Bridging the knowledge divide: a national cross-sectional analysis of urban-rural disparities in the stunting knowledge index among mothers/caregivers in Indonesia
This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Indonesia continues to encounter a significant public health crisis due to stunting. Maternal knowledge is the key driver of nutritional practices and outcomes. This study aimed to develop a maternal/caregiver stunting knowledge index (SKI), analyze disparities in urban-rural knowledge, and identify the key determinant factors driving these disparities.
Methods
This cross-sectional study analyzed data points from the 2024 Indonesian Nutritional Status Survey, involving 294,538 mothers and caregivers. Data regarding stunting knowledge, sociodemographic, maternal, and caregiver characteristics, child characteristics, area of residence, and variables related to maternal health knowledge and utilization were collected via interviews conducted with mothers. Height-for-Age Z-score (HAZ) was calculated using age and height measurements. Knowledge level was measured using a constructed SKI categorized as low, moderate, or high. Chi-square tests and multivariate logistic regression were used to identify the predictors of low SKI and risk(s) for stunting.
Results
The prevalence of low SKI scores was significantly higher in rural areas, and maternal education was the strongest determinant of low SKI scores in both regions. In rural areas, limited access to formal health information (e.g., lack of a Maternal and Child Health book) was a major predictor. Low SKI was associated with higher odds for stunting risk (urban: adjusted odds ratio [aOR], 1.617; rural: aOR, 1.576), whereas high awareness of the long-term impacts of stunting provided a significant protective effect.
Conclusion
Urban-rural knowledge disparities were concentrated among caregivers with lower education and limited health service utilization. Interventions should be tailored regionally, targeting caregivers with low education levels and emphasizing the long-term consequences of stunting to promote preventive behaviors.
Nutrition is a fundamental pillar of universal health coverage [1]. One of the most significant global nutritional challenges is stunting, a chronic undernutritional condition characterized by a child’s height being less than the standard(s) for age. Stunting is caused by prolonged nutritional deficit, recurrent infection(s), and suboptimal caregiving practices [2,3]. Its long-term consequences are extensive, including impaired cognitive development, reduced economic productivity in adulthood, and increased risk for noncommunicable diseases (NCDs) [4].
Stunting is widely recognized as a critical and persistent public health concern in Indonesia. Although the national prevalence decreased from 24.4% in 2021 to 19.8% in 2024, this figure remained far from the national target of 14.0% by 2024. Furthermore, there is a striking disparity in the prevalence of stunting between urban and rural areas, with significantly higher rates observed in rural settings [5].
A critical factor influencing stunting is the level of maternal and caregiver knowledge regarding proper nutrition and child health. As primary caregivers, mothers play a crucial role in implementing protective feeding and care practices, including health and nutrition services [1,6]. These protective care behaviors directly reflect the mother’s knowledge and ability to implement practices that support her child’s growth and development.
Caregiver knowledge concerning parenting, nutrition, and child health is strongly associated with feeding practices, child nutritional status, and success of stunting prevention interventions [7,8]. Previous studies have indicated that knowledge levels are influenced by factors such as maternal/caregiver education level, socioeconomic status (SES), access to media, and frequency of visits to healthcare facilities. Knowledge levels are often affected by unequal access to information and health services in urban and rural regions. Generally, mothers in urban areas exhibit better knowledge, higher nutritional literacy, and superior access to information than those in rural areas due to differences in education levels, healthcare access, and media exposure [9].
Although stunting disparities are widely documented, in-depth analyses focusing on disparities in the caregiver stunting knowledge index (SKI) as a proximal determinant remain limited, particularly at the national scale. Most existing research has focused solely on overall prevalence or general risk factors. The SKI is a comprehensive measure used to assess the depth and breadth of an individual’s understanding of stunting. The novelty of the present investigation is the application of a constructed SKI designed to measure understanding across four dimensions: definition, causes, future impact, and prevention efforts. The use of this index is essential to accurately identify the specific domains in which knowledge deficits exist.
Accordingly, this study aimed to analyze disparities in SKI among mothers/caregivers of children <5 years of age between urban and rural areas, and to identify the specific determinant factors driving these disparities using data obtained from the 2024 Indonesian Nutritional Status Survey (INSS). The findings are anticipated to provide empirical evidence to inform the development of targeted educational strategies customized for residential contexts (i.e., urban vs. rural), thereby accelerating national stunting prevention efforts.
Methods
Ethics statement
This survey was approved by the Health Research Ethics Commission (Komisi Etik Penelitian Kesehatan) of Health Polytechnic Ministry of Health Jakarta II (Politeknik Kesehatan Kementerian Kesehatan Jakarta II; Poltekkes Jakarta II), Jakarta, Indonesia (No. DP.04.03/I/KE/230/2024). The data were further analyzed with approval from the Health Development Policy Agency (Badan Kebijakan Pembangunan Kesehatan), Ministry of Health of the Republic of Indonesia. The requirement for informed consent was waived due to the use of deidentified data and the retrospective nature of the study.
Study population and data collection
The present investigation used a descriptive analytical design and secondary data from the 2024 INSS. A national scale survey was conducted using a cross-sectional design. The target population comprised all mothers or primary caregivers from households in Indonesia that included children <5 years (0–59 months) of age. The INSS used a multistage cluster sampling design, systematically selecting a sample from 34,500 census blocks across 514 districts/cities nationwide. With 10 households selected from each census blocks, the survey generated an overall sample size of 345,000 households with children <5 years of age.
To establish the final analytical sample, the following rigorous inclusion criteria were applied: only mother–caregiver pairs with complete, non-missing data across all essential variables (knowledge index components, sociodemographics, and maternal health utilization) were retained. Anthropometric data quality was ensured by screening for biological plausibility. Records with a Height-for-Age Z-score (HAZ) falling below −6 standard deviations (SDs) or above +6 SDs were considered to be extreme outliers and, subsequently, excluded from the analysis. After a comprehensive data cleaning and screening process, the final analytical sample comprised data from 294,538 eligible mothers/caregivers. While the excluded participants exhibited a slightly higher proportion of rural residences and lower secondary education than the final sample, the large size of the remaining analytical sample ensured that national representativeness remained robust.
Study variables
The instrument used during the 2024 INSS was a comprehensive, structured questionnaire administered by trained field staff. All variables used in this analysis were directly extracted from that instrument. The operational definitions of the key variables are described below.
Stunting knowledge index
The SKI, which measures mothers’ understanding of stunting, served as the dependent variable in this study. The SKI is determined using a 20-item questionnaire addressing four dimensions of knowledge: definition; determinants of stunting/underlying causes; prevention strategies; and potential long-term impacts of stunting.
The SKI was constructed through a rigorous process involving scoring, weighting, and categorization of the participant responses. Each question or statement in the knowledge instrument was classified into one of two groups for scoring: positive statements (correct information, i.e., those that were consistent with scientific facts about stunting); and negative statements (misconceptions/incorrect information, i.e., statements that were inconsistent with scientific facts about stunting).
A weighted score was applied to the available responses to ensure that the option reflecting the most accurate knowledge received the highest score. Responses were assessed using a 5-point Likert scale as follows: 4 (strongly agree), 3 (agree), 2 (undecided), 1 (disagree), and 0 (strongly disagree). The scoring applied a differential rule: for positive statements, a score of 4 was given for “strongly agree” and 0 for “strongly disagree;” and, for negative statements, the scoring was reversed, a score of 4 was given for “strongly disagree” and 0 for “strongly agree.”
The instrument consisted of 20 questions/statements, distributed across four specific dimensions: definition of stunting (six questions), determinants of stunting/underlying causes (seven questions), prevention strategies (four questions), and potential long-term impacts of stunting (three questions). The respondent’s total score is the sum of the weights/scores obtained from all 20 questions. Based on a five-point scale (maximum 4 points per question), the maximum score for overall knowledge was 80. The dimension-specific maximum scores were as follows: definition (score, 24); determinants of stunting/underlying causes (score, 28); prevention strategies (score, 16); and stunting impact (score, 12). The final SKI percentage was calculated using the following equation and used to classify knowledge level:
Cutoff points were established based on the quartile values of the SKI percentage distribution and were used to classify the respondents’ knowledge level into three distinct categories: low, moderate, and high. In the absence of a standardized clinical threshold for stunting literacy, the bottom 25% (lower quartile) was designated as “low SKI” to identify the most vulnerable segment relative to the national average. This distribution-based approach aligns with public health surveillance practices that target interventions and resource allocation.
Anthropometric and demographic variables
Anthropometric data, specifically HAZ for the child, served as the primary criterion for data inclusion. The HAZ was calculated using the World Health Organization Child Growth Standards. Data points with extreme HAZ values (below –6 SDs or above +6 SDs) were excluded during the data cleaning process to ensure data quality and reliability. Children were categorized as stunted if HAZ <–2 SD, and not stunted if HAZ ≥–2 SD.
Other demographic and socioeconomic variables extracted and defined were as follows: maternal/caregiver characteristics, age (young, <25 years; middle, 25–34 years, older, ≥35 years); education level (high, completed undergraduate up to postgraduate degrees; middle, completed high school up to D1–D3 diploma; and low did not attend school, completed or not completed primary school, completed junior high school); employment status (working in formal sectors, informal sectors, not working); marital status (married, widowed/divorced); and family SES (low, middle, high). Child characteristics included age (0–5, 6–11, 12–23, or 24–59 months), sex (male or female), birth order (first-born, laterborn), and area of residence (urban or rural).
Additional variables relating to maternal health knowledge and utilization included attendance at antenatal care (ANC) classes during pregnancy (yes, no), possession of the Maternal and Child Health (MCH) book (yes, no), receiving explanations regarding the content of the MCH book from a healthcare provider (yes, no), frequency of ANC visits during the child’s pregnancy (≥6 times, <6 times), and receiving nutrition counseling within the past year (yes, no).
Data analysis
All data management and statistical analyses were performed using IBM SPSS ver. 25.0 (IBM Corp.). To ensure national representativeness, sample weights were incorporated into the analysis. Complex analysis was used for data analysis.
A two-stage analytical approach was used. Descriptive statistics are used to express categorical variables (including SKI, maternal/child characteristics, and area of residence) as frequency and weighted percentage. Second, for inferential analysis, the chi-square test was used for bivariate analysis to determine the association between the area of residence and SKI category. Variables exhibiting an association with SKI (P<0.25) were then included in a multivariate logistic regression model to identify independent predictors while adjusting for potential confounders. The results are presented as adjusted odds ratio (AOR) with corresponding 95% confidence interval (CI). Differences with P<0.05 were considered to be statistically significant.
Results
Sociodemographic and child characteristics
The present study analyzed cross-sectional data from the 2024 INSS, encompassing 294,538 mothers/caregivers. Sociodemographic profiles exhibited key differences according to area (Table 1). Most mothers in both regions were 25–34 years of age, although rural areas had a higher proportion of younger mothers (<25 years of age). While most mothers in both areas were unemployed (approximately 66%–67%) and married, there was a stark educational gap: mothers with high education levels were significantly more prevalent in urban than in rural settings (12.9% vs. 7.1%), while the number of those with low education levels was approximately 1.6 times as high in rural areas as in urban areas (58.3% vs. 37.0%). SES also varied, with urban households predominantly in the high SES category (46.7%), whereas rural households were mainly in the low SES category (47.3%). Child characteristics were largely similar (e.g., sex and birth order). The prevalence of stunting was markedly higher among rural children (19.6%) than that among urban children (15.3%), underscoring the severity of the issue in rural contexts, which is likely linked to the observed educational and economic disparities.
Distribution of maternal stunting knowledge accuracy
The knowledge accuracy rate varied significantly between regions (Table 2). For example, regarding the misconception that ‘an obese child is definitely not stunted’, the accuracy rate was significantly higher among urban caregivers (42.5%) than that among rural caregivers (36.5%), indicating that urban respondents were more likely to correctly disagree with this assumption. Furthermore, a substantial proportion (urban, 51.4%; rural, 58.6%) believed that a short stature was acceptable if the child appeared to be healthy.
Caregivers in rural areas exhibited a significantly higher rate of misinformation regarding causes, with 65.1% incorrectly citing hereditary factors as the primary cause compared with 51.6% in urban areas. Knowledge of prevention methods (exclusive breastfeeding and animal-source food) was generally high, although understanding of the long-term impacts on NCDs was only moderate (approximately 50% awareness).
SKI distribution
The assessment of caregivers’ knowledge of stunting was further aggregated into SKI categories of low, moderate, and high. The distribution of the SKI according to area of residence (i.e., urban vs. rural) across both the total index and its four constituent dimensions (definition, causes, prevention, and impact) is illustrated in Figure 1. Overall, the largest proportion of caregivers in both urban (43.2%) and rural (43.7%) areas was represented by the moderate-knowledge category.
However, a critical difference emerged between these extremes. In urban areas, the proportion of caregivers with high knowledge was higher than that of the caregivers with low knowledge. Conversely, in rural areas, caregivers with low knowledge outnumbered those with high knowledge.
In both the urban and rural high knowledge groups, the strongest mastery was observed in the stunting prevention dimension (urban, 57.9%; rural, 54.9%) and stunting definition (urban, 44.9%; rural, 37.7%). The lowest scores in the high SKI group for both regions were found for the stunting causes dimension (rural, 17.5%; urban, 20.7%), indicating that this area is the most challenging, even for knowledgeable caregivers. Among caregivers with low SKI, the greatest knowledge deficits were concentrated in the definition of stunting (rural, 45.4%; urban, 37.1%), suggesting a fundamental lack of conceptual understanding in this segment.
Predictors of SKI
Bivariate analysis of 10 independent variables revealed that nearly all were significantly associated with SKI (P<0.05) (Table 3). The most robust predictors of low SKI across urban and rural areas included maternal education (yielding the largest odds ratio), household SES, and access to formal health knowledge (including MCH book engagement, ANC visits, and nutrition counseling). Maternal employment also exhibited a substantial association, particularly in rural settings. In summary, lower maternal knowledge about stunting was most strongly predicted by socioeconomic and educational disadvantages, followed closely by poor utilization of formal health-information channels.
Multivariate analysis confirmed that maternal education remained the most influential determinant of low SKI across both urban and rural settings (Table 3). Specifically, low education levels exhibited the highest risk for low SKI (urban: AOR, 2.823; 95% CI, 2.579–3.090; rural: AOR, 1.843; 95% CI, 1.659–2.048). The lower the educational level, the higher the likelihood of a low SKI score.
In rural areas, access to health knowledge variables emerged as strong independent predictors of low SKI and lacked an explanation of MCH book content (AOR, 1.761; 95% CI, 1.592–1.948), <6 ANC visits (AOR, 1.749; 95% CI, 1.630–1.877), not receiving nutrition counseling (AOR, 1.671; 95% CI, 1.551–1.799), and nonparticipation in pregnancy classes (AOR, 1.603; 95% CI, 1.501–1.712).
Furthermore, maternal employment status was linked to a higher likelihood of low SKI in rural areas, especially for mothers working in informal sectors (AOR, 1.855; 95% CI, 1.648–2.088). This highlights the fact that rural knowledge disparities are concentrated among caregivers with lower educational levels and limited health service utilization.
Relationship between SKI and stunting in children
Overall, SKI exhibited a significant inverse association with stunting risk (P<0.05), confirming that lower maternal knowledge was associated with higher odds for stunting risk in both urban and rural settings (Table 4). Compared to mothers with high SKI (reference category), mothers in the low SKI category faced an elevated risk for having a stunted child (urban: AOR, 1.617; rural: AOR, 1.576). Mothers with moderate SKI also exhibited increased risk (urban: AOR, 1.211; rural: AOR, 1.296).
A notable finding of this study was that the AOR for the association between a low SKI and stunting was remarkably consistent across both regions (AOR, approximately 1.6). This indicates that the effect size of maternal knowledge on children’s nutritional status is similar regardless of the residential setting. However, despite the similar magnitude of risk, the actual prevalence of stunting remained significantly higher in rural areas (19.6%) than that in urban centers (15.3%).
This apparent paradox suggests that the rural-urban stunting gap is not necessarily driven by a stronger ‘potency’ of knowledge in rural settings, but rather by a higher concentration and distribution of risk factors. Rural areas exhibited a disproportionately higher prevalence of both low maternal education and SKI scores. Therefore, the elevated stunting rates in rural Indonesia are likely a reflection of a greater density of caregivers who fall into high-risk categories combined with systemic barriers to access health information. These results highlight that, while the mechanism linking knowledge to stunting is universal, the public health burden is geographically skewed because of the unequal distribution of educational and informational resources.
Specific findings from the analysis according to knowledge dimension indicated that low knowledge of the definition of stunting was associated with higher odds for stunting (urban: AOR, 1.069; rural: AOR, 1.122). Similarly, low- and moderate-knowledge regarding the determinants of stunting exhibited a significant, although minor, association with stunting status. Conversely, knowledge of the long-term consequences of stunting was consistently associated with lower odds for stunting across both areas in the low- and moderate-knowledge categories. This suggests that even basic awareness of long-term consequences is linked to preventive health behaviors. The strongest association with reduced odds for stunting was found in the moderate-knowledge of the impact category (urban: AOR, 0.854; rural: AOR, 0.855).
Discussion
This study identified a clear association between maternal knowledge and the risk for stunting within a complex conceptual framework. Our findings demonstrated that a low SKI was independently associated with stunting, even after adjusting for potential confounders such as maternal education and household SES. This correlation suggests that, while education and economic stability provide the necessary foundation for childcare, specific knowledge regarding stunting (its causes, prevention, and long-term impact) contributes to a unique link with protective outcomes that are not entirely captured by formal schooling alone.
However, given the cross-sectional nature of this national survey, these findings must be interpreted as associations rather than as established causal pathways. Although the conceptual framework suggests a potential trajectory in which higher education improves knowledge, leading to better nutritional practices and reduced stunting, this study did not formally perform a mediation or decomposition analysis to quantify these indirect effects. Therefore, we have refined our terminology to avoid implicit causal claims, focusing instead on concentrated disparities among vulnerable groups. Future longitudinal research is warranted to further determine whether interventions specifically targeting the ‘knowledge divide’ can directly trigger a reduction in the prevalence of stunting independent of broader socioeconomic improvements.
This study used a newly developed, multidimensional SKI using large-scale, nationally representative data from the 2024 INSS to analyze disparities in caregiver knowledge regarding stunting between urban and rural Indonesia, and to identify the specific upstream determinants of these disparities. Our findings indicated that lower overall maternal knowledge was significantly associated with childhood stunting across both residential settings, reinforcing the essential role of knowledge in protective caregiving behaviors. Crucially, the analysis revealed clear disparities in both the quality of knowledge and the primary barriers to attaining it, demanding context-specific policy responses.
Our results demonstrated a significant association between caregiver knowledge and stunting. The overall prevalence of stunting in rural areas (19.6%) was markedly higher than that in urban areas (15.3%), which mirrors the distribution of SKI. While the largest share of caregivers in both areas fell into the moderate SKI category (approximately 43%), the higher proportion of mothers in the low SKI group in rural areas, coupled with lower educational attainment, reflected greater underlying vulnerability. This finding is highly consistent with cross-sectional studies from other low- and middle-income countries, which also highlight rural residence as an independent correlate of stunting, often observed alongside limited parental knowledge [10,11].
Statistical results regarding vulnerability were reinforced using multivariate analysis. Mothers in the low SKI category had significantly higher odds for having a stunted child than those with high SKI, both in urban (AOR, 1.617) and rural (AOR, 1.576) areas. This association is comparable with findings in regions such as South Asia and Sub-Saharan Africa, where inadequate maternal knowledge regarding complementary feeding and sanitation practices increases the odds for stunting by >50% [12]. Consistent with these findings, research from Rwanda identified a strong correlation between low maternal knowledge and inadequate feeding practices, highlighting significant gaps in the causes and consequences of stunting, particularly among vulnerable rural populations [13].
The consistency of AOR values across these contexts suggests that the association with maternal knowledge of stunting is stable across regions and is not significantly modified by residential characteristics. Maternal knowledge is a key proximal factor linked to nutritional status, representing a core component of care practices and the sociocultural environment outlined in the UNICEF Conceptual Framework [14].
Multivariate analysis identified a hierarchical structure among the factors associated with low SKI. Consistent with the existing literature, maternal education is the strongest upstream factor linked to knowledge levels [15], demonstrating the highest AORs for low knowledge across both urban (AOR, 2.823) and rural (AOR, 1.843) areas. The wide educational gap between urban and rural mothers (58.3% in rural areas vs. 37.0% in urban areas) is closely reflected in the overall SKI disparity. This suggests that educational attainment plays a fundamental role in caregivers’ capacity to acquire, process, and act on complex nutritional and health information, representing the core and structural barriers to improved knowledge.
However, its association with other factors was highly context dependent, reflecting the heightened vulnerability of the rural population. For the same employment status (unemployed and working in the informal sector), rural residences exhibited a significantly higher likelihood of a low knowledge index compared with urban areas (AOR, approximately 1.70–1.85 in rural areas vs. AOR, approximately 1.05–1.10 in urban areas). This suggests that, in rural contexts, a lack of formal employment may be linked to limited social networks, access to information, and/or financial capacity to prioritize health education [16]. Poor engagement with formal health systems, specifically inadequate ANC visits, lack of MCH handbook explanations, and absence of nutrition counseling, were significantly associated with higher odds for low SKI, but only among rural caregivers (AOR, 1.603–1.761).
The findings underscore the complex interplay between distal structural barriers and proximal service failures that perpetuates the ‘knowledge divide’ in Indonesia. While maternal education serves as a fundamental structural factor, its association with knowledge level is significantly moderated by the quality of interaction with the healthcare system. Our analysis revealed that, even among mothers with higher formal education, the absence of an explicit explanation of the MCH handbook and inadequate ANC counseling were significantly associated with low SKI.
This interaction suggests that formal education provides the cognitive capacity to process information; however, healthcare services, specifically the ‘active’ delivery of information through MCH book explanations, act as the essential vehicle for translating that capacity into functionally stunting literacy. In rural areas, the concurrence of lower educational attainment and less intensive health counseling is associated with the ‘double burden’ of information poverty. Therefore, addressing this knowledge gap is supported by a two-pronged strategy: long-term investment in female education; and immediate systemic improvements in the quality of ANC services. Healthcare providers must move beyond the ‘passive’ distribution of MCH handbooks toward an ‘active’ educational role, ensuring that the critical window of pregnancy is used to bridge the knowledge divide, regardless of the mother’s formal educational background.
This suggests that rural communities encounter a double burden of systemic socioeconomic barriers compounded by deficiencies in the dissemination and use of vital health information by local services. Contact with trained health professionals (both pre- and postnatal care) and frequent interactions are strongly associated with accurate knowledge, highlighting critical gaps in rural service delivery. Furthermore, a negative relationship was found between having received information about stunting >6 months before and knowledge of stunting, suggesting that consistent, frequent contact from trained health representatives is necessary to alleviate misconceptions and maintain knowledge over time [17,18].
This novel application of multidimensional SKI revealed qualitative disparities in knowledge quality. Despite a general understanding of the definition and prevention methods (e.g., exclusive breastfeeding), critical misconceptions persisted. A substantial portion of caregivers incorrectly associated stunting with nonmodifiable factors such as dwarfism or genetics. This condition was also reported in previous studies. According to the Health Belief Model (HBM), mothers who consider the cause of stunting to be hereditary and beyond their control may be less likely to engage in preventive behaviors [19]. This misinterpretation is supported by the finding that the ‘Causes’ dimension of the SKI was the weakest area for all caregiver groups. Furthermore, the persistent belief that ‘short stature is acceptable if the child appears healthy’ highlights a dangerous normalization of stunting (urban, 51.4%; rural, 58.6%), which is linked to lower rates of early detection and intervention. This suggests that interventions should move beyond simple counseling to encompass evidence-based and culturally contextualized social and behavioral change communication to challenge entrenched social norms [14].
A more granular analysis of the SKI subdomains revealed that knowledge regarding the long-term consequences of stunting, such as permanent cognitive deficits and the potential for NCDs in adulthood, exhibited a consistent association with lower stunting odds (AOR <1.0) even at moderate levels of awareness. This finding suggests that perceived severity and awareness of ‘irreversible harm’ may serve as a more potent motivational factor for preventive health behaviors than the mastery of technical nutritional knowledge (e.g., specific nutrient requirements or biological determinants).
In the HBM framework, the high perceived seriousness of the consequences of stunting may be linked to a stronger internal cue to action, leading caregivers to adhere more strictly to available nutritional guidance. Consequently, public health communication strategies should move beyond merely disseminating technical facts about stunting and place greater emphasis on the tangible long-term life outcomes of the child. This ‘impact-centric’ approach could bridge the gap between knowledge and practice, especially in resource-limited rural settings where technical literacy may be lower, but the motivation to ensure a child’s future potential remains a universal priority.
The most salient finding for the intervention design was the finding that high knowledge regarding the long-term consequences of stunting, such as cognitive deficits and the potential for NCDs, provided a significant and consistent association with lower stunting odds across all low and moderate categories (AOR, <1.0; specifically between 0.854 and 0.891). This association appears to transcend technical knowledge levels (e.g., only 10%–12% of respondents knew about the role of vegetables). This suggests that a serious comprehension of irreversible future harm, such as reduced cognitive capacity or adult illness, is a more potent motivational factor for positive behavioral change than simply understanding immediate, technical preventive steps.
These findings strongly support our recommendation that stunting prevention strategies move beyond generic national campaigns and adopt context-specific educational customization. Educational interventions must prioritize women with low educational levels and those in informal/unemployed sectors because maternal education is the strongest upstream factor associated with knowledge levels. Given that knowledge of the impact of stunting is a strong protective factor, public messaging should strategically focus on reinforcing irreversible long-term consequences (cognitive impairment and the potential for NCDs) to enhance motivation, rather than solely focusing on the technical aspects of nutrition or immediate feeding practices. Intervention delivery must be intensified in rural areas by strengthening the formal health channels that are significantly associated with knowledge levels. Specifically, this means ensuring that health personnel consistently provide full, comprehensive explanations of the MCH book, and that ANC visits meet the minimum required frequency to maintain consistent contact and information delivery. Initiatives should also consider leveraging existing social networks, such as neighbors and friends, to disseminate accurate information because poor-quality social information is negatively associated with knowledge [17].
A major strength of this study was the use of data from the 2024 INSS, which enabled a robust, national-level, generalization using complex survey sampling and ensures that the findings are contemporary. Furthermore, the use of the constructed multidimensional SKI provides a depth of understanding far beyond simple awareness, assessing comprehension across four critical dimensions (definition, causes, prevention, and long-term consequences). This enabled us to pinpoint specific qualitative deficits and crucially revealed a novel finding of the protective effect associated with knowledge of the long-term impact of stunting on health outcomes. A key limitation was the cross-sectional nature of the INSS data, which prevents the establishment of a causal relationship between knowledge changes and stunting status over time.
In conclusion, results of the present study indicate that low maternal knowledge (i.e., low SKI) is a significant and independent factor associated with higher odds for child stunting across all urban and rural areas of Indonesia. This finding highlights a critical disparity, where a high prevalence of rural stunting is observed alongside systemic disadvantages in caregiver knowledge acquisition. Knowledge disparities were concentrated among caregivers with lower education levels and limited health service utilization. However, poor access to formal health knowledge channels (inadequate ANC visits and lack of MCH book explanations) emerged as a factor uniquely concentrated among rural caregivers, indicating double vulnerability stemming from socioeconomic barriers and failures in information dissemination. Crucially, the analysis revealed that knowledge regarding the long-term impact of stunting (cognitive and NCD risks) had a significant protective effect (AOR, <1.0), suggesting that this is a potent motivational factor for behavioral change. National stunting prevention efforts must be tailored to the regional context, prioritizing low-educated women, and strategically leveraging messaging regarding the irreversible long-term impacts of stunting to strengthen the motivation for preventive action.
Notes
Conflict of interest
No potential conflict of interest relevant to this article was reported.
Acknowledgments
We thank Ministry of Health of Indonesia for providing data on this study. We also acknowledge the use of Google Gemini for language editing and proofreading support. The AI was utilized to improve the clarity and readability of the manuscript, while the final content and intellectual contributions remain solely the responsibility of the author(s).
Funding
None.
Data availability
The data sets used in this study were obtained from the Centre of Data and Information of the Indonesian Ministry of Health.
Author contribution
Conceptualization: RR, N. Data curation: RR, BS, AYA, NF. Formal analysis: RR, AYA, BS, NF. Methodology: RR, N, DHT, DSKP, Siti Masitoh. Writing–original draft: RR, N, BS, AYA, S, NF, YA, Syaiful Mizan. Writing–review & editing: RR, N, BS, AYA, NF. Final approval of the manuscript: all authors.
Figure. 1.
Proportion of knowledge index according to stunting among mothers/caregivers based on four dimensions of knowledge, according to area of residence. Values are presented as %.
Table 1.
Sociodemographic characteristics of the study samples, based on urban and rural areas of residence
a)Percentages for socioeconomic status are calculated based on total available cases including missing values.
Table 2.
Distribution of maternal stunting knowledge accuracy across urban and rural areas
Knowledge item/dimension
Urban
Rural
Total
Definition of stunting
Stunted children are children who are short or dwarfed
41,535 (23.8)
21,767 (18.2)
63,302 (21.5)
A stunted child is a child with a short body size due to prolonged malnutrition
127,449 (72.9)
83,297 (69.5)
210,747 (71.6)
A stunted child is a child with a lower body weight compared to a normal child of the same age
18,982 (10.9)
14,498 (12.1)
33,481 (11.4)
A stunted child is a child with a shorter height compared to a normal child of the same age
131,878 (75.5)
89,162 (74.4)
221,040 (75.0)
Stunting can be detected if infants and toddlers have their weight and height monitored routinely
150,017 (85.8)
97,412 (81.3)
247,429 (84.0)
An obese child is definitely not stunted
74,304 (42.5)
43,664 (36.5)
117,968 (40.1)
Factors causing stunting
Child stunting is more often caused by hereditary factors than by nutritional and health factors
84,493 (48.4)
41,768 (34.9)
126,261 (42.9)
The cause of child stunting is a lack of vegetable consumption during the mother’s pregnancy and the toddler period
39,613 (22.7)
23,410 (19.5)
63,022 (21.4)
The cause of child stunting is a lack of consumption of animal protein side dishes, such as eggs, fish, chicken, and meat, during the mother’s pregnancy and the toddler period
121,587 (69.6)
78,480 (65.5)
200,066 (67.9)
The cause of child stunting is the child’s difficulty in eating and small meal portions
112,674 (64.5)
81,563 (68.1)
194,237 (65.9)
The cause of child stunting is frequent illness or recurrent illness
99,841 (57.1)
74,089 (61.9)
173,930 (59.1)
The cause of child stunting is inappropriate child care and feeding practices
108,850 (62.3)
74,316 (62.0)
183,166 (62.2)
Children who live in a dirty environment or an unclean environment are at higher risk of stunting
111,867 (64.0)
76,303 (63.7)
188,169 (63.9)
How to prevent stunting
Child stunting can be prevented by having pregnant mothers and toddlers consume more vegetables
22,181 (12.7)
14,664 (12.2)
36,845 (12.5)
Child stunting can be prevented by having pregnant mothers and toddlers consume more animal protein side dishes, such as eggs, fish, chicken, and meat
142,088 (81.3)
88,059 (73.5)
230,147 (78.1)
Child stunting can be prevented by breastfeeding exclusively, which means only giving breast milk from birth until the age of 6 months
134,065 (76.7)
88,735 (74.1)
222,800 (75.6)
Pregnant mothers who do not consume iron tablets are at risk of giving birth to babies with a birth weight of less than 2.5 kg, which can lead to the risk of child stunting
108,023 (61.8)
72,875 (60.8)
180,898 (61.4)
Impact of stunting
Child stunting can have an impact on a child’s developmental and intellectual impairment
134,670 (77.1)
87,051 (72.7)
221,721 (75.3)
Child stunting can have an impact in adulthood, making them easily susceptible to hypertension, diabetes mellitus
92,621 (53.0)
61,456 (51.3)
154,077 (52.3)
A stunted child is not a problem; the important thing is that the child is healthy
84,909 (48.6)
49,542 (41.4)
134,451 (45.6)
Values represent the knowledge accuracy rate (%), defined as the percentage of respondents who provided the correct response for each item. For positive statements, this indicates an ‘agree/strongly agree’ response; for negative statements (misconceptions), this indicates a ‘disagree/strongly disagree’ response, signifying that the respondent correctly identified the misconception.
Table 3.
Multivariable analysis to predict low knowledge of stunting, based on urban and rural areas of residence
Factor
Urban
Rural
COR (95% CI)
P-value
AOR (95% CI)
P-value
COR (95% CI)
P-value
AOR (95% CI)
P-value
Maternal age (y)
<25
1.158 (1.064–1.261)
0.001
1.031 (0.935–1.136)
0.543
0.954 (0.880–1.035)
0.257
0.954 (0.864–1.052)
0.344
25–34
0.797 (0.748–0.828)
<0.001
0.884 (0.837–0.933)
<0.001
0.782 (0.742–0.825)
<0.001
0.892 (0.843–0.945)
<0.001
≥35
Reference
Reference
Maternal education level
Low education
3.688 (3.394–4.008)
<0.001
2.823 (2.579–3.090)
<0.001
2.953 (2.691–3.241)
<0.001
1.843 (1.659–2.048)
<0.001
Middle education
2.235 (2.072–2.412)
<0.001
1.872 (1.730–2.025)
<0.001
1.783 (1.628–1.952)
<0.001
1.326 (1.201–1.465)
<0.001
High education
Reference
Reference
Maternal employment status
Not working
1.693 (1.579–1.816)
<0.001
1.051 (0.975–1.133)
0.193
2.650 (2.387–2.942)
<0.001
1.703 (1.516–1.914)
<0.001
Working in informal sectors
1.650 (1.516–1.795)
<0.001
1.096 (1.004–1.196)
0.041
3.000 (2.687–3.349)
<0.001
1.855 (1.648–2.088)
<0.001
Working in formal sectors
Reference
Reference
Maternal marital status
Widowed/divorced
1.310 (1.131–1.518)
<0.001
1.028 (0.882–1.197)
0.727
1.330 (1.089–1.625)
0.005
1.176 (0.951–1.454)
0.135
Married
Reference
Reference
Socioeconomic status
Low
1.977 (1.862–2.099)
<0.001
1.550 (1.457–1.649)
<0.001
1.961 (1.835–2.095)
<0.001
1.470 (1.369–1.577)
<0.001
Middle
1.547 (1.451–1.649)
<0.001
1.321 (1.235–1.412)
<0.001
1.361 (1.268–1.461)
<0.001
1.182 (1.096–1.276)
<0.001
High
Reference
Reference
Attendance at ANC classes during pregnancy
No
1.762 (1.668–1.863)
<0.001
1.596 (1.504–1.694)
<0.001
2.133 (2.006–2.267)
<0.001
1.603 (1.501–1.712)
<0.001
Yes
Reference
Reference
Healthcare provider giving explanations regarding the content of the MCH book
No
1.849 (1.719–1.989)
<0.001
1.476 (1.370–1.591)
<0.001
2.501 (2.268–2.759)
<0.001
1.761 (1.592–1.948)
<0.001
Yes
Reference
Reference
Frequency of ANC visits during the child’s pregnancy
< 6 times
2.413 (2.263–2.573)
<0.001
1.974 (1.841–2.115)
<0.001
2.433 (2.277–2.601)
<0.001
1.749 (1.630–1.877)
<0.001
≥6 times
Reference
Reference
Receiving nutrition counseling within the last year
1. Gbratto-Dobe SA, Segnon HB. Is mother’s education essential to improving the nutritional status of children under five in Côte d′Ivoire? SSM Health Syst 2025;4:100056.
2. Victora CG, Christian P, Vidaletti LP, Gatica-Dominguez G, Menon P, Black RE. Revisiting maternal and child undernutrition in low-income and middle-income countries: variable progress towards an unfinished agenda. Lancet 2021;397:1388-99.
3. Dewey KG. Reducing stunting by improving maternal, infant and young child nutrition in regions such as South Asia: evidence, challenges and opportunities. Matern Child Nutr 2016;12 Suppl 1:27-38.
4. De Sanctis V, Soliman A, Alaaraj N, Ahmed S, Alyafei F, Hamed N. Early and long-term consequences of nutritional stunting: from childhood to adulthood. Acta Biomed 2021;92:e2021168.
5. Badan Kebijakan Pembangunan Kesehatan. Survei Status Gizi Indonesia (SSGI) 2024 Dalam Angka. Badan Kebijakan Pembangunan Kesehatan; 2025.
6. Rezaeizadeh G, Mansournia MA, Keshtkar A, Farahani Z, Zarepour F, Sharafkhah M, et al. Maternal education and its influence on child growth and nutritional status during the first two years of life: a systematic review and meta-analysis. EClinicalMedicine 2024;71:102574.
8. Dwi Ananta S, Krisnana I, Tri Lestari W. The relationship between maternal knowledge level and feeding practices with dietary patterns in stunted children. J EduHealth 2023;14:1393-9.
9. Wati L, Nasution NA, Aurallia N, Nashirah S, Rizki M, Harahap R, et al. Factor analysis of maternal knowledge on the incidence of stunting. Contagion 2022;4:143-52.
10. Jemide JO, Ene-Obong HN, Edet EE, Udoh EE. Association of maternal nutrition knowledge and child feeding practices with nutritional status of children in Calabar South Local Government Area, Cross River State, Nigeria. Int J Home Sci 2016;2:293-8.
11. Mauludyani AV, Khomsan A. Maternal nutritional knowledge as a determinant of stunting in West Java: rural-urban disparities. Amerta Nutr 2022;6:8-12.
12. Yaya S, Oladimeji O, Odusina EK, Bishwajit G. Household structure, maternal characteristics and children's stunting in Sub-Saharan Africa: evidence from 35 countries. Int Health 2022;14:381-9.
13. Mporanyi T, Condo J, Absolomon G, Mporanyi M, Rubuga FK, Munyanshongore C. Influence of maternal knowledge of causes and consequences of stunting on infant and young child feeding practices in Rwanda. Sci Rep 2025;15:44481.
15. Rah JH, Cronin AA, Badgaiyan B, Aguayo VM, Coates S, Ahmed S. Household sanitation and personal hygiene practices are associated with child stunting in rural India: a cross-sectional analysis of surveys. BMJ Open 2015;5:e005180.
16. World Health Organization. Health workforce [Internet]. World Health Organization; 2025 [cited 2025 Nov 11]. Available from: https://www.who.int/teams/health-workforce
17. Haines AC, Jones AC, Kriser H, Dunn EL, Graff T, Bennett C, et al. Analysis of rural Indonesian mothers’ knowledge, attitudes, and beliefs regarding stunting. Med Res Arch 2018;6:1-13.
18. Saleh A, Syahrul S, Hadju V, Andriani I, Restika I. Role of maternal in preventing stunting: a systematic review. Gac Sanit 2021;35 Suppl 2:S576-82.
19. Hall C, Bennett C, Crookston B, Dearden K, Hasan M, Linehan M, et al. Maternal knowledge of stunting in rural Indonesia. Int J Child Health Nutr 2018;7:139-45.
Bridging the knowledge divide: a national cross-sectional analysis of urban-rural disparities in the stunting knowledge index among mothers/caregivers in Indonesia
Figure. 1. Proportion of knowledge index according to stunting among mothers/caregivers based on four dimensions of knowledge, according to area of residence. Values are presented as %.
Graphical abstract
Figure. 1.
Graphical abstract
Bridging the knowledge divide: a national cross-sectional analysis of urban-rural disparities in the stunting knowledge index among mothers/caregivers in Indonesia
Characteristic
Urban
Rural
Total
Maternal characteristic
Mother’s age (y)
<25
17,302 (9.9)
17,164 (14.3)
34,467 (11.7)
25–34
97,539 (55.8)
63,353 (52.9)
160,891 (54.6)
≥35
59,911 (34.3)
39,269 (32.8)
99,180 (33.7)
Mother’s occupational status
Working in formal sectors
26,307 (15.1)
7,681 (6.4)
33,988 (11.6)
Working in informal sectors
32,772 (18.8)
31,796 (26.5)
64,568 (21.9)
Not working
115,673 (66.2)
80,309 (67.0)
195,982 (66.5)
Mother’s education level
High education
22,626 (12.9)
8,468 (7.1)
31,094 (10.5)
Middle education
87,382 (50.0)
41,508 (34.7)
128,890 (43.8)
Low education
64,745 (37.0)
69,810 (58.3)
134,555 (45.7)
Mother’s marital status
Married
171,207 (98.0)
117,883 (98.4)
289,090 (98.2)
Widowed/divorced
3,545 (2.0)
1,903 (1.6)
5,448 (1.8)
Socioeconomic statusa)
Low
46,181 (26.4)
56,604 (47.3)
102,784 (34.9)
Middle
30,093 (17.2)
22,977 (19.2)
53,070 (18.0)
High
81,641 (46.7)
32,147 (26.8)
113,787 (38.6)
Missing
16,838 (9.6)
8,059 (6.7)
24,897 (8.5)
Child’s characteristics
Sex
Boy
89,083 (51.0)
61,318 (51.2)
150,401 (51.1)
Girl
85,669 (49.0)
58,468 (48.8)
144,137 (48.9)
Child’s age (mo)
0–5
9,892 (5.7)
6,888 (5.8)
16,780 (5.7)
6–11
16,583 (9.5)
11,435 (9.5)
28,019 (9.5)
12–23
35,130 (20.1)
24,189 (20.2)
59,319 (20.1)
24–59
113,146 (64.7)
77,274 (64.5)
190,420 (64.7)
HAZ index
Stunting
26,662 (15.3)
23,519 (19.6)
50,181 (17.0)
Not stunting
148,090 (84.7)
96,267 (80.4)
244,357 (83.0)
Birth order
First-born
49,958 (28.6)
34,204 (28.6)
84,162 (28.6)
Later-born
124,794 (71.4)
85,583 (71.4)
210,376 (71.4)
Knowledge item/dimension
Urban
Rural
Total
Definition of stunting
Stunted children are children who are short or dwarfed
41,535 (23.8)
21,767 (18.2)
63,302 (21.5)
A stunted child is a child with a short body size due to prolonged malnutrition
127,449 (72.9)
83,297 (69.5)
210,747 (71.6)
A stunted child is a child with a lower body weight compared to a normal child of the same age
18,982 (10.9)
14,498 (12.1)
33,481 (11.4)
A stunted child is a child with a shorter height compared to a normal child of the same age
131,878 (75.5)
89,162 (74.4)
221,040 (75.0)
Stunting can be detected if infants and toddlers have their weight and height monitored routinely
150,017 (85.8)
97,412 (81.3)
247,429 (84.0)
An obese child is definitely not stunted
74,304 (42.5)
43,664 (36.5)
117,968 (40.1)
Factors causing stunting
Child stunting is more often caused by hereditary factors than by nutritional and health factors
84,493 (48.4)
41,768 (34.9)
126,261 (42.9)
The cause of child stunting is a lack of vegetable consumption during the mother’s pregnancy and the toddler period
39,613 (22.7)
23,410 (19.5)
63,022 (21.4)
The cause of child stunting is a lack of consumption of animal protein side dishes, such as eggs, fish, chicken, and meat, during the mother’s pregnancy and the toddler period
121,587 (69.6)
78,480 (65.5)
200,066 (67.9)
The cause of child stunting is the child’s difficulty in eating and small meal portions
112,674 (64.5)
81,563 (68.1)
194,237 (65.9)
The cause of child stunting is frequent illness or recurrent illness
99,841 (57.1)
74,089 (61.9)
173,930 (59.1)
The cause of child stunting is inappropriate child care and feeding practices
108,850 (62.3)
74,316 (62.0)
183,166 (62.2)
Children who live in a dirty environment or an unclean environment are at higher risk of stunting
111,867 (64.0)
76,303 (63.7)
188,169 (63.9)
How to prevent stunting
Child stunting can be prevented by having pregnant mothers and toddlers consume more vegetables
22,181 (12.7)
14,664 (12.2)
36,845 (12.5)
Child stunting can be prevented by having pregnant mothers and toddlers consume more animal protein side dishes, such as eggs, fish, chicken, and meat
142,088 (81.3)
88,059 (73.5)
230,147 (78.1)
Child stunting can be prevented by breastfeeding exclusively, which means only giving breast milk from birth until the age of 6 months
134,065 (76.7)
88,735 (74.1)
222,800 (75.6)
Pregnant mothers who do not consume iron tablets are at risk of giving birth to babies with a birth weight of less than 2.5 kg, which can lead to the risk of child stunting
108,023 (61.8)
72,875 (60.8)
180,898 (61.4)
Impact of stunting
Child stunting can have an impact on a child’s developmental and intellectual impairment
134,670 (77.1)
87,051 (72.7)
221,721 (75.3)
Child stunting can have an impact in adulthood, making them easily susceptible to hypertension, diabetes mellitus
92,621 (53.0)
61,456 (51.3)
154,077 (52.3)
A stunted child is not a problem; the important thing is that the child is healthy
84,909 (48.6)
49,542 (41.4)
134,451 (45.6)
Factor
Urban
Rural
COR (95% CI)
P-value
AOR (95% CI)
P-value
COR (95% CI)
P-value
AOR (95% CI)
P-value
Maternal age (y)
<25
1.158 (1.064–1.261)
0.001
1.031 (0.935–1.136)
0.543
0.954 (0.880–1.035)
0.257
0.954 (0.864–1.052)
0.344
25–34
0.797 (0.748–0.828)
<0.001
0.884 (0.837–0.933)
<0.001
0.782 (0.742–0.825)
<0.001
0.892 (0.843–0.945)
<0.001
≥35
Reference
Reference
Maternal education level
Low education
3.688 (3.394–4.008)
<0.001
2.823 (2.579–3.090)
<0.001
2.953 (2.691–3.241)
<0.001
1.843 (1.659–2.048)
<0.001
Middle education
2.235 (2.072–2.412)
<0.001
1.872 (1.730–2.025)
<0.001
1.783 (1.628–1.952)
<0.001
1.326 (1.201–1.465)
<0.001
High education
Reference
Reference
Maternal employment status
Not working
1.693 (1.579–1.816)
<0.001
1.051 (0.975–1.133)
0.193
2.650 (2.387–2.942)
<0.001
1.703 (1.516–1.914)
<0.001
Working in informal sectors
1.650 (1.516–1.795)
<0.001
1.096 (1.004–1.196)
0.041
3.000 (2.687–3.349)
<0.001
1.855 (1.648–2.088)
<0.001
Working in formal sectors
Reference
Reference
Maternal marital status
Widowed/divorced
1.310 (1.131–1.518)
<0.001
1.028 (0.882–1.197)
0.727
1.330 (1.089–1.625)
0.005
1.176 (0.951–1.454)
0.135
Married
Reference
Reference
Socioeconomic status
Low
1.977 (1.862–2.099)
<0.001
1.550 (1.457–1.649)
<0.001
1.961 (1.835–2.095)
<0.001
1.470 (1.369–1.577)
<0.001
Middle
1.547 (1.451–1.649)
<0.001
1.321 (1.235–1.412)
<0.001
1.361 (1.268–1.461)
<0.001
1.182 (1.096–1.276)
<0.001
High
Reference
Reference
Attendance at ANC classes during pregnancy
No
1.762 (1.668–1.863)
<0.001
1.596 (1.504–1.694)
<0.001
2.133 (2.006–2.267)
<0.001
1.603 (1.501–1.712)
<0.001
Yes
Reference
Reference
Healthcare provider giving explanations regarding the content of the MCH book
No
1.849 (1.719–1.989)
<0.001
1.476 (1.370–1.591)
<0.001
2.501 (2.268–2.759)
<0.001
1.761 (1.592–1.948)
<0.001
Yes
Reference
Reference
Frequency of ANC visits during the child’s pregnancy
< 6 times
2.413 (2.263–2.573)
<0.001
1.974 (1.841–2.115)
<0.001
2.433 (2.277–2.601)
<0.001
1.749 (1.630–1.877)
<0.001
≥6 times
Reference
Reference
Receiving nutrition counseling within the last year
No
1.803 (1.696–1.917)
<0.001
1.568 (1.472–1.670)
<0.001
2.058 (1.918–2.209)
<0.001
1.671 (1.551–1.799)
<0.001
Yes
Reference
Reference
Child’s birth order
Later-born
1.159 (1.106–1.214)
<0.001
1.019 (0.960–1.081)
0.541
1.184 (1.124–1.247)
<0.001
1.087 (1.018–1.162)
0.001
First-born
Reference
Reference
Factor
Stunting
Urban
Rural
AOR (95% CI)
P-value
AOR (95% CI)
P-value
Total knowledge index of stunting
Low
1.617 (1.453–1.800)
<0.001
1.576 (1.411–1.760)
<0.001
Moderate
1.211 (1.127–1.301)
<0.001
1.296 (1.196–1.403)
<0.001
High
Reference
Reference
Knowledge about definition of stunting
Low
1.069 (1.004–1.137)
0.037
1.122 (1.053–1.196)
<0.001
Moderate
1.047 (0.977–1.123)
0.195
1.072 (1.000–1.148)
0.051
High
Reference
Reference
Knowledge about factors causing stunting
Low
1.128 (1.027–1.238)
0.012
1.088 (0.991–1.194)
0.077
Moderate
1.087 (1.014–1.165)
0.018
1.084 (1.004–1.170)
0.038
High
Reference
Reference
Knowledge about how to prevent stunting
Low
0.921 (0.867–0.979)
0.008
1.005 (0.942–1.071)
0.890
Moderate
0.967 (0.909–1.029)
0.285
0.967 (0.908–1.030)
0.300
High
Reference
Reference
Knowledge about impact of stunting
Low
0.891 (0.832–0.955)
0.001
0.885 (0.833–0.940)
<0.001
Moderate
0.854 (0.802–0.910)
<0.001
0.855 (0.806–0.907)
<0.001
High
Reference
Reference
Table 1. Sociodemographic characteristics of the study samples, based on urban and rural areas of residence
Values are presented as number (%).
HAZ, Height-for-Age Z-score.
Percentages for socioeconomic status are calculated based on total available cases including missing values.
Table 2. Distribution of maternal stunting knowledge accuracy across urban and rural areas
Values represent the knowledge accuracy rate (%), defined as the percentage of respondents who provided the correct response for each item. For positive statements, this indicates an ‘agree/strongly agree’ response; for negative statements (misconceptions), this indicates a ‘disagree/strongly disagree’ response, signifying that the respondent correctly identified the misconception.
Table 3. Multivariable analysis to predict low knowledge of stunting, based on urban and rural areas of residence