Maternal Adiposity is Associated with Fat Mass Accretion in Female but not Male Offspring During the First 2 Years of Life
Abstract
Objective
This study investigated which antenatal and postnatal factors determine offspring adiposity during the first 2 years of life.
Methods
Participants were mother and child pairs (N = 224). Offspring percent fat mass (%FM) was obtained using quantitative nuclear magnetic resonance at 11 time points between ages 0.5 and 24 months. Independent variables included race, age, gestational weight gain, first-trimester %FM, delivery mode, gestational measures of resting energy expenditure, respiratory exchange ratio, physical activity, serum cytokines and lipids, and dietary intake for the mothers, as well as sex, birth weight and length, breastfeeding duration, and physical activity at age 2 years for the children. Linear mixed models were used to construct the best-fitted models for the entire cohort and for each sex.
Results
Maternal %FM (P = 0.006), high-density lipoprotein (HDL) (P < 0.001), and breastfeeding duration (P = 0.023) were positively associated with female offspring adiposity, whereas maternal dietary fiber intake (P = 0.016) had a negative association. Birth weight (P = 0.004), maternal HDL (P = 0.013), and breastfeeding duration (P = 0.015) were all positively associated with male offspring adiposity.
Conclusions
Antenatal and postnatal factors differentially impact male and female offspring adiposity during the first 2 years of life.
Study Importance
What is already known?
- Maternal obesity influences offspring body composition in a sexual dimorphic manner.
- Obesity-induced metabolic and inflammatory factors as well as maternal diet and physical activity have been independently associated with offspring body composition.
What does this study add?
- The study identifies maternal first-trimester fat mass and serum high-density lipoprotein (HDL) cholesterol, infant sex, and breastfeeding duration as significant determinants of offspring adiposity from birth to age 2 years.
- Maternal first-trimester adiposity and dietary fiber intake during pregnancy are associated with adiposity only in girls.
- Early-life adiposity in boys is associated with their birth weight but not determined by maternal fat mass.
How might these results change the direction of research?
- Further research is needed to understand how maternal HDL impacts offspring body composition and metabolism later in life.
- These results provide possible windows of opportunity to mitigate obesity in offspring through the development of intervention strategies aimed at reducing maternal cholesterol and adiposity and improving maternal diet prior to or during pregnancy.
Introduction
The prevalence of obesity among women of childbearing age is > 20% (1). In addition, there is a persistent increase in obesity rates among the pediatric population, with 18.5% of children in the United States classified as having obesity (BMI-for-age percentile > 95%) (2). Clinical and animal studies have shown that maternal obesity during gestation can program fetal metabolism that results in increased risk for developing obesity (3, 4). Infants born to mothers with overweight and obesity experience increased adiposity at birth (5) and accelerated postnatal weight gain (6) and carry a 2-fold risk for developing obesity by 2 years of age (7).
BMI is not a true measure of adiposity and may obfuscate the contribution of excess adipose tissue in women with similar BMI to offspring body composition (8). In addition to excess maternal fat mass (FM), there may be additional maternal factors that predict offspring adiposity, such as excessive gestational weight gain (GWG) (9), poor quality diet through gestation (10), and physical activity (PA) (11). Furthermore, obesity-induced metabolic disturbances such as increased inflammation (12) and dyslipidemia (13) have been independently associated with infant adiposity.
Identification of maternal and postnatal factors that determine offspring body composition is critical in order to aid in the design of successful prevention strategies, as the prevalence of prepregnancy overweight and obesity is increasing (1). To overcome this gap in knowledge, we aimed to identify determinants of offspring adiposity during the first 2 years of life in children born to healthy mothers with noncomplicated pregnancies. We further investigated the role of sex in offspring body composition, as prior publications have demonstrated a sexual dimorphism in programming of obesity (6). We hypothesized that maternal adiposity, serum lipids, insulin resistance (estimated by homeostatic model assessment-insulin resistance [HOMA-IR]), GWG, and PA during gestation would be associated with offspring adiposity during the first 2 years of life in a sexual dimorphic manner.
Methods
Participants
Participants were women and their offspring enrolled in a longitudinal cohort, the Growing Life, Optimizing Wellness study (GLOWING) (NCT01131117) at the Arkansas Children’s Nutrition Center, Little Rock, Arkansas. Women were enrolled either before pregnancy (n = 37) or within the first 10 weeks of gestation (n = 187). Inclusion criteria included BMI of 18.5 to 35 at enrollment, second parity, singleton pregnancy, ≥ 21 years old, and conception without assisted fertility treatments. Exclusion criteria included preexisting or ongoing medical conditions, including gestational diabetes, complications during pregnancies, use of medications during pregnancy known to influence fetal growth, smoking during pregnancy, alcohol consumption in any amount, and being a professional athlete. Infants born healthy and at term (≥ 37 weeks gestation) were eligible for postnatal visits. All study procedures were determined to be in compliance with the ethical research standards set by the Institutional Review Board of the University of Arkansas for Medical Sciences. Written consent was obtained and signed by the participants before any study procedures.
Body composition
Whole-body air displacement plethysmography (BodPod; Cosmed, Concord, California) was used to determine maternal FM (kilograms) at enrollment (prepregnancy or 4-10 weeks gestation) under standard conditions. The van Raaij density model was used to determine body composition parameters in pregnant mothers to account for differences from the pregravid state (14). Infant body composition was assessed utilizing quantitative nuclear magnetic resonance (QMR) (EchoMRI-AH small; Echo Medical System, Houston, Texas) at 2 weeks and at 1 to 9, 12, 18, and 24 months of age using mathematical adjustment as previously published (15). Briefly, the following adjustments were applied: if QMRFM ≤ 5.0845 kg, then adjusted QMRFM = 0.1160 (QMRFM [kilograms])2 + 0.5690 (QMRFM [kilograms]); if QMRFM > 5.0845 kg, then adjusted QMRFM = 0.8962 (QMRFM [kilograms]) + 1.3337.
Anthropometry
Anthropometric measures were obtained at each study visit using standardized methods. Maternal body weight was measured to the nearest 0.1 kg on a tared scale (Perspective Enterprises, Portage, Michigan). Standing height was measured using a standard wall-mounted stadiometer to the nearest 0.1 cm (Tanita Corp., Tokyo, Japan) at enrollment.
Objectively assessed PA
PA was assessed with accelerometry (Actical; Philips Respironics Co. Inc., Bend, Oregon). The Actical was placed on the participant’s ankle on the nondominant side and worn continuously for up to 7 days. To be included in the analyses, each participant needed to record at least 3 valid days of accelerometer data. Total activity counts (AC) per day were summed over the valid wear period and then divided by the total number of valid days worn to derive average total AC per day. For mothers, data were collected at 2 time points per trimester and averaged to yield gestational AC. In children, data were collected within a week of the 24-month visit. The use of the Actical to objectively assess PA in young children has been previously validated (16). We used the cut point of 0.083 kcal/min/kg (6.0 metabolic equivalent) to evaluate vigorous AC.
Maternal serum cytokines, lipids, and HOMA-IR
Blood was drawn from the antecubital vein following an overnight fast at 4 to 10, 12, 24, and 36 weeks of gestation. Serum was processed and stored at −80°C until analysis. Serum glucose (millimoles per liter), triglycerides (milligrams per deciliter), HDL (milligrams per deciliter), and low-density lipoprotein (LDL) (milligrams per deciliter) levels were measured using an RX Daytona (Randox Laboratories-US limited; Kearneysville, West Virginia) clinical analyzer. Serum insulin (milliunits per liter), IL-6 (picograms per milliliter), and leptin (picograms per milliliter) levels were measured using electrochemiluminescence immunoassays (Meso Scale Discovery, Rockville, Maryland). All samples were run in duplicate and averaged for use in subsequent analyses. Results from the 4- to 10-week and 12-week period were averaged to represent the first trimester. HOMA-IR was calculated from fasting glucose and insulin concentrations for the first trimester.
Resting energy expenditure and respiratory exchange ratio
Resting energy expenditure (REE) (kilocalories per day) and respiratory exchange ratio (RER) (VCO2/VO2) were derived from oxygen consumption and carbon dioxide production using indirect calorimetry in mothers during gestation after an overnight fast (Moxus; AEI Technologies, Bastrop, Texas). REE was calculated using the abbreviated Weir equation as follows: REE (kilocalories per day) = ([3.94 × VO2] + [1.11 × VCO2]). Participants were instructed not to exercise or consume caffeine for 12 hours before the measurement. After 10 minutes of absolute rest (adaptation period), respiratory exchanges were continuously measured for 20 more minutes. Heart rate was constantly monitored using a Polar telemetry system (Polar Electro, Bethpage, New York) to monitor the participant’s safety and evaluate the participant’s steady state. A 10-minute steady state period was selected to evaluate REE and RER. Data were collected at 4 to 10, 12, 24, and 30 weeks gestation. Values from 4 to 10 and 12 weeks gestation were averaged to yield the first-trimester estimates in REE and RER. Values from 24 weeks gestation were used as estimates of REE and RER for the second trimester and finally values from 30 weeks gestation were used as estimates for REE and RER for the third trimester.
Dietary intake
Dietary intakes were assessed using 3-day food records analyzed with the Nutrition Data System for Research software 2018 (Nutrition Coordinating Center, University of Minnesota, Minneapolis, Minnesota) following the interview guidelines from the Nutrition Coordinating Center. Participants were instructed to record all food, beverage, medication, and supplement intake as they ate throughout the day for 2 weekdays and 1 weekend day. Participants were provided with a booklet to evaluate size or volume of intake. Participants recorded the volume and size of intake for each component of their meal, including drinks, medications, or solid foods on the food records. The records were reviewed by a trained nutritionist to clarify any information and query any additional intake that may have been missed. The 3-day food records were collected at 4 to 10 and 12 weeks of gestation and every 6 weeks thereafter. Nutrient intakes were averaged for each trimester and reported as kilocalories, percent kilocalories, or grams per day. The Healthy Eating Index, which measures diet quality to assess how well intake aligns with key recommendations of the Dietary Guidelines for Americans, was computed according to the current guidelines (10, 17). Sensitivity analyses were performed to evaluate results while excluding women who underreported energy intake using a validated cutoff of energy intake greater than 1.35 × basal metabolic rate (BMR) (18).
Other covariates
At enrollment, participants self-reported their race, age, date of last menstrual period, and estimated delivery date. Delivery mode, infant’s birth weight, and infant’s sex were self-reported by the participant at the 2-week postnatal visit. Gestational age at birth was calculated from the date of their last menstrual period or from the estimated due date obtained by early ultrasound if the date of their last menstrual period was unknown. GWG was calculated by subtracting the participant’s weight at the first visit from weight at the final prenatal visit (36 weeks of gestation). Breastfeeding duration of infants was determined based on 3-day food records completed by mothers at 2 weeks and at 1 to 9, 12, 18, and 24 months of age. At each study visit, dietary records were reviewed with the mothers, during which breastfeeding patterns were identified. Duration was calculated by subtracting the date breastfeeding was initiated (the infant’s date of birth) from the date breastfeeding was stopped. Thus, the length of breastfeeding may include mixed feeding patterns or inclusion of solid foods.
Statistical analysis
Variables measured in the interval scale are summarized as means and SDs, whereas variables recorded in the ordinal or nominal scale are summarized as counts and percentages. The association between offspring percent FM (%FM) during the first 2 years of life and each maternal/infant characteristic was assessed using regressions, whereas mixed models were used to model offspring %FM during the first 2 years of life and all characteristics simultaneously using a time-varying covariate approach when measures were available at each trimester (e.g., REE, cytokines). The most parsimonious model was constructed using least absolute shrinkage and selection operator linear regression. In light of the sexual dimorphism observed in animal models and clinical studies in response to maternal obesity (6, 19, 20), models were further built independently for males (n = 128) and females (n = 96). Data management and analysis were performed using Stata software version 14.2 (StataCorp., College Station, Texas) and R version 3.6.1 (R Foundation, Vienna, Austria), whereas the recursive partitioning was performed using JMP Pro version 13.0 statistical software (SAS Institute, Inc., Cary, North Carolina).
Results
Between 2011 and 2014, 290 eligible women with BMI between 18 and 35 were enrolled in the GLOWING study at the Arkansas Children’s Nutrition Center. Of the enrolled participants, 14 voluntarily withdrew from the study, 13 were lost to follow-up, 6 did not get pregnant, 15 developed gestational diabetes, and 11 suffered miscarriage or stillbirth. Seven babies were born prematurely. In total, 224 mother and child pairs were included in the current study, of which 37 were recruited prior to pregnancy. A total of 102 participants included in the study were classified as normal weight, 81 had overweight, and 41 were classified with obesity. Thirty participants have partial offspring %FM data between the ages of 2 weeks and 24 months as they were lost to follow-up or withdrew from the study during that time frame. Phenotypic results of mother and infant pairs that were used in subsequent linear regression analyses are listed in Table 1 and Supporting Information Table S1. Infant FM measures obtained from QMR are presented in Supporting Information Table S2.
| Maternal variables | ||
| Race | ||
| Caucasian | 195 (87.1%) | |
| Non-Caucasian | 29 (12.9%) | |
| Delivery method | ||
| Vaginal induced | 65 (29.0%) | |
| C-section | 78 (34.8%) | |
| Vaginal | 81 (36.2%) | |
| Age at delivery (y) | 30.2 (3.6) | |
| Weight before gestation week 10 (kg) (n = 37) | 66.8 (9.2) | |
| Fat mass (% fat) a | 35.7 (7.4) | |
| Fat mass index (fat mass [kg]/height [m]2) | 9.6 (3.4) | |
| RER (gestation) | 0.85 (0.04) | |
| REE (kcal/d, gestation) | 1,528.7 (184.5) | |
| Gestational weight gain (kg) | 11.9 (4.2) | |
| Gestational weight gain (g/wk) | 0.43 (0.17) | |
| Gestational age (wk) | 39.2 (0.9) | |
| Physical activity counts (avg/d, ×1,000) (gestation) | 309.1 (89.8) | |
| Serum (gestation) | ||
| HOMA-IR a | 1.7 (0.9) | |
| Leptin (pg/L) | 26.5 (22.4) | |
| IL-6 (pg/mL) | 0.42 (0.35) | |
| Triglycerides (mg/dL) | 136.9 (66.3) | |
| HDL (mg/dL) | 63 (15.6) | |
| LDL (mg/dL) | 95 (44) | |
| Dietary intake (gestation) | ||
| Energy (kcal/d) | 1,947 (447.8) | |
| Fat (% kcal/d) | 35.6 (4.9) | |
| Carbohydrate (% kcal/d) | 50.5 (5.9) | |
| Protein (% kcal/d) | 15.5 (2.8) | |
| Dietary fiber (g/d) | 17.5 (5.9) | |
| Added sugar (g/d) | 62.9 (31.5) | |
| Healthy Eating Index | 48.6 (8.2) | |
| Males | Females | |
| Offspring variables | ||
| n | 128 (57.1%) | 96 (42.9%) |
| Birth weight (kg) | 3.6 (0.5) | 3.4 (0.5) |
| Birth length (cm) | 51.5 (2.6) | 50.5 (2.6) |
| Breastfeeding duration (d) | 244.7 (222.6) | 241.5 (212.7) |
| Physical activity at 2 years of age | ||
| Total activity counts (avg/d, ×1,000) | 497.9 (117.8) | 436.7 (110.9) |
| Sedentary activity counts (avg/d, ×1,000) | 4.9 (7.6) | 4.7 (7.5) |
| Vigorous activity counts (avg/d, ×1,000) | 14.0 (23.5) | 14.6 (62.8) |
- Data are presented as means (SDs), counts, or percentages.
- a Designates variables accounted for in first trimester only.
- RER, respiratory exchange ratio; REE, resting energy expenditure; HOMA-IR, homeostatic model assessment-insulin resistance; IL-6, interleukin 6; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
Maternal gestational serum HDL and LDL (β = 0.049, P < 0.0001 and β = 0.013, P = 0.0009, respectively) were significantly, independently, and positively associated with offspring %FM (Table 2). Male sex had a significant negative independent association with child %FM (β = −1.206, P = 0.0136), whereas infant birth weight (β = 1.304, P = 0.0083) and length of breastfeeding (β = 0.004, P = 0.0013) were positively and independently associated with child adiposity during the first 2 years of life.
| β | 95% CI | P | |
|---|---|---|---|
| Maternal variables | |||
| Race | |||
| Caucasian | Reference | ||
| Non-Caucasian | 0.003 | −1.452 to 1.457 | 0.9971 |
| Delivery method | |||
| Vaginal induced | Reference | ||
| C-section | −0.216 | −1.421 to 0.989 | 0.7253 |
| Vaginal | 0.321 | −0.876 to 1.518 | 0.599 |
| Age at delivery (y) | −0.03 | −0.165 to 0.105 | 0.6657 |
| Fat mass (% fat) a | 0.063 | −0.002 to 0.127 | 0.0591 |
| RER (gestation) | −2.006 | −6.392 to 2.380 | 0.3700 |
| REE (kcal/d, gestation) | 0.0010 | −0.0004 to 0.0024 | 0.1622 |
| Gestational weight gain (kg) | 0.05 | −0.066 to 0.166 | 0.3958 |
| Gestational age (wk) | 0.347 | −0.179 to 0.872 | 0.1961 |
| Physical activity counts (avg/d, ×1,000) (gestation) | 0.003 | −0.003 to 0.008 | 0.3161 |
| Serum (gestation) | |||
| HOMA-IR a | 0.314 | −0.274 to 0.902 | 0.2957 |
| IL-6 (pg/mL) | −0.078 | −0.841 to 0.686 | 0.8417 |
| Leptin (pg/mL) | 0.002 | −0.013 to 0.017 | 0.8189 |
| Triglycerides (mg/dL) | 0.003 | −0.003 to 0.008 | 0.2871 |
| HDL (mg/dL) | 0.049 | 0.027 to 0.071 | < 0.0001 |
| LDL (mg/dL) | 0.013 | 0.005 to 0.021 | 0.0009 |
| Dietary intake (gestation) | |||
| Energy (kcal/d) | −0.0003 | −0.0009 to 0.0004 | 0.4185 |
| Total fat (% kcal/d) | 0.071 | −0.052 to 0.194 | 0.2602 |
| Total carbohydrate (% kcal/d) | −0.076 | −0.176 to 0.024 | 0.1341 |
| Total protein (% kcal/d) | 0.115 | −0.092 to 0.322 | 0.2775 |
| Total dietary fiber (g/d) | −0.031 | −0.082 to 0.019 | 0.2234 |
| Added sugars (by total sugars) (g/d) | 0.002 | −0.009 to 0.014 | 0.6799 |
| Healthy Eating Index | 0.011 | −0.024 to 0.045 | 0.5426 |
| Offspring variables | |||
| Sex | |||
| Female | Reference | ||
| Male | −1.206 | −2.164 to −0.248 | 0.0136 |
| Birth weight (kg) | 1.304 | 0.336 to 2.271 | 0.0083 |
| Birth length (cm) | 0.129 | −0.055 to 0.314 | 0.1696 |
| Breastfeeding duration (d) | 0.004 | 0.001 to 0.006 | 0.0013 |
| Physical activity at 2 years of age | |||
| Total activity counts (avg/d, ×1,000) | −0.004 | −0.009 to 0.000 | 0.0806 |
| Sedentary activity counts (avg/d, ×1,000) | 0.214 | −0.504 to 0.931 | 0.5598 |
| Vigorous activity counts (avg/d, ×1,000) | −0.0002 | −0.0125 to 0.0121 | 0.9758 |
- a Designates variables accounted for in first trimester only.
- RER, respiratory exchange ratio; REE, resting energy expenditure; HOMA-IR, homeostatic model assessment-insulin resistance; IL-6, interleukin 6; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
The best-fitted model revealed that only maternal %FM (β = 0.08, P = 0.022), maternal gestational HDL (β = 0.05, P ≤ 0.001), birth weight (β = 1.27, P = 0.016), offspring sex (β = −1.40, P = 0.004), and breastfeeding duration (β = 0.004, P = 0.001) were determinants of offspring adiposity during the first 2 years of life (Table 3).
| Variable | β | 95% CI | P |
|---|---|---|---|
| Maternal fat mass (% fat) (first trimester) | 0.08 | 0.01 to 0.14 | 0.022 |
| Maternal HDL (mg/dL) (gestation) | 0.05 | 0.03 to 0.07 | < 0.001 |
| Sex | |||
| Female | Reference | ||
| Male | −1.40 | −2.35 to −0.45 | 0.004 |
| Birth weight (kg) | 1.27 | 0.24 to 2.30 | 0.016 |
| Breastfeeding duration (d) | 0.004 | 0.001 to 0.006 | 0.001 |
- HDL, high-density lipoprotein.
Given the sexual dimorphism findings in our previous study (6), the analysis was then stratified by child sex. For girls, maternal %FM (β = 0.12, P = 0.006), maternal HDL (β = 0.07, P < 0.001), maternal intake of dietary fiber during pregnancy (β = −0.09, P = 0.016), and length of breastfeeding (β = 0.004, P = 0.023) were positively associated with female adiposity during the first 2 years of life (Table 4). Graphs representing the estimated marginal means of female %FM for each of the significant variables in the final models are presented in Supporting Information Figures S1 and S2. Of note, the maternal FM effect on offspring body composition in males was reversed compared with females. In girls, the maternal FM effect was significant at age 2 weeks and throughout the first 2 years of life, which suggests that differences are set during gestation. Sensitivity analyses conducted on women who reported energy intake at a physiological level (energy intake/REE > 1.35, n = 98) (18) found a similar negative association of offspring FM and maternal intake of dietary fiber for the entire cohort (β = −0.07, P = 0.046). For boys, birth weight (β = 1.95, P = 0.004), maternal serum HDL (β = 0.04, P = 0.013), and breastfeeding duration (β = 0.004, P = 0.015) were positively associated with male adiposity during the first 2 years of life (Table 5). Sensitivity analyses investigating FM index rather than %FM on both maternal and offspring compartments found similar results, suggesting a role of maternal HDL and dietary fiber intake as well as birth weight on offspring FM index during the first 2 years of life.
| Variable | β | 95% CI | P |
|---|---|---|---|
| Maternal fat mass (% fat) (first trimester) | 0.12 | 0.03 to 0.21 | 0.006 |
| Maternal HDL (mg/dL) (gestation) | 0.07 | 0.04 to 0.10 | < 0.001 |
| Average total dietary fiber (g/d) | −0.09 | −0.16 to −0.01 | 0.016 |
| Breastfeeding duration (d) | 0.004 | −0.001 to 0.007 | 0.023 |
- HDL, high-density lipoprotein.
| Variable | β | 95% CI | P |
|---|---|---|---|
| Maternal HDL (mg/dL) (gestation) | 0.04 | 0.01-0.06 | 0.013 |
| Birth weight (kg) | 1.95 | 0.63-3.26 | 0.004 |
| Breastfeeding duration (d) | 0.004 | 0.001-0.007 | 0.015 |
- HDL, high-density lipoprotein.
Discussion
In the current study, maternal adiposity, maternal serum HDL, infant birth weight, sex, and breastfeeding duration were identified as independent predictors of offspring adiposity during the first 2 years of life. When stratified by sex, maternal %FM was the strongest predictor of adiposity in female offspring, whereas birth weight was the strongest predictor in male offspring. In both sexes, maternal HDL and breastfeeding duration were associated with greater adiposity during the first 2 years of life. Interestingly, maternal dietary fiber intake was negatively associated with female offspring adiposity.
Previous publications have pointed out that maternal total cholesterol is critical for neonatal development and birth weight outcomes (21). While others have shown an association between maternal LDL (13, 22) or HDL (13, 23, 24) and infant birth weight, limited studies have investigated the association between maternal HDL and LDL and offspring body composition. Studies have found no association between maternal LDL or HDL levels during pregnancy and offspring adiposity using sum of skinfolds at birth (24), 6 months, or 2 years of age (13). However, skinfold measurements are a crude measure of adiposity and are poorly correlated to body FM in children (25). In the presented cohort, a positive association was found between maternal serum HDL and offspring adiposity during the first 2 years. Further studies are needed to elucidate how maternal HDL contributes to metabolic programming and fat accretion in offspring.
A novel finding of this study is that maternal dietary fiber intake during gestation is negatively associated with female offspring adiposity. Very few studies have specifically looked at dietary fiber intake and its contribution to offspring whole-body adiposity, and none have evaluated its effect in a sexual dimorphic manner(26, 27). In agreement with our findings, a study by Hallam et al. (28) found that female offspring of rats exposed to a high fiber diet in utero had lower birth weight and body fat at 22 weeks of age (adolescence equivalent in humans) and lower levels of hepatic triacylglycerol, indicating an effect on hepatic lipid storage. Other research has found positive associations with maternal diet quality determined by the Healthy Eating Index during pregnancy and neonatal adiposity 72 hours after birth (29) and an inverse association between a higher quality diet during pregnancy and 3 months postpartum and infant body fat (percent) at 6 months of age (30). However, we did not find any significant effect of maternal Healthy Eating Index on offspring adiposity during the first 2 years of life. It is possible that maternal diet quality has the largest effect on fetal and early postnatal growth. Further research is sorely needed to delineate the mechanisms that mediate the interaction between maternal obesity and diet intake during different gestational periods and offspring adiposity risk later in life. Our attempt to account for underreporting in sensitivity analyses using a cutoff of 1.35 × BMR strengthens our findings. However, this cutoff is not intended for pregnant participants, and therefore results should be cautiously interpreted.
In the current study, we found a positive relationship between breastfeeding duration and child adiposity during the first 2 years of life in both sexes. While exclusively breastfeeding is known to elicit a protective effect against obesity later in life (31), this study agrees with findings by Sauder et al. (9) demonstrating exclusively breastfed babies had increased FM at 5 months of age.
Past research has identified a sexual dimorphism between maternal weight status during pregnancy and infant body composition with boys having less FM or body weight than girls as maternal BMI increases (6, 9). Interestingly, in the presented results, maternal %FM was only a significant predictor of adiposity during the first 2 years of life in female offspring, suggesting maternal obesity possibly elicits differential developmental effects on fat accretion by sex. Similar results were found by Henriksson et al. (32), who found a positive association between maternal %FM (measured by air displacement plethysmography) at 32 weeks gestation and %FM of 1-week-old infant girls but not boys.
Our lack of association between GWG and offspring adiposity are not in agreement with numerous studies linking GWG to offspring body composition outcomes. This can partially be explained by the GLOWING study design, which included a behavioral intervention in order to prevent excessive GWG (33). The intervention was successful in mitigating excessive GWG and therefore reducing the range of GWG in GLOWING participants with normal weight and class 1 obesity, which may explain the lack of association between GWG and offspring adiposity during the first 2 years of life. Offspring PA at 2 years tended to be negatively correlated with their adiposity (P = 0.08). This is in agreement with a study by Collings et al. (34) that showed an inverse association between PA and adiposity in children 11 months to 5 years of age. These findings emphasize the importance of PA beginning in early childhood to possibly reverse or mitigate adverse adiposity accrual that occurs in early infancy.
Our study is strengthened by its prospective longitudinal design with detailed phenotypic data of mothers during pregnancy as well as longitudinal measurements of offspring FM using a single methodology. The enrollment of second parity mothers lessens variation in the cohort caused by differences in birth weight between parities (35). This study accounted for gestational PA, dietary intake, and metabolic rate (REE) that may influence offspring FM accrual. Our study is not without limitations. Only ~87% of participants were able to complete all study visits at 2 years of age, resulting in a small reduction in sample size. The BMI range of 18.5 to 35 prohibits the generalization of the findings to class II and III obesity. The generalizability of the findings is also limited by the following 2 additional factors: (1) the majority of participants were Caucasian (87.1%) from the Central Arkansas region (although the racial distribution reflected the demographics of families living in the 4 predominant counties [Pulaski, Saline, Faulkner, Lonoke] in Central Arkansas), and (2) recruitment focused on healthy women only.
In conclusion, offspring adiposity during the first 2 years of life is influenced by maternal adiposity, maternal serum HDL, duration of breastfeeding, and infant sex, with maternal adiposity being the strongest determinant of adiposity in female offspring and birth weight being the strongest predictor of adiposity in males. Given the complexity of early-life influences on infant body composition, continued longitudinal assessment of adiposity into adolescence and adulthood is needed to determine the impact of maternal adiposity on child obesity risk. Development of intervention strategies aimed at reducing maternal adiposity prior to conception may provide protection against obesity later in life by reducing the influences of maternal adiposity on offspring metabolism.
Acknowledgments
The authors thank the Arkansas Children’s Nutrition Center clinical team and the participants for their time and dedication.
Funding agencies
This work was supported by USDA-ARS Project 6026-51000-010-05S. AA and MEHL were also supported by National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases R01 DK107516.
Disclosure
The authors declared no conflict of interest.
