Abstract
Context
An obesogenic perinatal environment contributes to adverse offspring metabolic health. Previous studies have been limited by lack of direct adiposity measurements and failure to account for potential confounders.
Objective
Examine the joint associations of maternal midpregnancy body mass index (BMI) and glycemia with direct adiposity measures in 10-14 year old offspring.
Design and Setting
International, epidemiological study: Hyperglycemia and Adverse Pregnancy Outcome (HAPO) and HAPO Follow-up Study, conducted between 2000-2006 and 2013-2016, respectively.
Participants and Main Outcome Measures
In 4832 children, adiposity measures for body mass index (BMI), body fat with air displacement plethysmography, skinfold thickness, and waist circumference were obtained at mean age 11.4 years.
Results
Maternal BMI and glucose, as continuous and categorical variables, were the primary predictors. In fully adjusted models controlling for child age, sex, field center, and maternal characteristics, maternal BMI had significant, positive associations with all childhood adiposity outcomes, while maternal glycemia had significant, positive associations with childhood adiposity outcomes except BMI. In joint analyses, and compared with a nonobese, nongestational diabetes mellitus (GDM) reference group, maternal obesity and GDM were associated with higher odds (maternal obesity odds ratio; OR [95% confidence interval; CI], GDM OR [95% CI]; combined OR [95% CI]) of childhood overweight/obese BMI (3.00 [2.42-3.74], 1.39 [1.14-1.71], 3.55 [2.49-5.05]), obese BMI (3.54 [2.70-4.64], 1.73 [1.29-2.30], 6.10 [4.14-8.99]), percent body fat >85th percentile (2.15 [1.68-2.75], 1.33 [1.03-1.72], 3.88 [2.72-5.55]), sum of skinfolds >85th percentile (2.35 [1.83-3.00], 1.75 [1.37-2.24], 3.66 [2.55-5.27]), and waist circumference >85th percentile (2.52 [1.99-3.21], 1.39 [1.07-1.80], 4.18 [2.93-5.96]).
Conclusions
Midpregnancy maternal BMI and glycemia are independently and additively associated with direct adiposity measures in 10-14 year old children. The combination of maternal obesity and GDM is associated with the highest odds of childhood adiposity.
Keywords: adiposity, maternal hyperglycemia, maternal BMI, offspring bodyfat, childhood obesity
In developed countries, nearly 50% of women entering pregnancy are either overweight or obese (1), and this trend is a growing concern in low- and middle-income countries (2). Commensurate with this trend are the worldwide increasing rates of hyperglycemia in pregnancy (3). The long-term impact of maternal obesity and hyperglycemia on developmental programming may have major implications for long-term metabolic health of the offspring (4). Unraveling the joint contributions of maternal body mass index (BMI) and glycemia during pregnancy to offspring adiposity is of paramount importance to direct prevention strategies.
Maternal obesity and gestational diabetes mellitus (GDM) are recognized as important contributors to increasing childhood obesity rates (5-7). However, previous studies have methodological limitations, including the use of maternal weight recall rather than direct measurements to assess maternal BMI and/or failure to adequately account for maternal glycemia. In addition, studies were generally limited to measurement of childhood BMI (8) rather than more direct measures of adiposity, which better reflect risks for adverse metabolic health outcomes such as glucose intolerance, dyslipidemia, hypertension, and fatty liver disease (9). Only a few studies have reported longitudinal data initiated from the index pregnancy to beyond the early childhood years (10,11). Yet, metabolic disease is more likely to persist into adulthood when present in adolescence compared with early childhood (9,12).
The Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study (13) and its Follow-up Study (HAPO FUS) (14) were observational studies that included a multinational, racially, and ethnically diverse population of women and their children. Given the uniqueness of this cohort and measurements available, the present study examined the joint associations of maternal BMI and glycemia with 10-14 year old offspring BMI and more precise, direct measures of adiposity. Other HAPO FUS publications have focused on independent associations of maternal glucose measured midpregnancy, modeled both as GDM versus no GDM (14) or along the continuum of observed glucose values (15), with offspring adiposity, adjusting for maternal BMI. The findings presented herein present joint modeling analyses of maternal glycemia and BMI, for cohesive interpretation of associations of both maternal predictors with 10-14 year old offspring adiposity outcomes.
Patients and Methods
The HAPO Study and HAPO FUS Study methods have been published (13,14). Protocols were approved by the institutional review board at each field center. All participants provided written informed consent and assent, where required.
Participants
In the HAPO Study, conducted from 2000-2006, pregnant women underwent a study visit at approximately 28 weeks’ gestation that included measurements of height, weight, blood pressure, and a fasting 75-g oral glucose tolerance test (OGTT) with blood sampled at fasting, 1-hour and 2-hour (13). Participants and providers were blinded to the OGTT results unless specific glucose thresholds were exceeded; then unblinding occurred. The remaining blinded participants were not treated. Demographic data including self-identified racial/ethnic group, smoking, and alcohol use were collected from standardized questionnaires.
Offspring of HAPO mothers were recruited to the HAPO FUS during the years 2013-2016, as previously described (14). HAPO FUS child participants were eligible if they were born with gestational age at delivery ≥37 weeks and had no major malformations. The child HAPO FUS visit included multiple anthropometric measures of childhood adiposity.
Child anthropometrics
Training and certification procedures were established for all study research personnel for both the HAPO Study and HAPO FUS. For the HAPO FUS, centralized training and maintenance of certification was conducted by the coordinating center (14). During the HAPO FUS visit, the child’s weight was measured to the nearest 0.1 kg using a calibrated scale, height was measured twice with a stadiometer to the nearest 0.5 cm, and waist circumference was measured twice at the top of the iliac crest to the nearest 0.5 cm. Calibrated calipers (Harpenden, Baty, UK) were used to measure skinfolds twice at 3 sites, triceps, subscapular, and suprailiac, to the nearest 0.1 mm. A third measurement was obtained if results differed by >0.5 kg for weight, >1.0 cm for height and waist circumference, and >1.0 mm for skinfolds. Body composition was measured by the method of air displacement plethysmography (Bod Pod, Cosmed, Italy), which provided fat mass, lean mass, and percent body fat.
Outcomes and predictors
Continuous childhood adiposity outcomes included BMI (weight in kilograms divided by height in meters squared), BMI z-score (age and sex specific) (16), sum of the 3 skinfold measurements, body fat percentage from the Bod Pod using the Lohman equation (17), waist circumference, fat mass calculated as the product of weight and body fat percentage, and lean mass calculated as the difference between total weight and fat mass. Dichotomous outcomes were childhood overweight/obese and obese status according to the International Obesity Task Force using Asian-specific cutoffs for self-reported Asian children and international cutoffs for all others (16), and body fat percentage, sum of skinfolds, and waist circumference >85th percentiles according to quantile regression adjusted for age, sex, and field center.
The primary predictors were maternal BMI and glycemia measured during pregnancy, as both continuous and categorical variables. For the BMI categorical analyses, maternal BMI was divided into the 4 prepregnancy categories established by the World Health Organization (WHO) and incorporated by the Institute of Medicine: underweight, normal range, overweight, and obese (18). Since BMI was measured midpregnancy (approximately 28 weeks’ gestation), comparable category limits were obtained from a regression model of OGTT BMI and gestational age at the OGTT to yield the following categories (kg/m2): <22.6 (underweight), 22.6-28.4 (normal range), 28.5-32.9 (overweight), and ≥33 (obese), as previously described (19). Maternal glycemia using the sum of glucose z-scores at fasting, 1-hour and 2-hour during the HAPO OGTT was analyzed across the continuum and GDM status during the HAPO pregnancy using International Association of Diabetes and Pregnancy Study Groups/WHO criteria (20) defined the categorical glycemia predictor.
Statistical analyses
HAPO FUS data were summarized using frequencies and counts for categorical variables and mean and standard deviation (SD) for continuous variables. Histograms and boxplots were examined to determine the shape of distributions and to identify potential outlying observations. For dichotomous outcomes, multiple logistic regression was used to estimate odds ratios (OR) and 95% confidence intervals (CI), and modified least-squares regression with Huber-White robust standard errors was used to estimate risk differences with 95% CI (21). Multiple linear regression was used for continuous outcomes; results are reported as regression coefficients (adjusted mean differences, ie, β estimates) with 95% CI. Joint models for pairs of maternal predictors from pregnancy were specifically considered as follows: continuous maternal BMI and glucose sum of z-scores; maternal obesity and GDM; and maternal BMI categories and GDM. Continuous maternal BMI and glucose sum of z-scores were modeled as main effects with statistical interaction terms between the 2 predictors also evaluated. Reported estimates of OR, risk differences, and adjusted mean differences for joint analyses of maternal obesity and GDM were calculated by creating 4 categories based on observed HAPO data: (1) no obesity, no GDM; (2) obesity, no GDM; (3) no obesity, GDM; and (4) obesity, GDM. Regression models were then fit using the no obesity, no GDM category as referent. Regression models were also fit treating maternal obesity and GDM as main effects and evaluating a statistical interaction term between the 2 predictors. A similar approach was used for joint analyses of maternal BMI categories (underweight, normal weight, overweight, obese) and GDM (yes or no) by creating separate categories for all 8 potential category combinations and treating the normal weight, non-GDM combination as the reference group. Regression models were also fit treating maternal BMI categories and GDM as main effects and evaluating statistical interaction terms between them. Covariate adjustment models were considered for all outcomes, with variables identified according to study design, known potential confounders, and adjustments used in HAPO analyses (13,14). Covariate adjustments were examined as follows: Model 1, offspring age, sex, and field center (for childhood adiposity outcomes that included offspring age, sex and/or field center in their definition, these variables were not included in Model 1); Model 2, Model 1 + maternal variables measured at the pregnancy OGTT (age, height, mean arterial pressure, parity (0 or 1+), smoking (yes/no), alcohol (yes/no), gestational age, and any family history of diabetes). The fully adjusted model refers to Model 2. Field center served as a proxy for maternal ethnicity due to substantial racial/ethnic similarity within each center.
Linear regression model fit was assessed by scatterplots of residuals versus fitted values, histograms and qqplots of residuals, and DFbeta statistics. Adjusted R2 values were used to gauge variability explained in linear models, with partial R2 values calculated to examine contributions of individual predictors. Quadratic terms and restricted cubic splines estimated with the rms R package (22) were used to assess linearity between the continuous predictor and the log odds of the outcome for logistic regression models and continuous outcomes for linear regression models. Statistical significance was determined according to P < .05. Analyses presented here are considered secondary for the HAPO FUS and are not corrected for multiple comparisons. All analyses were conducted in R (3.3.1) (23).
Results
Participants
Table 1 displays the characteristics of the 4832 mothers during the original HAPO Study and their children during the FUS. The HAPO Study visit occurred at mean gestational age of 27.6 weeks. Across maternal BMI categories during pregnancy, 12.8% of mothers were considered underweight, 52.7% normal weight, 21.9% overweight, and 12.5% obese. The frequency of GDM using IADPSG/WHO criteria (20) was 14.1%.
Table 1.
Characteristics of mothers during pregnancy oral glucose tolerance test visit and the offspring at follow-up. N = 4832.
| Maternal characteristics | Mean (SD) |
|---|---|
| Age (years) | 29.9 (5.7) |
| Height (cm) | 161.6 (6.5) |
| Weight (kg) | 70.8 (17.1) |
| Body mass index (BMI) (kg/m2) | 27.5 (4.9) |
| Gestational age (weeks) | 27.6 (1.7) |
| Mean arterial pressure (mmHg) | 80.5 (8.0) |
| Fasting plasma glucose (mg/dL) | 81.0 (6.6) |
| 1-hour plasma glucose (mg/dL) | 133.1 (30.2) |
| 2-hour plasma glucose (mg/dL) | 110.4 (23.0) |
| Glucose sum of z-scores | -0.43 (2.3) |
| Race/Ethnicity | N (%) |
| White, non-Hispanic | 2287 (47.3) |
| Black, non-Hispanic | 775 (16.0) |
| Hispanic | 507 (10.5) |
| Asian | 1176 (24.3) |
| Other | 87 (1.8) |
| GDM frequency | 683 (14.13) |
| Any prenatal smoking | 245 (5.1) |
| Any prenatal alcohol use | 406 (8.4) |
| Parity (any prior delivery >20 weeks) | 2485 (51.4) |
| Family history of diabetes mellitus | 1077 (22.3) |
| BMI categories (kg/m2) | |
| Underweight (<22.6) | 619 (12.8) |
| Normal weight (22.6-28.4) | 2547 (52.7) |
| Overweight (28.5-32.9) | 1060 (21.9) |
| Obese (>33.0) | 606 (12.5) |
| Childhood characteristics | Mean (SD) |
| Age (years) | 11.4 (1.2) |
| Height (cm) | 148.6 (10.2) |
| Weight (kg) | 43.3 (13.3) |
| BMI z-score | 0.48 (1.24) |
| Sum of skinfolds (mm) | 39.2 (21.5) |
| Lean body mass (kg) | 33.3 (7.4) |
| Fat mass (kg) | 10.2 (7.8) |
| Percent body fat (Bod Pod) | 21.2 (10.5) |
| Waist circumference (cm) | 70.3 (11.9) |
| N (%) | |
| Sex (% female) | 2367 (49.0) |
| Overweight/obese | 1441 (30.2) |
| Obese | 535 (11.2) |
| Percent bodyfat >85th percentile | 698 (15.0) |
| Sum of skinfolds >85th percentile | 706 (15.0) |
| Waist circumference >85th percentile | 718 (15.0) |
Joint associations of maternal BMI and glucose with childhood adiposity measures
Maternal BMI as a continuous predictor had consistent, positive associations with all continuous childhood adiposity outcomes (Table 2). In Model 2, maternal BMI higher by 1 kg/m2 was associated with child’s BMI higher by 0.31 kg/m2. Model adjustments for nonglucose maternal variables measured at OGTT had minimal effect on β estimates. When jointly modeled with maternal BMI, maternal glucose sum of z-scores was not significantly associated with childhood BMI or BMI z-score but was positively and significantly associated with the more direct measures of childhood adiposity: percent body fat, sum of skinfolds, and waist circumference. No significant interactions between maternal BMI and maternal glucose sum of z-scores were observed, nor were departures from linearity. Partial R2 estimates were consistently higher for the maternal BMI predictor than for the maternal sum of glucose z-scores predictor for all outcomes.
Table 2.
Associations of maternal body mass index (BMI) and glucose with continuous measures of childhood adiposity
| Maternal BMI | Maternal glucose sum of z-scores | ||||||
|---|---|---|---|---|---|---|---|
| Childhood outcome | Adjusted mean difference (95% CI) | P value | Partial R2 | Adjusted mean difference (95% CI) | P value | Partial R2 | Total adjusted R2 |
| BMI (kg/m2) | |||||||
| Model 1 | 0.31 (0.28-0.33) | <.001 | 0.10 | 6 × 10-4 (-0.05 to 0.05) | .9 | 1.6 × 10-4 | 0.22 |
| Model 2 | 0.31 (0.28-0.33) | <.001 | 0.09 | 0.03 (-0.02 to 0.08) | .2 | 7.0 × 10-5 | 0.23 |
| BMI z-score | |||||||
| Model 1 | 0.09 (0.08-0.10) | <.001 | 0.10 | 1.0 × 10-4 (-0.015 to 0.015) | .9 | 1.7 × 10-4 | 0.17 |
| Model 2 | 0.09 (0.08-0.10) | <.001 | 0.09 | 8 × 10-3 (-0.008 to 0.02) | .3 | 2.9 × 10-6 | 0.18 |
| Body fat % | |||||||
| Model 1 | 0.55 (0.49-0.62) | <.001 | 0.06 | 0.23 (0.10-0.35) | <.001 | 0.002 | 0.15 |
| Model 2 | 0.55 (0.48-0.61) | <.001 | 0.05 | 0.29 (0.16-0.42) | <.001 | 0.003 | 0.16 |
| Sum of skinfolds (mm) | |||||||
| Model 1 | 1.20 (1.07-1.33) | <.001 | 0.06 | 0.40 (0.13-0.67) | .004 | 0.001 | 0.12 |
| Model 2 | 1.19 (1.05-1.33) | <.001 | 0.05 | 0.53 (0.25-0.81) | <.001 | 0.002 | 0.13 |
| Waist circumference (cm) | |||||||
| Model 1 | 0.74 (0.68-0.81) | <.001 | 0.08 | 0.13 (-0.005 to 0.27) | .05 | 4.1 × 10-4 | 0.24 |
| Model 2 | 0.74 (0.67-0.82) | <.001 | 0.07 | 0.22 (0.08-0.36) | .003 | 0.001 | 0.25 |
Adjusted mean differences (ie, β estimates) represent differences in each outcome for maternal BMI higher by 1.0 kg/m2 and maternal glucose sum of z-scores higher by 1 unit.
Model 1: Age, child sex, field center.
Model 2: Model 1 + Maternal variables measured at the oral glucose tolerance test: age, mean arterial pressure, parity (0 or 1+), smoking (yes/no), alcohol (yes/no), any family history of diabetes, height, gestational age
Joint associations of maternal obesity and GDM with childhood adiposity outcomes
Maternal obesity and GDM were both positively associated with all continuous childhood adiposity outcomes, with the combined category of maternal obesity and GDM demonstrating the largest adjusted mean difference for each outcome (Table 3). For example, in Model 2, maternal obesity alone was associated with child’s BMI higher by 2.95 kg/m2, and maternal GDM alone was associated with child’s BMI higher by 0.54 kg/m2, and in combination maternal obesity and GDM were associated with child’s BMI higher by 3.46 kg/m2 than children born to mothers without obesity or GDM. Model 2 adjustments for maternal variables measured at OGTT had minimal effect on the beta estimates.
Table 3.
Associations of maternal obesity and gestational diabetes mellitus (GDM) with continuous measures of childhood adiposity
| Maternal obesity only | Maternal GDM onlya | Maternal obesity and GDM | |||||
|---|---|---|---|---|---|---|---|
| Childhood outcome | Adjusted mean difference (95% CI) | P value | Adjusted mean difference (95% CI) | P value | Adjusted mean difference (95% CI) | P value | Adjusted R2 |
| BMI (kg/m2) | |||||||
| Model 1 | 3.23 (2.84-3.62) | <.001 | 0.59 (0.22-0.95) | <.001 | 3.77 (3.12-4.41) | <.001 | 0.17 |
| Model 2 | 2.95 (2.54-3.35) | <.001 | 0.54 (0.17-0.91) | .001 | 3.46 (2.80-4.11) | <.001 | 0.18 |
| BMI z-score | |||||||
| Model 1 | 0.82 (0.70-0.93) | <.001 | 0.17 (0.06-0.27) | <.001 | 0.99 (0.80-1.18) | <.001 | 0.11 |
| Model 2 | 0.72 (0.60-0.84) | <.001 | 0.14 (0.03-0.25) | <.001 | 0.89 (0.69-1.08) | <.001 | 0.12 |
| Body fat % | |||||||
| Model 1 | 5.55 (4.56-6.53) | <.001 | 1.57 (0.67-2.48) | <.001 | 7.90 (6.28-9.52) | <.001 | 0.12 |
| Model 2 | 4.83 (3.82-5.84) | <.001 | 1.40 (0.48-2.32) | <.001 | 7.09 (5.45-8.73) | <.001 | 0.16 |
| Sum of skinfolds (mm) | |||||||
| Model 1 | 11.86 (9.77-13.96) | <.001 | 3.70 (1.78-5.62) | <.001 | 16.78 (13.33-20.22) | <.001 | 0.08 |
| Model 2 | 10.42 (8.27-12.58) | <.001 | 3.37 (1.43-5.31) | <.001 | 15.20 (11.70-18.70) | <.001 | 0.09 |
| Waist circumference (cm) | |||||||
| Model 1 | 7.88 (6.81-8.95) | <.001 | 2.13 (1.15-3.11) | <.001 | 9.91 (8.15-11.67) | <.001 | 0.20 |
| Model 2 | 7.17 (6.07-8.27) | <.001 | 1.98 (0.99-2.96) | <.001 | 9.06 (7.28-10.84) | <.001 | 0.21 |
Adjusted mean differences (ie, β estimates) represent differences in each outcome for each maternal obesity/GDM category relative to the referent category of nonobese mothers without GDM.
Model 1: Age, sex, field center.
Model 2: Model 1 + Maternal variables measured at the OGTT: age, mean arterial pressure, parity (0 or 1+), smoking (yes/no), alcohol (yes/no), any family history of diabetes, height, gestational age.
No maternal obesity and no GDM was the reference category for all analyses.
aDefined by International Association of Diabetes and Pregnancy Study Groups criteria as ≥1 of the following results from a 75-g oral glucose tolerance test (1) ≥92 mg/dL for fasting plasma glucose level; (2) ≥180 mg/dL for 1-hour plasma glucose level; or (3) ≥153 mg/dL for 2-hour plasma glucose level.
Table 4 reports OR and risk differences for dichotomous childhood outcomes of obese BMI, overweight/obese BMI, and body fat percentage, sum of skinfolds, and waist circumference greater than the 85th percentile in statistical models comparing categories of maternal obesity only, maternal GDM only, and both maternal obesity and GDM with the reference group of mothers without obesity or GDM. Outcome frequencies in the HAPO FUS for each category are also displayed. Compared with the reference group, maternal obesity and GDM were associated with higher odds (maternal obesity OR [95% CI], GDM OR [95% CI]; combined OR [95% CI]) of childhood overweight/obese BMI (3.00 [2.42-3.74], 1.39 [1.14-1.71], 3.55 [2.49-5.05]), obese BMI (3.54 [2.70-4.64], 1.73 [1.29-2.30], 6.10 [4.14-8.99]), percent body fat >85th percentile (2.15 [1.68-2.75], 1.33 [1.03-1.72], 3.88 [2.72-5.55]), sum of skinfolds >85th percentile (2.35 [1.83-3.00], 1.75 [1.37-2.24], 3.66 [2.55-5.27]), and waist circumference >85th percentile (2.52 [1.99-3.21], 1.39 [1.07-1.80], 4.18 [2.93-5.96]). Model adjustments for maternal variables measured at the OGTT modestly attenuated associations for maternal obesity and had little effect on GDM associations.
Table 4.
Associations of maternal obesity and gestational diabetes mellitus (GDM)a with dichotomous measures of childhood adiposity
| No maternal obesity, no GDMc | Maternal obesity only | Maternal GDM only | Maternal obesity and GDM | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Childhood outcome | Frequency (%) of outcome | Frequency (%) of outcome | Odds ratio (95% CI) P value | Risk difference (95% CI) P value | Frequency (%) of outcome | Odds ratio (95% CI) P value | Risk difference (95% CI) P value | Frequency (%) of outcome | Odds ratio (95% CI) P value | Risk difference (95% CI) P value | C statistic |
| Overweight/ obese BMIb | 922 (25.3) | 250 (55.3) | 179 (33.8) | 90 (59.2) | |||||||
| Model 1 | 3.45 (2.80- 4.25)<.001 | 0.22 (0.19- 0.25)<.001 | 1.42 (1.16- 1.73)<.001 | 0.08 (0.03- 0.13).001 | 4.07 (2.89- 5.74)<.001 | 0.24 (0.18- 0.29)<.001 | 0.68 | ||||
| Model 2 | 3.00 (2.42- 3.74)<.001 | 0.20 (0.17- 0.24)<.001 | 1.39 (1.14- 1.71).001 | 0.08 (0.03- 0.13)<.001 | 3.55 (2.49- 5.05)<.001 | 0.22 (0.17- 0.28)<.001 | 0.69 | ||||
| Obese BMIb | 284 (7.8) | 121 (26.8) | 73 (13.8) | 57 (37.5) | |||||||
| Model 1 | 4.05 (3.14- 5.23)<.001 | 0.10 (0.08- 0.13)<.001) | 1.74 (1.31- 2.30)<.001 | 0.05 (0.01- 0.09).010 | 6.98 (4.82- 10.12)<.001 | 0.17 (0.13- 0.22)<.001 | 0.74 | ||||
| Model 2 | 3.54 (2.70- 4.64)<.001 | 0.09 (0.07- 0.11)<.001 | 1.73 (1.29- 2.30)<.001 | 0.05 (0.01- 0.09).012 | 6.10 (4.14- 8.99)<.001 | 0.16 (0.12- 0.21)<.001 | 0.75 | ||||
| Body fat percent >85th percentile | 439 (12.4) | 114 (25.7) | 85 (16.4) | 60 (39.7) | |||||||
| Model 1 | 2.48 (1.97- 3.13)<.001 | 0.11 (0.09- 0.14)<.001 | 1.39 (1.08- 1.78)<.001 | 0.04 (–0.0002 to 0.08).051 | 4.55 (3.23- 6.40)<.001 | 0.17 (0.12- 0.22)<.001 | 0.68 | ||||
| Model 2 | 2.15 (1.68- 2.75)<.001 | 0.10 (0.07- 0.13)<.001 | 1.33 (1.03- 1.72)<.001 | 0.04 (–0.0003 to 0.08).08 | 3.88 (2.72- 5.55)<.001 | 0.16 (0.11- 0.21)<.001 | 0.63 | ||||
| Sum of skinfolds (mm) >85th percentile | 437 (12.1) | 116 (26.6) | 101 (19.5) | 52 (35.6) | |||||||
| Model 1 | 2.68 (2.12- 3.38)<.001 | 0.11 (0.08- 0.13)<.001 | 1.80 (1.41- 2.28)<.001 | 0.09 (0.04- 0.14)<.001 | 4.18 (2.95- 5.92)<.001 | 0.17 (0.12- 0.21) <.001 | 0.70 | ||||
| Model 2 | 2.35 (1.83- 3.00)<.001 | 0.10 (0.07- 0.12)<.001 | 1.75 (1.37- 2.24)<.001 | 0.09 (0.04- 0.13)<.001 | 3.66 (2.55- 5.27)<.001 | 0.16 (0.11- 0.20)<.001 | 0.64 | ||||
| Waist circumference (cm) >85th percentile | 441 (12.0) | 130 (29.0) | 86 (16.3) | 61 (40.7) | |||||||
| Model 1 | 2.99 (2.38- 3.75)<.001 | 0.13 (0.10- 0.15)<.001 | 1.43 (1.11- 1.84)<.001 | 0.05 (0.01- 0.09).016 | 5.02 (3.58- 7.05)<.001 | 0.18 (0.13- 0.22)<.001 | 0.70 | ||||
| Model 2 | 2.52 (1.99- 3.21)<.001 | 0.11 (0.09- 0.14)<.001 | 1.39 (1.07- 1.80)<.001 | 0.05 (0.01- 0.09).019 | 4.18 (2.93- 5.96)<.001 | 0.16 (0.11- 0.21)<.001 | 0.64 |
aDefined by International Association of Diabetes and Pregnancy Study Groups criteria as ≥ 1 of the following results from a 75-g oral glucose tolerance test (1) ≥ 92 mg/dL for fasting plasma glucose level (2), ≥180 mg/dL for 1-hour plasma glucose level, or (3) ≥ 153 mg/dL for 2-hour plasma glucose level.
bDefined by International Obesity Task Force with body mass index-based cutoffs specific to sex and the child’s age in months. Model 1: Age, sex, field center. Model 2: Model 1 + maternal variables measured at the OGTT: age, mean arterial pressure, parity (0 or 1+), smoking (yes/no), alcohol (yes/no), any family history of diabetes, height, gestational age.
cNo maternal obesity and no GDM was the reference category for all analyses.
In regression models with main effects and interaction terms for maternal obesity and GDM, no statistically significant interaction terms were observed for any of the continuous or dichotomous childhood outcomes.
Joint associations of maternal BMI category and GDM with childhood adiposity outcomes
Joint associations for maternal BMI categorized as underweight, normal range, overweight, and obese adjusted for 28 weeks of pregnancy (19) and maternal GDM (20) were evaluated for all dichotomous childhood adiposity outcomes. In the fully adjusted model, with maternal normal weight, non-GDM as referent, ORs were estimated for each childhood adiposity outcome in the presence and absence of GDM (Fig. 1). Maternal underweight BMI without GDM was associated with lower odds of all childhood adiposity outcomes. However, for maternal underweight BMI and GDM, the 95% CI suggested no difference in childhood adiposity outcomes when compared with the referent category. Among normal weight women, GDM was associated with more frequent childhood overweight/obese BMI, obese BMI, and sum of skinfolds and waist circumference greater than the 85th percentile, but not body fat percentage greater than the 85th percentile. Maternal overweight without GDM was positively associated with all childhood adiposity outcomes; with GDM, point estimates of ORs for obese BMI, and body fat percentage and sum of skinfolds greater than 85th percentile trended higher. Maternal obesity without GDM was positively associated with all childhood adiposity outcomes; with GDM, ORs trended higher for each childhood adiposity outcome.
Figure 1.
Joint associations of maternal body mass index (BMI) category and gestational diabetes mellitus (GDM) (absent or present) with childhood adiposity Outcomes. Forest plots displaying ORs of childhood adiposity outcomes. (A) Overweight/obesity. (B) Obesity. (C) Body fat >85th percentile. (D) Sum of skin folds >85th percentile. (E) Waist circumference >85th percentile, according to maternal BMI category with normal range, non-GDM as the referent category. Results displayed are from the fully adjusted model: field center, child age, sex, ethnicity, and maternal variables at pregnancy oral glucose tolerance test (age, height, mean arterial pressure, parity (0 or 1+), smoking, alcohol, gestational age, and any family history of diabetes). Circles represent normal glucose and triangles represent GDM using International Association of Diabetes and Pregnancy Study Groups criteria.
Joint association of maternal BMI category and GDM with childhood body composition
Maternal BMI categories and GDM were also jointly examined for associations with childhood body composition outcomes, lean and fat mass, which together equal body weight (Fig. 2). In fully adjusted models, with maternal normal weight, non-GDM as referent, maternal underweight without GDM was associated with lower (adjusted mean difference in kg, 95% CI) total weight (-3.67, -4.65 to -2.68), lean mass (-1.67, -2.19 to -1.16), fat mass (-1.99, -2.64 to -1.35), and body fat% (-2.86, -3.75 to -1.97). Maternal overweight without GDM was associated with a higher total weight (3.93, 3.12-4.73), lean mass (1.64, 1.22-2.06), fat mass (2.28, 1.76-2.81), and body fat percentage (3.02, 2.30-3.75). Maternal obesity without GDM was associated with higher total weight (9.31, 8.27-10.35), lean mass (3.69, 3.15-4.23), fat mass (5.62, 4.95-6.29), and body fat percentage (6.04, 5.11-6.98). GDM was associated with higher total weight (0.91, 0.003-1.82), fat mass (0.82, 0.23-1.41), and body fat percentage (1.19, 0.38-2.01) across all BMI categories but was not significantly associated with lean body mass (0.09, -0.38 to 0.56).
Figure 2.
Joint association of maternal body mass index (BMI) category and gestational diabetes mellitus (GDM) (absent or present) and childhood body composition. Adjusted mean difference of childhood lean body mass and fat mass according to maternal BMI category and GDM absence (A) or presence (B), with normal range maternal BMI as the referent category. Results displayed are from the fully adjusted model: field center, child age, sex, ethnicity, and maternal variables at pregnancy oral glucose tolerance test (age, height, mean arterial pressure, parity (0 or 1+), smoking, alcohol, gestational age, and any family history of diabetes).
Discussion
This report considers maternal BMI and glycemia, measured midpregnancy, as joint predictors of childhood adiposity outcomes at mean age 11.4 years. Children born to mothers with both obesity and GDM had the highest odds of adverse adiposity outcomes: overweight/obese BMI, obese BMI, and body fat percentage, sum of skinfolds, and waist circumference greater than the 85th percentile.
Maternal BMI and glycemia modeled together had significant, positive associations with direct measures of childhood adiposity: percent body fat, sum of skinfolds, and waist circumference. Maternal BMI was significantly associated with childhood BMI and BMI z-score; however, maternal glycemia across the continuum did not demonstrate significant associations with childhood BMI or BMI z-score. These results substantiate the finding that childhood BMI may not be a good proxy for direct measures of adiposity. Although BMI is used as a screening tool for obesity (24), in children and specific ethnic populations, BMI does not always correlate well with adiposity (25). In this study, direct childhood adiposity measures are reported for a racially and ethnically diverse cohort. Results of this study provide an important contribution to the literature in demonstrating both the independent and joint associations of maternal BMI and glucose with childhood adiposity, improving upon previous reports of childhood obesity outcomes as measured by BMI (7,26).
The original HAPO Study demonstrated that the combination of untreated GDM and maternal obesity was associated with higher odds for birth weight, newborn percent body fat, and cord C-peptide >90th percentile compared with the odds for either GDM or obesity alone, suggesting an additive association of GDM and obesity with newborn outcomes (27). The present study has now demonstrated that maternal obesity and untreated GDM also have additive associations with childhood adiposity outcomes, including obesity as well as percent body fat, sum of skinfolds, and waist circumference >85th percentile, thus demonstrating that additive associations with adiposity outcomes persist into childhood. Moreover, in statistical models in which each predictor was treated as a main effect and a potential interaction term was examined, the interaction term was not statistically significant across the outcomes examined. This indicates that the association of maternal obesity during pregnancy with childhood adiposity does not vary by maternal GDM status, and vice versa.
The childhood body composition differences in the joint analyses of maternal BMI category and absence or presence of GDM represent novel findings and confirm that associations of maternal BMI with childhood adiposity exist across the continuum of maternal BMI. Higher overall weight was evident for children of overweight and obese women, on average 3.9 kg and 9.3 kg, respectively, compared with children of normal weight women, whereas children of underweight mothers had lower weight by -3.7 kg on average. For children of overweight and obese mothers, the extra weight consisted of both lean mass and fat mass. Across categories of higher maternal BMI, child percent body fat was higher, demonstrating that the relative amount of fat mass compared with lean mass was higher in children of overweight and obese women than in normal weight women. The presence of GDM was related to higher childhood body fat by about 1% across all BMI categories but was not associated with lean body mass. The excess adiposity demonstrated among children born to overweight and obese mothers with and without GDM may well be a precursor to future adult obesity and metabolic health risks (9,12).
Our data indicate that the combination of maternal obesity and GDM places an exposed child at highest risk of increased adiposity. These findings support the developmental origins of health and disease model which postulates that intrauterine developmental programming has lasting effects on offspring metabolic health (28). While genetic susceptibility to obesity is well established (29,30), shared genes do not explain the rapid rate at which obesity prevalence has increased among all age groups. Rather, intrauterine programming of maternal obesity on offspring obesity has been demonstrated among siblings born prior to or after their mother’s weight reduction following bariatric surgery, in which younger siblings had less obesity and insulin resistance than their older siblings (31). The higher obesity rate observed in the older siblings born prior to their mother’s weight reduction, even with genetics and postnatal environment similar to their younger siblings, provides evidence that the metabolic milieu of an obese intrauterine environment contributes to obesity development in offspring, perhaps through epigenetic programming (32). Earlier sibling studies among the Pima Indian population in which children born prior to or after development of maternal diabetes equally support intrauterine programming effects (6).
A limitation of this study is the lack of research measurements of maternal prepregnancy BMI or gestational weight gain. In HAPO, measured weight was obtained midpregnancy at approximately 28 weeks’ gestation, and maternal BMI categories were estimated to model pregnancy weight status (19). As previously reported, the correlation between prepregnancy BMI and BMI during the HAPO study was 0.92 (19). A second limitation is that mothers with a fasting plasma glucose level >105 mg/dL or 2-hour plasma glucose level >200 mg/dL at the HAPO OGTT were unblinded for treatment and excluded from statistical analysis. The present study demonstrated higher partial correlations for maternal BMI than maternal glycemia, indicating that, in these data, maternal BMI explains more variability in the outcomes than maternal glycemia. However, the absence of high glucose values in the HAPO data set likely impacted the estimates of partial correlation for maternal glycemia. Data on paternal BMI are lacking, presenting another limitation. Finally, an important limitation in interpreting the findings is that this was an observational study. Interventional studies with long-term offspring follow-up are necessary to determine whether maternal BMI reductions and treatment of gestational hyperglycemia will reduce offspring obesity and adiposity.
This study improves on the methodological limitations of previous reports that relied on recalled maternal height and weight and data obtained from chart reviews. Moreover, other studies were initiated decades before the current obesity epidemic (26,33), whereas HAPO is a contemporary cohort in which all children were born after the year 2000 when obesity rates in developed countries dramatically increased (34). Unlike many previous reports, this study includes detailed measures of maternal midpregnancy glycemia, which also contributes to childhood adiposity (14, 15). Other major strengths are the prospective direct measurements of BMI during pregnancy, the multiple, direct measurements of childhood adiposity, and adjustment for confounding maternal variables.
The current report establishes that maternal obesity and GDM in pregnancy are jointly associated with the highest odds of childhood adiposity, but not all children of obese and/or GDM mothers were overweight or obese in the HAPO FUS. Thus, there is opportunity to further understand mechanisms whereby these adolescents are resistant to the development of obesity. Other contributors to childhood obesity development may include nutrition during early life, the gut microbiome, and paternal factors, data which are not available for the present analysis.
Conclusions
Mid-pregnancy maternal BMI and glycemia are independently and additively associated with direct adiposity measures in 10-14 year old children. The combination of maternal obesity and GDM is associated with the highest odds of childhood adiposity.
Acknowledgments
The article was written on behalf of the HAPO Follow-up Study Cooperative Research Group: Field Centers: Bangkok: Chaicharn Deerochanawong, Thadchanan Tanaphonpoonsuk (Rajavithi Hospital), Sukeeta Binratkaew Uraiwan Chotigeat, Wanee Manyam, (Queen Sirikit National Institute of Child Health); Barbados: Martinette Forde, Andre Greenidge, Kathleen Neblett, Paula Michele Lashley, Desiree Walcott (Queen Elizabeth Hospital/School of Clinical Medicine and Research, University of the West Indies, Barbados); Belfast: Katie Corry, Loraine Francis, Jo-anne Irwin, Anne Langan, David R McCance, Maureen Mousavi, (Belfast Health and Social Care Trust), Ian Young (Queen’s University Belfast); Bellflower: Jennifer Gutierrez, Jennifer Jimenez, Jean M Lawrence, David A Sacks, Harpreet S Takhar, Elizabeth Tanton (Kaiser Permanente Southern California); Chicago: Wendy J Brickman, Jennifer Howard, Jami L Josefson, Lauren Miller (Ann and Robert H Lurie Children’s Hospital/Northwestern University Feinberg School of Medicine); Cleveland: Jacqui Bjaloncik, Patrick M Catalano, Ajuah Davis, Michaela Koontz, Larraine Presley, Shoi Smith, Amanda Tyhulski (MetroHealth Medical Center/Case Western Reserve University); Hong Kong: Albert Martin Li, Ronald C Ma, Risa Ozaki, Wing Hung Tam, Michelle Wong, Cindy Siu Man Yuen (The Chinese University of Hong Kong/Prince of Wales Hospital); Manchester: Peter E Clayton, Aysha Khan, Avni Vyas (Royal Manchester Children’s Hospital, Central Manchester University Hospitals NHS Foundation Trust, Manchester Academic Healthy Sciences Centre/School of Medical Sciences, Faculty of Biology, Medicine & Health, University of Manchester); Michael Maresh (St. Mary’s Hospital, Central Manchester University Hospitals NHS Foundation Trust, Manchester Academic Health Sciences Centre); Petah-Tiqva: Hadasse Benzaquen, Naama Glickman, Alona Hamou, Orna Hermon, Orit Horesh, Yael Keren, Yael Lebenthal, Shlomit Shalitin (Schneider Children’s Medical Center of Israel); Toronto: Kristina Cordeiro, Jill Hamilton, Hahn Y Nguyen, Shawna Steele (The Hospital for Sick Children, University of Toronto); Coordinating Centers: Fei Chen, Alan R Dyer, Wenyu Huang, Alan Kuang, Maria Jimenez, Lynn P Lowe, William L Lowe, Jr, Boyd E Metzger, Michael Nodzenski, Anna Reisetter, Denise Scholtens, Octavious Talbot, Paul Yim (Northwestern University Feinberg School of Medicine); Consultants: David Dunger, Alicia Thomas; NIDDK: Mary Horlick, Barbara Linder, Aynur Unalp-Arida; NICHD: Gilman Grave.
Financial Support: The HAPO Study was funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Institute of Diabetes and Digestive and Kidney Diseases (R01-HD-34242 and R01-HD-34243). The HAPO FUS was funded by the National Institute of Diabetes and Digestive and Kidney Diseases and the Eunice Kennedy Shriver National Institute of Child Health and Human Development (1U01-DK-094830).
Glossary
Abbreviations
- BMI
body mass index
- CI
confidence interval
- GDM
gestational diabetes mellitus
- HAPO
Hyperglycemia and Adverse Pregnancy Outcome
- FUS
Follow-up Study
- OGTT
oral glucose tolerance test
- OR
odds ratio
Additional Information
Disclosure Summary: The authors have no financial relationships or conflicts of interest relevant to this article to disclose.
Data Availability: The datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.
References
- 1. Chu SY, Kim SY, Bish CL. Prepregnancy obesity prevalence in the United States, 2004-2005. Matern Child Health J. 2009;13(5):614-620. [DOI] [PubMed] [Google Scholar]
- 2. Finucane MM, Stevens GA, Cowan MJ, et al. ; Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group (Body Mass Index) National, regional, and global trends in body-mass index since 1980: systematic analysis of health examination surveys and epidemiological studies with 960 country-years and 9·1 million participants. Lancet. 2011;377(9765):557-567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hod M, Kapur A, McIntyre HD; FIGO Working Group on Hyperglycemia in Pregnancy; FIGO Pregnancy and Prevention of early NCD Committee Evidence in support of the International Association of Diabetes in Pregnancy study groups’ criteria for diagnosing gestational diabetes mellitus worldwide in 2019. Am J Obstet Gynecol. 2019;221(2):109-116. [DOI] [PubMed] [Google Scholar]
- 4. Catalano PM. Obesity and pregnancy-the propagation of a viscous cycle? J Clin Endocrinol Metab. 2003;88(8):3505-3506. [DOI] [PubMed] [Google Scholar]
- 5. Hillier TA, Pedula KL, Schmidt MM, Mullen JA, Charles MA, Pettitt DJ. Childhood obesity and metabolic imprinting: the ongoing effects of maternal hyperglycemia. Diabetes Care. 2007;30(9):2287-2292. [DOI] [PubMed] [Google Scholar]
- 6. Dabelea D, Hanson RL, Lindsay RS, et al. Intrauterine exposure to diabetes conveys risks for type 2 diabetes and obesity: a study of discordant sibships. Diabetes. 2000;49(12):2208-2211. [DOI] [PubMed] [Google Scholar]
- 7. Patro Golab B, Santos S, Voerman E, Lawlor DA, Jaddoe VWV, Gaillard R; MOCO Study Group Authors Influence of maternal obesity on the association between common pregnancy complications and risk of childhood obesity: an individual participant data meta-analysis. Lancet Child Adolesc Health. 2018;2(11):812-821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Heslehurst N, Vieira R, Akhter Z, et al. The association between maternal body mass index and child obesity: a systematic review and meta-analysis. PLoS Med. 2019;16(6):e1002817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Freedman DS, Mei Z, Srinivasan SR, Berenson GS, Dietz WH. Cardiovascular risk factors and excess adiposity among overweight children and adolescents: the Bogalusa Heart Study. J Pediatr. 2007;150(1):12-17.e2. [DOI] [PubMed] [Google Scholar]
- 10. Patel S, Fraser A, Davey Smith G, et al. Associations of gestational diabetes, existing diabetes, and glycosuria with offspring obesity and cardiometabolic outcomes. Diabetes Care. 2012;35(1):63-71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Perng W, Gillman MW, Mantzoros CS, Oken E. A prospective study of maternal prenatal weight and offspring cardiometabolic health in midchildhood. Ann Epidemiol. 2014;24(11):793-800.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Simmonds M, Llewellyn A, Owen CG, Woolacott N. Predicting adult obesity from childhood obesity: a systematic review and meta-analysis. Obes Rev. 2016;17(2):95-107. [DOI] [PubMed] [Google Scholar]
- 13. HAPO Study Cooperative Research Group. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med 2008;358(19):1991-–2002.. [DOI] [PubMed] [Google Scholar]
- 14. Lowe WL Jr, Scholtens DM, Lowe LP, et al. ; HAPO Follow-up Study Cooperative Research Group Association of gestational diabetes with maternal disorders of glucose metabolism and childhood adiposity. JAMA. 2018;320(10):1005-1016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Lowe WL Jr, Lowe LP, Kuang A, et al. ; HAPO Follow-up Study Cooperative Research Group Maternal glucose levels during pregnancy and childhood adiposity in the Hyperglycemia and Adverse Pregnancy Outcome Follow-up Study. Diabetologia. 2019;62(4):598-610. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Cole TJ, Lobstein T. Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity. Pediatr Obes. 2012;7(4):284-294. [DOI] [PubMed] [Google Scholar]
- 17. Lohman TG. Assessment of body composition in children. Pediatric Exercise Science 1989;1:19-30. [DOI] [PubMed] [Google Scholar]
- 18. Rasmussen KM, Yaktine AL, eds. Weight Gain During Pregnancy: Reexamining the Guidelines; Washington, DC: Institute of Medicine; 2009. [PubMed] [Google Scholar]
- 19. HAPO Study Cooperative Research Group. Hyperglycaemia and Adverse Pregnancy Outcome (HAPO) Study: associations with maternal body mass index. BJOG 2010;117(5):575-584. [DOI] [PubMed] [Google Scholar]
- 20. International Association of Diabetes in Pregnancy Study Groups Consensus Panel, Metzger BE, Gabbe SG, Persson B, et al. International association of diabetes and pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy. Diabetes Care 2010;33(3):676- 682. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Cheung YB. A modified least-squares regression approach to the estimation of risk difference. Am J Epidemiol. 2007;166(11):1337-1344. [DOI] [PubMed] [Google Scholar]
- 22. Regression Modeling Strategies. [computer program]. R package version; 51-2; 2018. https://cran.r-project.org/web/packages/rms/index.html. Accessed January 2019. [Google Scholar]
- 23. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing; https://www.R-project.org. 2016. Accessed January 2019. [Google Scholar]
- 24. Navder KP, He Q, Zhang X, et al. Relationship between body mass index and adiposity in prepubertal children: ethnic and geographic comparisons between New York City and Jinan City (China). J Appl Physiol (1985). 2009;107(2):488-493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Vanderwall C, Randall Clark R, Eickhoff J, Carrel AL. BMI is a poor predictor of adiposity in young overweight and obese children. BMC Pediatr. 2017;17(1):135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Hughes AR, Sherriff A, Lawlor DA, Ness AR, Reilly JJ. Incidence of obesity during childhood and adolescence in a large contemporary cohort. Prev Med. 2011;52(5):300-304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Catalano PM, McIntyre HD, Cruickshank JK, et al. ; HAPO Study Cooperative Research Group The hyperglycemia and adverse pregnancy outcome study: associations of GDM and obesity with pregnancy outcomes. Diabetes Care. 2012;35(4): 780-786. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Gluckman PD, Hanson MA, Cooper C, Thornburg KL. Effect of in utero and early-life conditions on adult health and disease. N Engl J Med. 2008;359(1):61-73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Stunkard AJ, Foch TT, Hrubec Z. A twin study of human obesity. JAMA. 1986;256(1):51-54. [PubMed] [Google Scholar]
- 30. da Fonseca ACP, Mastronardi C, Johar A, Arcos-Burgos M, Paz-Filho G. Genetics of non-syndromic childhood obesity and the use of high-throughput DNA sequencing technologies. J Diabetes Complications. 2017;31(10):1549-1561. [DOI] [PubMed] [Google Scholar]
- 31. Smith J, Cianflone K, Biron S, et al. Effects of maternal surgical weight loss in mothers on intergenerational transmission of obesity. J Clin Endocrinol Metab. 2009;94(11):4275-4283. [DOI] [PubMed] [Google Scholar]
- 32. Guénard F, Deshaies Y, Cianflone K, Kral JG, Marceau P, Vohl MC. Differential methylation in glucoregulatory genes of offspring born before vs. after maternal gastrointestinal bypass surgery. Proc Natl Acad Sci U S A. 2013;110(28): 11439-11444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Olsen LW, Baker JL, Holst C, Sørensen TI. Birth cohort effect on the obesity epidemic in Denmark. Epidemiology. 2006;17(3):292-295. [DOI] [PubMed] [Google Scholar]
- 34. Ogden CL, Carroll MD, Lawman HG, et al. Trends in Obesity Prevalence Among Children and Adolescents in the United States, 1988-1994 Through 2013-2014. Jama. 2016;315(21): 2292-2299. [DOI] [PMC free article] [PubMed] [Google Scholar]