Background and Aims: Assess the effects of a supervised aerobic exercise intervention during pregnancy on region-specific one-month-old neonatal body composition and fat distribution.

Methods: Eligibility criteria included pregnant women <16 weeks of gestation, singleton pregnancy, aged 18 – 40 years, and BMI between 18.5 kg∙m 2 and 35.0 kg∙m 2. Eligible pregnant women (n=35) were randomly assigned to one of two groups: a moderate-intensity aerobic exercise (n=18) or a non-exercising group (n=17) consisting of light stretching and breathing techniques. Women exercised for 24 or more weeks during pregnancy.

Results: While no group-wide differences were observed in visceral fat, prenatal aerobic exercise significantly reduce body fat and BMI in offspring born to mothers with overweight or obesity, with distinct sex-specific adaptations in lean mass and adiposity. Prenatal aerobic exercise tended (p ≤ .10) to reduce one-month morphometrics in infants born to mothers with overweight or obesity. Trends for infant sex-specific responses to exercise were noted in a smaller subgroup. Female offspring to exercisers showed decreased (-3%; p = .01) body fat compared to control, but no difference in lean body mass percentage or body mass index. On the contrary, male offspring to exercisers showed a 3 kg∙m2 lower (p=.01) BMI than control and showed a non-significant general increase (+3%) in lean body mass. Infant sex, and participation in prenatal aerobic exercise were both consistent significant predictors for morphometric outcomes.

Conclusion: This study shows that improvements in one-month body composition or fat distribution in infants of prenatal aerobic exercising mothers are not region-specific.

Maintaining maternal health during the gestational period has been shown to promote healthy development of infants [1, 2] and physical activity contributes to gestational and offspring wellbeing [3, 4]. Prenatal exercise interventions can adjust factors such as body composition in offspring [5–9]. Further, infant body composition is associated with health outcomes in childhood [10] and later in life [11], and both show a positive response to prenatal exercise [12–15]. However, the effect of body composition on health is more complicated than once thought.  In non-gravid adults, increase in subcutaneous, visceral, and intramuscular adiposity, are independently associated with long-term health outcomes [16, 17]. It has been shown that healthy infants carry a high amount of subcutaneous adipose tissue, as compared to visceral [11, 18, 19]. Despite the strong association of visceral fat and metabolic health in adults [16, 17, 20], visceral adiposity in infants remains understudied. Investigations of distribution of adipose tissue in developing infants have been called for [19] considering that increases in body fat in the visceral and subcutaneous compartments could underly neonatal health decrements [21]. While our group has shown that prenatal aerobic exercise reduces skinfold thickness and body fat % at 1-month in infants [12], changes in the subcutaneous and visceral distribution of adipose tissue in response to prenatal exercise remain unknown. The examination of different infant body composition metrics in response to prenatal exercise will elucidate any improvements in offspring body composition and promote the assessment of long-term health in this population. Additionally, it is important to probe for specific benefits in infants born to mothers with overweight or obesity [22].  Given the former, the purpose of this study was to assess the effects of a supervised aerobic exercise intervention during pregnancy on one-month-old neonatal body fat distribution and composition. We hypothesized that one-month-old infants born to aerobic exercise-trained women would exhibit lower visceral and subcutaneous body fat and decreased infant body fat % compared to one-month-old infants born to non-exercising women.

Study Design and Participants This study was a prospective, partially blinded, two-arm randomized controlled trial aiming to determine the relationship between maternal aerobic exercise on one-month infant body fat location and amount compared to non-exercise exposed infants [23]. Women were recruited from local obstetric clinics via brochures, flyers, word-of-mouth, and social media. Inclusion criteria for this study were as follows: 1) singleton pregnancy ([16 weeks of gestation), 2) between 18 and 40 years of age, 3) pre-pregnancy BMI between 18.5 and 35.0 kg/m2, 4) physician clearance to participate in an exercise program, and 5) able to communicate fluently in English and be contacted via phone or email. Women were excluded from the study if they: 1) had pre-existing medical conditions (e.g., diabetes), or comorbidities known to affect fetal development (e.g., lupus), 2) were taking medications known to affect fetal development or pregnancy outcomes, or 3) were using tobacco, alcohol, or other recreational drugs. Informed consent was obtained from each participant prior to study enrollment. This study was approved by the East Carolina University Institutional Review Board and registered on the Clinicaltrials.gov website (Registry #NCT03517293). This research study was implemented without patient involvement. Patients were not invited to comment on any aspect or phase of the study (e.g., study design, recruitment, data collection). All women consented prior to participation. Between 2015 and 2018, 162 pregnant women were assessed for eligibility. Of these, 140 were randomized to either the aerobic exercise group (n=70) or a non-exercising control group (n=70). Of those randomized, 15 participants did not receive their assigned intervention due to group refusal (n=14) and a miscarriage (n=1); during the pregnancy intervention, 26 pregnant women were lost-to-follow-up consequent to leaving the geographical area, no time for participation, bed rest or an unknown reason. Of the remaining 99 pregnant women (aerobic=55 and control=44) completing the intervention, two were excluded for drug use, 56 declined or had incomplete neonatal measures, and six had preterm births. The final analytical sample was 35 pregnant women (aerobic=18 and control=17). While limited by a sample size of n=35, the high analytical sensitivity of MRI allows for precise quantification of tissue changes, warranting smaller cohorts than traditional field studies that are reliant on skinfold estimations. 

Pre-Intervention and Testing Randomization 

Prior to the exercise intervention, participants completed a submaximal exercise treadmill test to determine individualized target heart rate (THR) ranges of moderate intensity. Peak oxygen consumption (VO2peak) was estimated via the modified Balke protocol previously validated and replicated by Mottola et al. for pregnant women [24]. After completing this test, participants were randomized to an aerobic exercise or non-exercising control group determined via computerized sequencing (GraphPad software). 

Exercise Intervention 

Each exercise session began with a 5-minute warm-up and ended with a 5-minute cool-down (treadmill speed <3.0 mph). The aerobic exercise groups participated in individual supervised, moderate-intensity (60-80% VO2peak) exercise sessions for 50 minutes, 3 times per week. Intensity was monitored via previously validated THR in pregnant women [24, 25]. Aerobic exercise was performed using treadmills, elliptical machines, stationary bicycles, rowing and/or stair-stepping equipment. The non-exercising group were offered the opportunity to attend low-intensity (<60% VO2peak) stretching and breathing sessions that focused on standing or sitting stretches of major muscle groups while incorporating breathing techniques. Supervised exercise and stretching/breathing sessions took place at one of two university-affiliated gyms. 

Maternal Exercise Adherence and Dose

Exercise session attendance was tracked via an electronic record in REDCap [26] and calculated by dividing the number of sessions attended by the total number of possible sessions within the participants’ gestational period. Participants were considered “exercise adherent” if their attendance was ≥ 80% of possible exercise sessions. Maternal exercise intensity in metabolic equivalents (METs) was determined based on the published compendium of physical activity for the exercise performed in each session [27, 28]. Average maternal exercise dose throughout pregnancy, expressed as MET∙min∙wk-1, was quantified (frequency X duration of session) then multiplied by the intensity (METs) level of their exercise. The total for all weeks was summed and averaged for the pregnancy exercise dose or MET∙min∙wk-1 value [27].

Neonatal Magnetic Resonance Imaging [MRI]

Neonatal MRIs occurred at the one-month postnatal visits. The MRI sequence used was based on a previous study in infants [29]. In brief, the protocol uses a rapid T1-weighted spin-echo sequence (repetition time of 600 ms, echo time of 16 ms). Scans were done from clavicle-to-knee, taking approximately six minutes. Prior to each scan, infants were fed, swaddled, placed on the MRI table, and provided ear protection for safety. Mothers or staff remained in the room to comfort the infant, if needed. The MRI scan was performed with trained radiology technicians at East Carolina MRI. Images were analyzed using Mimics software (Version 21, Materialise NV, Leuven, Belgium). Images were obtained with a slice and interslice thickness of 5 mm that has been widely used in body composition studies.  Body composition and distribution calculations were performed by two blinded, pediatric radiologists using a Slice-O-Matic (version 5.0, Tomovision, Magog, Quebec, CA). 

Neonatal Morphometrics

At one-month of age, neonatal weight (kg), length (cm), body fat (%), abdominal, head, and mid-upper arm circumferences (mm)] and fat-free mass (g)]were obtained. Weight and length were measured using a standard, calibrated infant scale and horizontal stadiometer, respectively. If the horizontal stadiometer was not accessible, height was obtained using a Gulick tape measure while the infant was laying supine. Abdominal, head, and mid-upper arm circumferences were collected using a Gulick body measuring tape and following standard anthropometric protocols. The standard BMI calculation was used for one-month infants: BMI=((weight (kg))÷([height (m^2 )])). Weight to Length ratio was also calculated in kg/cm. Skinfold thickness was measured via calibrated Lange calipers at three designated anatomical sites on the right side of the infant’s body: biceps, triceps and subscapular. Values for these sites were summed to determine the sum of skinfold thickness. The skinfold thickness data was then used to calculate percent body fat (BF%) using the following equation by Slaughter et al. [1988] [30]: 

Lean Mass and Lean Mass volume were then converted to a percentage based on body size.

Visceral (VAT) and subcutaneous (SAT) adipose tissues were assessed from a single slice of MRI images, located at L2-L3 for female and L1-L2 for males, based on the study of O’Connor et. al. [31]. Images were analyzed using Mimics software (Version 21, Materialise NV, Leuven, Belgium). Using pre-established Hounsfield unit (HU)]thresholds, adipose tissue was segmented using values of -190 to -30 HU [32, 33]. Then VAT and SAT were segmented from each other using the inner boundary of the abdominal muscle wall [31]. Directly measured body composition variables were VAT and SAT at cross-sectional areas (cm2). These values were then indexed for height and were expressed in units of cm2/m2 [34, 35]. Visceral to subcutaneous fat was expressed as a ratio (VAT/SAT) [36] as a correlate of cardiometabolic risk. All MRI measurements were performed by two investigators, AV and AP, blinded to group allocation. Based on infant age and sex, z-scores were calculated for the following neonatal measures BMI, Weight to length ratio, lean body mass, sum of skinfolds, body fat %, visceral fat %, subcutaneous fat %.

Maternal and Neonatal Covariates

Maternal age, gravida, parity, pre-pregnancy weight and height, gestational weight gain (GWG), race (white, black, other minority), gestational age, infant birth weight and infant sex were abstracted from various sources including pre-screening eligibility questionnaires as well as maternal and neonatal electronic health records. Gestational weight gain was calculated using the standard expression: GWG (lbs)=(〖weight〗_(at delivery)- 〖weight〗_(before pregnancy)). Weight at delivery [lbs.] was extracted from electronic health records while weight before pregnancy (lbs) was collected via the pre-screening eligibility questionnaires. In cases where weight at delivery was unavailable, the last recorded study weight at 36 weeks was used. Pre-pregnancy BMI (healthy BMI 18.0-24.49; Overweight 24.5-29.99, Obese >30) was calculated using height (m), measured by stadiometer, and weight (kg) collected from the pre-screening eligibility questionnaire via the following established equation [37]: 

BMI= ((weight (kg)) ÷ ([height (m2)])).

Two-sample t-tests were performed to test between-group differences in maternal and neonatal descriptive characteristics; Mann Whitney U tests were used for gravida and parity due to non-normality. Primary morphometric outcomes were neonatal visceral fat %, subcutaneous fat %, lean mass %, body fat %; while secondary outcomes were neonatal BMI, BMI z-score, weight to length z-score, skinfold thicknesses of the biceps, triceps and subscapular skinfolds, sum of skinfold thickness, % body fat; head, abdominal and mid-upper arm circumferences, and VAT/SAT.  Analyses were performed initially using the intention-to-treat approach, including all participants with complete data, and next using the per protocol approach by including ‘exercise adherent’ participants, defined as those attending ≥ 80% of total possible exercise sessions. For the per protocol analysis, we then performed two stratified analyses: 1) by maternal pre-pregnancy BMI (<25 BMI=Healthy Weight, >25 BMI=Overweight/Obese), and 2) by infant sex (Males, Females). Based on significant findings, we then presented data with the Per Protocol, Overweight/Obese participants and stratified by infant sex. Lastly, multiple linear regression models were run to determine if maternal exercise measures (group, intensity, duration, dose) predicted the neonatal morphometric outcomes after controlling for potential covariates: infant sex, maternal age, pre-pregnancy BMI, race, parity, and gestational weight gain. Statistical analyses were performed using SAS, version 9.4 (SAS Inc., Cary, NC, USA). A two-tailed significance level of 0.05 based on the t-distribution was adopted for all statistical tests. Corrections for multiple comparisons were not applied to maintain power for this exploratory infant MRI data.

Participant Recruitment and Retention

A total of 162 pregnant women were assessed for eligibility, with 140 randomized to either the aerobic exercise (n=70) or control group (n=70). Major reasons for attrition included group refusal (n=15), loss-to-follow-up (n=26), and incomplete neonatal measures (n=56), primarily due to the infant MRI protocol. The final analytical sample for the intention-to-treat (ITT) analysis was 35 (Aerobic=18; Control=17). For the per-protocol (PP) analysis, three participants were excluded due to low adherence (<80%), resulting in a sample of 32 pregnant women (Aerobic=15; Control=17).

Study Population

Maternal descriptive characteristics are summarized in Table 1. Women in the aerobic and non-exercising groups were, on average, 30 years of age, had a BMI ~26 kg∙m2, and delivered full-term infants. As intended by the study design, significant between-group differences were observed for exercise intensity and average weekly exercise dose in both intention-to-treat and per protocol analyses (Table 1).

Table 1: Maternal demographics, by intervention group and exercise adherence.

Neonatal Morphometric Analyses

In the primary analysis of the full cohort (ITT and PP)] there were no statistically significant between-group differences in one-month infant morphometrics, including visceral adipose tissue (VAT)] subcutaneous adipose tissue (SAT)] or total body fat percentage (Table 2). When the PP sample was stratified by maternal pre-pregnancy BMI, infants born to mothers with overweight or obesity (OWOB)]in the aerobic group exhibited a trend toward lower BMI z-scores (p = .10)]and a trend toward increased lean mass percentage (p = .09) compared to controls (Table 3). Further stratification by infant sex within the OWOB cohort revealed distinct dimorphic responses to prenatal exercise (Table 4)] The female infants of aerobic exercisers demonstrated significantly lower sum of skinfolds and total body fat percentage (p < .05) compared to controls. No significant differences were found for BMI or lean body mass. Male infants of the OWOB aerobic exercisers showed a significantly lower BMI (p = .01) and exhibited a trend toward increased lean body mass with lower body fat %, compared to male infants of controls (Table 4). No significant trends for the ratio of VAT/SAT were seen in exercisers vs. control.

Table 2: One-Month Infant body morphometric data by intervention group and exercise adherence.

Table 3: PER PROTOCOL: One-Month Infant morphometric data by Maternal Pre-Pregnancy BMI and intervention.

Table 4: PER PROTOCOL: Morphometric data from One-Month Infants of Overweight-Obese Women.Regression Analyses

Regression analyses were conducted to determine the ability of maternal exercise variables (group, intensity, duration, attendance, dose) to predict neonatal morphometric outcomes. Infant sex (female) was a consistent predictor of neonatal morphometric outcomes, as well as group allocation. As shown in Table 5, performance of prenatal aerobic exercise reduced one-month infant body fat 2.3 %. Exercise group allocation, as well as the dose (MET*min/wk) and duration of exercise training had a trend (p<.10) for a reduction of infant subcutaneous fat, and skinfold sum. No statistically significant effects of exercise were observed for infant visceral fat distribution, or the ratio of visceral-to-subcutaneous fat distribution.

Table 5: Regression coefficients for effects of prenatal aerobic exercise on neonatal morphometrics.

Overall Findings

In this study, we hypothesized that one-month infants born to aerobic-trained women would exhibit lower visceral body fat, lower subcutaneous fat, as well as decreased total body fat compared to the infants of the non-exercising counterparts. The data summarized in Table 1 delineates a clear relationship between exercise intensity and tissue adaptation. While these findings confirm that all exercise contributes to overall health, the recorded metrics suggest that achieving specific metabolic thresholds acts as a primary catalyst for the morphometric changes observed via MRI. This distinction highlights that while general activity is beneficial, intensity-dependent stimuli are likely the necessary driver for the specific morphological adaptations noted throughout the training season.  Specifically, our main findings are as follows: 1) infants, especially females, of aerobic exercising OWOB women have decreased skinfolds and body fat, 2) aerobic exercise predicts infant morphometric changes in a dose-dependent manner, and 3) there were no significant changes in infant body fat distribution between groups.

Comparison to Previous Findings

Similar to previous findings, this report shows that prenatal aerobic exercise results in positive adaptions on infant body fat [12]. Broadly, one-month infant BMI, body fat %, and skinfold measurements were all decreased in offspring from aerobically exercising mothers. Aerobic exercise seemed especially effective in the offspring of OWOB mothers and could therefore be recommended in this population during gestation to help offset the negative health outcomes in the infant. Importantly, aerobic exercise is relatively easy to implement as part of a healthy pregnancy; walking or cycling at moderate intensity for 150 minutes per week was sufficient to improve infant morphometrics. Average exercise intensity in metabolic equivalents (METs), and exercise dose throughout pregnancy (MET*min per week) approached an effect on numerous one-month infant morphometric measurements. The potential effect of exercise intensity and dose should be investigated further, as it will inform decisions on target intensities and durations of exercise training bouts for future recommendations made to pregnant women. Previous randomized control trials have also reported a significant relationship between maternal exercise and infant adiposity, reporting that infants born to women who exercised aerobically during pregnancy had lower adiposity at birth, one-month of age, and 6-months of age, when assessed using skinfold measurements [12, 13, 40].  Our lab has shown that women who exercise aerobically throughout mid to late pregnancy have infants with lower body fat % compared to infants of women who did not exercise [12]. In comparison, due to equipment and technician availability, the present study had a smaller sample size of control participants, which is a limitation of the present study and potential reason for these discrepancies. Alternatively, women enrolled in the present study were “healthier” with lower pre-pregnancy BMIs and thus, may have diluted some of the effect of aerobic exercise on infant morphometric outcomes [12, 40]. These differences suggest that changes in pre-pregnancy body weight may influence changes to postnatal total infant body fat.  Prior to the stratification of data, there were no significant differences in infant body fat % at one month in infants born to women who participated in prenatal aerobic exercise compared to no exercise. Observational research has found that newborns of women who participated in physical activity in mid to late pregnancy had lower body fat % assessed via air displacement plethysmography compared to those who did not [15, 38]. Similarly, those who participated in moderate-intensity physical activity in late pregnancy had newborns with increased lean mass assessed via dual-energy x-ray absorptiometry [39]. Interestingly, despite no significant differences in neonatal body fat distribution, we observed that female infants of aerobic exercising OW/OB women have decreased skinfold measurements and body fat %. However, no differences were observed between groups for female infant lean body mass or BMI. In contrast, male offspring of exercisers show decreased BMI, with trends for increased lean body mass and decreased body fat %. 

Collectively, these are important, novel, sex-specific findings that support aerobic exercise as a therapeutic strategy to offset the altered body composition observed in OW/OB offspring. Broadly, our findings suggest a reduction in OW/OB offspring risk for developing overweight or obesity, and most likely the cardiometabolic complications associated with such. Future studies should include a more diverse selection of exercise types and intensity levels to parse out other effective means for improving body composition in OW/OB offspring. Research is also needed to determine how these findings may be physiologically significant with respect to childhood health risk later in life.

There are several strengths of the current study. First, to the best our knowledge, this is the first study to evaluate the effects of supervised aerobic exercise at recommended levels during pregnancy on one-month-old neonatal body fat composition and distribution. The findings of this study, specifically regarding neonatal adiposity, strengthen and extend the growing evidence base for promoting aerobic exercise throughout pregnancy. Second, we employed a prospective, randomized controlled trial study design, providing the strongest design for determining causality from the exercise intervention. In addition to strengths, there are several potential limitations. First, our sample consisted of ‘apparently healthy’ pregnant women, reducing the generalizability of our findings to those with complicated pregnancies. Second, the sample size collected for our study was underpowered for neonatal weight, circumferences, and fat-free mass. The former notwithstanding, the present study was able to draw meaningful conclusions regarding the influence of exercise during pregnancy on infant body composition. There is an increased risk of Type 1 errors due to multiple comparisons; however, given the novel nature of MRI-measured infant fat distribution, we prioritized discovery and effect size over conservative alpha adjustments. Third, other potentially important covariates associated with differences in infant body composition were not included, for example, dietary patterns and breastfeeding. Finally, aerobic exercise of moderate intensity was the only intervention studied, and future studies will further specify which types and intensities of exercise may also benefit offspring.

While the current study demonstrates clear structural improvements in one-month neonatal body fat distribution following supervised maternal aerobic exercise, the underlying biochemical mechanisms warrant further investigation. Future longitudinal cohorts should aim to pair localized tissue imaging with maternal and neonatal plasma analyte profiling. Specifically, investigating the role of metabolic and anti-aging genetic markers, such as Sirtuin 1 (SIRT1), could elucidate how maternal exercise drives these morphometric adaptations. SIRT1 expression, which increases in response to aerobic exercise, is highly associated with enhanced lipid metabolism, cellular longevity, and the regulation of metabolic homeostasis [41, 42]. Tracking changes in maternal SIRT1 levels throughout pregnancy alongside neonatal adiposity would provide vital insight into the systemic epigenetic and metabolic pathways bridging maternal physical activity and infant body composition. Consequently, integrating biomolecular screening into future exercise interventions remains a critical recommendation for validating these physiological outcomes across diverse global populations [43].

In conclusion, we found that participation in moderate-intensity aerobic exercise during pregnancy could influence the relative increase in bodyweight in male and female offspring of OW/OB mothers. Additionally, positive changes in body composition, i.e., reduction in body fat and increase in lean body mass % were seen in male offspring. Aerobic exercise did not lead to changes in infant visceral or subcutaneous body fat distribution. However, male- and female-specific improvements in lean mass, skinfolds, and body fat % associated with prenatal aerobic exercise were observed. Furthermore, the percentage of visceral adiposity trended toward a reduction in male offspring from exercising mothers. These findings further confirm that participation in approximately 150-minutes per week of aerobic exercise during pregnancy is safe and effective for baby and show positive changes to infant body composition, especially for infants of women with overweight or obesity.

Acknowledgements

We would like to thank the pregnant women and their babies for their participation and time in this study. 

Data Availability

De-identified data will be made available via secure electronic file transfer upon request, from 01/01/2023 until 01/01.2028.

Conflicts of Interest

The authors report no conflicts of interest.

Funding Statement

This study was funded in part by the American Heart Association (AHA grant #15GRNT22470029) and by ECU internal funds.

Author Contributions: 

Breanna WISSEMAN: Writing – original draft, Writing – review & editing; Alex CLAIBORNE: Writing – review & editing; Writing – review & editing; Filip JEVTOVIC: Writing – review & editing; Ali VAHDATI: Analysis & curation of data; Samantha M. MCDONALD: Writing – original draft; Edward NEWTON: Project administration, Writing – review & editing; Devon KUEHN: Project administration, Writing – review & editing; George A. KELLEY: Project administration, Writing – review & editing; Tara Whiton: Writing – review & editing, Linda E. MAY: Conceptualization, Methodology, Data curation, Formal analysis, Funding acquisition, Investigation, Project administration, Resources, Software, Supervision, Validation, Writing – review & editing.