The double burden of malnutrition and its impact on academic achievement and cognitive function in students of rural India

ABSTRACT

Background:

The study explores the interrelationship between nutritional status, cognitive function (IQ), and academic performance in rural primary school children, recognizing these as critical and interconnected health parameters.

Objectives:

To investigate the relationship among nutritional condition, cognitive ability, and academic outcomes in children aged 6–8 years in rural Jharkhand, India.

Methods:

A total of 560 children (280 boys and 280 girls), aged between 6 and 8 years, from four rural primary schools in Jharkhand, were included in the study. Nutritional condition was assessed using Body Mass Index (BMI), and cognitive function was evaluated using the Raven Progressive Matrices test.

Results:

A high rate of malnutrition was observed: 28.03% of children were undernourished and 28.75% were severely malnourished, with a higher prevalence among boys. Additionally, 4.47% of the children were found to be overweight or obese, indicating a growing double burden of malnutrition in rural areas. A positive association was found between BMI and IQ score (r = 0.41, P ≤ 0.01), indicating that better nutritional status is linked with improved cognitive function. Alarmingly, 50.71% of the children were categorized as “intelligently impaired” based on IQ scores, with only 3.22% scoring in the above-average range. Logistic regression revealed strong associations between academic performance and IQ categories, emphasizing the key role of cognitive function in educational outcomes.

Conclusion:

The study highlights the intricate relationship between malnutrition, cognitive development, and academic performance. It underscores the urgent need for integrated interventions targeting both under- and over-nutrition, while prioritizing cognitive enhancement and educational support in resource-constrained rural settings.

Introduction

Malnutrition, encompassing both undernutrition and overnutrition, is a major public health issue affecting millions of children worldwide. Undernutrition, marked by stunting, wasting, and micronutrient deficiencies, hampers physical and cognitive development, while overnutrition leads to obesity-related health risks (Black et al., 2013[1]; UNICEF, 2019).[2]

Cognitive function, often measured through Intelligence Quotient (IQ), is significantly influenced by nutritional status. Studies show that malnourished children tend to have lower IQ scores due to impaired brain development, affecting memory, concentration, and problem-solving skills (Prado and Dewey, 2014[3]; Walker et al., 2007[4]). The early years are particularly crucial, as inadequate nutrition during this period can lead to long-term cognitive deficits (Levitsky and Strupp, 1995).[5]

Academic performance is closely linked to both IQ and nutritional status. Well-nourished children tend to perform better in school, exhibit higher attention spans, and have improved learning outcomes compared with their malnourished peers (Grantham-McGregor et al., 2007).[6] In contrast, malnutrition is associated with absenteeism, poor classroom engagement, and lower test scores, creating barriers to educational success and future opportunities (Nyaradi et al., 2013).[7]

Addressing malnutrition through targeted interventions in schools, including nutritional programs and health education, is essential to fostering both cognitive development and academic achievement among students, particularly in resource-limited settings. Research across diverse geographical contexts has consistently demonstrated the intricate relationship between nutritional status and cognitive functioning. A comprehensive study by Ghosh et al. (2015)[8] revealed significant correlations between undernutrition and cognitive development among school children in Kolkata. Their research demonstrated that socioeconomic status and nutritional deficiencies interact to substantially influence cognitive capabilities, underscoring the multifaceted nature of nutritional challenges. Complementary research by Adedeji et al.[9] (2017) in Nigeria provided further insights, examining the direct relationship between malnutrition and intelligence quotient (IQ) among primary school pupils. Their findings highlighted the potential long-term neurological consequences of nutritional deficiencies during critical developmental stages. Building on these perspectives, Aurora et al.[10] (2021) explored the specific effects of stunting on children’s cognitive potential. Their research emphasized how early nutritional interventions could potentially mitigate the negative impacts on intellectual development. Although educational research often focuses on systemic interventions, Mariano et al. (2024)[11] remind us of the broader contextual factors that influence educational outcomes, suggesting the need for comprehensive approaches to understanding student performance.

Understanding the complex interplay between nutrition, socioeconomic factors, and cognitive development remains crucial for developing targeted interventions that can support children’s academic and intellectual potential which forms the basis for this research work.

As it has already been established by previous researches that malnutrition poses a significant public health challenge in India, especially in rural areas, where children increasingly experience a two-way burden—both undernutrition and over nutrition. Undernutrition, reflected in high rates of wasting and stunting, remains widespread, but there is also a growing concern about overweight and obesity, even in low-resource

environments (Ghosh et al., 2016[12]; Ghosh et al., 2015[8]). The situation is particularly severe, with India accounting for more than half of global child deaths linked to undernutrition, amounting to an estimated 5.6 million deaths annually (UNICEF, 2019).[2] The complex nature of malnutrition, involving inadequate dietary intake and frequent infections, as explained by Benson et al. (2008)[13], manifests in various forms, including growth stunting, developmental delays, anemia, and in severe cases, blindness. Anthropometry has emerged as a valuable, affordable tool in resource-limited settings for assessing nutritional status, with Body Mass Index (BMI) being one of the most widely used measures (World Health Organization, 1995[14]; de Onis et al.,[15] 2003; Cole et al.,[16] 2000; Das et al.,2024).[17] Beyond physical health, malnutrition significantly impacts cognitive development, a crucial factor in shaping a child’s overall health and life course (Prado et al., 2014[3]; Ghosh et al.,2024).[18] Recent research has increasingly focused on understanding the intricate relationship between childhood development and cognitive abilities, with early childhood experiences, particularly those related to nutrition, providing a critical foundation for future success in school, long-term health outcomes, and overall well-being (Grantham-McGregor et al., 2007[6]; Ghosh, 2024).[19] Although malnutrition alone may not cause permanent brain damage, it is believed to interact with various environmental factors to negatively influence a child’s cognitive behaviour (Walker et al., 2007,[4] Levitsky and Strupp, 1995[6]). Numerous studies have highlighted the importance of nutrition in this area, with findings suggesting a link between childhood IQ scores and both obesity and BMI values (Yu et al., 2010;[20] Kanazawa, 2015;[21] Liang et al., 2015).[22] However, India faces a dual challenge: while undernutrition remains a significant concern, overweight and obesity are emerging as public health issues globally (NCD; Risk Factor Collaboration, 2016)[23] and in India (Ranjani et al., 2016).[24] Previous studies have reported on the prevalence of both undernutrition and overweight/obesity in children of West Bengal (Dey et al., 2017[25]; Das et al., 2011[26]; Ghosh et al., 2016[12]; Dutta et al., 2012[27]; Shakil et al.,2024),[28] but a crucial gap remains in understanding the potential association between BMI and IQ in this specific sample taken.

This research attempts to bridge a knowledge gap by investigating rural primary school children’s nutrition in Jharkhand, India, using the prism of BMI measurements and examining its potential correlation with IQ. The research focus is apparently healthy children between 6 and 8 years old in four rural primary schools in Shimulpur, Salka Kumarhut, and Ramnagar. Through investigating the relationship between nutrition, intelligence, and school performance, we hope to create useful information for the formulation of targeted programs and policies.

This study aims to (i) assess the nutritional status of rural primary school children in Jharkhand using covariates like IQ, age, BMI, gender, Z-score, academic performance (ii) Evaluate cognitive function using the Raven Progressive Matrices test; (iii) Examine the relationship between nutritional status, cognitive performance (IQ), and academic achievement using logistic regression (iv) provide evidence to support targeted nutritional and educational interventions in resource-poor settings. The results could help us better understand how malnutrition affects children’s development in underdeveloped areas and support initiatives to improve the health and prospects of Indian children living in rural areas. Through a holistic approach that addresses both undernutrition and over-nutrition in cognitive development, this research hopes to identify the key problems of rural school children in India and offer more effective and inclusive solutions.

Materials and Methodology

Study design

We have done a random sampling method in recruiting participants, with some inclusion criteria in mind. A stratified random sampling method was used to ensure a representative distribution of students across different age groups, gender and socioeconomic backgrounds. Initially, primary schools in rural Jharkhand were categorized into different strata based on geographic location and school strength. In every school again, age and gender are the strata. Within each stratum, students were randomly selected, ensuring proportionate representation across gender and age groups. This approach minimizes selection bias and enhances the generalizability of the findings. The inclusion criteria were designed to make sure that the participants were (i) apparently healthy children with no chronic diseases or physical disabilities, and (ii) those who volunteered to participate in the study. By adhering to these pre-specified criteria, the researchers aimed to obtain a sample that was representative of the broader population of school children, thus eliminating confounding variables that could influence the result of the study.

Study subject

This study was conducted in four rural primary schools of Shimulpur, Salka Kumarhut, and Ramnagar regions of Jharkhand, India. The study was conducted on 560 children with 280 boys and 280 girls with ages between 6 to 8 years. The study was conducted from August 2022 to March 2023.

The sample size of 560 children was determined based on several considerations:

• i) Feasibility and Representativeness: The four selected rural primary schools provided a readily accessible population that could ensure equal gender representation (280 boys and 280 girls) within the target age range (6–8 years).

• ii) Statistical Power: A sample size of 560 is consistent with similar studies in the field (e.g. Ghosh et al., 2020)[29] and is sufficient to detect meaningful associations between nutritional status, cognitive function, and academic performance.

• iii) Resource Constraints: Given the logistical and practical constraints in rural settings, this number provided a balance between scientific rigor and field feasibility.

Ethical consideration

This study was approved by the All-India Institute of Hygiene and Public Health’s Ethical Committee in Kolkata. It is one of the partial projects published earlier (Ghosh et al., 2020).[29] During the school surveys, guardians’ meetings were conducted with the headmaster and the parents, who had their children accompany them. These sessions were intended to give a clear description of the study’s purpose. Furthermore, written informed consent was also taken from the students’ mothers.

Statistical methods

Descriptive statistics for all the continuous variables were calculated. The study also employed the Chi-square test of association in order to assess associations between multiple factors (i.e. age, gender, Z-score, BMI, and academic achievement) and IQ ranges. All these statistical approaches enabled the researchers to explore any possible associations among nutritional status measures and cognitive performance in the target population.

Logistic regression

Logistic Regression is a widely used statistical and machine learning algorithm for binary and multi-class classification tasks.

In logistic set up, The dependent variable is IQ independent variables included age, BMI, gender, Z-score, and academic performance. These predictors were chosen based on both theoretical considerations and previous empirical evidence suggesting their relevance to cognitive function and academic outcomes (Grantham-McGregor et al., 2007[6]; Black et al., 2013).[1] The final models were selected by evaluating statistical significance (P values) and model performance metrics.

Logistic Regression is widely used in fields like medical diagnosis, credit risk assessment, and fraud detection due to its interpretability, efficiency, and ability to handle linearly separable data. However, it may struggle with complex relationships requiring non-linear decision boundaries, where models like Decision Trees or Neural Networks perform better.

Results and Findings

The results provide insights into various associations between demographic factors, academic performance, and IQ scores among rural school children in Jharkhand, India. In the present study, 50% of the participants scored below 20 on the IQ scale, 40% scored between 20 and 80, and 10% scored above 80. Regarding academic performance, 10% of participants were rated as excellent, 17% as very good, 16% as good, 12% as average, and 44% fell into the below-average category. A strong relationship is found between age and IQ category (P < 0.001), which indicates that age plays a role in cognitive function. However, no association was observed between gender and IQ category (P = 0.45), indicating that cognitive function may not differ significantly between males and females in this population [Table 1]. Interestingly, neither BMI nor Z-score showed a significant association with IQ category (P = 0.32 for both), implying that body mass may not directly influence cognitive function in this group. The most striking findings were the strong associations between academic performance (both good and bad) and IQ category (P < 0.001 for both), indicating moderate to strong associations. This suggests a strong association between intellectual capacity and academic performance, such that children with greater IQs are likely to have greater academic performance. These results highlight the complex interplay between various determinants of cognitive capacity and academic performance in rural Indian children, necessitating an integrated strategy in health and education interventions.

Table 1.

Association among different variables (α=0.01)

Variable pair	Chi square Test Statistic	P	Decision
Age and IQ	65.817	7.5852*10–13	Association present
Gender and IQ	0.563	0.4527	No association
Z Score and IQ	317.87	0.322	No association
BMI and IQ	341.67	0.423	No association
Academic performance and IQ	333.09	0.000	Strong Association

Predictive analysis

This graph displays the outcome of a number of logistic regression models that examine predictors of cognitive capacity (IQ scores) among rural Indian school children in Jharkhand, India [Figure 1]. The models employ varying sets of predictors such as age, BMI, gender, Z-score, and measures of academic performance. The results are summarized in the following [Tables 2-7]. The different logistic models results are given in [Tables 2-7].

Figure 1.

ROC curves of the Logistic Regression models

Table 2.

Logistic Regression of IQ on Age, BMI, Gender, Z-score, Academic performance (BAP and GAP)

Covariate	Coefficient Estimate (LR)	Standard Error (s.e)	Test Statistic (Z)	P
Intercept	-0.02585	0.3025	-0.854	0.393
Age	0.003931	0.2218	0.177	0.859
BMI	-2.947*10–8	2.590*10–8	-1.138	0.255
Gender Category	0.3713	0.4314	0.861	0.389
Z-score	2.947*10–8	2.590*10–8	1.138	0.255
Bad academic performance (BAP)	-4.945	0.7703	-6.420	1.36*10–8***
Good Academic Performance (GAP)	5.079	1.033	4.916	8.81*10–8***

Table 7.

The accuracy measures of all the models

	Accuracy	Precision	Recall	F1 Score	Auc
Logistic_0	0.9	0.8985507	0.9538462	0.9253731	0.9859341
Logistic_1	0.93	0.9027778	1	0.9489051	0.9828571
Logistic_2	0.64	0.6526316	0.9538462	0.775	0.6397802
Logistic_3	0.91	0.9117647	0.9538462	0.9323308	0.9841758
Logistic_4	0.65	0.65625	0.9692308	0.7826087	0.596044

Table 3.

Logistic Regression of IQ on Age, BMI and Academic Performance (BAP, GAP)

Covariate	Coefficient Estimate (LR)	Standard Error (s.e)	Test Statistic (Z)	P
Intercept	-0.16934	0.24082	-0.703	0.482
Age	-0.01460	0.21886	-0.067	0.947
BMI	0.09475	0.22315	0.425	0.671
Bad academic performance (BAP)	-4.77337	0.75332	-6.336	2.35*10–8***
Good Academic Performance (GAP)	5.05099	1.03079	4.900	9.58*10–8***

* and *** implies significance of a covariate in the regression model by its p-value. *implies small p-value and the corresponding covariate is significant. ***implies very small p-value and the corresponding covariate is highly significant

Table 4.

Logistic Regression of IQ on Age, BMI and Z-Score

Covariate	Coefficient Estimate (LR)	Standard Error (s.e)	Test Statistic (Z)	P
Intercept	-0.6.605	9.640e-02	-6.853	7.25*10–12***
Age	-0.2436	9.802e-02	-2.486	0.0129*
BMI	1.761*10–8	1.053*10–8	1.672	0.0946
Z-Score	-1.761*10–8	1.053*10–8	-1.672	0.0946

* and *** implies significance of a covariate in the regression model by its p-value. *implies small p-value and the corresponding covariate is significant. ***implies very small p-value and the corresponding covariate is highly significant

Table 5.

Logistic Regression of IQ on Age, BMI, Z-score and Academic Performance (BAP, GAP)

Covariate	Coefficient Estimate (LR)	Standard Error (s.e)	Test Statistic (Z)	P
Intercept	-0.1098	0.247	-0.445	0.657
Age	0.02500	0.2214	0.113	0.910
BMI	-3.027*10–8	2.583*10–8	-1.172	0.241
Z-score	3.027*10–8	2.583*10–8	1.172	0.241
Bad academic performance	-4.904	0.7673	-6.391	1.65*10–8***
Good Academic Performance	5.065	1.033	4.905	9.35*10–8***

Table 6.

Logistic Regression of IQ on Age, BMI and Gender

Covariate	Coefficient Estimate (LR)	Standard Error (s.e)	Test Statistic (Z)	P
Intercept	-0.57131	0.12670	-4.509	6.5*10–6***
Age	-0.23793	0.09781	-2.433	0.014992*
BMI	0.32886	0.09782	3.362	0.000774 ***
Gender	-0.19716	0.19350	-1.019	0.308254

The most striking finding across all models is the strong and highly significant association between academic performance and IQ category. Both Bad Academic Performance (BAP) and Good Academic Performance (GAP) show very low P values (P < 10^–6) in models where they are included, indicating a robust relationship with cognitive function and portraying them as significant variables affecting IQ. Age shows a significant negative association with IQ category in some models (e.g. Logistic_2 and Logistic_4), suggesting that older children in this sample tend to have lower IQ scores. This could potentially reflect improvements in education or nutrition for younger cohorts.

BMI was used as an indicator of nutritional status. Children with low BMI values were categorized as “wasted” or “severely wasted” based on established WHO criteria (World Health Organization, 1995). This is viewed as the undernutrition category. Conversely, children with BMI values above the recommended range were classified as overweight/obese which can be viewed as overnutrition.

This classification enabled us to identify both undernourished and overnourished children, thereby highlighting the double burden of malnutrition in the study population. BMI is found to be a highly significant variable to predict IQ in Logistic_6 model.

BMI and Z-score show varying results across models, with some suggesting a potential relationship with IQ category, but the high standard errors and varying significance levels make these findings less reliable. The analysis revealed a significant correlation between BMI and Z-scores and IQ, indicating that better nutritional status is generally associated with higher cognitive function. However, when adjusted for other factors using logistic regression, the effect size diminished, suggesting that while nutrition plays a crucial role, additional variables—such as socioeconomic status, parental education, and school environment—also influence cognitive outcomes. This contrast highlights the complex interplay between nutrition and cognitive development, where direct correlations may be moderated by broader contextual factors. Gender does not seem to be a strong predictor of IQ category in these models, aligning with the earlier Chi-square test results. In terms of model performance, Logistic_1 (including Age, BMI, Bad Academic Performance, and Good Academic Performance) shows the highest accuracy (93%) and perfect recall (1.0), suggesting it may be the most effective at predicting IQ category. Logistic_0 and Logistic_3, which are both derived from more variables, also work extremely well with high accuracy and AUC scores. These results are consistent with the high intercorrelations between the various factors and cognitive ability in this sample, with academic achievement being a very strong predictor of IQ classification.

The logistic regression models demonstrated robust performance, with the best-performing model (including age, BMI, and academic performance) achieving an accuracy of 93% and an Area Under the ROC Curve (AUC) of 0.98. The precision and recall values are 0.89 and 0.95 respectively. These metrics indicate a high level of model fitness. The pseudo R² values is around 0.4 which indicates a good proportion of variance being explained in logistic model specially in a socio-demographic study. The significant P values (P < 0.001) for key predictors such as academic performance suggest that the independent variables collectively account for a substantial portion of the variation in IQ categories. This supports the conclusion that factors such as academic performance, along with age and BMI, are strong predictors of cognitive function.

Discussion

The findings of this research in rural primary school children of Jharkhand, India are indicative of a complex relationship among nutritional status, cognitive capacity, and school achievement. The prevalence of approximately 28.03% of the children were classified as undernourished (≈157 children) and 28.75% were classified as severely malnourished (≈161 children). (Benson et al., 2008).[13] These rates, although troubling, are comparatively more favorable than those observed in some previous research in India and West Bengal, suggesting perhaps geographical variation in nutritional status (Sinha et al., 2018;[30] Dasgupta et al., 2014).[31]

The positive correlation between Intelligence Quotient (IQ) scores and Body Mass Index (BMI) (r = 0.41, P ≤ 0.01) observed here is evidence of the potential relationship between nutritional status and cognitive development, as earlier observed in studies investigating the role of nutrition in shaping brain function (Nyaradi et al., 2013).[7] However, the very high proportion of children (50.71%) classified as “intellectually impaired” is evidence of severe issues with the overall cognitive development of this population. This observation combined with the very skewed gender distribution of high IQ scores in favor of males is evidence of socio-cultural determinants on cognitive development and learning access (Rindermann et al., 2007).[32]

Around 4.47% of the children were overweight/obese (≈25 children) which indicates over nourishment. These figures emphasize the significant burden of both undernutrition and overnutrition in the sample in the research population, confirms the growing double burden of malnutrition in the developing nations where underweight or undernutrition goes hand in hand with growing prevalence of overweight and obesity (Popkin et al., 2012).[33] The trend, however smaller than in some other states of India, is particularly problematic because obesity that occurs during childhood is more likely to persist in adulthood, 30% of obesity that progresses during childhood, and 50-80% of the overweight children ending up as obese adults (Singh et al., 2008).[34]

Logistic regression results demonstrate a strong association between IQ levels and academic performance, emphasizing the crucial role of cognitive abilities in educational success. These findings align with previous research highlighting the impact of cognitive development on learning outcomes, particularly in resource-limited settings.

In addition, the high prevalence of malnutrition observed in this study suggests a pressing need for targeted nutritional interventions. Undernutrition has been widely linked to impaired cognitive function, which, in turn, affects academic performance. This study reinforces existing evidence that improving nutritional status can enhance cognitive abilities and educational outcomes among children.

Furthermore, gender disparities in malnutrition rates, with a higher prevalence among boys, call for further investigation into dietary habits, physical activity levels, and socio-cultural factors influencing nutritional status. Addressing these disparities through school-based nutrition programs and awareness campaigns could contribute to better health and academic performance.

Future direction and limitation

Follow-up studies must aim at longitudinal research to determine causal links between nutritional status, cognitive ability, and scholarship in children in rural India. It is essential to investigate effective, culturally tailored interventions that solve simultaneously undernutrition, over-nutrition, and cognitive development. Investigations into micronutrients, especially those that influence brain development, may yield important information. In addition, research on gender disparities in nutritional status and cognitive outcomes is needed to inform targeted interventions. However, this study has several limitations to consider: its cross-sectional design limits causal inferences; the sample size and geographical focus may not be representative of all rural Indian children; potential confounding factors such as socioeconomic status, parental education, and home environment were not fully accounted for; and the use of BMI as the sole anthropometric measure may not capture all aspects of nutritional status. Future studies should address these limitations by employing more comprehensive assessment tools, including additional anthropometric measures, biomarkers of nutritional status, and more detailed cognitive assessments, while also considering a wider range of potential confounding variables.

Conclusion

This study of rural primary school children in Jharkhand, India, unveils a complex landscape of nutritional and cognitive challenges. The high prevalence of wasting (28.03%) and severe wasting (28.75%), coupled with emerging overweight/obesity (4.47%), highlights the double burden of malnutrition. The positive correlation between BMI and IQ scores (r = 0.41, P ≤ 0.01) and the alarming rate of children classified as “intelligently impaired” (50.71%) stresses the intricate relationship between nutritional status and cognitive development. Strong associations between academic performance and IQ categories further emphasize this connection. These findings collectively call for urgent, comprehensive interventions that address both ends of the malnutrition spectrum while focusing on cognitive development and educational support. Future research should delve deeper into the causal relationships between these factors and explore effective strategies to break the cycle of malnutrition, poor cognitive development, and academic underachievement in resource-limited settings, ultimately aiming to improve the overall health and prospects of rural Indian children.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.