Maternal Overnutrition in Pregnancy Predicts Greater Body Mass Index Gains in Offspring by Early Midlife: A Systematic Review and Meta‐Analysis

ABSTRACT

Maternal overnutrition during pregnancy predisposes children to increased adiposity later in life. This systematic review and meta‐analysis examined how this association persists into adulthood and whether it differs by sex. MEDLINE, EMBASE, Scopus, and CENTRAL were searched for human studies published in English before May 2025. From 9191 screened records, 29 studies (430,764 mother–offspring pairs) provided quantitative data on the link between maternal pre‐pregnancy body mass index (ppBMI) or gestational weight gain (GWG) and adiposity in adult offspring. The meta‐analysis found significant positive associations between maternal ppBMI and offspring BMI, body fat percentage, or waist circumference. Each 1 kg/m² increase in maternal ppBMI was linked to a 0.4 kg/m² higher BMI and 0.35% more body fat in offspring. Maternal overweight/obesity predicted significantly higher offspring BMI or waist circumference, especially in females. Excessive GWG increased the odds of overweight/obesity in the offspring by 35% compared with adequate GWG. This correlation was stronger in early midlife (mean age 30–45) than in young adults. In early midlife offspring, each additional kilogram of GWG was linked to a 0.14 kg/m² higher BMI, independent of ppBMI. These findings underscore the importance of nutritional interventions before and during pregnancy to reduce obesity risk in adulthood.

Abbreviations

BIA

bioelectrical impedance analysis

BMI

body mass index

CI

confidence interval

DEXA

dual‐energy X‐ray absorptiometry

GWG

gestational weight gain

IOM

US Institute of Medicine

MD

weighted mean difference

OR

odds ratio

ppBMI

pre‐pregnancy body mass index

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses

PROSPERO

International Prospective Register of Systematic Reviews

QUIPS

Quality in Prognosis Studies

WC

waist circumference

1. Introduction

The prevalence of overweight and obesity has increased dramatically over the past decades; therefore, today nearly 3 billion people are affected worldwide, with the highest rates observed among middle‐aged adults [1]. Obesity among women of reproductive age and during pregnancy is rising in parallel with trends in the general population, and through its intergenerational effects, it compromises offspring health span and life expectancy, underscoring the public health need for prevention [2, 3]. In addition to the known role of lifestyle factors (e.g., eating habits, smoking, and sedentary lifestyle), there is growing evidence that the modulation of fetal gene expression by maternal overnutrition just before or during pregnancy also increases the risk of obesity/overweight in the offspring via intrauterine developmental programming [4]. Maternal pre‐pregnancy body mass index (ppBMI) or excessive gestational weight gain (GWG) are used as indicators of a high‐calorie nutritional environment during pregnancy. Previous meta‐analyses have confirmed the association of high ppBMI or excessive GWG with obesity/overweight in children independent of socioeconomic and lifestyle‐related factors [3, 5, 6, 7, 8]. However, it is still not clarified whether this association remains significant in adulthood. Results of the available studies are controversial [9, 10, 11]. Only one meta‐analysis included a subgroup of two studies in adult population and suggested increased relative risk of obesity in the offspring of mothers with excessive GWG [6]. The changes in the association during aging of the offspring have not yet been studied except for only a few follow‐up studies on ppBMI, but their results are also inconclusive [12, 13, 14]. This would be of particular interest for the development of age‐related obesity in the middle‐aged population. A comprehensive meta‐analysis did not observe sex differences in the associations of ppBMI or GWG with the risk of overweight/obesity across childhood [8]. A possible sex difference of the association in adulthood is still unclear; the available data are controversial [15, 16, 17]. Therefore, we aimed to systematically review the literature and perform a meta‐analysis about the impact of the high ppBMI or excessive GWG (independent of ppBMI) on overweight/obesity in adult offspring. We also aimed to analyze the potential influence of offspring's age and sex on these associations. We hypothesized positive correlations of both ppBMI and GWG with risk for overweight/obesity in adult offspring, and that these associations are independent of offspring age or sex.

2. Methods

2.1. Search Strategy

Our search was conducted in MEDLINE, EMBASE, Scopus, and CENTRAL databases until May 1, 2025. The protocol was registered onto the Prospective Register of Systematic Reviews (PROSPERO: CRD42017058858). Search terms were (offspring OR son OR daughter) AND (adult*) AND (maternal OR gestation* OR pregestat* OR pre‐gestat* OR prenat* OR perinat* OR pregnan* OR prepregnan* OR pre‐pregnan* OR preconception* OR pre‐conception* OR periconception* OR peri‐conception* OR pregravid* OR pre‐gravid* OR antenatal* OR mother) AND (obesity OR overnutrit* OR over‐nutrit* OR overweight OR adiposity OR obese OR “body weight” OR “weight gain” OR “weight change” OR “weight growth” OR “body mass” OR “body mass index” OR BMI OR “skinfold thickness” OR “body composition” OR “body fat” OR “fat mass” OR “lean mass” OR “fat‐free mass”) with English language and human filters. Duplicates were removed with EndNote X7.0.2. The cited and citing (by Google Scholar search engine) articles were also screened for eligible articles. No supplementary information was obtained from investigators of the original studies; only published data were used.

2.2. Study Selection

Using meta‐analysis of observational studies in epidemiology (PRISMA, checklist: Table S1) guidelines [18], two researchers (S.E. and E.P.) conducted the screening and data extraction independently. After screening the titles and abstracts, all potentially relevant studies were retrieved for full‐text evaluation. Disagreements were resolved by a third investigator (M.B.). To be included, each article had to provide data about maternal overnutrition just before or during singleton pregnancy along with data about their adult offspring's (at least 18 years of age) overnutrition. We excluded studies that provided data only of mothers with eclampsia or gestational diabetes mellitus and did not give relevant data of healthy mothers. However, we included articles that published data of mixed populations that applied proper adjustment for these pathological conditions. Further exclusion criteria were animal experiments, non‐English studies, studies providing maternal data only after birth, twin and preterm pregnancies. We extracted the following data from each study: first author, publication year, country, study setting with inclusion and exclusion criteria, maternal age at delivery, timing of the maternal measurement, number of participants, offspring's age and sex, maternal data about nutritional status just before or during pregnancy, data about offspring's overnutrition, data (with corresponding measures of precision, e.g., standard deviation) describing the association between maternal data and their adult offspring's overnutrition, and adjusted covariates. The values adjusted for most confounders were extracted. If duplicate studies were found within the same data source, the larger population was selected.

2.3. Risk of Bias Assessment

Risk of bias for each study was assessed by the Quality in Prognosis Studies (QUIPS) tool [19]. Two independent investigators (S.E. and P.M.) performed the quality assessment separately. Disagreements were resolved by a third author (E.P.). QUIPS covers six main domains. For each item of the domains, “yes,” “no,” “partly,” or “unclear” was used. An overall rating for each domain was assigned as carrying “low,” “moderate,” or “high” risk of bias defined in the footnote of Table S2.

2.4. Data Analysis

We used the random effects model to summarize the individual study findings. Due to this weighting method, studies with low participant numbers or big variations were assigned lower weights during the analysis. We presented our findings on forest plots reporting the average effect size, its 95% confidence interval (CI), and the results of the statistical tests. With regard to the association between maternal nutritional state and offspring's parameters, we analyzed either unstandardized regression coefficients or odds ratios (ORs) or weighted mean differences (MDs). If the available data allowed, we performed subgroup analyses by maternal BMI (body mass index) categories, offspring's age, and sex. Several studies provided multiple data items using the only control group of the given article thereby presenting overlapping populations. These cannot be used simultaneously in the same analysis. In order to avoid overrepresentation of such populations, we performed sensitivity analyses. In other cases, if it was possible, common OR or MD values were calculated to form a merged group.

To assess heterogeneity, the Q test and the I ² statistics were calculated. The I ² statistic represents the percentage of effect size heterogeneity that cannot be explained by random chance. Heterogeneity could be interpreted as moderate (30%–60%), substantial (50%–90%), or considerable (above 75%) [20]. We considered the Q test significant if p < 0.1. For all other tests, statistical significance was considered if p value < 0.05.

To test the presence of publication bias (small‐study effect), we assessed the symmetry of the funnel plots visually. All statistical analyses were performed with R program (R core Team 2022, V4.1.3) using meta (Version 6.5–0) and metafor (Version 4.4–0) packages.

3. Results

3.1. Search, Selection, and Characteristics of the Studies

The systematic search produced 20,650 records. After removing duplicates and screening titles, abstracts, and full‐text papers for eligibility, 37 studies were included in the qualitative synthesis (Figure 1), eight of which were not appropriate for meta‐analysis [14, 21, 22, 23, 24, 25, 26, 27]. The reason for their exclusion is shown in Table S3. The remaining 29 studies were analyzed (Table 1) [9, 10, 11, 12, 13, 15, 16, 17, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48]. These included studies were published from 2001 to 2024. Five articles were retrospective cohort [13, 17, 32, 34, 41], one interventional [10], one cross‐sectional [31], and the remaining 22 articles were prospective cohort studies. Data of 430,764 mother–offspring pairs were involved in our analysis. The number of participants (mother–offspring pairs) ranged 37–146,894 per study. Caucasians were the dominant ethnic group, mostly from Europe and the United States. The offspring's mean age at outcome assessment ranged 18–72 years. The majority of the studies contained pooled data of males and females. Seven articles [15, 16, 17, 33, 40, 42, 48] analyzed both sexes separately providing independent data items. One study [40] reported results of two birth cohort studies (ESTER and AYLS) labeled with numbers 1 and 2 in parentheses in the figures, respectively. The included studies reported ppBMI (assessed just before pregnancy or at the first antenatal visit) or maternal BMI prior to delivery or GWG (total or maximum) as indicators of maternal overnutrition and BMI or waist circumference (WC) or body fat as offspring's adiposity measures. Pre‐pregnancy body weight was (at least partly) self‐reported for the assessment of ppBMI in 14 studies [12, 13, 15, 16, 28, 29, 30, 31, 34, 35, 36, 37, 45, 48]. In all other cases, the maternal parameters were obtained from direct measurements or registries. In four articles [15, 31, 37, 38], definition of excessive GWG was based on ppBMI, by GWG categories according to the US Institute of Medicine (IOM) guidelines [49]. In other cases, GWG exceeded 16 kg in the excessive group [9, 13, 39], which is considered to be excessive independent of ppBMI. The BMI values were categorized into three groups: overweight/obesity (ppBMI > 25 kg/m²), overweight (25 < ppBMI ≤ 30 kg/m²), and obesity (ppBMI > 30 kg/m²). Multiple data items of a given article based on different maternal and offspring's BMI categories are labeled with numbers and letters in parentheses in the figures. Using continuous variables, regression coefficients were assessed to describe the relationship between maternal and offspring data.

FIGURE 1.

Flowchart of the study selection process.

TABLE 1.

Characteristics of the included studies.

Author, year, country, reference number	Sample size	Maternal characteristics			Offspring characteristics
		Age a (year)	Prevalence of overnutrition (%)	Mean value of the nutritional state a	Age a (year)	Male (%)
Bhat, 2015, Australia, P [28]	957	29.1 95% CI (28.8; 29.5)	ND	ppBMI: 22.3 kg/m2 95% CI (22.0; 22.5)	20	47.8
Brunton, 2021, Canada, P [29]	3217	29 ± 4.56	ppBMI = 25–29.9 kg/m2: 15.2; ppBMI > 30 kg/m2: 5	ppBMI: 22.8 ± 3.7 kg/m2	18	44.8
Bucci, 2016, Finland, P [10]	37	ND	BMI > 28.1 kg/m2: 45.9	ND	71.9 ± 0.8	0
Chaparro, 2017, Sweden, P [30]	113	27.4 ± 3.9	ppBMI > 25 kg/m2: 11.5	ppBMI: 22.3 ± 2.7 kg/m2	21.21 ± 1.9	53
de Oliveira Nascimento Freitas, 2023, Brazil, C [31]	124	ND	ppBMI > 25 kg/m2: 10.5 excessive GWG: 30	23–24	28 (24–31)	0
Derraik, 2015, Sweden, R [32]	24,197	26.2 ± 5.0	ppBMI = 25–29.9 kg/m2: 13.9; ppBMI > 30 kg/m2: 3.1	ND	22.4 ± 2.3	0
Eriksson, 2015, Finland, P [33]	1186	28.6 ± 5.5	BMI > 28.1 kg/m2: 65.7	BMI prior to delivery: 26.5 ± 2.6 kg/m2	62	50
Eshriqui, 2021, Brazil, R [34]	150	19–35	ppBMI > 25 kg/m2: 14.7	ND	22 (20–29)	0
Ester, 2017, USA, P [35]	929	ND	ND	ppBMI: 22.4 kg/m2 GWG: 9.4 kg	44.1 ± 1.9	0
Hochner, 2012, Israel, P [36]	1256	28.38 ± 5.47	ND	ppBMI: 23.9 ± 3.7 kg/m2 GWG: 11.0 ± 4.6 kg	32	49.5
Houghton, 2016, USA, P [37]	1035	26.0 ± 5.7	ppBMI > 25 kg/m2: 23.6; excessive GWG: 9.8	ppBMI: 26.8 ± 4.2 kg/m2 In the subset for GWG‐analysis: ppBMI < 25 kg/m2	44.1 ± 1.9	0
Hrolfsdottir, 2015, Denmark, P [38]	308	29.0 ± 4.9	ppBMI > 25 kg/m2: 0; excessive GWG: 29	ppBMI with adequate GWG: 21.0 ± 1.5 kg/m2 ppBMI with excessive GWG: 21.2 ± 1.4 kg/m2 GWG: 14.4 ± 4.9 kg	20	39.3
Iliadou, 2010, Sweden, P [39]	86,075	25–29	ppBMI = 25–30 kg/m2: 12.6; ppBMI > 30 kg/m2: 2	ND	18	100
Kaseva, 2018, Finland, P [40]	(1) ESTER study: 325 (2) AYLS study: 375	ND	(1) ESTER study: ppBMI > 25 kg/m2: 13.5 (2) AYLS study: ppBMI > 25 kg/m2: 29	(1) ESTER study: 22.2 ± 1.8 kg/m2 (2) AYLS study: 23.1 ± 2.4 kg/m2	(1) ESTER study: 23.6 ± 1.1 (2) AYLS study: 25.3 ± 0.6	(1) ESTER study: 50 (2) AYLS study: 45.8
Koupil, 2008, Sweden, R [41]	1103	ND	ND	ppBMI: 21.9 ± 2.9 kg/m2	18.2	100
Laitinen, 2001, Finland, P [42]	5598	ND	ppBMI = 25–29.9 kg/m2: 19.3; ppBMI > 30 kg/m2: 3.4	ppBMI: 23.05 ± 2.9 kg/m2	31	45.7
Lawlor, 2011, Sweden, P [43] “A”	146,894	27.8 ± 5.0	ppBMI = 25–30 kg/m2: 12.5; ppBMI > 30 kg/m2: 1.9	ppBMI: 21.9 ± 3.1 kg/m2	18	100
Lawlor, 2011, Sweden, P [11] “B”	125,748	28.0 ± 5.0	ppBMI > 25 kg/m2: 0	ppBMI: 21 ± 3 kg/m2 GWG: 14.2 ± 4.2 kg	18	100
Mamun, 2009, Australia, P [15]	2271	25.5 ± 5.0	Excessive GWG: 34	ppBMI with adequate GWG: 21.1 ± 3.4 kg/m2 ppBMI with excessive GWG: 22.7 ± 3.8 kg/m2	21	50
Mingrone, 2008, Italy, R [17]	67	24.9 ± 3.6	ppBMI > 30 kg/m2: 77	ND	23.8 ± 4.5	41.8
Ounjaijean, 2021, Thailand, P [44]	566	26.6 ± 4.6	ppBMI > 25 kg/m2: 9.4	ppBMI: 21.55 ± 1.77 kg/m2	20.6 ± 0.5	41.2
Rath, 2016, Australia, P [45]	1015	21–33	ppBMI = 25–30 kg/m2: 11; ppBMI > 30 kg/m2: 6	ND	22	50.8
Rerkasem, 2024, Thailand, P [46]	391	26.3 ± 4.4	ppBMI > 25 kg/m2: 8.2	ppBMI: 21.3 ± 2.5 kg/m2	20.5 ± 0.5	45
Rooney, 2011, USA, P [47]	453	ND	ppBMI > 30 kg/m2: 10.2	ND	19–20	44.6
Schack‐Nielsen, 2010, Denmark, P [9]	1566	25.5 ± 6.4	ppBMI > 25 kg/m2: 9; excessive GWG (> 16 kg): 19	ppBMI with optimal GWG (10–12 kg): 21.5 ± 2.4 kg/m2 ppBMI with excessive GWG (> 16 kg): 21.6 ± 2.8 kg/m2	42	46.6
Schoppa, 2019, USA, P [12]	701	ND	ppBMI > 25 kg/m2: 9.6	ppBMI: 23.06 ± 3.03 kg/m2	66 ± 9	47
Stuebe, 2009, USA, R [13]	19,501	25.70 ± 4.72	ppBMI > 25 kg/m2: 7	ppBMI with optimal GWG (7–9 kg): 21.1 ± 2.5 kg/m2 ppBMI with excessive GWG (> 16 kg): 21.21 ± 2.73 kg/m2	36–56	100
Tequeanes, 2009, Brazil, P [48]	2377	20–34	ppBMI = 25–30 kg/m2: 17.2; ppBMI > 30 kg/m2: 4.9	ND	23	53
Zalbahar, 2016, Australia, P [16]	2229	26.1 ± 5.0	ppBMI > 25 kg/m2: 15.8	ppBMI: 22.0 ± 3.9 kg/m2	21	50

Abbreviations: C: cross sectional, GWG: gestational weight gain, ND: no data, P: prospective study, ppBMI: pre‐pregnancy body mass index, R: retrospective study.

^a Values are expressed in mean ± standard deviation or 95% CI or range or mean (interquartile range).

3.2. Study Quality

Results of the risk of bias assessment are found in Table S2. The domains “Prognostic factor measurement” (maternal nutritional state) and “Outcome measurement” (offspring's nutritional state) and “Statistical analysis and reporting” were evaluated as low risk of bias in all studies. Eleven studies carried moderate risk of bias [11, 28, 29, 35, 39, 42, 43, 44, 45, 47, 48], and four studies carried high risk of bias [10, 12, 17, 33] in the domain “Study confounding” because they did not report how the important cofactors were accounted for in the analysis. All other ones showed a low risk of bias in this domain. In 44% of the studies, the high dropout rate caused selection bias (“Study attrition”). Two of these latter studies [12, 33] showed the highest risk of bias because they were rated as high risk in two domains (“Study Attrition” and “Study Confounding”). Three studies carrying the lowest risk of bias were rated as low risk in all domains [36, 38, 41]. Overall, 48% of the included studies (n = 14) were judged to have a low or moderate risk of bias across all assessed domains. High selection bias was identified in 38% of the studies (n = 11) due to high dropout rates.

3.3. Association Between Maternal ppBMI Versus Overweight or Obesity in the Offspring

The analysis of studies that investigated the OR for overweight/obesity (BMI > 25 kg/m²) in the adult offspring showed significantly higher odds in offspring of mothers with ppBMI above 25 kg/m² compared with those of normal‐weight mothers (Figure 2A). Maternal obesity was associated with significantly higher odds than maternal overweight (4.04‐fold vs. 2.55‐fold increase, p < 0.001). Similar results were obtained when ORs for obesity (BMI > 30 kg/m², Figure 2B) in offspring were analyzed: 5‐fold versus 2.96‐fold increase (p = 0.009). Odds of being overweight/obese or obese did not differ in offspring of mothers either with obesity (comparison of upper panels in Figure 2A vs. B, p = 0.092) or overweight (comparison of lower panels in Figure 2A vs. B, p = 0.363).

FIGURE 2.

Forest plot of the odds ratios (ORs) for overweight/obesity (body mass index [BMI] above 25 kg/m², A) and obesity (BMI above 30 kg/m², B) in the adult offspring of mothers with pre‐pregnancy BMI (ppBMI) above 25 kg/m² compared with those of normal‐weight mothers (ppBMI below 25 kg/m²). The effects of maternal obesity (upper panels) and overweight (lower panels) were compared by subgroup. Gray squares show OR values with the area reflecting the weight assigned to the individual studies. Horizontal bars indicate 95% confidence intervals (CIs). Diamonds show the overall point estimates with 95% CI.

With regard to the offspring's age, we compared the OR for obesity/overweight in subgroups of studies in which the mean age of participants was below or above 30 years. We could perform these analyses for two different maternal ppBMI categories: 26–29 and > 25 kg/m², but not for maternal obesity because only one study on offspring aged over 30 years was available [37]. Our result (Figure S1) showed significantly higher odds in both age groups of offspring of mothers with ppBMI above 25 kg/m² compared with those of normal‐weight mothers. The OR values for the older and younger subgroups were similar (3.36‐fold vs. 2.86‐fold increase in odds, respectively; p = 0.602), suggesting that the association persists with age in adulthood. Similarly, no age‐related difference was found in the analysis of OR in offspring of mothers with overweight (Figure S2). Several studies provided ORs for multiple offspring's BMI categories (overweight, obesity or BMI above 25) as compared with the only control group of the given article (i.e., offspring of normal‐weight mothers). Therefore, these OR data items of the given article contain overlapping populations, and they cannot be used simultaneously in the same plot. Additional sensitivity analyses using the OR data items from the same article alternatively (each analysis contains only one such OR value from each article) showed similar results (Tables S4 and S5).

Analysis of association between ppBMI and offspring's BMI as continuous variables showed significant positive correlation (Figure 3). Subgroup analysis based on offspring's mean age (Figure 3A) showed higher regression coefficient in the group of offspring older versus younger than 30 years, suggesting a trend of an age‐related strengthening of the association, but this difference was not statistically significant. This result indicates that the effect of maternal exposure does not weaken with age. The available data allowed subgroup analysis based on sex of the offspring (Figure 3B). Regression coefficient for female offspring was higher than in male ones, but the direct comparison of the male and female subgroups failed to show significant difference (p = 0.597). Thus, the analysis of the regression coefficients indicates a 0.4 kg/m² increase in offspring's BMI for every 1 kg/m² gain in maternal ppBMI, independent of offspring's age and sex.

FIGURE 3.

Forest plot of the regression coefficients describing the association between maternal pre‐pregnancy body mass index (ppBMI) and offspring's BMI. Subgroup analyses were performed based on the age (A) and the sex (B) of the offspring. Gray squares show regression coefficient values with the area reflecting the weight assigned to the individual studies. Horizontal bars indicate 95% confidence intervals (CIs). Diamonds show the subtotal and overall point estimates with 95% CI (m = men, w = women).

Six articles contained offspring's mean BMI values as outcomes in relation to different maternal ppBMI categories. Four of them [17, 40, 42, 48] provided two or four independent data items according to offspring's sex and maternal ppBMI. Our meta‐analysis on weighted MD indicated significantly higher mean BMI in offspring of mothers with obesity or overweight compared with those of normal‐weight mothers in all three subgroups according to maternal ppBMI categories (Figure S3). Direct comparison of the first two subgroups also confirmed that maternal obesity was associated with significantly higher increase in mean BMI of offspring than maternal overweight: MD: 5.29, 95% CI (2.04; 8.54) versus 1.50, 95% CI (1.13; 1.87) kg/m², subgroup difference: p = 0.023. In contrast to offspring of normal‐weight mothers, mean BMI of offspring with maternal obesity or overweight or overweight/obesity exceeded 25 kg/m² suggesting that maternal overnutrition results in overweight of the offspring (mean BMI values: 27.80, 25.28, and 25.55 kg/m², respectively). The available data allowed analysis of sex differences. Because the degree of maternal overnutrition significantly affected the MD in offspring's BMI between mothers with elevated ppBMI and those with normal weight, we further analyzed the sex‐specific MD by calculating paired male‐minus‐female MD differences in offspring's BMI from studies that included data of both sexes within the same maternal ppBMI categories (ppBMI = 26–29 kg/m² or above 30 kg/m²). We used MD values relative to peers whose mothers had normal weight. Analysis of these paired differences provided significant negative estimate indicating smaller increase of BMI in men than in women (Figure 4). Compared with the offspring of normal‐weight mothers, maternal overnutrition was associated by 1.17 kg/m² higher BMI increase in women than in men (p = 0.018). Sensitivity analysis by excluding a small study in which 77.6% of mothers had pre‐pregnancy obesity, even morbid obesity and 72% of the offspring had overweight or obesity, or morbid obesity [17], also provided significant result (p = 0.033) and the heterogeneity remained substantial: I ² = 66%, p = 0.033. In contrast to the analysis of continuous variables (cf. Figure 3B), our categorization‐based analysis yielded significant result showing impact of sex on the association between maternal ppBMI and offspring's BMI.

FIGURE 4.

Forest plot of the paired male‐minus‐female differences (weighted mean differences, MDs, based on data pairs from the same studies) of MD in body mass index (BMI) of offspring with maternal obesity or overweight compared with the corresponding offspring's group of normal‐weight mothers. Gray squares show MD values with the area reflecting the weight assigned to the individual studies. Horizontal bars indicate 95% confidence intervals (CIs). Diamond shows the overall point estimate with 95% CI (the numbers in parentheses indicate which values of the given article was used in the analysis. 1: pre‐pregnancy BMI = 26–29 kg/m², 2: pre‐pregnancy BMI above 30 kg/m²).

The BMI value does not necessarily reflect body composition and the degree of adiposity; therefore, we also analyzed studies which provided data about body composition of the young adult offspring using DEXA (dual‐energy X‐ray absorptiometry) [30, 31, 34, 46] or BIA (bioelectrical impedance analysis) [40]. Our results were in accordance with those on BMI. Offspring of mothers with ppBMI above 25 kg/m² showed significantly higher body fat percentage than those of normal‐weight mothers: MD 3.05%, 95% CI (2.07; 4.03), p < 0.001 (Figure S4). Analysis of regression coefficients (Figure S5) revealed that each 1 kg/m² increase in ppBMI was associated with an additional 0.35% body fat (95% CI [0.19; 0.52], p < 0.001).

Measurement of the WC is a widely used tool to define visceral obesity. Five articles [17, 31, 40, 42, 46] reported offspring's mean WC values as outcomes in different maternal ppBMI categories and provided 10 independent data items. Analysis of these data showed significantly higher mean WC in offspring of mothers with high ppBMI compared with those of normal‐weight mothers: MD: 9.43 cm, 95% CI (2.88; 15.98), p = 0.005 (Figure S6). The available data did not allow statistical analysis of sex differences in the same manner as for offspring's BMI. However, in studies including both sexes within the same maternal ppBMI categories, comparison of mean differences in WC between male and female offspring revealed a consistent trend: Maternal overnutrition was associated with a greater increase in WC among females than males (Table S6).

Two studies [10, 33] investigated offspring's mean BMI in relation to maternal BMI prior to delivery and provided three independent data items. This maternal parameter depends on both ppBMI and GWG. In accordance with our results on ppBMI, analysis of these data also showed significantly higher mean BMI in offspring of mothers with BMI above 25 kg/m² compared with those of normal‐weight mothers: MD: 1.15 kg/m², 95% CI (0.67; 1.63), p < 0.001 (Figure S7).

3.4. Association Between Maternal GWG Versus Offspring's BMI

The analysis of studies that investigated OR for obesity or overweight in adult offspring showed significantly higher, 1.35‐fold increased odds in offspring of mothers with excessive GWG than in those of mothers who gained weight within the recommended range (adequate GWG) (Figure 5). Because the mean ppBMI was in the optimal range (with a prevalence of maternal overweight or obesity below 15%) in all included datasets, the enhanced odds could therefore be attributed to excessive GWG when ppBMI was appropriate. Our subgroup analysis based on the mean age of the offspring indicated higher OR in the older group, but this difference between offspring younger versus older than 30 years was not significant. Two studies of the older subgroup [9, 13] provided ORs for multiple offspring's BMI categories as compared with the only control group of the given article. Therefore, we performed additional sensitivity analyses using the OR data items from the same article alternatively and obtained similar results suggesting a maintained high risk with age in the adulthood (Table S7).

FIGURE 5.

Forest plot of the odds ratios (ORs) for overweight/obesity in the adult offspring of mothers with excessive gestational weight gain (GWG) compared with those of mothers who gained weight within the recommended range (adequate GWG). Subgroup analysis was performed based on the mean age of the offspring. Gray squares show OR values with the area reflecting the weight assigned to the individual studies. Horizontal bars indicate 95% confidence intervals (CIs). Diamonds show the subtotal and overall point estimates with 95% CI. The number “2” and letter “b” in parentheses indicate that OR for obesity (body mass index [BMI] above 30 kg/m²) was used in the analysis from the given articles. In other cases OR for BMI above 25 kg/m².

Analysis of association between GWG and offspring's BMI as continuous variables also showed significant positive correlation (Figure 6). The regression coefficient indicates 0.09 kg/m² increase in BMI of offspring with every 1‐kg gain in maternal GWG. Interestingly, our subgroup analysis based on the mean age of the offspring revealed a marked difference. The group of offspring over 30 years of age (mean age 30–45 years) showed significantly higher regression coefficient than that of younger ones: 0.14 versus 0.06, respectively (p = 0.033). Because all values were adjusted for ppBMI, our results suggest that BMI of the offspring in early midlife increases by 0.14 kg/m² with every 1‐kg gain in maternal GWG independent of maternal ppBMI. These findings underline the significant contribution of GWG to the development of adiposity in offspring of even normal‐weight mothers, especially in the middle‐aged population.

FIGURE 6.

Forest plot of the regression coefficients describing the association between maternal gestational weight gain (GWG) and offspring's body mass index. Subgroup analysis was performed based on the mean age of the offspring. Gray squares show regression coefficient values with the area reflecting the weight assigned to the individual studies. Horizontal bars indicate 95% confidence intervals (CIs). Diamonds show the subtotal and overall point estimates with 95% CI.

3.5. Publication Bias

Visual inspection of the funnel plots suggested no small‐study effect. Thus, the results of our analyses appear not to be influenced by small‐study effect.

4. Discussion

4.1. Main Findings

Our meta‐analysis is the first large‐scale quantitative analysis with substantial statistical power demonstrating the long‐term impact of ppBMI and GWG separately on the BMI of the adult offspring, despite inconsistencies in the published literature. Our results showed significant positive correlation between maternal overnutrition—both just before pregnancy (ppBMI) and during pregnancy (GWG)—and offspring's BMI. Each 1 kg/m² gain in maternal ppBMI results in a 0.4 kg/m² rise in the adult offspring's BMI and 0.35% body fat, with an additional 0.09 kg/m² increase in BMI per 1 kg of maternal weight gained during pregnancy. These correlations did not diminish with age; rather, they showed an increasing trend. Compared with the offspring of normal‐weight mothers, maternal obesity was associated with significantly higher increase in mean BMI of offspring and higher OR for obesity or overweight in offspring than maternal overweight (5.29 vs. 1.5 kg/m², 5‐fold vs. 2.96‐fold increase, respectively). Compared with normal maternal weight, maternal ppBMI above 25 kg/m² was associated with significantly higher mean BMI and WC in the adult offspring—particularly in females. We found a 1.17 kg/m² greater increase in BMI among women than in men. In addition, the correlation between ppBMI and offspring BMI tended to be stronger in females than in males. Excessive GWG was associated with 35% higher OR for obesity/overweight in the offspring than adequate GWG. The correlation between maternal GWG and offspring's BMI was significantly stronger in offspring in early midlife (ages 30–45) than in young adults (ages 18–30). In offspring during early midlife, each 1‐kg increase in maternal GWG was associated with a 0.14 kg/m² higher BMI, independent of maternal ppBMI.

4.2. Interpretation of the Results With Regard to the Offspring Age

Previous meta‐analyses focused only on children and found significant associations between maternal ppBMI or GWG and offspring BMI, which became stronger throughout childhood as age increased [5, 8]. A follow‐up study from childhood to adulthood also found that the association between ppBMI and offspring BMI strengthened with age [27]. The only meta‐analysis that reported a subgroup including two studies in adults demonstrated a higher risk ratio for obesity in the offspring of mothers with excessive GWG. However, they observed a somewhat decreased obesity risk from childhood to adolescence, followed by a slight increase in adulthood [6]. These differences were not significant. Only a few longitudinal studies have analyzed, at multiple adult ages, the impact of maternal ppBMI or GWG on offspring BMI in the same population, but their findings are inconsistent. One such study found a weaker association between maternal adiposity (assessed by ppBMI and GWG) and having obesity or overweight among offspring in middle adulthood (ages 36–56) than at age 18 [13]. Another study described a similar positive correlation between ppBMI and offspring BMI at ages 20 and 55 [14]. Over longer intervals between in utero exposure and the assessment of offspring weight, other factors influencing adiposity—such as diet and physical activity—would attenuate the strength of the association between maternal and offspring adiposity. On the contrary, a more recent study on change in BMI over time in adulthood (33–66 years) in relation to maternal pre‐pregnancy overweight or obesity demonstrated a more rapid weight acceleration in early and mid‐adulthood among individuals exposed to maternal obesity or overweight in utero—compared with those with mothers with normal weight [12]. Before the age of 50 (but not thereafter), the association persisted even after adjusting for inherited genetic predisposition to obesity and lifestyle factors such as physical activity, diet, smoking, and education. A difference in slope of BMI change was observed at 50 years of age: The excess increase in BMI per year among the exposed compared with unexposed between 18 and 50 years of age was significant (0.07 kg/m² per year), but the difference in BMI change over time among exposed compared with unexposed after 50 years of age was not significantly different [12].

Our meta‐analysis findings align with those reported in the latter study. We also demonstrated that offspring exposed to maternal pre‐pregnancy overweight or obesity had significantly higher odds of overweight or obesity in adulthood compared with those unexposed. Our results indicated a significant positive correlation between maternal ppBMI and offspring BMI not only in younger but also in mid‐adulthood. Despite the long interval between maternal exposure and the outcome in the offspring, the impact of maternal overnutrition did not diminish with age. The somewhat higher regression coefficient observed in offspring over 30 compared with younger adults suggests a trend of an age‐related strengthening of the association. When comparing our results with those of a meta‐analysis conducted in children [8], a clear age‐related increasing trend can be observed—particularly in relation to maternal overweight. From early childhood (ages 2–5) to late childhood (ages 10–18), the reported OR for overweight or obesity in the offspring increased from 1.66 to 2.28. In adulthood, we observed even higher values, ranging from 2.55 to 2.96. A substantial part of the maternal ppBMI–offspring BMI association is explained by offspring genotype. In addition to genetic predisposition, maternal pre‐pregnancy obesity reflects maternal fat accumulation and low‐grade inflammation, which may substantially impair intrauterine metabolic programming mechanisms leading to a long‐term effect on adult BMI. Emerging evidence suggests that in utero epigenetic mechanisms, including DNA methylation alterations and shifts in the gut microbiome, play a role in the development of obesity in offspring [50].

The impact of ppBMI is known to be augmented by a limited but significant effect of the amount of weight gain during pregnancy. A previous meta‐analysis conducted in children observed that excessive GWG was related to a 39%–72% higher risk of overweight/obesity throughout childhood with the strongest effect at later ages [8]. According to their results, offspring weight status was associated with increasing GWG across the full range, not just when it exceeds the IOM recommendations. They concluded that the additional effect of GWG was smaller in mothers who had obesity before pregnancy compared with those with normal or overweight ppBMI [8]. Compared with the corresponding OR values of the aforementioned meta‐analysis, a study investigating the impact of excessive GWG—stratified by maternal ppBMI—on middle‐aged adult daughters reported consistently higher ORs for having offspring with overweight or obesity [37]. This study also observed a diminished additional effect of GWG among daughters with maternal pre‐pregnancy obesity. Specifically, among mothers with overweight who exceeded GWG guidelines, the odds of having a daughter with overweight or obesity were significantly elevated, showing a 2.6‐fold increase compared with peers with similar ppBMI who stayed within recommendations. In contrast, for individuals with obesity who exceeded GWG guidelines, the OR was not statistically significant [37]. In our meta‐analysis focusing on adult offspring, we could not stratify GWG effects by maternal ppBMI category. Normal‐weight mothers with adequate GWG served as the reference group in all included datasets. Because the average maternal ppBMI in the included cohorts fell within the normal range (20–25 kg/m²) or we selected the dataset of normal‐weight mothers, our results could be compared with the trends of those reported for children of normal‐weight mothers in the earlier meta‐analysis [8]. Within the normal‐weight ppBMI category, their data from both binary and linear regression models followed a similar pattern during childhood: increasing from early childhood (ages 2–5) to mid‐childhood (ages 5–10), followed by a slight decline in late childhood (ages 10–18). The corresponding OR values indicated a 28%, 36%, and 30% increase in odds related to excessive GWG, respectively. Extending this trend, our estimates showed a further age‐related increase in odds during adulthood. Excessive GWG was associated with elevated odds of overweight or obesity in offspring—particularly after age 30, where the odds rose by 36%, compared with a 34% increase in younger adults (ages 18–30), although this difference did not reach statistical significance. In our analysis using continuous variables, the regression coefficient was small in young adulthood, though it remained statistically significant. Interestingly, after age 30 (early midlife)—the onset of typical age‐related weight gain—the coefficient increased markedly from 0.04 to 0.14, indicating a substantial strengthening of the association between maternal GWG and offspring BMI with age. In summary, these findings suggest that the increasing trend during childhood is followed by a slight decline in adolescence, possibly due to a diminishing influence of shared lifestyle factors as children grow older. Thereafter, the association between maternal GWG and offspring BMI strengthens considerably. Evidence suggests that childhood BMI accounts for only part of the GWG–adult BMI association, implying that excessive GWG may program a persisting susceptibility to obesogenic environments, independent of shared family characteristics [9, 37]. Thus, this increasing strength of the association with age might reflect the impact of intrauterine programming mechanisms becoming more apparent when a young adult gets older. These intrauterine effects may contribute to long‐term alterations in appetite regulation, neuroendocrine pathways affecting adipose tissue development and function, or energy metabolism, thereby promoting obesity in later life [51]. Our findings underscore the substantial role of maternal overnutrition during pregnancy in influencing offspring adiposity during this life stage, potentially contributing to the high obesity burden among middle‐aged adults.

4.3. Interpretation of the Results With Regard to the Offspring Sex

Previous research has demonstrated sex differences in the development of weight and body fat during childhood [52]. However, an earlier meta‐analysis found no sex‐specific differences in the association between maternal ppBMI and offspring BMI in children, likely because BMI does not distinguish between fat and lean mass [8]. Indeed, another meta‐analysis found a positive link between maternal ppBMI and fat mass in children [7], whereas a retrospective study revealed sex‐specific differences in offspring body composition up to age 6, with maternal adiposity having a stronger impact on boys than on girls [53]. Our meta‐analysis confirmed significant direct associations of ppBMI with both general and visceral adiposity in adult offspring. Offspring of mothers who had overweight or obesity before pregnancy showed 3% higher body fat and nearly 10 cm greater WC than those born to mothers with normal weight. The association with fat mass index, examined in three included studies, was consistently significant, showing a 0.017–0.03 kg/m² increase for every 1 kg/m² rise in ppBMI [31, 34, 46]. Our results are consistent with the findings of an earlier study that we could not include in the meta‐analysis due to the incomplete outcome data. They also found that greater WC, body fat percentage, and fat mass index of young adult offspring were associated with both higher maternal ppBMI and GWG, independent of cofactors such as maternal age, smoking status, or social class [54]. Only one of the included studies examined the relationship between maternal GWG and offspring body composition in young women, and it found that each additional kilogram of GWG was linked to 3.5% more body fat and a 0.22 kg/m² higher fat mass index [31]. Regarding the influence of sex on the association between maternal ppBMI and offspring's BMI in adulthood, our meta‐analysis found a more pronounced effect among females. Regression coefficients were generally higher in women than in men. Compared with peers with normal ppBMI, the BMI increase associated with elevated maternal ppBMI was significantly greater in daughters than in sons (within the same maternal ppBMI categories). To further assess visceral adiposity, we analyzed offspring's WC, which was greater—especially in females—among those born to mothers who had overweight or obesity before pregnancy, suggesting heightened sensitivity to maternal adiposity in women. Only one study investigated sex‐specific differences in the relationship between maternal ppBMI and adult offspring body composition [30]—and consistent with our findings on BMI and WC, the association was significant in daughters but not in sons (see Figure S5). Accordingly, offspring visceral adipose tissue assessed by DEXA was significantly correlated with maternal ppBMI in a study of females [34], but not in a mixed population [46].

Sex‐specific differences in epigenetic responses have been well documented in animal studies, with female offspring showing greater susceptibility to maternal nutritional influences during pregnancy [55, 56, 57]. Females appear more sensitive to intrauterine conditions, possibly due to heightened responsiveness to placental changes. The offspring's sex affects both placental development and its adaptability to adverse stimuli [57, 58]. This sexual dimorphism may be explained by sex hormone–driven differences in epigenetic gene regulation, contributing to varying disease susceptibility between males and females [59]. In line with these findings, our results suggest that in female offspring, the genetic predisposition to epigenetic influences in utero becomes evident only in adulthood—not in childhood. Consequently, maternal obesity leads to a greater increase in adiposity, BMI, WC, and fat mass in daughters than in sons. This transgenerational association suggests that maternal overnutrition during pregnancy, through its impact on offspring, may trigger a vicious cycle that accelerates the obesity epidemic across generations. Preventing obesity at the onset of a woman's reproductive life may therefore reduce the risk of obesity in her children, particularly in daughters.

4.4. Strengths and Limitations

Our study is the first meta‐analysis that demonstrates the quantitative long‐term impact of maternal overnutrition on adiposity (assessed by BMI, WC, or body fat percentage) in adult offspring. The main strength of our study is that we could separately analyze the effects of ppBMI and those of GWG independent of ppBMI. Further strengths include the investigation of the influence of offspring sex and age, the high number of participants and exclusion of twin and preterm pregnancies. In addition, all included studies dealt with pathological pregnancies such as preeclampsia and maternal diabetes mellitus appropriately (applying either exclusion or proper adjustments for these conditions). Limitations include the involvement of five retrospective studies, which may carry some risk of bias and use of self‐reported pre‐pregnancy weight in half of the studies. The main conclusion regarding age‐dependent impact of GWG on BMI was not biased by the retrospective studies. Nevertheless, self‐reported and measured maternal weight have been reported to be strongly correlated [60]. Pre‐pregnancy weight was self‐reported as a continuous value and collected in early pregnancy in the included studies, minimizing participant‐driven misclassification and long‐term recall bias. Available evidence indicates that any remaining error is small, with self‐reported weight differing from measured weight by less than 1 kg/m² of BMI on average [60, 61]. In addition, four included studies provided evidence of no systematic bias associated with self‐reported weight, supporting the validity of self‐reported maternal pre‐pregnancy weight through subset or correlation analyses [12, 16, 29, 35]. The high (> 20%) loss to follow‐up documented in prospective studies may introduce selection bias, as follow‐up participation is typically selective, with dropouts more likely to have lower socioeconomic status (SES) and higher ppBMI [11, 39, 43, 48]. This pattern would plausibly lead to attenuation rather than overestimation of the association between maternal overnutrition and adult offspring adiposity, supporting that the observed significant associations are robust rather than exaggerated. Further limitations include lack of investigation of specific confounders known to influence the association between maternal overnutrition and offspring adiposity: maternal age, smoking habits, SES, and breast feeding [9, 31, 34] because only 11 studies considered the first three cofactors and only one breast feeding in the analyses. Only 17% of the included studies adjusted for offspring socioeconomic factors, reflecting the limited availability of such data in long‐term follow‐up studies. As maternal overnutrition is strongly socially patterned and maternal SES shows substantial intergenerational continuity with offspring SES, adjustment for maternal SES is likely to account for much of the potential socioeconomic confounding [62]. Nevertheless, residual confounding by offspring SES cannot be fully excluded.

With regard to our conclusive results on ppBMI, the heterogeneity of the data was substantial only in the analysis of the MD in offspring BMI. This indicates the presence of other necessary determining factors in the background. Data on regression coefficients showed considerable heterogeneity except for the subgroup of offspring older than 30 years and for that of females suggesting the influence of age and sex on the association between ppBMI and offspring BMI. In the analyses of GWG, subgroup analyses by age reduced heterogeneity to a non‐significant level, especially in the offspring over 30 years old, supporting the significant role of offspring age in the association between GWG and offspring BMI in adulthood.

4.5. Implication for Practice and Research

Although the additional effect of GWG beyond that of maternal ppBMI is modest, both remain important modifiable risk factors for offspring weight status, with substantial implications at the population level. Because BMI‐specific recommendations are lower for individuals with overweight or obesity, those who exceed the guidelines tend to have higher ppBMI than those who adhere to them. Most individuals with overweight gain more weight during pregnancy than recommended, and excessive GWG is more common among individuals with obesity than among those with a normal BMI [49].

Our findings highlight that optimizing preconceptional body weight and controlling GWG are crucial, as even a 1 kg/m² increase in maternal ppBMI results in measurable increases in adult offspring BMI and body fat percentage, whereas excessive GWG independently raises the risk of obesity. These results underscore the importance of implementing targeted nutritional and lifestyle interventions during pregnancy to reduce long‐term obesity risk at the population level. This aligns with the principles of translational medicine, which emphasize that accelerating the translation of scientific evidence into societal benefit is highly important [63, 64]. Therefore, future research and public health initiatives should prioritize the prevention of maternal overweight and excessive GWG to secure long‐term health benefits for future generations.

5. Conclusion

Our meta‐analysis is the first large‐scale quantitative analysis with substantial statistical power to demonstrate that the quantitative positive association between maternal overnutrition (high ppBMI and GWG) and offspring obesity/adiposity not only persists into adulthood but is not attenuated after the age of 30; instead, it shows a strengthening trend. The impact of maternal obesity in adulthood exceeds the risk observed earlier in childhood. Moreover, our novel findings indicate that the correlation between maternal GWG and offspring BMI is significantly stronger in early midlife (30–45 years) than in young adulthood (18–30 years), independent of ppBMI. In addition, we are the first to show that the effect of high maternal ppBMI on adult offspring BMI is sex‐dependent, with a greater BMI increase observed in female than in male offspring. Our findings emphasize the importance of preconceptional weight management and interventions targeting women of reproductive age to achieve a normal weight. These observations suggest persisting susceptibility to obesogenic environments induced by intrauterine malprogramming due to maternal overnutrition and highlight the importance of avoiding excessive weight gain during pregnancy even in normal‐weight women, potentially contributing to the obesity burden in middle‐aged adults.

Author Contributions

S.E. and E.P. carried out the design and developed the search strategy. S.E. and E.P. conducted the searches, screening, and data extraction. P.M. carried out the data analysis, and M.B. and P.H. advised on the interpretation of the results. S.E. drafted the manuscript, and E.P. oversaw the study as the principal investigator. All authors made significant contributions to the study and approved the final manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: PRISMA 2020 checklist.

Table S2: Quality in Prognosis Studies (QUIPS) tool.

Table S3: Characteristics of the excluded studies and the reason of their exclusion.

Table S4: Sensitivity analysis to Figure S1.

Table S5: Sensitivity analysis to Figure S2.

Table S6: Comparison of mean differences in waist circumference between male and female offspring.

Table S7: Sensitivity analysis to Figure 5.

Figure S1: Forest plot of the odds ratios (ORs) for overweight or obesity in the adult offspring of mothers with pre‐pregnancy body mass index (ppBMI) above 25 kg/m² compared with those of normal‐weight mothers (ppBMI below 25 kg/m²).

Figure S2: Forest plot of the odds ratios (ORs) for overweight or obesity in the adult offspring of mothers with overweight (pre‐pregnancy body mass index [ppBMI] = 26–29 kg/m²) compared with those of normal‐weight mothers (ppBMI below 25 kg/m²).

Figure S3: Forest plot of the weighted mean differences (MDs) in body mass index (BMI) between groups of offspring of mothers with pre‐pregnancy BMI (ppBMI) above 25 kg/m² and those of normal‐weight mothers (ppBMI below 25 kg/m²).

Figure S4: Forest plot of the weighted mean differences (MDs) in body fat (%) between groups of offspring of mothers with pre‐pregnancy body mass index (ppBMI) above 25 kg/m² and those of normal‐weight mothers (ppBMI below 25 kg/m²).

Figure S5: Forest plot of the regression coefficients describing the association between maternal pre‐pregnancy body mass index (ppBMI) and offspring's body fat (%).

Figure S6: Forest plot of the weighted mean differences (MDs) in waist circumference between groups of offspring of mothers with pre‐pregnancy body mass index (ppBMI) above 25 kg/m² and those of normal‐weight mothers (ppBMI below 25 kg/m²).

Figure S7: Forest plot of the weighted mean differences (MDs) in body mass index (BMI) between groups of offspring of mothers with BMI prior to delivery above 25 kg/m² and those of normal‐weight mothers (BMI below 25 kg/m²).

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.