Altitude of area of residence as a risk factor for stunting in children aged 0–60 months: A systematic review and meta-analysis

Abstract

Background

Altitude has not been a factor considered in stunting incidence. Several studies have found that higher elevations increase a child's risk of stunting. However, many cases of stunting also occur in lowland areas. Accurate evidence is needed to justify the influence of altitude on stunting incidence globally, so that policies and interventions can be more specific.

Purpose

This study aims to evaluate existing evidence regarding the effect of altitude on stunting cases in children 0–60 month.

Methods

This design was a meta-analysis. We search for relevant articles from 2014 to 2024 from Pubmed, Science Direct, Sage Journal, Scopus, and Oxford academic. Two independent reviewers extracted data from the selected studies, including baseline information, strategies, screening processes, inclusion and exclusion criteria, data extraction, study quality evaluation, and statistical analysis. The Joanna Briggs Institute conducted the critical appraisal. All data analyses were performed using RevMan 5.3 with multiple logistic regression analysis.

Results

The electronic search yielded a total of 805 articles, and 5 articles met the inclusion criteria for meta-analysis. Children living in highland areas have a 2.91 times higher risk of stunting compared to those living in lowland areas (OR = 2.91; 95 % CI: 2.44–3.48). Highland areas face challenges in limited access to health care, low education, less diverse food, and poverty.

Conclusions

This study provides suggestions for increasing the number of health workers and health services, premarital education for childcare, and the provision of nutritional supplements for children.

1. Introduction

As many as 150.2 million children under the age of 5 experience stunting [1]. Stunting is a key target of the Sustainable Development Goals (SDGs), specifically Goal 2, which aims to eliminate hunger and all forms of malnutrition by 2030 and achieve food security. WHO actively supports health programs to address stunting in various countries, especially in developing nations [2]. However, cases of stunting persist due to unsupportive policies and economic conditions [3,4]. Only one-third of developing countries are currently on track to reduce the number of children affected by stunting by 50 % by 2030. Furthermore, a quarter of the countries categorized as “on track” are still unable to make significant progress in addressing stunting. These estimates highlight the need for more intensive and context-specific efforts if the world is to meet the global target of reducing the number of children experiencing stunting to 89 million by 2030 [5,6].

Children are defined as stunted if their height-for-age z-score is less than −2 standard deviations (stunted) and less than −3 standard deviations (severely stunted). Stunting can hurt children's cognitive development [7]. Stunted children may experience impaired brain development, mental retardation, poor concentration, low memory retention, and a higher risk of chronic diseases such as hypertension and diabetes. The consequences of stunting affect not only children and their families but also the overall quality of a country's human resources [8]. Nevertheless, these impacts can be mitigated through multisectoral intervention approaches [9].

Low economic and educational conditions have a significant impact on the incidence of stunting [10]. Research has shown that the causes of stunting include poor nutritional status during pre-pregnancy, pregnancy, and breastfeeding periods, inadequate food supply, low micronutrient content, contamination of food and water, infectious diseases, and poor environmental conditions [9,11,12].

Based on the biopsychosocial theoretical model related to stunting incidents proposed by Engel 1977, a person's health condition is the result of the interaction between biological, psychological, and social factors. Stunting conditions in children can be influenced by pre-pregnancy and early life periods such as nutritional deficiencies, starvation, fetal distress, hypoxia, and infectious diseases [13]Social and environmental conditions also greatly influence children's growth and development as stated in the Environmental Health Theory, also known as the ecological determinants of health (EDOH) which reports that ecological factors certainly include climate change, stratospheric ozone depletion, loss of biodiversity, changes in hydrological systems and freshwater supplies, land degradation, and pressure on food production systems will greatly affect health outputs including stunting [14].

Although environmental factors are one of the key determinants of stunting, However altitude has not been a factor considered in stunting incidence, intervention programs often overlook the complex interplay of environmental variables [15]. Several studies have investigated geographic variations in the prevalence of stunting. In Rwanda, for example, the prevalence and determinants of stunting vary significantly across regions [16]. A similar situation is observed in Ethiopia, where the prevalence of stunting varies significantly across zones, highlighting the need for location-specific strategies [17]. The altitude of the residential area has been significantly associated with stunting status in Ethiopia [18]. Higher altitudes are associated with a higher prevalence of stunting. The highest rates of stunting are found in the highlands of Amhara state, while the lowest is in the lowlands of Somalia. Similarly, the highlands of Ecuador show a significant increase in stunting prevalence. Conversely, in the lowlands of Southwest Ethiopia, stunting prevalence is also on the rise. Cases of stunting are more prevalent in densely populated lowland urban areas. In Rishikesh, an urban lowland area, the prevalence of stunting is 22.9 % higher than in rural highland areas [19]. Interestingly, the number of families with stunted toddlers in both lowland and highland regions shows minimal difference [20]. These findings highlight the variation in stunting prevalence based on altitude, yet only a few studies have specifically investigated the effect of altitude on stunting in children [15]. The diversity of results requires a conclusion that is able to interpret most of the population so as to justify the influence of altitude on stunting incidence globally, so that policies and interventions can be more specific.

Based on studies in various developing countries or countries with low middle economies, the implementation of policies to overcome stunting is carried out without consideration of geographical conditions. The implementation of the policy is applied equally regardless of the specific needs of each region as is the case in Indonesia, Nepal, and Ethiopia. In Indonesia, specific and sensitive nutrition policies are implemented nationally. Nutrition-specific is a direct intervention related to nutrition such as supplementation and supplementary feeding, while Nutrition-sensitive is an indirect but affecting aspect of nutrition such as sanitation, maternal education, access to clean water, reproductive health, poverty reduction [21,22]. Nepal implements a policy in the form of a Multi-Sectoral Nutrition Plan (MSNP) to overcome stunting [23]. The policy is an evidence-based and cost-effective nutrition intervention that integrates national and community priorities. Its implementation includes improving the nutritional status of mothers, adolescents and children, increasing food production through home gardening, maintenance of ungags, nutrition education, and improving hygiene including clean water [24]. The Ethiopian government's policy to address stunting is contained in a commitment known as the “Seqota Declaration (SD)”. The policy is operationalized through multisectoral by involving the Ministries of Health, Agriculture, Education, Water Irrigation and Energy, to the Ministry of Manpower and the Ministry of Social Culture and Tourism. This policy also focuses on specific and sensitive nutritional interventions [25].

However, some challenges often arise such as inequality between regions/geographies and between social economic groups and capacity at the local level so that many urban or lowland areas already have high food and education sufficiency, but get assistance with facilities that are not suitable for example getting vitamin assistance and the obligation to attend classes for pregnant women. The UNICEF-WHO-WB Joint Child Malnutrition Estimates (JME) found that only about one-third of all countries have a good progress target to halve the number of children affected by stunting in the world by 2030 and a quarter of countries that fall into the “on track” category do not allow to have good progress in handling stunting [26].

Intervention programs should differentiate between stunting cases in children living in highland and lowland areas.This research is essential because intervention programs must consider the specific geographic and environmental conditions of the child's residence. A systematic review (SR) and meta-analysis (MA) are needed to assess the impact of altitude on stunting in children aged 0–60 months. This study aims to evaluate existing evidence regarding the effect of altitude on stunting cases in children within this age group.

2. Method

This study employs a systematic review and meta-analysis for studies that used the dependent variable of stunting incidence in children aged 0–60 months, which were conducted in highland and lowland areas. We have restrictions on the date published (2014–2024).

2.1. Population, Concept, Context (PCC) Approach

The target population includes cases in children aged 0–60 months. The main concept of this research is stunting (chronic linear growth disorder). Stunting is the condition when height for age is < −2 SD from WHO growth standards, which is an indicator of children's long-term nutritional and health status. The research context includes geographical areas with variations in regional heights which are divided into 2 categories: lowland (less than or equal to 864 m above sea level/masl) and highland (more than 864 m above sea level).

2.2. Selection criteria

This published systematic literature review follows the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) (Fig. 1). The databases including PubMed, Science Direct, Sage Journal, Scopus and Oxford Academic. The keywords and queries we used in 2 type for all database: (1) stunting AND highland AND lowland.(2) stunting and altitude.

Fig. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)

We included all articles in both English and Indonesian. Specifically for Indonesian articles, we applied strict methodological screening to ensure study quality. We excluded articles not translated into English due to our limitations in language translation and to avoid misinterpretation.

2.3. Filtering process

Screening was conducted by reading the title and abstract to assess the relevance of the article's topic. If the article was found to be relevant to the topic, the full text was then reviewed for data extraction.

2.4. Eligibility criteria

The inclusion criteria for this study were: (1) original research, (2) articles published between 2014 and 2024, (3) studies targeting children aged 0–60 months, (4) a focus on stunting cases, and (5) research conducted in both highland and lowland areas. The exclusion criteria for this study were: (1) repeated or duplicate articles, (2) systematic review and meta-analysis articles, (3) studies involving children over 5 years of age, (4) cases related to wasting or other forms of malnutrition besides stunting, (5) articles that were not open access, and (6) studies that did not specify research locations in highland and lowland areas.

2.5. Data extratction

The information extracted from the articles during the systematic review included the year of publication, title, study type, age of the sample, altitude of the study area, statistical values for odds or odds ratios (OR) using logistic regression, and the 95 % confidence interval. Data extraction from the selected studies was conducted using Microsoft Excel.

2.6. Evaluation of study quality and statistical analysis

To assess the methodological quality of the articles included in the systematic review, the Joanna Briggs Institute's assessment tool was used. The analysis was conducted using the Review Manager program (version 5.3). The influence or relationship between residential altitude and child stunting cases was assessed using odds ratios (OR) and confidence intervals (CI).

3. Result

The electronic search yielded a total of 805 articles: 419 from the ScienceDirect database, 97 from PubMed, 107 from Sage Journals, 97 from Scopus, and 85 from Oxford Academic. Among these, 127 duplicate articles were identified and excluded. After screening the titles and abstracts, 484 articles were found to be unrelated to the study topic, and 176 articles did not meet the case criteria. As a result, 18 articles were considered eligible for inclusion in the systematic review. Of these, five articles met the inclusion criteria for the meta-analysis (Fig. 1).

The five selected articles were published in the following years: Roba et al. (2016), Lock et al. (2018), Mohammed et al. (2020), Mrema et al. (2021), and Tesfaye et al. (2022). These studies were conducted in highland and lowland areas across Africa and Asia. Four studies were conducted in Africa, three in Ethiopia, and one in Tanzania. Additionally, one study was conducted in Asia, specifically in Nepal. A critical appraisal was conducted to ensure that the selected articles were of high quality. The appraisal was conducted by the Joanna Briggs Institute's guidelines for cross-sectional studies. The majority of the articles included met the criteria for critical appraisal (Table 1). (See Table 2.)

Table 1.

Critical Appraisal.

No	Critical Appraisal		Authors
		Teji Roba et al, 2016	Locks et al, 2018	Moham-med et al, 2020	Mrema et al, 2021	Tesfaye et al, 2022
1	Were the criteria for inclusion in the sample clearly defined?	Yes	Yes	Yes	Yes	Yes
2	Were the study subjects and the setting described in detail?	Yes	Yes	Yes	Yes	Yes
3	Was the exposure measured in a valid and reliable way?	Yes	Yes	Yes	Yes	Yes
4	Were objective, standard criteria used for measurement of the condition?	Yes	No	Yes	Yes	No
5	Were confounding factors identified?	No	Yes	No	Yes	No
6	Were strategies to deal with confounding factors stated?	No	No	No	No	No
7	Were the outcomes measured in a valid and reliable way?	Yes	Yes	Yes	Yes	Yes
8	Was appropriate statistical analysis used?	Yes	Yes	Yes	Yes	Yes

Table 2.

Basic Characteristics of studies taken .

No	Researchers and years	Year	Study Design	Data Types	Location	Sample	Altitude (masl)		Results
							Highland	Lowlands
1	Teji Roba et al	2016	Cross-sectional	Primary	District Babile, Enderta and Hintalowajirat (Ethiopia)	Children  age 6–23 months	2000	950	Stunting cases are higher in lowland areas during pre-harvest periods (OR = 0,94; 95 % CI; 0,54-1,63). A major contributing factor for child malnutrition may be poor feeding practices. Health information strategies focused on both Infant and Young Child Feeding (IYCF) practices and dietary diversity of mothers could be a sensible approach to reduce the burden of child malnutrition in rural
2	Locks et al	2018	Cross-sectional	Primary	Kapilvastu and Ancham District (Nepal)	12–23 months	3820	300–610	In Kapilvastu, a lowland ecological area, stunting cases are lower compared to the Ancham region, which is classified as a highland (OR = 0,82; 95 % CI; 0,75-0,91). This result is affected with sex, age and caste, maternal education, household food insecurity level and asset tertile, source of drinking water, type of toilet, whether anyone in the household
3	Mohammed et al	2020	Cross-sectional	Secondary	Ethiopia	<12 months up to >23 months	2000-2499	864	The factors can increasing stunting cases such as: altitude ≥2500 masl (Adjusted Odds Ratio/AOR = 1,41), dietary diversity less than 4 (AOR = 1,38), Meal frequency less than 3 times (AOR = 1,31), Breastfeeding duration less than 12 (AOR = 3,39),low birth weight (AOR = 1,34), Toilet facility unimproved (AOR = 1,58), Maternal weight less than 50 kg (AOR = 1,54), Rural area (AOR = 2,57), and poverty (AOR = 1,68).
4	Mrema JD et al	2021	Cross-sectional	Primary	Kilo district (Tanzania)	Children aged 6–59 months	865	437	Prevalence of stunting and underweight was higher in the highland compared to the lowland areas (p 〈0,001). Significant determinants of underweight were areas of residence (AOR = 4,21), age of the children (AOR = 5,85), and child birth weight (AOR = 4.98), while determinants of stunting were the area of residence (AOR = 2,77), maternal age (AOR = 0,33), sex of a child (AOR = 1,89), and child birth weight (AOR = 3,29).
5	Tesfaye et al	2022	Cross-sectional	Primary	Hararghe Zone, Eastern Ethiopia	Children  age 24–60 months	2300	125	The factors can increasing stunting cases such as: lack of maternal education [AOR = 3,39], women's empowerment [AOR = 3,48] and fourth antenatal care visit [AOR = 4,2], practicing hand washing [AOR = 0,46], living in mid-land [AOR = 1,94] and low-land[AOR = 0,27] agro-ecological zones, PSNP membership [AOR = 1,82], childhood illness [AOR = 8,41], non-exclusive breastfeeding [AOR = 3,6], inadequate minimum dietary diversity [AOR = 4,7], child's sex [AOR = 1,73] and age (24–59 months) [AOR = 3,2] were independent predictors of stunting.

In this study, a critical appraisal was conducted to ensure that the selected articles were of high quality. The appraisal was conducted using the Joanna Briggs Institute (JBI) checklist for cross-sectional studies. The majority of articles met the critical appraisal criteria (Table 1). The results indicated that most of the five articles fulfilled the requirements for critical appraisal, including clear inclusion criteria, detailed explanations of the study subjects and settings, exposure measurements, measurable outcomes, and appropriate statistical analyses.

However, in the critical appraisal question regarding the identification of confounding factors and their handling strategies, none of the articles addressed confounding variables. Additionally, two articles by Locks et al. (2018) and Tesfaye et al. (2022) did not include measurement standards for distinguishing highland and lowland altitudes. Despite this, all articles met the criteria for measurable outcomes and appropriate statistical analysis.

All articles used have similarities in the research design, using cross sectional. The order of articles selected based on the year of publication was the most articles by Teji Roba et al. (2016), Locks et al. (2018), Mohammed et al. (2020), Mrema J D et al. (2021), and Tesfaye et al. (2022). Four studies used primary data and 1 study used secondary data, an article by Mohammed et al. (2020) which used secondary data from the results of the Ethiopian Demographic and Health Survey (DHS). The research targets in all selected articles have met the research criteria children aged 0–60 months.

3.1. The altitude limits of lowland and highland areas

Three studies were conducted in Ethiopia, one studies in Tanzania, and one in Nepal. Research with primary data conducted in Ethiopia, the location consists of various districts, namely Babile, Enderta, Hintalowajirat, and East Hararghe. The studies were conducted in areas with varying altitudes, including both highland and lowland. The altitudes differed, the lowland areas being below 1.000 masl (meters above sea level) and highland areas above 1.000 masl, reaching up to 3.000 masl in some cases. Plateaus are defined as being above 700 masl, although in this SR and Meta study, detailed observations were made of the height limits of plateaus and lowlands listed in each article. The article by Teji Roba shows that the lowlands are at a maximum altitude of 950 masl and the altitude of the highlands is at 2000 masl. There is a slight difference related to the lowland standard in the article when compared to the general standard, namely the highland is above 700 masl. Likewise in the article by Mohammed et al. (2020) which classifies lowland areas with a limit of 864 masl. In the other three articles, the lowland boundary corresponds to the general definition, which is in the range of 125–610 masl, which is found in the articles Locks et al. (2018), Mrema et al. (2021), and Tesfaye et al. (2022). The three articles use a minimum highland classification ranging from an altitude of 865 masl to >3000 masl. However, this classification is not a significant difference because it has the same reference, namely in Highland it is above 1000 masl and in lowland it is below 1000 masl.

3.2. Prevalence of stunting based on residential area altitude

The prevalence of stunting shows different numbers at each height. There are two articles that state that stunting cases occur more in low-lying areas, namely in the articles Locks et al. (2018) and Tesfaye et al. (2022). Furthermore, the other three articles state that the prevalence of stunting is more and at high risk in children living in highlands, articles by Teji Roba et al. (2016), Mohammed et al. (2020), and Mrema et al. (2021). Each article includes other variables besides the height limit of the area. As is the case in the article Teji Roba et al. (2016) accompanied by pre- and post-harvest conditions, in the article Mrema et al. (2021) it is equipped with the variable age of mothers in highlands and lowlands who have stunted children.

3.3. Meta-analysis

Five articles were included in the meta-analysis. The results of the meta-analysis (Fig. 2), represented by diamond plots, show a significant relationship between residential altitude and stunting among children aged 0–60 months. Children living in highland areas have a 2.91 times higher risk of stunting compared to those living in lowland areas (OR = 2.91; 95 % CI: 2.44–3.48). The heterogeneity value of this meta-analysis was 92 %. (See Fig. 3.)

Fig. 2.

Results from Review Manager (RevMan) in the form of a Table Image on Random-Effect Model.

Fig. 3.

Results from Review Manager (RevMan) in the form of a Table Image on Fixed-Effect Model.

In the results of the research Locks et al. stated that children with mothers who routinely check their children at health facilities at least once per month have a lower risk of stunting compared to children with mothers who do not regularly check their children at health care facilities. Children living in high-altitude areas tend not to be often taken to health facilities, in the results of research by Mohammed et al., it is stated that stunting cases in high areas occur due to a lack of food diversity, the age of the mother who tends to be young, maternal education, economic factors, height and weight of the mother. These results have similarities with Mrema et al. research which also states that stunting in highlands is caused by lack of access to health facilities, lack of food diversity, and also the young age of mothers.

The forest plot (Fig. 2) shows the statistical results of each article included in the meta-analysis. The research article by Lock et al. (2018) had the narrowest Confidence Interval (CI), at 2.27, indicating the highest level of confidence. A funnel plot was created to visually illustrate the risk of statistical bias in the study (Fig. 4). The data show that four studies are positioned close to the midline, indicating a lower risk of bias. Only one study had a point far from the midline, suggesting a higher risk of bias. Overall, the distribution in the funnel plot indicates a low risk of publication bias.

Fig. 4.

Results from Review Manager (RevMan) in the form of a Graph (Forest Plot).

4. Discussion

Highlands are formed due to erosion and sedimentation or can also result from large former calderas that are buried by material from the surrounding mountain slopes. The temperature in the highlands is cooler or tends to be colder. Therefore, highland areas always feel fresh [27,28]. A plateau is a large plain located in a high or mountainous area. Highlands or plateaus are generally situated at an altitude above 864 m above sea level. This systematic literature review and meta-analysis examine the relationship between the altitude of the area where children live and the incidence of stunting. Several key points emerge from the results of this study. First, the results of the meta-analysis show a relationship between the altitude of the area and the incidence of stunting in children aged 0–60 months.Second, the characteristics among the selected articles use the variable of altitude, with data that varies; therefore, it needs to be observed and compared until clear limits are established for defining lowland and highland areas.

Stunting cases in the highlands are associated with birth conditions, availability of water sources and toilets, history of infection, maternal health status, and family economic status [18]. Previous research also reported stunting in highland areas may also be related to chronic hypoxia, which disrupts child growth and causes physiological disorders in respiration, metabolism, and blood circulation [29].

- Highland and Lack of Health Facilities.

Communities living in highland and remote areas face challenges in accessing health services, education, and economic opportunities [30]. As in Indonesia, many highland areas have inadequate educational infrastructure, road, and health service centers, such as in Papua [31,32]and even on the island of Java, which is the center of government [33]. Starting from difficult road access, distant schools, to minimal health facilities. This has an impact on the quality of public knowledge, which is still low, and the quality of public health is lacking. In regions such as the North Midlands and the Mountainous Region of Vietnam, as well as in lowland areas, significant gaps exist in health service delivery that impact child health outcomes, including high rates of stunting in mountainous areas [34]. In Papua New Guinea, the highland area shows the highest number of children experiencing stunting and underweight, at 71 % and 33 %, respectively. This high prevalence is associated with low socioeconomic conditions, including a lack of basic household facilities and limited access to health services [24]. Stunting cases are higher in highlands due to the lack of community nutrition improvement services through routine nutritional surveillance and improvements in the quality of health services, including sanitation and environmental management, which are crucial for preventing stunting [35].

5. Highland and poverty

Poverty, limited infrastructure, and access to markets and resources hinder households' ability to improve their agricultural and nutritional practices [36]. An integrated, community-based approach is needed, leveraging technology and innovative solutions to improve infrastructure and service delivery [37]. Promoting economic growth in highland areas is essential to building a decent society with a modern standard of living. Comprehensive public policies in addressing multidimensional barriers are essential for good access to quality education and health services [38].

Similarly, in South Africa, stunting rates are high in areas with poor WASH infrastructure, especially among orphans living in areas with inadequate services [39]. Previous research has reported that there is an increase in stunting cases in highlands areas, such as Ethiopia [18], Ecuador [40] and some developing countries [41]Infrastructure improvements in the highlands area, especially in developing countries, should be accompanied by improving the quality of education and health programs so that the community has more awareness of disease prevention, especially in preventing stunting [42]. Children living in poor environments, such as near inadequate toilets and whose feces are disposed of unsafely, are at higher risk of stunting, this is the case in the highlands of Ethiopia [43]. Families with low socioeconomic status often struggle to provide consistent and nutritious food for their children, which exacerbates the risk of stunting [44]. These factors, combined with economic hardship, create an environment where stunting is more prevalent [45].

6. Highland and unwanted pregnancy/high-risk pregnancy

According to [2], stunting cases are significantly related to maternal age. Several studies have reported the incidence of unwanted pregnancies in highland populations in African countries such as Bolivia [46], Papua New Guinea [47], and even evidence from Indonesian regions such as Papua [48].In addition, high-risk pregnancies in the highlands are more common than in the lowlands [12]. This is in line with research on the rate of early marriage in rural and highland areas [43,49,50]. Low levels of education among girls and their families are strongly associated with high rates of early marriage [39]. Educated girls are less likely to marry at a young age [51]. The results of research conducted by [52] stated that early marriage is significantly related to the possibility of increased stunting in children under the age of five. The Indonesian Demographic and Health Survey (IDHS) study reported that early marriage in Indonesia carries the risk of giving birth to small babies and increases stunting [53].

7. Highland and food consumption or food diversity

Generally, highlands have good agricultural potential. This is supported by several factors such as fertile soil, water availability, and suitable climate [54]. Although the highlands have good agricultural potential, nutritional problems, including stunting in children aged 0–60 months are increasingly at high risk in this region. This is due to many causal factors. Children in the lowlands consume a more diverse diet than children in highland areas [55]. Increased hemoglobin (Hb) levels are a normal adaptive response in highland residential areas [56]. Slow body growth may also be caused by hypoxia, influenced by ecological, socio-cultural, and genetic factors specific to highland regions [57]. This is possible through the mechanism of impaired health organ function in the child's body, so that the child's growth is disrupted, which has an impact on stunting, for example due to disorders in the lungs. According to [30] the number of stunting cases in highland areas is high because these areas often have farms where the environment is exposed to pesticides. High pesticide exposure, especially in shallot farming, increases the risk of stunting in these regions.

However, in general, the research still needs to be confirmed regarding the sample size and the mechanism for assessing pesticide exposure to strengthen the research hypothesis. Better access to protein-rich and diverse foods, which are essential for optimal growth, puts children living in lowland areas at lower risk of stunting. Compared to children who consume traditional diets, children who consume protein-rich diets have a lower risk of stunting [58]. This is evidenced by a study conducted in Kenya [59]. Children in lowland areas are less prone to stunting due to increased access to nutritious food, improved environmental conditions, and higher socio-economic status. Furthermore, they also face a lower risk of parasitic infections that can hinder growth. Together, these elements lead to healthier growth and a reduced likelihood of stunting in lowland regions [18].

Although urban conditions in lowland areas in terms of food diversity availability are met, infrastructure facilities are more complete than in highland areas, but stunting cases still exist. Research carried out by [60] states that stunting cases in urban areas are caused by inadequate parental care. In lowland urban areas, most mothers work actively, which results in a lack of parenting patterns related to the nutrition of their infants or toddlers, so that children grow up with a lack of nutrition [61]. The discussion of the results of systematic literature review and meta-analysis research shows a significant relationship between the altitude of the residential area with the stunting of children 0–60 months. There are differences in the factors that cause stunting in highland and lowland areas. The results of this research further prove the need to consider the difference in stunting handling and prevention intervention programs between highland and lowland areas.

8. Strength and limitation

This meta-analysis has several limitations, including heterogeneity and publication bias [62]. The studies analyzed in this meta-analysis differ significantly in their methods of measuring altitude, population characteristics (child age, socioeconomic status, and environmental conditions), and outcome variation, such as the presence of studies reporting a high prevalence of stunting in lowland areas. This high heterogeneity can affect the generalizability of the meta-analysis results. Studies reporting significant results may be more likely to be published than studies with non-significant results, potentially leading to overestimation of the relationship between altitude and stunting. Although publication bias analysis (e.g., funnel plots or Egger's test) were conducted, the potential for undetected bias remains.

Most of the studies analyzed were from developing countries in specific mountainous regions (e.g., the Andes, Himalayas, East Africa), so the results may not be generalizable to other regions with different geographic, environmental, or socioeconomic characteristics. Intermediary factors, such as chronic hypoxia, soil quality (which affects food production), or changes in agricultural practices in highlands, have not been widely explored in primary studies, so the biological and environmental mechanisms linking altitude and stunting are not fully understood [63].

9. Conclusion

The altitude of residential areas is significantly associated with the risk of stunting in children aged 0–60 months. Early childhood living in highland areas is more at risk of stunting than children living in lowland areas. This research can provide evidence to health policy-making institutions so they do not apply uniform stunting interventions across different areas.

The altitude factor is crucial to consider so that stunting management programs can be tailored separately for highland and lowland areas. This research limits altitude classification to two categories: highlands and lowlands. Risk factors related to differences in social, economic, and health characteristics, as well as other specific factors, between the highlands and lowlands, were not analyzed in this study. Generally, stunting is more prevalent in highland areas than in lowland areas. However, the degree of difference may vary depending on specific regional and environmental factors.

We suggest several policy for highland area such as: (1) increasing the number of health workers and health services, especially midwives, nutritionists, and health promotion workers. (2) training and assistance for high-nutrient agriculture (vegetables, tubers, legumes), as well as integrated farming systems (agroforestry), that specifically aimed at pregnant women and families with children under 2 years old. (3) “Anti-stigma family planning social campaigns through community and religious leaders to increase public acceptance of family planning programs and Increased premarital and pre-pregnancy counseling, emphasizing the importance of maternal nutritional preparedness before pregnancy”. (4) Logistics support and distribution of high-protein foods (eggs, fish, powdered milk) to highland areas. Microfinance and cooperative programs to help manage small local food businesses.

Funding info

No funding was received for this study.

CRediT authorship contribution statement

Dina Fitriana Rosyada: Writing – review & editing, Writing – original draft, Visualization, Methodology, Formal analysis, Data curation, Conceptualization. Toto Sudargo: Writing – review & editing, Writing – original draft, Formal analysis, Data curation. Indah Kartika Murni: Writing – review & editing, Writing – original draft, Visualization, Supervision, Methodology, Data curation, Conceptualization.

Ethics approval and consent to participate

Review articles do not require ethical approval.

Declaration of competing interest

All authors declare that they have no conflicts of interest.

Data availability

Data extraction for all included studies is presented in the supplementary file Included.SR.MA.DialoguesInHealth.xlsx.