Double burden of malnutrition by mid-childhood in an Indian birth cohort: a longitudinal study (2010–2021) Authors: Birsen Yilmaz, Venkata Raghava Mohan, Sushil John, Gagandeep Kang, Felix Jebasingh, Nihal Thomas, Mahendra Sonawane, Ullas Kolthur-Seetharam, Beena Koshy Published: 2026 ABSTRACT SummaryBackgroundThe double burden of malnutrition (DBM) is characterized by concurrent malnourishment and obesity and is largely prevalent in low- and middle-income countries. This study evaluated the longitudinal patterns of body mass index (BMI) from birth to 9 years of age in an urban Indian birth-cohort: Malnutrition and Enteric Disease (MAL-ED) cohort, to understand the emergence of DBM and the role of early-life factors.MethodsThe MAL-ED study recruited 251 children from urban slums in Vellore, Tamil Nadu, India, between 2010 and 2012, with further follow-ups conducted until 9 years of age. Standardized anthropometric measurements were performed at the ages of 2, 3, 4, 5, 7, and 9 years, with additional BMI calculations performed at each time point. After normality checks, appropriate statistical analyses were performed for individual variables, and later-life BMI was modeled using multinomial regression.FindingsApproximately 45% of children were stunted at 2 years of age. By 7 years, 26.3% of children were classified as thin with 5.2% of children classified as overweight/obese. By 9 years, the prevalence of underweight was 21.6%, while overweight/obesity increased to 14.6%. Maternal BMI was a predictor of childhood thinness, particularly at ages 5 and 9.InterpretationDBM though present in early childhood, became more pronounced after 5 years of age and intensified by 9 years. Low maternal BMI was associated with the BMI trajectory of the offspring. Hence, age-specific and maternal health interventions are crucial for addressing the risk of DBM in vulnerable children.FundingThis research was supported by the 10.13039/100000865Bill and Melinda Gates Foundation (grant no. OPP 47075) for the original MAL-ED study, DBT/Wellcome Trust India Alliance fellowship (grant no. IA/CPHI/19/1/504611) for the follow-up and TIFR-DAE funding (RTI4003; 19P-0911). Research in contextEvidence before this studyRecent studies have highlighted the growing complexity of the double burden of malnutrition (DBM) in India and other low- and middle-income countries. Analyses of nationally representative datasets have shown that undernutrition and overweight increasingly coexist within the same populations and even within mother–child dyads, reflecting socioeconomic and nutritional transitions. We searched PubMed, Scopus, and Web of Science databases for studies published until July 2025 using appropriate combinations of the following search terms: “double burden of malnutrition,” “triple burden of malnutrition,” “body mass index,” “children,” “longitudinal cohort,” “India,” “maternal BMI,” “birth weight,” and “childhood obesity.” Reference lists of relevant reviews and key articles were also screened to identify additional studies. We included observational and longitudinal studies that examined malnutrition patterns or body mass index (BMI) trajectories in children, and excluded case reports, commentaries, and studies lacking quantitative anthropometric data. While several studies have reported the prevalence of the DBM among children at various ages, longitudinal data are lacking on the trajectory of BMI from birth through childhood and adolescence. Furthermore, existing research, including from the Malnutrition and Enteric Diseases (MAL-ED) cohort, indicates that early stunting often begins in the first 1000 days, which is a critical period for growth and development, while obesity mostly emerges later. However, without longitudinal data, the transition between states of malnutrition, as well as the influence of early-life factors and the long-term consequences of these transitions on health outcomes, remain unclear. Therefore, longitudinal studies are essential for understanding how BMI evolves over time in the context of malnutrition, particularly in rapidly urbanizing and economically transitioning settings, such as India.Added value of this studyTo the best of our knowledge, this is the first study to report the onset of DBM in children by mid-childhood in India using data from the MAL-ED cohort. The persisting association between maternal and childhood BMI concurs with the findings in the literature and highlights the importance of promoting maternal health in addition to mitigating DBM in early childhood. The observed associations provide insights into the emerging DBM but should be interpreted in the context of the cohort size and exploratory analyses.Implications of all the available evidenceIn this study DBM emerges in mid-childhood after 5 years of age, with a worsening trend between 7 and 9 years of age. Low maternal BMI had a lasting effect on the BMI trajectory of the offspring. Further studies are needed to understand complex societal changes, including diet patterns, physical activity schedules, and broader ecological systems. Research in context Evidence before this study Recent studies have highlighted the growing complexity of the double burden of malnutrition (DBM) in India and other low- and middle-income countries. Analyses of nationally representative datasets have shown that undernutrition and overweight increasingly coexist within the same populations and even within mother–child dyads, reflecting socioeconomic and nutritional transitions. We searched PubMed, Scopus, and Web of Science databases for studies published until July 2025 using appropriate combinations of the following search terms: “double burden of malnutrition,” “triple burden of malnutrition,” “body mass index,” “children,” “longitudinal cohort,” “India,” “maternal BMI,” “birth weight,” and “childhood obesity.” Reference lists of relevant reviews and key articles were also screened to identify additional studies. We included observational and longitudinal studies that examined malnutrition patterns or body mass index (BMI) trajectories in children, and excluded case reports, commentaries, and studies lacking quantitative anthropometric data. While several studies have reported the prevalence of the DBM among children at various ages, longitudinal data are lacking on the trajectory of BMI from birth through childhood and adolescence. Furthermore, existing research, including from the Malnutrition and Enteric Diseases (MAL-ED) cohort, indicates that early stunting often begins in the first 1000 days, which is a critical period for growth and development, while obesity mostly emerges later. However, without longitudinal data, the transition between states of malnutrition, as well as the influence of early-life factors and the long-term consequences of these transitions on health outcomes, remain unclear. Therefore, longitudinal studies are essential for understanding how BMI evolves over time in the context of malnutrition, particularly in rapidly urbanizing and economically transitioning settings, such as India. Added value of this study To the best of our knowledge, this is the first study to report the onset of DBM in children by mid-childhood in India using data from the MAL-ED cohort. The persisting association between maternal and childhood BMI concurs with the findings in the literature and highlights the importance of promoting maternal health in addition to mitigating DBM in early childhood. The observed associations provide insights into the emerging DBM but should be interpreted in the context of the cohort size and exploratory analyses. Implications of all the available evidence In this study DBM emerges in mid-childhood after 5 years of age, with a worsening trend between 7 and 9 years of age. Low maternal BMI had a lasting effect on the BMI trajectory of the offspring. Further studies are needed to understand complex societal changes, including diet patterns, physical activity schedules, and broader ecological systems. Introduction The double burden of malnutrition (DBM) is an emerging concern in low- and middle-income countries (LMICs), including India. DBM refers to the co-existence of undernutrition (stunting and wasting) and overnutrition (overweight and obesity) or diet-related non-communicable diseases (NCDs) within individuals, households, and populations. Although childhood undernutrition could lead to obesity/NCDs later in life, DBM can also manifest within the same country or region with subpopulations who have stunting, wasting, and obesity.1 The World Health Organization (WHO) has estimated that approximately 231 million children under 5 years are affected by malnutrition globally, including overweight and obesity.2 However, information is lacking on the emergence of DBM in longitudinal cohorts in different LMIC settings, particularly in India. Information on this topic can guide not only the timing of intervention to prevent DBM but also the adoption of culturally relevant practices for better implementation. Studies and surveys have provided cross-sectional data on stunting/wasting and obesity within the same populations.3,4 Stunting prevalence in India decreased from 48% to 38% between 2006 and 2016; however, child overweight/obesity remained stable at approximately 7%–8% during this period according to the 2006 and 2016 National Family Health Surveys (NFHS) (n = 311182, age = 0–5 years).3 The factors associated with DBM include low birth weight (LBW; i.e., birth weight <2500 g)5 and maternal body mass index (BMI).6 A systematic review and meta-analysis was conducted to understand the relationship between LBW and the risk of stunting in children aged <60 months across various LMICs, including India.7 The quantitative synthesis of 15 studies showed a significant association between LBW and an increased risk of stunting. Additionally, the risk was relatively higher in Asian countries than in African countries.7 Similar findings were observed in the 2015–2016 and 2019–2021 NFHS.5 A secondary data analysis of the NFHS-5 study revealed a significant relationship between LBW and childhood malnutrition. Children born with LBW had a higher risk of stunting, wasting, and being underweight than those born with normal birth weight (NBW). The analysis also reported that LBW contributed to 14.8%, 10.4%, and 9.6% cases of stunting, wasting, and underweight, respectively, among Indian children.5 Haque et al. (2022) analyzed Demographic and Health Survey data from eight countries to examine the associations between maternal BMI and children's anthropometric outcomes. The results showed that mothers who were underweight were more likely to have children with stunting, underweight, and wasting.8 According to the secondary analysis of data from the NFHS-4 study, mothers with low maternal BMI (underweight) showed a significantly higher risk (6%) of having children with stunting than did those with normal BMI.9 Despite growing recognition of malnutrition, longitudinal evidence describing how early-life factors influence later BMI patterns in children living in urban low-income settings remains limited. Data from the longitudinal cohort studies are important for understanding how nutritional status evolves throughout childhood and for identifying early determinants of later health outcomes. To address this gap, the present study utilized the existing Malnutrition and Enteric Diseases (MAL-ED) dataset for India, which provides a comprehensive longitudinal dataset. These data allowed us to evaluate the trajectory of children's anthropometric measures over time and to gain a deeper understanding of how the DBM emerges within an Indian setting. The MAL-ED cohort for India was established in an urban slum community in Vellore, a setting which is particularly relevant for studying early-life nutritional challenges due to its socioeconomic constraints and high burden of infectious diseases. Within this cohort, the prevalence of LBW was 17%, and approximately 45% of children were stunted by the age of two years, reflecting substantial early-life growth deficits. In the light of this, we hypothesized that the emergence of DBM in this population is associated with maternal BMI as well as birth weight patterns, and that both may play significant roles in predicting children's BMI trajectories. Methods The MAL-ED India cohort, was established in urban slum communities in Vellore, Tamil Nadu, India, was part of a multi-country study conducted in eight LMICs that examined the effect of early-life enteric infections on the growth and development of children.10,11 The study design and recruitment protocols are detailed elsewhere.11 Sex of the child (male or female) was recorded at birth from hospital records. The inclusion criteria were as follows: willingness of the caregiver to receive home visits twice weekly, no plans to move out of the study area for at least 6 months after the first enrollment, and healthy infants that were enrolled in the study within 17 days of birth. The exclusion criteria were as follows: maternal age <16 years, enrollment of another child from the same family in the MAL-ED study, non-singleton children, lack of signed consent from the child's guardian, birth weight <1500 g, and any indication of serious diseases in the child.11 Two hundred and fifty-one newborns (45% male, 55% female) were recruited between 2010 and 2012 from densely populated areas after screening 301 pregnant women and obtaining parental consent until 9 years of age. The sample size for the cohort was determined to enable the analysis of primary objectives of the MAL-ED study and was calculated as minimum 200 infants per site. Prior to enrollment, a population survey was conducted in the study area to enumerate women of reproductive age. Based on this survey, it was estimated that approximately 200 infants would be born within the MAL-ED study area during the planned 2-year enrollment period. Accordingly, enrollment was structured to recruit approximately 10 infants per month over two years. This recruitment strategy resulted in the enrollment of 251 infants into the Vellore MAL-ED birth cohort.10,11 These children were closely monitored through home visits until the age of 2 years, and data on sociodemographic factors, health status, and growth measurements were collected. Follow-up visits were conducted at ages 3, 4, 5, 7, and 9 years (Supplementary Figure S1). Anthropometric measurements were collected following standardized protocols and guidance from the study investigators. For children aged ≤2 years old, recumbent length was measured using a calibrated infantometer (to the nearest 0.1 cm). For children aged >2 years old, the standing height was measured using a stadiometer (to the nearest 0.1 cm). Body weight was measured to the nearest 10 g using a computerized scale after removal of shoes/sandals, socks, and heavy clothing. Further details on measurement techniques, standardization procedures, and follow-up have been reported earlier.10,12 All scales and stadiometers were calibrated before each measurement session. Each measurement was taken in duplicate, and if the two readings differed by more than a predefined threshold (0.1 kg for weight or 0.5 cm for length/height), a third measurement was taken. The mean of the two closest values was used for analysis. Furthermore, data collectors underwent rigorous centralized training before data collection and participated in periodic standardization exercises to minimize inter- and intraobserver variability. Field supervisors reviewed measurement data regularly for consistency and adherence to protocol. The body mass index (BMI) was calculated as weight divided by height squared (kg/m2). The Multicentre Growth Reference Study criteria were used to assess anthropometry in children measured at or before the age of 5 years.13,14 WHO AnthroPlus software was used to calculate the Z-scores for BMI-for-age from data collected at ages 7 and 9 years. The Z-score cut-offs for underweight, normal, overweight, and obesity were defined according to the WHO classification system.13,14 Maternal body weight and height were also measured using the same standardized procedures described above. Maternal BMI values were classified using Asia–Pacific BMI cut-offs [underweight (<18.5 kg/m2), normal (18.5–22.9 kg/m2), overweight (23.0–24.9 kg/m2), and obese (≥25.0 kg/m2)]15 Follow-up assessments were completed for 228 (46% boys, 54% girls), 212 (46% boys, 54% girls), 195 (45% boys, 55% girls), and 205 (47% boys, 53% girls) children at 2, 5, 7, and 9 years of age, respectively. The primary reason for the attrition or nonparticipation was family migration away from the study site (Supplementary Figure S1). All measurement and follow-up protocols adhered to the standardized MAL-ED procedures outlined in previously published papers from the same cohort.11,12,16 Paper-based data were collected by trained fieldworkers and subsequently entered into the double-entry database system maintained by the central data coordination center of the MAL-ED Network located at the Fogarty International Center, National Institutes of Health. During the initial part of the birth cohort, data entry and verification followed the standardized MAL-ED data management protocol to ensure intersite comparability and minimize transcription errors. Any discrepancies between the two entries were automatically flagged and resolved by cross-checking of original source documents. For the current analysis, the data collected at different time points were consolidated. Children were excluded from the analysis if birth weight data were missing and/or if maternal BMI information was unavailable. For the calculation of BMI trends across follow-up age groups, children who had at least one record available in each age category were included in the analyses. Data from children with at least one missing data point were excluded from further analyses. Statistical analysis Statistical analyses were performed using IBM SPSS Statistics for Windows, Version 22.0 (IBM Corp., Armonk, NY, USA) and Microsoft Excel (Microsoft Corp., Redmond, WA, USA). Visual (histogram and probability plots) and analytical methods (Kolmogorov–Smirnov/Shapiro–Wilk tests) were used to evaluate the normality of data distribution. In normally distributed variables, group comparisons were conducted using one-way analysis of variance (ANOVA). When overall differences were identified, post-hoc pairwise comparisons with Bonferroni correction were performed to account for multiple comparisons. Fisher's exact test was used to examine the relationships between categorical variables. Pearson's correlation analysis was used to assess the association between maternal BMI and birth weight. Prior to analysis, assumptions of the test, including approximate normality and linearity, were assessed using graphical methods and scatterplots. Multinomial logistic regression models were used to test the predictors of BMI in later life. The models were adjusted for birth weight and child sex. Results are presented as odds ratios (ORs) with 95% confidence intervals (CIs) to describe the magnitude and precision of associations. Statistical significance was set at p < 0.05. Ethics statement The study was approved by the Institutional Review Board (IRB) of Christian Medical College Vellore (IRB no. 6769; dated February 25th, 2009 for the original cohort and follow-up, and IRB no. 11821; dated January 30th, 2019 for the 9-year follow-up) and adhered to the Declaration of Helsinki. Written informed consent was obtained from parents at each recruitment stage, with additional child assent collected for the follow-up at 9 years.10 Role of funding source The sponsors/funders had no role in the study design, data collection and analysis, preparation of the manuscript, and decision to publish the manuscript. Results A total of 251 children were enrolled at birth, 205 (82%) of whom completed follow-up at 9 years of age. The MAL-ED India cohort was population representative with subsequent follow-ups not differing from the original birth-cohort characteristics: these details are already published.10,12 For both males and females, BMI at 4–8 months showed an increasing trend compared with BMI at birth, followed by a declining trend up to 7 years of age (Fig. 1). This trend reversed at 9 years of age with an increase in BMI (15.48 ± 3.09 kg/m2 for males and 15.35 ± 3.13 kg/m2 for females; Fig. 1).Fig. 1Body mass index values across follow-up age groups. Body mass index values across follow-up age groups. The age-dependent change in BMI distribution was analyzed to assess the prevalence of severe thinness (less than −3 standard deviations [SD]), underweight (−2 to −3 SD), normal weight, overweight (+2 to +3 SD), and obesity (more than +3 SD) in children. The results showed that a high percentage of children maintained a normal BMI until the age of 5 years (86.6%) (Supplementary Table S1 and Fig. 2). At 7 years of age, 26.3% had thinness, and 5.2% had overweight/obesity. At 9 years, 21.6% had thinness, and 14.6% had overweight/obesity. Overnutrition in terms of overweight and obesity was evident by the age of 7 years and worsened by the age of 9 years, a trend that is indicative of DBM in this cohort.Fig. 2Distribution of body mass index categories by age (top: all children; center: male children; bottom: female children). Distribution of body mass index categories by age (top: all children; center: male children; bottom: female children). The prominent changes in BMI within these age windows could be a result of either growth spurts or stunting, potentially confounding the interpretations. To address this, the height of children at age 9 years was plotted against their corresponding BMI (Supplementary Figure S2). The results indicated that a higher BMI was not associated with stunting, thus suggesting the presence of “true” obesity. Age-dependent changes in BMI were tabulated across age groups (Table 1 and Supplementary Table S2) which showed that 16.3% and 11.2% of children with thinness at birth achieved a normal BMI by ages 4 and 9 years, respectively. Among the children with a normal BMI at birth, 12.2% had overweight/obesity by the age of 9 years (Table 2).Table 1Body mass index comparison between two different age groups.Severe thinnessUnderweightNormalOverweightObeseBMI group at birthBMI group at 4 years of age Severe thinness0.0 (0)1.3 (2)5.0 (8)0.0 (0)0.0 (0) Underweight0.6 (1)1.9 (3)11.3 (18)0.0 (1)0.0 (0) Normal0.0 (0)3.8 (6)71.1 (113)2.5 (4)0.6 (1) Overweight0.0 (0)0.0 (0)1.3 (2)0.6 (1)0.0 (0) Obese0.0 (0)0.0 (0)0.0 (0)0.0 (0)0.0 (0)BMI group at birthBMI group at 9 years of age Severe thinness1.0 (2)2.4 (5)2.9 (6)0.0 (0)0.0 (0) Underweight3.4 (7)2.9 (6)8.3 (17)1.0 (2)0.5 (1) Normal2.0 (4)10.7 (22)51.2 (105)6.8 (14)5.4 (11) Overweight0.0 (0)0.0 (0)1.0 (2)0.0 (0)0.5 (1) Obese0.0 (0)0.0 (0)0.0 (0)0.0 (0)0.0 (0)BMI group at 4 years of ageBMI group at 9 years of age Severe thinness0.7 (1)0.0 (0)0.0 (0)0.0 (0)0.0 (0) Underweight2.6 (4)2.6 (4)2.0 (3)0.0 (0)0.7 (1) Normal2.0 (3)13.2 (20)60.9 (92)7.9 (12)4.0 (6) Overweight0.0 (0)0.0 (0)0.7 (1)0.7 (1)2.0 (3) Obese0.0 (0)0.0 (0)0.0 (0)0.0 (0)0.7 (1)BMI: body mass index (kg/m2).Data are presented as % (n).Table 2Birth weight of children and body mass index in later life.UnderweightObesityLow birth weightNormal birth weightLow birth weightNormal birth weight1 year16.7 (6/36)9.1 (17/186)2.8 (1/36)2.2 (4/186)2 years8.1 (3/37)4.9 (9/185)5.4 (2/37)3.8 (7/185)3 years13.0 (3/23)9.2 (8/87)4.3 (1/23)2.3 (2/87)4 years12.0 (3/25)6.1 (8/131)0.0 (0/25)4.6 (6/131)5 years17.1 (6/35)8.3 (14/168)0.0 (0/35)3.0 (5/168)7 years41.4 (12/29)21.1 (32/152)0.0 (0/29)6.6 (10/152)9 years40.6 (13/32)17.3 (29/168)3.1 (1/32)16.7 (28/168)Data are presented as % (n). Table 1Body mass index comparison between two different age groups.Severe thinnessUnderweightNormalOverweightObeseBMI group at birthBMI group at 4 years of age Severe thinness0.0 (0)1.3 (2)5.0 (8)0.0 (0)0.0 (0) Underweight0.6 (1)1.9 (3)11.3 (18)0.0 (1)0.0 (0) Normal0.0 (0)3.8 (6)71.1 (113)2.5 (4)0.6 (1) Overweight0.0 (0)0.0 (0)1.3 (2)0.6 (1)0.0 (0) Obese0.0 (0)0.0 (0)0.0 (0)0.0 (0)0.0 (0)BMI group at birthBMI group at 9 years of age Severe thinness1.0 (2)2.4 (5)2.9 (6)0.0 (0)0.0 (0) Underweight3.4 (7)2.9 (6)8.3 (17)1.0 (2)0.5 (1) Normal2.0 (4)10.7 (22)51.2 (105)6.8 (14)5.4 (11) Overweight0.0 (0)0.0 (0)1.0 (2)0.0 (0)0.5 (1) Obese0.0 (0)0.0 (0)0.0 (0)0.0 (0)0.0 (0)BMI group at 4 years of ageBMI group at 9 years of age Severe thinness0.7 (1)0.0 (0)0.0 (0)0.0 (0)0.0 (0) Underweight2.6 (4)2.6 (4)2.0 (3)0.0 (0)0.7 (1) Normal2.0 (3)13.2 (20)60.9 (92)7.9 (12)4.0 (6) Overweight0.0 (0)0.0 (0)0.7 (1)0.7 (1)2.0 (3) Obese0.0 (0)0.0 (0)0.0 (0)0.0 (0)0.7 (1)BMI: body mass index (kg/m2).Data are presented as % (n). Body mass index comparison between two different age groups. BMI: body mass index (kg/m2). Data are presented as % (n). Table 2Birth weight of children and body mass index in later life.UnderweightObesityLow birth weightNormal birth weightLow birth weightNormal birth weight1 year16.7 (6/36)9.1 (17/186)2.8 (1/36)2.2 (4/186)2 years8.1 (3/37)4.9 (9/185)5.4 (2/37)3.8 (7/185)3 years13.0 (3/23)9.2 (8/87)4.3 (1/23)2.3 (2/87)4 years12.0 (3/25)6.1 (8/131)0.0 (0/25)4.6 (6/131)5 years17.1 (6/35)8.3 (14/168)0.0 (0/35)3.0 (5/168)7 years41.4 (12/29)21.1 (32/152)0.0 (0/29)6.6 (10/152)9 years40.6 (13/32)17.3 (29/168)3.1 (1/32)16.7 (28/168)Data are presented as % (n). Birth weight of children and body mass index in later life. Data are presented as % (n). Furthermore, we compared the progressive shift in BMI up to 9 years of age as a correlate of birth weight and calculated the prevalence of thinness (severe and thinness) and obesity (overweight and obese) among children with LBW or NBW. We observed that children with LBW had a higher prevalence of thinness at most ages, with the proportion increasing notably by 7 years (41.4%) and remaining high at 9 years (40.6%). While it was not surprising to find that the prevalence of thinness was consistently lower among children with NBW, the proportion of children with NBW who became thin increased dramatically at 7 and 9 years of age (21.1% and 17.3%, respectively). Children with NBW had the highest incidence of obesity in later years, probably because of dietary and lifestyle habits. Because previous reports have highlighted the associations between maternal BMI and birth outcomes, we assessed the relationship between maternal BMI and birth weight. The majority of children were born with NBW across all maternal BMI groups. Furthermore, mothers with underweight and those with obesity showed a significant difference in terms of the birth weight of their children. The birth weight of babies born to mothers in the underweight group was significantly lower than that of babies born to mothers in the obese group (mean birth weight [SD]: 2.77 kg [0.43] and 3.02 kg [0.47], respectively; p = 0.030) (Supplementary Figure S3). Maternal BMI and birth weight showed a positive correlation (r = 0.184, p = 0.005). Multinominal logistic regression analysis was used to investigate the BMI trajectory/shifts at ages 2, 5, and 9 years, and the unadjusted model revealed that children with LBW had a 2.607-times higher odds of having thinness at age 2 years (95% CI 1.189–5.717) and a 2.273-times higher odds at age 5 years (95% CI 1.039–4.973) than children with NBW (Supplementary Table S3). The effect of maternal BMI on children's BMI in later life was evaluated, and the unadjusted model revealed that each unit increase in the mother's BMI was associated with a decrease in the log-odds of the child having thinness at ages 5 (odds ratio [OR] = 0.863, p = 0.001) and 9 years (OR = 0.789, p = 0.000) (Supplementary Table S4). When adjusted for birth weight and sex (Table 3), higher maternal BMI was associated with lower log-odds of the child having thinness at ages 5 and 9 years. Each unit increase in maternal BMI corresponded to an estimated odds ratio of 0.878 at 5 years (95% CI 0.805–0.957) and 0.789 at 9 years (95% CI 0.709–0.878), suggesting an inverse relationship between maternal BMI and the likelihood of thinness in later childhood. These results suggest that lower maternal BMI increases the likelihood of the child being thin in later life and likely contributes to the emergence of DBM.Table 3Association between maternal body mass index and child body mass index trajectory in later life (adjusted for sex and birth weight).Age groupDependent variableIndependent variableBStd. errorWaldSigExp (B)95% confidence interval for Exp (B)Lower boundUpper bound2 yearsUnderweight in later lifeaIntercept1.1370.9621.3970.237–––Maternal BMI (kg/m2)−0.0350.0430.6830.4090.9650.8871.0505 yearsUnderweight in later lifeaIntercept4.9131.6858.5060.004–––Maternal BMI (kg/m2)−0.1300.0448.6700.0030.8780.8050.9579 yearsUnderweight in later lifeaIntercept7.8311.97215.7670.000–––Maternal BMI (kg/m2)−0.2370.05518.8400.0000.7890.7090.878aReference: normal body mass index group. BMI: body mass index. Table 3Association between maternal body mass index and child body mass index trajectory in later life (adjusted for sex and birth weight).Age groupDependent variableIndependent variableBStd. errorWaldSigExp (B)95% confidence interval for Exp (B)Lower boundUpper bound2 yearsUnderweight in later lifeaIntercept1.1370.9621.3970.237–––Maternal BMI (kg/m2)−0.0350.0430.6830.4090.9650.8871.0505 yearsUnderweight in later lifeaIntercept4.9131.6858.5060.004–––Maternal BMI (kg/m2)−0.1300.0448.6700.0030.8780.8050.9579 yearsUnderweight in later lifeaIntercept7.8311.97215.7670.000–––Maternal BMI (kg/m2)−0.2370.05518.8400.0000.7890.7090.878aReference: normal body mass index group. BMI: body mass index. Association between maternal body mass index and child body mass index trajectory in later life (adjusted for sex and birth weight). Reference: normal body mass index group. BMI: body mass index. Discussion DBM has emerged as an alarming global problem, leading to adverse effects at both individual and population levels. Given the paucity of information on the emergence of DBM in India with regard to onset age and maternal/neonatal predictors, we evaluated data from the MAL-ED cohort in India to obtain new insights into the longitudinal trajectory of BMI from birth to 9 years of age. Our results highlighted that DBM emerged in mid-childhood after 5 years of age, became evident at approximately 7 years of age, and further worsened at approximately 9 years of age. Our findings also confirmed the hypothesis that DBM is associated with both maternal BMI and birth weight patterns. Specifically, low maternal BMI and LBW were significantly linked to a higher risk of thinness in mid-childhood, highlighting the intergenerational and early-life origins of DBM. The United Nations Children's Fund (UNICEF) 2025 Child Nutrition Report highlights the importance of DBM in children and adolescents. The report estimated that 20% of children aged 5–19 years have either overweight or obesity, thus highlighting the burgeoning overnutrition concern.17 The report also estimated that 147 million children aged 5–9 years old are overweight, the highest proportion in the overall age group. Notably, the year 2025 has marked a historic turning point, because the global prevalence of obesity (9.4%) surpassed the prevalence of underweight (9.2%) among children and adolescents aged 5–19 years for the first time. LMICs were reported to experience the sharpest rise in overweight and obesity among school-age children and adolescents, reflecting a rapid nutritional transition in these populations. The birth cohort analysis in the current study adds further evidence on the emergence of DBM in mid-childhood in a longitudinal LMIC cohort, in which approximately 45% children showed stunting at 2 years of age. To our knowledge, this is the first evaluation of its kind in an Indian setting. Although a global consensus on the definitions of DBM is emerging, systematic reviews have highlighted that most studies on DBM used cross-sectional datasets.18,19 Recent estimates of DBM in India report a prevalence of 7.7% at the household level.20 In the under-5 age group, the prevalence of coexisting wasting and underweight was 8.1%, whereas the prevalence of concurrent wasting, stunting, and underweight was 6.5%.5 Analytic studies utilizing cross-sectional data on stunting/wasting and obesity within the same population on the Indian subcontinent have also shown similar DBM trends.3,4 An evaluation of sub-Saharan African countries revealed that DBM is present in both children under 5 years of age and in women of reproductive age, with an increasing trend of obesity/overweight in both groups.21 Longitudinal follow-up studies are sparse. A 31-year follow-up cohort study in Finland used Bayesian hierarchical modeling to map the trajectory of BMI and suggested that high BMI should be addressed before the age of 6 years.22 A review evaluating DBM emergence in children emphasized the associations of antenatal factors such as maternal height, BMI, and education, as well as sex and family income.23 Mysore cohort studies in India have highlighted the importance of DBM as a precursor for lifestyle diseases from the perspective of the Developmental Origins of the Health and Disease Hypothesis (DOHaD). However, the exact timing of DBM emergence has not been demonstrated.24 Our study showed that a significant percentage of children (14.6%) in the MAL-ED LMIC birth cohort had overweight or obesity at age 9 years. The Comprehensive National Nutrition Survey estimated that the prevalence of overweight and obesity ranges from 1.6% in young children (0–4 years) to 4.8% in adolescents (10–19 years).25 However, a recent meta-analysis of 21 studies reported a higher prevalence of overweight/obesity at 20.8%.26 This meta-analysis involved 18,6901 children aged 3–18 years from different parts of India, including rural and urban residences.26 We posit that the estimation of overweight or obesity should account for changes in height to ensure that higher BMI values reflect “true obesity,” similar to that in our analysis. Furthermore, given that weight gain during early developmental periods could be a significant predictor of increased adiposity in adulthood, we emphasize the importance of longitudinal follow-up measurements to identify the specific periods when DBM emerges. Etiological pathway evaluation has highlighted the interconnected environmental, physiological, and biological pathways, including nutritional intake patterns, food availability, inflammation, gut microbiome, and metabolic dysregulation.17,27 The DOHaD suggests that multi-generational maternal undernutrition and unfavorable intrauterine environments may program growth and metabolic profiles that adversely affect later childhood nutritional experiences.24 The recent UNICEF report has raised concerns about the easy access to cheap, unhealthy food in school environments and community settings and the misleading advertisements employed by the producers of such products.17 Even though reports have clearly established an association between maternal BMI and birthweight/neonatal growth,28,29 information is lacking on whether maternal BMI predicts DBM in mid-to-late childhood. Our analyses indicate that maternal BMI could be an important factor that drives the emergence of DBM. Children born to mothers with a low BMI had a significantly higher risk of having thinness, consistent with findings from another birth cohort in Vellore.30 However, unlike some of the findings from other populations, high maternal BMI did not seem to be associated with obesity between 5 and 9 years of age. In addition to indicating the differences between study groups/populations, the current study also suggests the potential role of other factors, such as the food environment, in obesity during mid-childhood. This study has a few limitations. First, the absence of dietary intake data is a significant limitation because such information could have provided critical insights into the dietary determinants influencing the observed BMI trajectories. Second, our study population was drawn from a single area in India. Given the heterogeneity in socioeconomic status, dietary patterns, cultural practices, and urbanization levels, our findings may not be generalizable to the broad and diverse populations of India. Future studies should include a more representative sample that encompasses urban, rural, and semi-urban settings across different regions of India. Furthermore, our analyses represent a secondary analysis of an existing longitudinal cohort, and the sample size was determined by the design of the parent MAL-ED study rather than being calculated specifically for the present research question. Multiple statistical tests were performed, which may increase the possibility of type I error. A further limitation of this study is that approximately one-fifth of the children enrolled at baseline were lost to follow-up over time. The primary reason for attrition was family migration away from the study area. Although this is common in long-term cohort studies, such loss to follow-up may introduce potential selection bias if children who migrated differed systematically from those who remained in the study. Although our study captures important anthropometric data, a more comprehensive evaluation that includes biochemical markers of nutrition and metabolic health, as well as body composition analysis, would strengthen the understanding of the observed trends and their potential implications for public health policies focused on adolescent health. The strength of this study lies in the availability of longitudinal data and the high quality of anthropometric measurements. In conclusion, our study demonstrated that DBM becomes more pronounced in mid-childhood after 5 years of age, with a worsening trend between 7 and 9 years of age. Low maternal BMI had a lasting effect on the BMI trajectory of the offspring, highlighting the intergenerational dimension of malnutrition. The findings of this study could be helpful to support the importance of strengthening maternal nutrition before and during pregnancy, as well as sustained growth monitoring beyond early childhood to detect emerging risks of both undernutrition and overweight. Addressing DBM will require integrated and context-specific strategies that consider changing dietary patterns, physical activity behaviors, and the broader food environment. Moreover, a multi-pronged strategy is needed to reduce the worsening DBM trend in children. These efforts should be implemented in early to mid-childhood, with a concomitant focus placed on maternal health and well-being in addition to the unhealthy food environment that is one of the main factors associated with DBM in children and adolescents. Contributors BK, VRM and GK contributed to the study conception and design. Material preparation, data collection, and analysis were conducted by BK, VRM and SJ. BY, BK and VRM directly accessed and verified the underlying data. All authors assisted with the overall proposal of the current paper, manuscript preparation, and corrections. All authors confirm that they had full access to all the data in the study and accept responsibility for submitting the manuscript for publication. Data sharing statement All MAL-ED data till 5 years are available at https://clinepidb.org/ce/app. The anonymized version of the MAL-ED data including MRI brain parcellation data is available in the Harvard Dataverse database. Our data can be accessed at the following link: https://doi.org/10.7910/DVN/KJGMAY. An anonymized dataset for this analysis is also enclosed. Declaration of interests The authors declare no relevant financial or non-financial interests.