Dietary intake diversity and its impact on child malnutrition among under-five children in Northern Nigeria: An analysis of demographic and health surveys 2013–2018 data

Research design

This study employs a cross-sectional design to examine the evolving trends and determinants of child malnutrition, specifically stunting, wasting, and underweight among children under five in Northern Nigeria. By analyzing secondary data from the NDHS for the years 2013 and 2018, the study captures changes over time without the need for primary data collection or intervention.

Sampling technique and population

The study focuses on children under five residing in Northern Nigeria, encompassing the North-Central, North-East, and North-West regions. The NDHS utilizes a two-stage stratified cluster sampling method. Initially, each state (including the Federal Capital Territory) is stratified by urban and rural areas. Subsequently, clusters (enumeration areas) are selected using the latest census data, followed by systematic household selection within each cluster. This approach yielded a representative sample for the region. Detailed descriptions of the sampling procedure are available in the NDHS reports.

Data source and collection

Data for this research were obtained from the NDHS datasets for 2013 and 2018. These nationally representative datasets provide extensive information on child health, including anthropometric measurements and household characteristics. Anthropometric data, such as weight, height, and age, were collected using calibrated instruments and processed into Z-scores for weight-for-age, height-forage, and body mass index (BMI)-for-age according to the World Health Organization (WHO) standards. In this study, stunting, underweight, and wasting are defined as Z-scores below -2 for height-for-age, weight-for-age, and BMI-for-age, respectively.

Data analysis

Data analysis was performed using IBM Statistical Package for the Social Sciences Statistics (Version 28.0). Summary measures, including mean Z-scores for weight-for-age, height-for-age, and BMI-for-age, were computed to assess the overall nutritional status of the child population. Logistic regression analyses were conducted to evaluate the associations between independent variables (e.g., maternal education, child’s age, socioeconomic status) and the dependent variable (child malnutrition). The relationships are presented as odds ratios (OR) with 95% confidence intervals (CI), and statistical significance was established at a P < 0.05.

Ethical considerations

This study employs secondary data from the DHS, which are publicly accessible and deidentified to protect participant confidentiality. Approval to access and utilize the DHS dataset was granted by the DHS program. Furthermore, this study adheres to ethical guidelines established by relevant institutions and aligns with international research ethics standards, including the principles outlined in the Declaration of Helsinki.

Sociodemographic characteristics of mother and child in Northern Nigeria

The study presents sociodemographic data from 39,720 mother-child pairs in Northern Nigeria, with 47.6% surveyed in 2013 and 52.4% in 2018. The majority of mothers were aged 25–29 years (28.3%), predominantly from the NorthWest (44.5%), and largely lived in rural areas (75.8%). Most had no formal education (63.4%) and practiced Islam (80.7%). Clerical and sales work was the most common occupation (45.4%), while 20.3% were unemployed. Poverty levels were high, with 31.7% in the poorest and only 7.7% in the richest category. Normal BMI was observed in 68.5% of mothers. Most households had 1–2 children (62.0%), and child sex was nearly evenly split (50.2% male and 49.8% female). Children were mainly aged 3–4 years (40.0%) and of average (44.5%) or large (39.4%) size at birth. A vast majority (74.0%) of births occurred at home, and most were single births (97.5%), with 65.7% being first-borns.

Nutritional intake among children in Northern Nigeria

Table 2 highlights the dietary patterns of children across three age groups in Northern Nigeria. Plain water is the most consumed item, with 26.6% of children under 1 year, 32.6% of children aged 1–2 years, and 28.3% of children aged 3–4 years consuming it. Consumption of liquid and fortified children’s foods is notably low across all age groups, with only 0.1% of children under 1 year, 0.3% of children aged 1–2 years, and 0.2% of children aged 3–4 years consuming these foods. Solid and semi-solid carbohydrate foods are more commonly consumed as children age, with 7.7% of children under 1 year, 20.6% of children aged 1–2 years, and 14.1% of children aged 3–4 years consuming them. Protein intake is minimal, reported at 0.2% for children under 1 year, 0.5% for those aged 1–2 years, and 0.2% for children aged 3–4 years. Similarly, fruits and vegetables are underrepresented in the diet, with consumption reported at 0.4% for children under 1 year, 1.3% for children aged 1–2 years, and 0.8% for children aged 3–4 years. Consumption of infant-specific foods is relatively stable but remains low, with 0.8% of children under 1 year, 1% of children aged 1–2 years, and 0.7% of children aged 3–4 years consuming such foods.

Patterns of nutritional status among children in Northern Nigeria (2013 and 2018)

Figure 1 illustrates the trends in malnutrition patterns among children in Northern Nigeria between 2013 and 2018. The prevalence of stunted malnutrition increased from 39.4% in 2013 to 44.7% in 2018, representing a 5.3 percentage point rise. Conversely, wasted malnutrition experienced a marked decline, reducing from 17% in 2013 to 8.1% in 2018, equating to a 52.4% reduction. Similarly, underweight malnutrition decreased from 30% in 2013 to 27.5% in 2018, reflecting a 2.5 percentage point decline.

Close

Figure 1:

Pattern of nutritional status/malnutrition among children in Northern Nigeria in 2013 and 2018.

Export to PPT

Prevalence of stunted, wasted, and underweight malnutrition among children in Northern Nigeria (2013 and 2018)

Figure 2 presents the prevalence of stunted, wasted, and underweight malnutrition among children in Northern Nigeria in 2013 and 2018. Overall, 59.2% of children were categorized as normal with respect to stunting, while 40.8% were stunted, indicating chronic malnutrition. For wasting, which reflects acute malnutrition, 85.5% of children were classified as normal, with 14.5% experiencing wasting. Similarly, in the case of underweight malnutrition, representing general malnutrition, 70.7% of children were considered normal, while 29.3% were malnourished.

Close

Figure 2:

Prevalence of stunted, wasted, and underweight malnutrition among children in Northern Nigeria (2013 and 2018).

Export to PPT

Malnutrition and nutritional intake diversity among children in Northern Nigeria

The results in Table 3 show a clear association between malnutrition and lower nutritional intake diversity among children in Northern Nigeria. Stunted children had reduced intake across all food categories, with only 38.2% consuming plain water, 39.7% consuming carbohydrates, and 29.8% consuming protein-rich foods, compared to 61.8%, 60.3%, and 70.2% of normal children, respectively. Wasted children showed the lowest diversity, with just 16.0% consuming plain water, 14.4% carbohydrates, 11.9% protein, and 12.7% fruits and vegetables, all markedly lower than the corresponding 84.0%, 85.6%, 88.1%, and 87.3% observed among normal children. Underweight children also had lower intake, with 28.7% consuming plain water, 27.6% carbohydrates, and 20.5% protein, while normal children had higher rates at 71.3%, 72.4%, and 79.5%, respectively. Similar disparities were noted for fruits, vegetables, and infant foods, indicating that malnourished children consistently consumed less diverse diets than their well-nourished peers.

Factors associated with malnutrition from stunted growth among children in Northern Nigeria

Tables 4 and 5 represent the multivariate analysis of factors associated with malnutrition among children under five in Northern Nigeria, which revealed significant trends. Children surveyed in 2018 had higher odds of stunting (adjusted OR [AOR] = 1.365) but lower odds of wasting (AOR = 0.445) and underweight (AOR = 0.891) compared to 2013. Maternal age influenced outcomes, particularly underweight, with mothers aged 20–24 and 25–29 having increased odds (AOR = 1.858 and 1.903), while those aged 30–34 had reduced odds of stunting (AOR = 0.863). Regional disparities were notable; children in the North-East had higher odds of stunting (AOR = 1.496), wasting (AOR = 1.253), and underweight (AOR = 1.472), with even greater risks in the North-West (stunting: AOR = 2.040; wasting: AOR = 1.701; underweight: AOR = 2.207). Rural residence increased wasting (AOR = 1.822) but reduced underweight odds (AOR = 0.917). Maternal education was protective against stunting and wasting, especially at higher levels (stunting: AOR = 0.557; wasting: AOR = 0.680), while higher education reduced underweight (AOR = 0.486), though primary and secondary levels increased it. Religion also played a role, with Muslim households linked to higher wasting (AOR = 1.659) and Christian households to underweight (AOR = 1.817). Employment in agriculture or manual labor raised the odds of stunting and underweight (e.g., manual: stunting AOR = 1.149; underweight AOR = 1.188).

Higher household wealth was associated with lower stunting (richest: AOR = 0.531) and underweight (AOR = 0.808). Maternal nutritional status was strongly predictive that children of underweight mothers had markedly higher odds of all forms of malnutrition (stunting: AOR = 1.872; wasting: AOR = 1.920; underweight: AOR = 2.698), with elevated risks also seen for children of mothers with normal and overweight BMIs. Households with more children (e.g., 3–4 children: stunting AOR = 1.110; wasting AOR = 1.955) had higher malnutrition odds, especially underweight with 5+ children (AOR = 1.999). Male children were more likely to be stunted (AOR = 1.177) and underweight (AOR = 1.135) but less likely to be wasted (AOR = 0.885). Older children (ages 1–2 and 3–4) had substantially higher stunting (AOR = 3.426 and 3.613) and underweight (AOR = 1.799 and 1.688) odds. Small or average birth size increased wasting and underweight (e.g., small size: wasting AOR = 2.219; underweight AOR = 2.632), while birth in public hospitals lowered wasting odds (AOR = 0.180) but increased underweight (AOR = 1.210). Twin births raised all malnutrition risks (e.g., underweight AOR = 1.921), whereas higher birth order (3rd+) reduced them (stunting AOR = 0.600). Finally, adequate dietary diversity was protective across all outcomes (stunting: AOR = 0.894; wasting: AOR = 0.075; underweight: AOR = 0.919)

Trends and patterns in child malnutrition

The nutritional status data indicate that nearly 41% of children are stunted, a marker of chronic malnutrition, with a troubling increase from 39.4% in 2013 to 44.7% in 2018. In contrast, wasting (reflecting acute malnutrition) decreased dramatically from 17.0% to 8.1% over the same period, and underweight prevalence showed a modest decline. These divergent trends are in line with other national analyses. For example, Tesfaw and Fenta^[17] reported stunting at 36.2% and wasting at 6.7% in their NDHS-based analysis, while Lawal et al.^[18] found stunting and wasting prevalence of 36.5% and 6.9%, respectively. Although some local studies, such as that by Usman et al.^[19] in Mil-Goma, Kaduna State reported even higher stunting rates (up to 59.7%). This divergence suggests that while interventions or economic improvements may have reduced the incidence of acute malnutrition, long-term chronic nutritional deficiencies persist, likely as a result of ongoing poor dietary quality, socioeconomic disadvantage, and limited access to quality health services. This finding suggests that while emergency interventions and improvements in acute care may reduce wasting, long-term deficiencies continue to drive stunting. This observation is also supported by Gething et al.,^[20] who noted that despite overall improvements in child growth failure across Africa, chronic undernutrition remains unevenly distributed, particularly in the northern regions.

Dietary intake diversity as a determinant of nutritional outcomes

A central finding of this study is the strong association between dietary intake diversity and child nutritional status. Malnourished children, whether stunted, wasted, or underweight, consistently exhibited lower consumption of essential food groups compared to their normal counterparts. These findings show that only 38.2% of stunted children consumed adequate amounts of plain water, carbohydrate foods, proteinous foods, fruits and vegetables, and infant foods. Similar patterns were evident among wasted and underweight children. This finding is well supported by studies such as Azupogo et al.^[21] in northern Ghana, which documented a significant positive association between dietary diversity scores and weight-for-height and weight-for-age Z-scores. Likewise, complementary feeding studies by Olatona et al.^[22] and Senbanjo et al.^[23] have demonstrated that poor feeding practices and limited dietary diversity are critical risk factors for malnutrition. These findings indicate that inadequate dietary diversity may be a key modifiable risk factor for malnutrition, emphasizing the need for nutrition education and interventions that promote varied and balanced diets among young children.

Multifactorial determinants of child malnutrition

Temporal trends from this study revealed that, compared to 2013, children surveyed in 2018 had significantly higher odds of stunting but lower odds of wasting and underweight. This mixed picture may reflect improvements in acute care or emergency feeding programs alongside a deterioration in factors contributing to chronic malnutrition.

Multivariate models revealed outcome-specific nuances, with some age groups showing increased odds for underweight outcomes. More strikingly, higher maternal education was consistently protective against stunting and wasting, highlighting the role of maternal knowledge and empowerment in ensuring proper child nutrition. This finding echoes the results of studies by Lawal et al.,^[18] Akombi et al.,^[24] and Reinbott and Jordan,^[25] all of which underpin the role of maternal education in ensuring better child nutrition.

Geographic disparities were also evident. Children from the North-East and North-West had significantly higher odds of all forms of malnutrition compared to those in the North-Central region, suggesting that regional differences in access to health services, food security, and cultural practices may influence nutritional outcomes. This finding is in agreement with the work of Alarape et al.^[26] and Ezeh et al.,^[27] who reported relative risks (e.g., OR = 2.15 in the North East and OR = 2.98 in the North-West, in Alarape et al.’s study) that underscore the higher burden of malnutrition in these regions. Socioeconomic status, as measured by the wealth index, exhibited a clear gradient: children from richer households were significantly less likely to be malnourished. Furthermore, maternal nutritional status emerged as a strong predictor, with children of underweight mothers at markedly higher risk of stunting, wasting, and underweight status, and it is an indication of the intergenerational cycle of malnutrition. A similar finding is robustly documented in studies by Adesuyi et al.^[28] and Edeh et al.^[29]

Child-level factors, including sex, age, birth size, and birth order, also contributed to nutritional risk. Older children (aged 1–4 years) had substantially higher odds of stunting and underweight than infants, possibly reflecting cumulative dietary deficits. In addition, children with lower birth weight or smaller size at birth, those delivered outside of optimal health settings, and twin births were more vulnerable to malnutrition. Importantly, even after adjusting for these myriad factors, adequate dietary intake diversity remained protective as it is found to reduce the odds of stunting, wasting, and underweight, reinforcing its critical role in child nutrition. Studies of Azupogo et al. well support this finding^[21] in northern Ghana, which documented a significant positive association between dietary diversity scores and weight-for-height and weight-for-age Z-scores. Likewise, complementary feeding studies by Olatona et al.^[22] and Senbanjo et al.^[23] have demonstrated that poor feeding practices and limited dietary diversity are critical risk factors for malnutrition.

Persistent rise in stunting and potential solutions

The continued rise in stunting, despite reductions in wasting and underweight, may be linked to long-term factors that emergency nutrition programs alone cannot address. These include persistent poverty, low maternal education, poor dietary diversity, inadequate sanitation, and limited access to quality health services, particularly in rural and conflict-affected areas.^[30] Seasonal food shortages, cultural feeding practices, and chronic food insecurity further contribute to insufficient nutrient intake during critical growth periods. Addressing stunting requires a shift from short-term relief to sustained, multi-sectoral strategies such as improving maternal and child nutrition through education, strengthening primary healthcare systems, investing in water, sanitation, and hygiene (WASH), and supporting local food production and market access.^[31] Coordinated action involving government, health workers, and community stakeholders is essential to reverse this trend.

Study limitations and mitigation strategies

Despite the robust design and comprehensive analysis, several limitations warrant acknowledgment. We implemented various strategies to minimize their impact:

1. Data availability: Although we initially planned to include data from the 2021 Nigeria Demographic and Health Survey (NDHS) to capture more recent trends, malnutrition-related data were unavailable for 2021. Consequently, our analysis is confined to the 2013 and 2018 NDHS datasets. To bridge this gap, we complemented our findings with secondary data from UNICEF and WHO reports.

2. Secondary data constraints: Relying on NDHS data means that some local nuances, such as detailed cultural dietary practices and specific household food security measures, were not captured. However, the use of standardized anthropometric measurements and established indicators (e.g., Z-scores for stunting, wasting, and underweight) ensured consistency. We further enriched our analysis through triangulation with other studies and reports.

3. Self-reported data: Some variables, including maternal occupation, education, and childbirth details, were based on self-reports, which may introduce recall bias. Nonetheless, the NDHS employs rigorous quality assurance procedures such as interviewer training and validation checks. Sensitivity analyses were also conducted to confirm the robustness of our findings.