Maternal pre-pregnancy BMI and neurodevelopmental outcomes in children aged 18–36 months: a nationwide cohort study in Korea

Abstract

Maternal pre-pregnancy obesity is a recognized risk factor for adverse developmental outcomes in offspring; however, the impact of maternal underweight remains poorly understood. This study investigated the association between maternal body mass index (BMI) and neurodevelopmental outcomes in Korean children during the first three years of life. We conducted a retrospective cohort study of 258,367 mother–child dyads born between 2014 and 2021 who completed the Korean Developmental Screening Test at 18–24 and 30–36 months. Maternal BMI, obtained from national health screenings within three years prior to delivery, was categorized using Asian-specific criteria. At 18–24 months, offspring of underweight, obese, and severely obese mothers showed significantly higher risks of developmental delay. The most pronounced risks were observed among children of severely obese mothers, particularly in self-care (RR 2.04, 95% CI 1.92–2.17), cognition (1.98, 1.87–2.10), and language (1.51, 1.43–1.59). At 30–36 months, risks associated with maternal underweight were attenuated, whereas delays persisted in children of obese and severely obese mothers. These findings suggest that the threshold for neurodevelopmental risk may be lower than previously recognized, encompassing the overweight category. Both pre-pregnancy underweight and obesity elevate developmental risks, underscoring the necessity of proactive weight management before conception.

Introduction

Maternal pre-pregnancy weight is acknowledged as a pivotal determinant of both maternal and perinatal health outcomes. Pre-pregnancy obesity is associated with an increased risk of adverse pregnancy outcomes, including preeclampsia, cesarean delivery, and preterm birth^1,2. Beyond perinatal complications, maternal pre-pregnancy obesity has been linked to long-term health consequences in offspring, such as impaired cardiovascular function and childhood obesity^3–5. Moreover, maternal pre-pregnancy obesity is associated with 1.74-fold higher healthcare utilization in offspring, resulting in an additional medical costs exceeding 1,770 CAD per individual during the first 18 years of life⁶.

In addition to cardiometabolic consequences, maternal pre-pregnancy body mass index (BMI) has been implicated in offspring neurodevelopment. Maternal obesity has been negatively associated with cognitive, language, and gross motor development beyond six months of age⁷. While, maternal underweight has also been associated with adverse developmental outcomes, potentially reflecting intrauterine nutritional insufficiency⁸. Notably, a representative U.S. cohort study demonstrated a U-shaped association between maternal pre-pregnancy BMI and early childhood development, highlighting the need to understand the impact of the full BMI spectrum on neurodevelopment⁹.

Despite these findings, several critical gaps remain. First, the association between maternal weight and specific developmental domains is not yet fully articulated; while obesity and underweight are frequently linked to delays in cognition and language, findings regarding motor development remain inconsistent⁷. Second, although postnatal catch-up growth by age two is known to influence neurobehavioral status¹⁰, the impact of catch-up growth during the adjacent critical window of 18–36 months on neurodevelopmental outcomes remains unclear. Furthermore, previous research has predominantly been conducted in Western populations, potentially limiting the generalizability of the findings to populations with different genetic, nutritional, and socio-cultural backgrounds.

To address these limitations, we utilized data from the National Health Screening Program for Infants and Children (NHSPIC) in South Korea. Established in 2007, the NHSPIC is a nationwide initiative designed to systematically monitor and manage developmental milestones. This program offers a robust longitudinal framework that allows for a comprehensive evaluation of the association between maternal BMI and childhood neurodevelopment at critical early stages.

Therefore, this study aimed to examine the association between maternal pre-pregnancy BMI and neurodevelopmental outcomes in children aged 18–36 months, using representative population-based data from Korea.

Methods

Study data sources and participants

​This retrospective cohort study employed data from the National Health Insurance Service-Health Screening Cohort (NHIS-HEALS) and the NHSPIC, provided by the National Health Insurance Service (NHIS). The NHIS conducts biennial health screenings for all beneficiaries aged ≥ 20 years and collects self-reported medical histories, anthropometric measurements, and laboratory results¹⁰. The NHSPIC is a nationwide child health surveillance program covering children aged 4–71 months through seven follow-up screenings at 4–6, 9–12, 18–24, 30–36, 42–48, 54–60, and 61–71 months. Each screening includes a caregiver-com​pleted health questionnaire, developmental assessment using the Korean Developmental Screening Test (K-DST), an oral health questionnaire, oral examinations and counseling, anthropometric and physical examinations, and anticipatory guidance provided by primary care physicians with a face to face encounter¹¹. Mother-child dyads were characterized by matching their mothers using the family’s insurance card number and delivery date¹². In addition, the NHIS claims database holds detailed records of diagnoses, medical procedures, and prescriptions for the entire population of South Korea, all coded according to the International Classification of Diseases, Tenth Revision (ICD-10). This database is widely regarded as being representative of the Korean population and is frequently used in epidemiological studies.

Among 2,285,943 births between January 1, 2014, and December 31, 2021, a total of 886,641 mother-child dyads were eligible after excluding those without maternal health screening data within three years preceding delivery (n = 1,399,302). Of these, 628,274 children who did not complete the K-DST at 18–24 and 30–36 months were excluded. Finally, 258,367 mother–child dyads were included in the final analysis (Fig. 1).

Fig. 1.

Flow diagram of participant selection.

Ethics statement

The study was conducted in accordance with the principles of the Declaration of Helsinki. Because this study used fully anonymized data that are publicly available, the study protocol was exempted from full review by the Institutional Review Board (IRB) of the Konkuk University Medical Center (KUMC: 2023-10-023). The requirement for informed consent was also waived by the same IRB, as the analysis involved only de-identified data and posed no potential risk to the participants.

Clinical variables

Maternal BMI was determined based on the most recent measurements obtained during a health screening conducted within three years prior to delivery. This window was selected to accommodate the biennial nature of the Korean National Health Screening Program while ensuring data reflected recent pre-pregnancy status. Measurements taken during the 10-month gestational period were excluded to avoid the influence of pregnancy-related weight gain.

Maternal BMI was calculated by dividing weight by height squared (kg/m²) and classified into five categories based on the Asia-Pacific guidelines: underweight (BMI < 18.5), normal weight (BMI 18.5–22.9), overweight (BMI 23.0–24.9), obese (BMI 25.0–29.9; obese class I), and severely obese (BMI ≥ 30.0; obese classes II and III)¹³. Hypertension was diagnosed based on claims with ICD codes I10–I15 and O10, the prescription of antihypertensive medication, or a measured blood pressure ≥ 140/90 mmHg. Other maternal comorbidities, including pregnancy-induced hypertension, pre-gestational and gestational diabetes mellitus, and depression, as well as delivery methods, were identified using ICD-10 code as previously described¹⁴. Regarding neonatal outcomes, preterm birth was defined as a gestational age of less than 37 weeks. Small-for-gestational-age (SGA) was identified by birth weight below the 10th percentile for gestational age, using the ICD-10 codes P05.0, P05.1, and P05.9. Congenital malformation was defined using the broad range of ICD-10 codes from Q00 to Q98.4. Finally, neonatal intensive care unit (NICU) admissions were recorded identified through the following procedure codes: AJ101, AJ111, AJ121, AJ131, AJ144, AJ161, AJ201, AJ211, AJ221, AJ231, AJ244, AJ261, AJ301, AJ311, AJ321, AJ331, AJ351, and AJ051–AJ054.

Korean developmental screening test (K-DST)

The NHSPIC schedules at specific intervals: the first visit occurs at 4–6 months of age, the second at 9–12 months, the third at 18–24 months, the fourth at 30–36 months, the fifth at 42–48 months, the sixth at 54–60 months, and the seventh at 61–71 months of age. These check-ups are based on chronological age rather than corrected age. The health screening program encompasses medical history, physical examination, anthropometric measurements, visual acuity screening, developmental screening using the K-DST, oral examination, and questionnaires with anticipatory guidance.

The K-DST is comprised of six domains: gross motor function, fine motor function, cognition, language, sociality, and self-care. Each domain contained eight questions, each of which was scored on a scale from 0 to 3, with a maximum attainable score of 24 points per domain¹⁵. Gross motor development is closely associated with neuromotor function and may be impaired when there are abnormalities in the development of the central or peripheral nervous or musculoskeletal system. Fine motor development involves the use of the arms, hands, and fingers for precise movements and object manipulation. Through the development of coordination and manual dexterity, children gain the ability to explore and interact with their environment. This ability is also closely associated with cognitive development. Cognitive function encompasses a wide range of abilities, including the visual and auditory perception of the environment, integration of sensory information, reasoning, comparison and classification, memory, imitation, number and spatial concepts, and problem-solving skills. Language development reflects age-appropriate expressive and receptive abilities, and is associated with both cognitive and social functioning. The social domain includes behaviors, such as imitation, understanding others’ emotions, rule-based play, and imaginative play. Self-care skills assess a child’s ability to perform age-appropriate activities of daily living independently, such as eating, toileting, dressing, undressing, and maintaining personal hygiene. The self-care category is applicable only to participants aged 18 months or older.

The K-DST was initially standardized in 2014 (n = 3,284) subsequently revised through re-standardization using a larger NHSPIC cohort (n = 3.06 million, 2015–2016)¹⁶. At the age of 18–24 months, children typically acquire the ability to walk backward, differentiate between larger and smaller objects among two items, and convey a desire to others to observe their actions through gestures or verbal cues. By the age of 30–36 months, children generally exhibit the ability to walk forward along a straight line, trace over dotted lines with a pencil, identify their gender, and state their first and last names upon inquiry. Additionally, they can feed themselves independently and wait for their turn during shared activities¹⁶. Based on previously analyzed standard deviation scores, scores above 1 standard deviation (SD) are classified as ‘advanced development,’ scores between − 1 and 1 SD as ‘appropriate for age,’ scores between − 2 and − 1 SD as ‘follow-up needed,’ and scores below 2 SD as ‘further testing needed.’ In this study, scores below 2 SD are categorized as ‘developmental delay.’

Statistical analysis

Descriptive statistics were presented as mean ± standard deviation for continuous variables and as numbers (proportions) for categorical variables. Baseline characteristics were assessed according to maternal BMI using analysis of variance (ANOVA) and chi-square tests. To adjust for underlying differences in baseline characteristics between the groups, we derived the propensity score (PS) to balance the baseline covariates between the comparator and reference groups, using a multivariable logistic regression model that included all specified covariates. Crude incidence rates (IRs) per 1,000 individuals with developmental delays were calculated. We then estimated the PS-weighted relative risks (RRs) of developmental delays and 95% confidence intervals (CIs) using a generalized linear model¹⁷. The reference group consisted of participants with a healthy weight. These analyses were adjusted for maternal age, comorbidities, multiple births, congenital malformation, birth weight, sex, SGA, preterm birth, delivery method, and admission to the intensive care unit after birth. All statistical analyses were conducted using SAS (version 9.4; SAS Institute Inc., Cary, NC, USA).

Results

The study included 258,367 participants categorized according to maternal pre-pregnancy BMI as underweight (n = 31,290, 12.1%), healthy weight (n = 154,981, 60.0%), overweight (n = 34,548, 13.4%), obese (n = 30,418, 11.8%), or severely obese (n = 7,130, 2.8%) (Table 1). Maternal age, chronic hypertension, pregnancy-induced hypertension, pre-gestational diabetes, and gestational diabetes increased significantly with higher BMI (p < 0.001). Maternal depression was more prevalent in underweight, obese, and severely obese groups. In relation to the children’s characteristics, an increase in maternal BMI was correlated with higher birth weight and an elevated likelihood of NICU admission, whereas gestational age and the rate of vaginal delivery decreased. The distribution of child sex was not associated with maternal BMI. The incidence of preterm birth was lower in the healthy weight and overweight groups, with the lowest prevalence of congenital malformation observed in the healthy weight group. Although the incidence of SGA was relatively low in the obese group, it was significantly higher among the children of underweight mothers and those of severely obese mothers.

Table 1.

Baseline characteristics of the study population according to the maternal pre-pregnancy categories.

BMI	Underweight	Healthy weight	Overweight	Obese	Severe obese	P-value
N (%)	31,290 (12.1)	154,981(60.0)	34,548(13.4)	30,418(11.8)	7130(2.8)
Maternal characteristics
Age, year	31.6 ± 3.7	32.2 ± 3.8	32.8 ± 4.1	33.0 ± 4.1	33.1 ± 4.3	< 0.001
Age (≥ 35 years)	6219(19.9)	39,910(25.8)	10,953(31.7)	10,300(33.9)	2588(36.3)	< 0.001
Chronic hypertension	445(1.4)	2523(1.6)	761(2.2)	1163(3.8)	682(9.6)	< 0.001
PIH	2673(8.5)	15,462(10.0)	4235(12.3)	4458(14.7)	1669(23.4)	< 0.001
Depression	581(1.9)	2638(1.7)	580(1.7)	606(2.0)	173(2.4)	< 0.001
Pregestational DM	66(0.2)	553(0.4)	311(0.9)	622(2.0)	446(6.3)	< 0.001
Gestational DM	4109(13.1)	22,093(14.3)	5921(17.1)	6436(21.2)	2059(28.9)	< 0.001
Children’s characteristics
Gestational age, week	35.6 ± 2.3	35.5 ± 2.3	35.3 ± 2.5	35.2 ± 2.5	34.8 ± 2.6	< 0.001
Sex, male	15,957(51.0)	79,252(51.1)	17,664(51.1)	15,415(50.7)	3555(49.9)	0.180
Birth weight, kg	3.09 ± 0.41	3.17 ± 0.43	3.23 ± 0.45	3.27 ± 0.47	3.31 ± 0.53	< 0.001
Multiple birth	998(3.2)	5715(3.7)	1239(3.6)	1086(3.6)	303(4.3)	< 0.001
Preterm birth	1365(4.4)	6554(4.2)	1464(4.2)	1382(4.6)	412(5.8)	< 0.001
Vaginal delivery	20,162(64.4)	89,601(57.8)	17,480(50.6)	13,333(43.8)	2304(32.3)	< 0.001
Congenital malformation	2613(8.4)	12,330(8.0)	2841(8.2)	2620(8.6)	698(9.8)	< 0.001
SGA	351(1.1)	1306(0.8)	253(0.7)	183(0.6)	59(0.8)	< 0.001
NICU admission	1800(5.8)	9065(5.9)	2202(6.4)	2191(7.2)	703(9.9)	< 0.001

Data are shown as the mean ± standard deviation or number (percentage). BMI body mass index, PIH pregnancy-induced hypertension, DM diabetes mellitus, SGA small for gestational age, NICU neonatal intensive care unit.

Table 2 presents the developmental assessment outcomes at 18–24 and 30–36 months, categorized by maternal pre-pregnancy BMI. At 18–24 months, children of mothers classified as underweight, obese, or severely obese demonstrated an increased risk of developmental delays across all six domains of the K-DST compared to children of mothers with a healthy weight. In the cognitive domain, it is noteworthy that children born to overweight mothers demonstrated a significantly increased risk of developmental delay at 18–24 months. At 30–36 months, the overall incidence of developmental delay was higher than at 18–24 months; however, the previously observed elevated relative risks in the underweight group were no longer statistically significant. In contrast, higher maternal BMI remained consistently associated with an increased risk of developmental delay across all six domains, with a progressive gradient corresponding to greater obesity. Interestingly, at 30–36 months, although the incidence of delays in cognition and language continued to rise, the relative risks were reduced compared to the healthy weight group.

Table 2.

The neurodevelopmental risks according to the maternal pre-pregnancy BMI.

BMI	N	18–24 months				30–36 months
		Event	IR	RR(95% CI)	p-value	Event	IR	RR(95% CI)	p-value
Gross motor
Underweight	31,290	248	0.79	1.127(1.039–1.223)	0.004	273	0.87	1.025(0.952–1.103)	0.517
Healthy weight	154,981	1040	0.67	Reference		1348	0.87	Reference
Overweight	34,548	239	0.69	1.022(0.940–1.112)	0.609	299	0.87	0.986(0.914–1.063)	0.706
Obese	30,418	251	0.83	1.197(1.104–1.299)	< 0.001	303	1.00	1.170(1.088–1.258)	< 0.001
Severe obese	7130	68	0.95	1.656(1.534–1.789)	< 0.001	89	1.25	1.239(1.152–1.332)	< 0.001
Fine motor
Underweight	31,290	240	0.77	1.125(1.033–1.224)	0.007	478	1.53	0.975(0.921–1.031)	0.369
Healthy weight	154,981	973	0.63	Reference		2360	1.52	Reference
Overweight	34,548	228	0.66	1.048(0.961–1.143)	0.290	540	1.56	1.023(0.967–1.082)	0.433
Obese	30,418	233	0.77	1.22(1.121–1.327)	< 0.001	562	1.85	1.223(1.158–1.290)	< 0.001
Severe obese	7130	77	1.08	1.886(1.745–2.039)	< 0.001	159	2.23	1.443(1.369–1.522)	< 0.001
Cognition
Underweight	31,290	423	1.35	1.133(1.064–1.206)	< 0.001	433	1.38	0.925(0.873–0.981)	0.009
Healthy weight	154,981	1771	1.14	Reference		2283	1.47	Reference
Overweight	34,548	433	1.25	1.085(1.018–1.156)	0.013	526	1.52	1.017(0.960–1.076)	0.572
Obese	30,418	441	1.45	1.263(1.187–1.343)	< 0.001	597	1.96	1.307(1.238–1.379)	< 0.001
Severe obese	7130	142	1.99	1.983(1.873–2.100.873.100)	< 0.001	178	2.50	1.609(1.527–1.695)	< 0.001
Language
Underweight	31,290	572	1.83	1.170(1.109–1.235)	< 0.001	631	2.02	0.922(0.878–0.968)	0.001
Healthy weight	154,981	2384	1.54	Reference		3253	2.10	Reference
Overweight	34,548	558	1.62	1.036(0.980–1.095)	0.216	766	2.22	1.042(0.994–1.093)	0.091
Obese	30,418	522	1.72	1.107(1.047–1.169)	0.003	854	2.81	1.322(1.264–1.383)	< 0.001
Severe obese	7130	163	2.29	1.505(1.429–1.585)	< 0.001	263	3.69	1.707(1.635–1.783)	< 0.001
Sociality
Underweight	31,290	345	1.10	1.163(1.085–1.246)	< 0.001	515	1.65	0.976(0.925–1.03)	0.378
Healthy weight	154,981	1428	0.92	Reference		2598	1.68	Reference
Overweight	34,548	340	0.98	1.049(0.977–1.127)	0.189	599	1.73	1.014(0.961–1.070)	0.608
Obese	30,418	344	1.13	1.223(1.141–1.311)	< 0.001	658	2.16	1.283(1.219–1.35)	< 0.001
Severe obese	7130	106	1.49	1.860(1.744–1.984)	< 0.001	184	2.58	1.530(1.456–1.608)	< 0.001
Self-care
Underweight	31,290	341	1.09	1.152(1.076–1.235)	< 0.001	540	1.73	0.995(0.943–1.049)	0.838
Healthy weight	154,981	1454	0.94	Reference		2630	1.70	Reference
Overweight	34,548	347	1.00	1.046(0.974–1.123)	0.217	582	1.68	0.964(0.913–1.017)	0.182
Obese	30,418	352	1.16	1.218(1.137–1.305)	< 0.001	635	2.09	1.219(1.158–1.283)	< 0.001
Severe obese	7130	110	1.54	2.042(1.918–2.174)	< 0.001	182	2.55	1.392(1.324–1.464)	< 0.001

BMI body mass index, incidence rates were calculated per 100 individuals. RR relative risk, CI confidence interval.

Discussion

This study provides evidence of a significant association between maternal pre-pregnancy BMI and neurodevelopmental delay in offspring aged 18–36 months. Maternal weight status was found to influence child development, with both obesity and underweight increasing the risk of delays across gross motor, fine motor, cognitive, language, social, and self-care domains compared to children of mothers with healthy weight. Importantly, maternal obesity was consistently associated with an elevated risk of developmental delay, demonstrating a clear dose–response relationship. Notably, our findings reveal that even maternal overweight status—often considered a transitional phase before clinical obesity—was associated with a significant risk of cognitive delay, particularly during the early 18–24 month period.

This finding is particularly significant given the rising global prevalence of overweight and obesity among women of childbearing age¹⁸. In South Korea, obesity rates among women in their 20 s and 30 s have nearly doubled over the past decade¹⁹. While previous studies in New York²⁰ and China²¹ have highlighted the risks associated with clinical obesity, our results extend these concerns to the overweight category, suggesting that the threshold for neurodevelopmental risk may be lower than previously recognized. The clinical implication of this finding is that the window for pre-pregnancy intervention should not be limited to obese individuals but should encompass overweight or underweight women as well⁸. From a public health perspective, this underscores the necessity of proactive weight management prior to conception to mitigate subtle yet significant developmental vulnerabilities in offspring. Even when results did not reach full statistical significance in all domains, the consistent upward trend in risk among overweight mothers emphasizes the need for early screening and monitoring of their children.

The relationship between maternal pre-pregnancy BMI and child developmental outcomes can be attributed to various biological and metabolic mechanisms. Maternal pre-pregnancy obesity is linked to an altered intrauterine environment due to increased inflammation, pro-inflammatory cytokines, and adipokines (such as IL-6, TNF-α, leptin, and insulin), along with reduced levels of beneficial hormones, such as adiponectin²². These inflammatory mediators and hormonal imbalances can traverse or affect the placenta, subjecting the developing fetus to biochemical stress. Moreover, epigenetic modifications of neurotrophic genes, including the brain-derived neurotrophic factor (BDNF) gene, as well as alterations in the microbiome and impaired serotonin and dopaminergic signaling, may directly enhance the permeability of the fetal blood-brain barrier, leading to cognitive, behavioral, and motor developmental delays^23–25. Furthermore, obese mothers are at an elevated risk of developing gestational diabetes and hypertension during pregnancy. These conditions may lead to fetal hypoxia, nutrient imbalance, or preterm delivery. The pro-inflammatory and metabolic disturbances associated with maternal obesity likely initiate a cascade of placental and fetal brain changes that may adversely affect early neurodevelopment²².

Conversely, inadequate maternal nutrition, characterized by insufficient intake of calories, proteins, and essential micronutrients, may result in placental insufficiency and restricted fetal growth, thereby potentially compromising brain development²⁶. However, our findings showed that the developmental risks in children of underweight mothers were attenuated by 30–36 months. This recovery may be attributed to postnatal catch-up growth. Previous studies have indicated that children born SGA who achieve catch-up growth by age two exhibit neurodevelopmental outcomes comparable to their peers^27,28. This suggests that the developmental delays associated with maternal underweight may be more related to nutritional deficits that can be mitigated through adequate postnatal environment and growth.

In contrast, the developmental risks associated with maternal obesity persisted through 36 months, suggesting a different underlying mechanism. Unlike the transient nutritional deficiencies of underweight, maternal obesity is associated with chronic low-grade systemic inflammation and lipo-toxicity, which can lead to permanent structural and functional changes in the developing fetal brain²⁹. Furthermore, epigenetic modifications and alterations in the hypothalamic-pituitary-adrenal axis induced by maternal obesity may have long-lasting effects that are less susceptible to postnatal compensatory growth³⁰. This divergence explains why the risk attenuation observed in the underweight group was not mirrored in the obesity group, highlighting the more persistent neurobiological impact of maternal overnutrition.

In our study, no significant sex-related differences were observed in the association between maternal BMI and child development outcome. However, previous studies, including those based on low-income, multiethnic birth cohorts, have reported an increased risk of poor developmental outcomes, particularly in male children³¹. These findings will be important for future studies to continue comparing results across populations and to investigate any differences by child sex or specific developmental domains, as some prior studies have suggested that cognitive and language domains are more sensitive to maternal BMI influences than gross motor skills⁸.

This study has several limitations. First, developmental outcomes were evaluated using routine screening tools during health check-ups, which employed brief developmental tests rather than comprehensive clinical assessments^9,32. Although these screening instruments are effective in identifying potential developmental delays, they may be susceptible to reporting bias, and may not fully capture the complexity of a child’s developmental status. Consequently, there is a possibility of misclassification: a child might initially screen positive for a delay, but later develop typically, or subtle deficits might remain undetected. Second, the analysis used pre-pregnancy BMI as recorded in the health records; however, BMI is an imperfect indicator of nutritional status and body composition. It does not account for fat distribution or micronutrient deficiencies, and there may be measurement errors in the self-reported weight and height data. Gestational weight gain was not measured in this study; therefore, we cannot exclude the possibility that inadequate or excessive gestational weight gain may influence neurodevelopmental outcomes among children born small for gestational age³³. However, existing literature suggests that perinatal factors such as gestational weight gain, gestational age, and birth weight do not fully mediate the observed associations³⁴. Furthermore, the influence of pre-pregnancy BMI may differ according to the geographical region and ethnic background. When employing region-specific BMI classifications, the relationship between maternal pre-pregnancy BMI and fetal health outcomes may differ across various continents³⁵. Third, this study could not fully account for all potential confounding variables due to the inherent limitations of the NHIS database. Specifically, detailed information on socioeconomic and educational factors was not available, which may influence both maternal BMI and child development through differences in nutrition and healthcare access³¹. Adjustments were made for some available confounders, such as maternal age, parity, or birth outcomes. Residual confounding by unmeasured factors such as parental education, home environment, or genetic predispositions cannot be ruled out³⁶. Postnatal influences, including reduced breastfeeding, diverse infant feeding practices, or limited physical interaction due to maternal health conditions, may also impact early development, although the study design primarily implicates prenatal factors³⁷. Furthermore, as the data were extracted based on pre-defined clinical BMI categories in accordance with the initial research protocol, we were unable to analyze BMI as a continuous variable. While continuous modeling could have refined the identification of optimal BMI cut-offs, our analysis was constrained by these pre-established data extraction parameters. Future studies with access to raw continuous values are warranted to validate these findings using advanced modeling techniques such as splines. Although we conducted multiple statistical comparisons across various developmental domains and time points, we did not apply formal correction methods. Given that these developmental domains are highly interrelated and measured longitudinally, such corrections could be overly conservative. To minimize the risk of Type I errors, we focused on highly significant associations and consistent clinical trends. However, the possibility of false-positive results cannot be entirely excluded, and our findings should be considered exploratory. Furthermore, our inclusion criteria required children to have completed K-DST screenings at both the 18–24 and 30–36 month intervals. Although this longitudinal design enabled consistent assessment of developmental trajectories over time, it may have introduced selection bias by excluding children who participated in only a single screening. As a result, the final analytic cohort represented approximately 11% of all births during the study period. While the large, nationwide nature of the cohort supports the statistical robustness of the observed associations, the potential for limited generalizability should be considered. Future studies incorporating the broader screened population, including children with partial screening data, may help to validate and extend our findings. Lastly, the prevalence of congenital malformations observed in our study was higher than that reported in previous literature³⁸. This is likely attributable to the nature of the NHIS, which incorporates ‘rule-out’ diagnoses recorded for reimbursement purpose or diagnostic evaluations. In addition, recent nationwide Korean studies have suggested that certain conditions, such as atrial septal defect (ASD), may be overestimated due to the inclusion of suspected cases or the misclassification of patent foramen ovale³⁸. Furthermore, our broad inclusion of ICD-10 codes ranging from Q00 to Q98.4 may have encompassed minor or clinically insignificant anomalies, thereby contributing to the higher prevalence compared with registry-based studies applying more stringent diagnostic validation or exclusion criteria. Most participants were assessed using the revised K-DST (2017), though a minor subset born in 2014 likely used the original edition. Despite minor item-level differences, both versions utilize an identical − 2 SD cutoff for developmental delay, maintaining the robustness of our comparative analysis.

In South Korea, current preconception public health initiatives, primarily managed by local public health centers, focus largely on folic acid supplementation and screening for infectious diseases³⁹. Although nutritional support programs are available, eligibility is generally limited to low-income pregnant or postpartum women, leaving a substantial policy gap in preconception weight management for the broader population of women of reproductive age. To address this gap, recent pilot initiatives have begun incorporating digital healthcare platforms to monitor BMI and provide lifestyle counseling for women of childbearing age; however, these efforts have not yet been integrated into a coordinated national strategy.

Based on our findings, we suggest that clinicians should routinely assess BMI in women of reproductive age and provide evidence-based counseling regarding optimal weight before conception. From a policy perspective, structured weight management strategies targeting women in the overweight category prior to pregnancy should be considered. Furthermore, linking preconception screening efforts with NHSPIC data may facilitate early identification of high-risk mother–child dyads and enable timely preventive strategies.

In conclusion, maternal pre-pregnancy BMI is a critical determinant of neurodevelopmental outcomes in offspring during the early years of life. Our findings demonstrate a significant association between maternal weight status and developmental delays in children aged 18–36 months, emphasizing the public health importance of achieving an optimal pre-pregnancy BMI to mitigate long-term neuropsychological risks. To further refine these clinical insights, future research should employ advanced statistical techniques, such as restricted cubic splines, to identify precise optimal BMI ranges and offer a more granular view of the non-linear relationship between maternal weight and offspring development. Additionally, longitudinal studies that incorporate gestational weight changes and comprehensive socio-environmental factors are warranted to clarify the persistent nature of these developmental trajectories and inform more targeted pre-conception interventions.

Author contributions

Conceptualization, JS, TEK, SHP, and HWP; methodology, JS, TEK, SHP, and HWP; formal analysis, SHP; writing—original draft preparation, JS; writing—review and editing, TEK, SHP, and HWP; and funding acquisition, JS. All the authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by Konkuk University Medical Center Research Grant 2025.

Data availability

The data supporting this article are accessible from the National Health Insurance Service Open Data Portal ([https://nhiss.nhis.or.kr/](https:/nhiss.nhis.or.kr)). The research management number is REQ2025040312-001, and accessed from June, 12, 2025.

Declarations

Competing interests

The authors declare no competing interests.

Ethic approval

All procedures involving human participants were conducted in accordance with the ethical standards of the Institutional Review Board (IRB) of Konkuk University Medical Center, which approved the exemption from ethical review because all data were fully anonymized and contained no identifiable information, as well as with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. The study was exempted from full IRB review by the IRB of Konkuk University Medical Center (KUMC: 2023-10-023).