume of left entorhinal cortex (L.EC, β = −1.36, t = −3.53, P = 0.04), and left parahippocampal gyrus (L.PHI, β = −1.50, t = −3.31, P = 0.04); maternal fat mass percentage at 36 weeks was significantly associated with neonatal cortical volume of L.EC (β = −1.76, t = −3.25, P = 0.04), L.PHI (β = −2.26, t = −3.53, P = 0.04), and right temporal pole (R.TP, β = −2.24, t = −3.19, P = 0.04, Fig. 3). There were no other significant associations between maternal fat mass percentage at any time point during pregnancy and neonatal cortical volume of any other regions (or thickness or surface area in any regions). Maternal fat mass percentage during pregnancy was negatively associated with neonatal cortical volume in multiple brain regions. Significant interaction (FDR-corrected P ≤ 0.05) was identified between neonatal sex and maternal fat mass at 12 weeks on cortical volume in left pars triangularis (L.PTR, β = −8.28, t = −3.46, P = 0.05). There were no other significant associations between the interaction of neonatal sex and maternal fat mass percentage at any time point during pregnancy on neonatal cortical volume of any other regions (or thickness or surface area in any regions). In the sex-specific analyses, boys were found to have larger ICV compared with girls (524.7 ± 40.8 cm3 versus 484.1 ± 41.7cm3, P < 0.001). Additionally, maternal fat mass percentage at 12 weeks was significantly associated with cortical volume of L.PTR in boys (β = −5.14, t = −2.55, P = 0.01), but not in girls (β = 1.94, t = 1.23, P = 0.22, Fig. 4). Sex-specific associations between maternal fat mass percentage at 12 weeks and neonatal cortical volume in left pars triangularis (L.PTR). Discussion Using direct measure of adiposity during pregnancy, our study identified significant negative associations between maternal adiposity and regional cortical brain development in neonates among several brain regions, including left entorhinal cortex, left parahippocampal, and right temporal pole. Specifically, the higher the mother’s body fat mass percentage at early and late pregnancy, the lower the regional cortical volume in the neonatal brain. Additionally, sex-dependent association between maternal fat mass and cortical volume in left pars triangularis was identified. Our study identified relationships between maternal adiposity and neonatal cortical volume in the left entorhinal, left parahippocampal cortices and right temporal pole. Interestingly, these cross-generational findings in mother/neonate dyads are consistent with research findings in individuals with obesity. For example, participants with higher BMI were found to have lower brain volume in entorhinal-parahippocampal regions (Vainik et al., 2018). A review paper also suggested consistent findings in the literature of the associations between obesity and a reduction of gray matter volume in temporal areas (Gómez-Apo et al., 2021). Our findings are also consistent with findings from previous studies which have reported associations between maternal obesity and changes in fetal/neonatal brain development (Lee et al., 2023; Takeoka et al., 2025). While the exact mechanism in humans remain unclear, these effects are possibly linked to mechanisms observed in animal studies such as neuroinflammation, increased oxidative stress, dysregulated insulin, dysregulated serotonergic and dopaminergic signaling, and/or perturbations in synaptic plasticity (Edlow, 2016; Cirulli et al., 2020). The regions in the temporal lobe identified with significant associations with maternal adiposity in the present study grow rapidly during the prenatal and early postnatal periods, and play an important role in the function of memory and spatial and emotional processing. For example, the subareal differentiation of entorhinal gyrus begins as early as gestational week 13 when area-specific neurons and fiber-rich strata develop synchronously (Kostović et al., 2004). The entorhinal cortex serves as a gateway for information flowing between the neocortex and the hippocampus. It is connected to the perirhinal cortex and plays a vital role in memory and navigation (Schultz et al., 2015). The parahippocampal gyrus is larger and more medially positioned during the second trimester of gestation (Kier et al., 1997), crucial for memory encoding and retrieval, spatial navigation, as well as for processing visual scenes (Aminoff et al., 2013). The temporal pole serve as an interface, particularly for emotion-related memories and for social perception (Nakamura et al., 2001). Research studies have indicated that maternal obesity can contribute to elevated risks of offspring cognitive, social behavioral, and emotional problems (Sanchez et al., 2017). Our results suggest a potential association between temporally driven exposure during pregnancy and alterations in regional brain developmental, which may contribute to modified functional outcomes later in life. Our results also suggested a sex-specific interaction in the relationship between maternal adiposity and neonatal cortical volume in left pars triangularis. The left pars triangularis demonstrated a negative correlation with obesity in both neonatal and middle-aged male cohorts (Na et al., 2021; Franz et al., 2019). Some studies also suggest that boys’ brain is more vulnerable to insult or negative consequences of exposure to maternal obesity. For example, studies found reductions in hippocampal volume in boys but not in girls (aged 7–11 years) with prenatal exposure to maternal obesity (Alves et al., 2020). Pre-pregnancy overweight and obesity were also identified to be associated with lower IQ among boys, but not girls, at 7 years (Widen et al., 2019). However, other studies have indicated that girls could be more susceptible to consequences of maternal obesity. For example, maternal obesity showed stronger association with infant growth pattern and adiposity in girls than boys in the first two years of life (Heard-Lipsmeyer et al., 2020; Oostvogels et al., 2017). It is noteworthy that in animal studies, sex-specific findings were also observed. For example, in mice, one study demonstrated that maternal obesity induced more dysregulated genes in the male fetal brain, with pathway analyses highlighting functions in pro-inflammatory/immune signaling, cell death and adipogenesis (Edlow et al., 2016). Maternal obesity also resulted in spatial cognitive deficiency in the male offspring, by influencing the fatty acid metabolite profiles in the prefrontal cortex (Zhu et al., 2018). Nevertheless, reasons for the discrepancy in sex-specific effects and the exact underlying mechanism for the sex-specific effects are still unclear, and more studies in the future will be needed to investigate these. Our study further contributes to the field’s understanding about sex-specific relationships between maternal obesity and offspring brain development, using fat mass as a true measure of adiposity/obesity. Limitations for our study include: This was an association study, and therefore no causal relationship can be established; The exact underlying mechanisms for the associations found in our study remain to be determined; Given that adiposity is a continuous condition across pregnancy, the trimester-specific associations should also be interpreted with caution; We only have one time point in MR imaging for the cohort at neonatal age at this time and longitudinal MR imaging at later ages will be imperative to determine whether the observed differences persist into later childhood. Conclusion Our study identified significant negative associations between maternal fat mass percentage during pregnancy and neonatal cortical volume in multiple temporal brain regions. Our results suggest that there may be a negative impact of increased adiposity during pregnancy on offspring’s cortical brain development. Funding This project was supported in part by National Institutes of Health [grant number 1R01HD099099] and USDA-ARS [grant number 6026-10700-001-000D]. Footnotes CRediT authorship contribution statement Xiaoxu Na: Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. Brooke L Yancey-Ward: Writing – original draft, Writing – review & editing. Yan Xue: Formal analysis, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. Charles M. Glasier: Resources, Writing – review & editing. Yongming Liu: Methodology, Software, Validation, Writing – original draft, Writing – review & editing. Xiawei Ou: Conceptualization, Data curation, Funding acquisition, Resources, Supervision, Writing – review & editing. Declaration of competing interest The authors have no conflicts of interest to declare. Data availability statement The data underlying this article will be shared on reasonable request to the corresponding author. References Associated Data Data Availability Statement The data underlying this article will be shared on reasonable request to the corresponding author.