Prevalence of stunting, wasting, and underweight among children aged 6 to 59 months and adolescents in pastoral and semi-pastoral areas of Ethiopia: a systematic review and meta-analysis Authors: Zekarias Markos, Yegle Asrat Tekletsadik, Ayenew Kassegn, Esubalew Muluneh Aligaz Published: 2025 ABSTRACT AbstractBackgroundIn Ethiopia’s pastoral areas, undernutrition continues to be a serious public health concern that has an impact on children’s development, growth, and long-term health. The purpose of this study was to estimate the pooled prevalence of stunting, wasting, and underweight in pastoral and semi-pastoral areas of Ethiopia.MethodsA thorough search was carried out using several important databases, such as Henari, PubMed, Google Scholar, Scopus, Web of Science, AJOL, and institutional repositories of Ethiopian universities. After removing duplicates, 800 records were left out of the 1,510 that were first identified. After screening by title, abstract, and full-text, 13 studies were found to be eligible and added to the final meta-analysis. A random-effects model was used to compute the pooled prevalence estimates.ResultsThis meta-analysis included 13 studies encompassing 7,769 children from pastoral and semi-pastoral regions of Ethiopia. The pooled prevalence of stunting was 39.3% (95% CI: 31.6%–47.1%), indicating a high burden of chronic malnutrition. The prevalence of wasting was 13.0% (95% CI: 9.0%–17.0%), while underweight affected 20.5% (95% CI: 14.7%–26.3%) of children. Subgroup analysis revealed regional variations, with the highest stunting rates observed in Benishangul-Gumuz and Afar regions. Considerable heterogeneity was found across studies (I² > 95%), and publication bias was detected for all three nutritional outcomes. Adjusted estimates using trim-and-fill analysis showed slightly lower prevalence’s.ConclusionUndernutrition remains a critical public health issue among children in Ethiopia’s pastoral and semi-pastoral areas, with stunting and underweight affecting nearly two in five and one in five children, respectively. These rates surpass both national and global averages, underscoring the urgent need for region-specific nutrition interventions and policies that address the unique vulnerabilities of these communities.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40795-025-01210-2.Keywords: Ethiopia, Pastoral, Semi pastoral, Stunting, Underweight, Wasting Background Malnutrition encompasses both under- and over nutrition. Undernutrition includes being underweight, wasted, or stunted. Undernutrition can result from either long-term or short-term inadequate nutrient consumption. Furthermore, it might be driven by food insecurity or an increase in disease prevalence [1]. In Ethiopia, between 29.2% and 42.7% of school-age children suffer from stunting, while the prevalence of wasting ranges from 8% to 21% [2, 3]. Factors associated with undernutrition in Ethiopia include age, sex, household wealth quintiles, school enrollment, dietary diversity score, recent fever, consumption of animal-based foods, household ownership and consumption of livestock products, education level of the primary food preparer, number of children in the household, and literacy status of the head of household [2, 3]. Children in poor nations have a high risk of sickness and mortality as a result of malnutrition. In low-income countries, malnutrition greatly increases the risk of childhood illness and mortality, and it can impair cognitive performance. These effects extend to household, community, and national levels, ultimately impacting economic productivity [4]. Malnutrition at this age exposes children to major short- and long-term health issues since school-age children grow up to 50% of their adult weight, 20% of their adult height, and 50% of their adult bone mass [5]. Agricultural and labor productivity has been found to be correlated with height, and a 1.4% decrease in production is linked to every 1% reduction in adult height caused by childhood stunting. Better health and nutrition enhance children’s mental and physical capacities, which in turn enhances their earning potential in the future [6]. Improved health and nutrition enhance children’s mental and physical capacities, thereby increasing their future earning potential. Animal-based foods, which are rich in essential minerals, proteins, and energy, were thought to be more readily available to Ethiopian pastoral children. As a result, they are thought to have a better nutritional status than children in rural communities [7]. But the truth is more nuanced. Despite the apparent nutritional potential, seasonal food insecurity, livestock productivity variability, environmental shocks, and restricted access to basic nutrition and health services make pastoralist children susceptible to undernutrition [7]. Pastoralism plays a vital role in household livelihoods, contributing more than half of household income and food consumption through a combination of livestock rearing, crop farming, and sustainable resource management [8]. Pastoralists use adaptive strategies such as resource conservation, livelihood diversification, mobility, and collective social organization. However, increasing pressure on these adaptive mechanisms has made the nutritional status of children in these communities a pressing public health concern requiring targeted research and policy action [9, 10]. Several studies have examined the prevalence of stunting, wasting, and underweight in Ethiopia, but to our knowledge, no previous meta-analysis has comprehensively focused on pastoral and semi-pastoral areas or separately examined children aged 6–59 months and adolescents in these contexts. Some earlier meta-analyses have included studies from pastoral settings, but they did not specifically analyze these regions as a distinct population subgroup. This unique focus is critical, as pastoralist communities face distinct environmental, cultural, and economic challenges that influence child nutrition outcomes. Understanding the pooled prevalence and determinants of undernutrition in Ethiopia’s pastoral and semi-pastoral areas is essential for guiding resource allocation, designing context-specific interventions, and strengthening health and nutrition programs. Therefore, this systematic review and meta-analysis aims to estimate the pooled prevalence of stunting, wasting, and underweight, and to identify their determinants among children aged 6–59 months and adolescents living in these settings. Methodology Study design This study is a systematic review and meta-analysis designed to estimate the pooled prevalence of wasting, stunting, and underweight at the pastoral regions of Ethiopia. The review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines2020 to ensure methodological rigor and transparency. The protocol was registered at PROSPERO database with protocol number of PROSPERO 2025 CRD420251059708. https://www.crd.york.ac.uk/PROSPERO/view/CRD420251059708. Search strategy Several electronic databases, such as Henari, PubMed, Google Scholar, Scopus, Web of Science, African Journals Online (AJOL), and institutional repositories of Ethiopian universities, were used to do a thorough literature search. Studies released between Relevant articles were identified using a combination of free-text keywords and Medical Subject Headings (MeSH) terms. Search phrases included terms such as “wasting,” “acute malnutrition,” “stunting,” “chronic malnutrition,” “undernutrition,” “underweight,” “prevalence,” “magnitude,” “pastoral,” “pastoralist,” “semi-pastoral,” “children under five,” “adolescent,” and “Ethiopia.” Boolean operators (AND, OR) were employed to effectively combine these search terms. In addition to electronic database searches, manual screening of reference lists from included studies and relevant reviews was performed to identify any additional eligible studies. The inclusion and exclusion criteria were framed using the CoCoPop (Condition, Context, Population) framework for systematic reviews of prevalence data. The justification for using CoCoPop over PICO/PECO is that our review focused on estimating the prevalence of a condition (malnutrition) in a specific population and context, rather than evaluating an intervention or exposure’s effect [11]. Condition (Co): Stunting, wasting, and underweight (forms of malnutrition). Context (Co): Pastoral and semi-pastoral areas of Ethiopia. Population (Pop): Preschool and school-aged children and adolescents aged 6 months to 19 years. Eligibility criteria The inclusion criteria for this review comprised observational studies that reported the prevalence of wasting, stunting, or underweight among children. Eligible studies were conducted specifically in the pastoral regions of Ethiopia, including Afar, Somali, and parts of Oromia and SNNPR. The study populations included children aged between 6 months and 19 years. Additionally, studies were required to provide sufficient data to allow calculation of prevalence estimates or proportions. Only studies published in English were considered. We included studies reporting prevalence estimates of stunting, wasting, and underweight among children and adolescents. Studies were excluded if they did not provide clear prevalence data or were conducted outside the designated pastoral regions. Furthermore, reviews, editorials, case reports, qualitative studies, and duplicate publications were excluded to maintain the focus on primary quantitative evidence. Studies lacking sufficient data to calculate prevalence or effect estimates were excluded. Study selection process The study selection process followed the PRISMA 2020 flow diagram (Fig. 1). All records identified from the database searches were imported into Zotero reference management software for deduplication. The deduplicated list of titles and abstracts was then screened by two independent reviewers (ZM and YA) against the eligibility criteria. Studies that clearly did not meet the criteria were excluded. The full text of the remaining articles was retrieved and assessed in detail for eligibility by the same two reviewers. The study selection process was conducted manually by the reviewers; no automation tools were used for screening or selection. Any disagreements between the reviewers at any stage of the selection process were resolved through discussion and consensus, with the involvement of a third reviewer (EM) when necessary. The reasons for excluding studies at the full-text stage were documented. PRISMA flow diagram Data extraction Two independent reviewers extracted data using a pre-defined data extraction form in Exce(ZM and YA)l. Extracted information included: Author(s) and year of publication, Study location and region, Study design and sample size, Age group of participants Reported prevalence of wasting, stunting, and underweight. Quality assessment The methodological quality of the included studies was assessed using the Joanna Briggs Institute (JBI) critical appraisal checklist for prevalence studies(Supplementary file 1). Each study was rated as high([7]– [8]), moderate([5]– [6]), or low quality( 641) and areas (Pastoral and semi pastoral). Publication bias was evaluated using funnel plots and Egger’s regression test, with a p-value < 0.05 indicating potential bias. Subgroup analysis was conducted for determining factors of under nutrition indicators. Result Study selection and characteristics A total of 1,510 records were identified through electronic database searches (Hinari, Google Scholar, Scopus, PubMed, and AJOL and other). The study populations encompassed the predefined age range of 6 months to 19 years, including adolescents. After deduplication was performed using Zotero software, 800 records remained. Of these, 500 records were excluded based on title and abstract review. 300 full-text articles were assessed for eligibility. After full-text review, 283 articles were excluded with documented reasons (e.g., irrelevant outcomes, non-pastoral settings, or insufficient data). Finally, thirteen studies met the inclusion criteria and were included in the systematic review and meta-analysis (Fig. 1). In addition to peer-reviewed articles, relevant grey literature, including theses and dissertations, was retrieved from Ethiopian institutional repositories. Bizuayehu Y’s [12] study was obtained from the Kidest Mariam College Repository, and Shiferaw et al.’s [13] work was accessed from the DSpace Institutional Repository. A total of thirteen studies were included in this meta-analysis to assess the prevalence and associated factors of stunting, wasting and underweight at pastoral and semi pastoral areas of Ethiopia [7, 12–23]. All included studies were community-based cross-sectional in design and conducted in various pastoral and semi-pastoral regions of Ethiopia, such as the Somali Regional State, Afar, and South Omo zones, between 2013 and 2023. The total number of participants across the 13 studies was 7,769. All studies assessed stunting using standard anthropometric measures. Stunting: Height-for-age Z-score (HAZ) < −2 standard deviations (SD) from the WHO median, indicating chronic malnutrition. Stunting: Height-for-age Z-score (HAZ) < −2 standard deviations (SD) from the WHO median, indicating chronic malnutrition. Wasting: Weight-for-height Z-score (WHZ) < −2 SD, reflecting acute malnutrition. Wasting: Weight-for-height Z-score (WHZ) < −2 SD, reflecting acute malnutrition. Underweight: Weight-for-age Z-score (WAZ) < −2 SD, indicating a combination of acute and/or chronic malnutrition. Underweight: Weight-for-age Z-score (WAZ) < −2 SD, indicating a combination of acute and/or chronic malnutrition. Table 1 summarizes the key characteristics of 13 studies conducted in various pastoral and semi-pastoral regions of Ethiopia, focusing on the nutritional status of children aged 6–59 months and adolescents aged 10–19 years. The included regions—SNNPR, Oromia, Benishangul-Gumuz, Afar, and Somalia—are predominantly rural and home to vulnerable populations often affected by food insecurity and limited healthcare access. Table 1Study characteristicsAuthorYearsamplestuntingwastingUnderweightRegionAge groupAreasKelkay & Seid (2019) [14]2019339186120SNNPR6–59 monthSemi PastoralTadesse et al. (2018) [15]20186451776999SNNPR6–23 monthSemi PastoralWana et al. (2023) [16]2023409180SNNPR6–59 monthSemi PastoralMegersa et al. (2021) [7]202153822153153Oromia6–59 monthSemi PastoralBizuayehu (2013) [12]20138182937595Oromia6–59 monthSemi PastoralShiferaw (2020) [13]202056623471154Benshangul Gumuz6–59 monthPastoralBeyene (2020) [23]20204222161968Benshangul Gumuz6–59 monthPastoralHadush et al. (2021) [18]2021736196Afar10–19 yearsPastoralGebre et al. (2019 [21]2019840362136208Afar6–59 monthPastoralToma et al. (2023) [19]202371743065Afar6–59 monthPastoralEngidaw & Gebremariam (2019) [20]201941540Somalia10–19 yearsPastoralShine et al. (2017) [17]2017770246Somalia6–59 monthPastoralGebreayohanes & Dessie (2022) [22]2022554248Afar6–59 monthPastoral Study characteristics The sample sizes of the studies range from 339 to 840, reflecting moderate to large-scale surveys across diverse communities. Most studies focused on children under five, while a few (Hadush et al. [18]., Engidaw & Gebremariam [20] targeted adolescents, highlighting a growing interest in malnutrition beyond early childhood. Stunting (chronic malnutrition) was assessed in all but one study and showed consistently high figures. The highest number of stunted cases was reported by Toma et al. [19]. (2023) with 430 cases out of 717 children in Afar, while Bizuayehu [12] (2013) and Gebre et al. (2019) [21] also reported high counts (293 and 362 cases, respectively). Wasting (acute malnutrition) data were reported in eight studies. Kelkay & Seid [14] (2019) and Gebre et al. (2019) [21] reported some of the highest absolute numbers (120 and 136, respectively), indicating the presence of ongoing or recent nutritional stress. Underweight data were less consistently reported but are available in eight studies. The highest number of underweight children was recorded by Gebre et al. (2019) (208 cases) [21], followed by Shiferaw (2020) [13](154 cases). The inclusion of both preschool children and adolescents provides a broader picture of the malnutrition burden in the region. Pooled prevalence of wasting, stunting and underweight at pastoral and semi pastoral areas of Ethiopia This meta-analysis evaluated the prevalence of undernutrition in children in Ethiopia’s pastoral and semi-pastoral areas shows high levels of stunting, wasting, and underweight with considerable heterogeneity among studies. The analysis employed a random-effects inverse-variance model, accounting for variability across studies by applying the DerSimonian-Laird method for estimating between-study variance. The pooled prevalence of wasting, based on 8 studies and using the random-effects model, was 13% (95% CI: 9% to 17%), indicating a considerable burden of acute malnutrition. However, there was substantial heterogeneity among the studies (p < 0.001; I² = 95.4%. (Fig. 2) Prevalence of wasting The pooled prevalence of stunting among children, based on 13 studies and using a random-effects model, was estimated at 39% (95% Confidence Interval: 31% to 47.0%), indicating a high burden of chronic malnutrition in the populations studied. However, there was substantial heterogeneity among the included studies (p < 0.001; I² = 98.3%), suggesting that most of the variability in stunting prevalence across studies is due to real differences rather than chance. Figure 3. Prevalence of stunting The pooled prevalence of underweight was found to be 20% with a 95% confidence interval (CI) ranging from 15% to 25%. This indicates that, on average, approximately one in five children included in these studies was underweight, although the true prevalence may vary widely depending on the population and context. (Fig. 4). Pooled prevalence of underweight Sub group analysis Subgroup analysis of wasting prevalence by region, age group, sample size and areas This subgroup analysis offers a detailed breakdown of the pooled prevalence of wasting among children and adolescents across different Ethiopian regions, age groups, and sample categories. The overall pooled effect size for wasting across all included studies is 13.0% (95% CI: 9.0%–17.0%), highlighting a significant burden of acute malnutrition. However, the analysis also reveals considerable heterogeneity among studies (I² = 95.4%, p < 0.001), suggesting variations in population characteristics, study design, or implementation contexts.(Table 2). Table 2Subgroup analysis -WastingSubgroupCategoryPooled Effect Size (95% CI)Heterogeneity (I²)p-value (Effect) Region SNNPR0.229 (− 0.013, 0.471)98.7%0.063Oromia0.094 (0.079, 0.110)0.0%0.000Benishangul-Gumuz0.085 (0.006, 0.163)95.4%0.035Afar0.126 (0.056, 0.196)94.6%0.000 Age Group 10–19 years0.091 (0.070, 0.112)—0.000≤ 5 years0.137 (0.089, 0.184)96.0%0.000 Sample Category 6410.112 (0.081, 0.143)87.1%0.000 Areas Pastoral0.105 0.055 0.15594.7%0.000Semi Pastoral0.158 0.085 0.23296.7%0.000 Subgroup analysis -Wasting SNNPR recorded the highest pooled prevalence of wasting at 22.9%, but with a broad confidence interval (1.3% to 47.1%) and extremely high heterogeneity (I² = 98.7%, p = 0.063), indicating substantial variation across studies. Oromia had a lower and more precise estimate of 9.4% (95% CI: 7.9%–11.0%), with no heterogeneity (I² = 0.0%, p < 0.001), suggesting strong consistency and reliability. Benishangul-Gumuz reported a pooled prevalence of 8.5% (95% CI: 0.6%–16.3%) with very high heterogeneity (I² = 95.4%), reflecting variability between the two included studies. Afar showed a moderate wasting prevalence of 12.6% (95% CI: 5.6%–19.6%) with high heterogeneity (I² = 94.6%, p < 0.001), indicating inconsistencies across studies. Children under 5 years exhibited a notably higher pooled prevalence of wasting (13.7%, 95% CI: 8.9%–18.4%), with high heterogeneity (I² = 96.0%, p < 0.001), suggesting acute malnutrition is more prevalent and varied among younger children. Adolescents (10–19 years) had a lower and more consistent prevalence of 9.1% (95% CI: 7.0%–11.2%, p < 0.001), with a tighter confidence interval, indicating relative consistency in findings. Sample category ≤ 641 showed a pooled wasting prevalence of 15.3% (95% CI: 6.2%–24.3%), with very high heterogeneity (I² = 97.7%), reflecting substantial variability within this group. Sample category > 641 had a lower and more consistent pooled prevalence of 11.2% (95% CI: 8.1%–14.3%), with reduced heterogeneity (I² = 87.1%). In the subgroup analysis for wasting, the pooled prevalence was 10.5% (95% CI: 5.5%–15.5%) in pastoral areas and 15.8% (95% CI: 8.5%–23.2%) in semi-pastoral areas, indicating a higher burden in semi-pastoral communities. Subgroup analysis of stunting prevalence by region, age group, and sample size This subgroup analysis provides insights into the pooled prevalence of stunting, a chronic form of malnutrition, across various Ethiopian regions, age groups, and sample categories. The overall pooled prevalence of stunting across all studies was 39.3% (95% CI: 31.4%–47.2%), underscoring the significant burden of chronic malnutrition in the country. However, substantial heterogeneity (I² = 98.3%, p < 0.001) suggests notable variations in study contexts or populations (Table 3). Table 3Subgroup Analysis - stuntingSubgroup VariableCategoryPooled Prevalence (95% CI)Heterogeneity (I²)Subgroup p-value Region SNNPR0.420 (0.256–0.584)97.5%0.000Oromia0.383 (0.331–0.434)73.6%0.000Benishangul-Gumuz0.462 (0.365–0.558)89.5%0.000Afar0.436 (0.296–0.576)98.4%0.00 0Somalia0.208 (− 0.011–0.426)99.0%0.063 Age Group 10–19 years0.433 (0.106–0.760)99.5%0.009≤ 5 years0.386 (0.305–0.466)98.0%0.000 Sample Category 6410.385 (0.299–0.471)97.7%0.000 Areas Pastoral0.393 (0.314–0.472)95.3%0.000Semi Pastoral0.405 (0.321- 0.489)98.9%0.000 Subgroup Analysis - stunting Benishangul-Gumuz recorded the highest pooled prevalence of stunting at 46.2% (95% CI: 36.5%–55.8%), signaling a severe chronic malnutrition issue despite high heterogeneity (I² = 89.5%). Afar followed closely at 43.6% (95% CI: 29.6%–57.6%), but extreme heterogeneity (I² = 98.4%) indicates inconsistent findings across studies. SNNPR exhibited a prevalence of 42.0% (95% CI: 25.6%–58.4%), accompanied by high variability (I² = 97.5%). Oromia reported a slightly lower but still concerning prevalence of 38.3% (95% CI: 33.1%–43.4%), with relatively lower heterogeneity (I² = 73.6%). Somali Region had the lowest pooled prevalence at 20.8% (95% CI: −1.1%–42.6%), though this estimate is statistically non-significant (p = 0.063) and highly variable (I² = 99.0%), likely due to limited or inconsistent data. Adolescents (10–19 years) had a slightly higher pooled prevalence of 43.3% (95% CI: 10.6%–76.0%), but the wide confidence interval and high heterogeneity (I² = 99.5%) suggest inconsistency and potential instability in estimates due to fewer studies in this category. Children aged ≤ 5 years had a pooled prevalence of 38.6% (95% CI: 30.5%–46.6%), with very high heterogeneity (I² = 98.0%). However, the estimate remains statistically significant (p < 0.001), reinforcing the widespread burden of stunting in young children. Sample category 641, which recorded 38.5% (95% CI: 29.9%–47.1%). Despite minimal differences in prevalence rates, both categories exhibited high heterogeneity (I² > 97%), suggesting that sampling characteristics alone do not account for much of the variability in stunting rates. In the subgroup analysis for stunting, the pooled prevalence was 39.3% (95% CI: 31.4%–47.2%) in pastoral areas and 40.5% (95% CI: 32.1%–48.9%) in semi-pastoral areas, with high heterogeneity in both groups (I² = 95.3% and 98.9%, respectively; p = 0.000). Subgroup analysis of underweight prevalence in ethiopia’s pastoral regions This subgroup analysis examines the pooled prevalence of underweight among children across different Ethiopian regions, age groups, and sample categories, highlighting significant geographic and methodological variations in malnutrition patterns.(Table 4). Table 4Subgroup analysis - underweightStratifierSubgroup/StudyEffect Size [95% CI]P VALUEI² (%) By Age Group ≤ 5 years (all0.205(0.147, 0.263)0.00195.5 By Region SNNPR0.153 [0.126, 0.181]OROMIA0.200 [0.035, 0.364]0.00198.2BENISHANGUL-GUMUZ0.216 [0.108, 0.325]0.00194.6AFAR0.248 [0.218, 0.277] By Sample Category 6410.172 [0.096, 0.248]0.00196.0 Areas Pastoral0.227 [0.164 0.290]0.00090.6%Semi Pastoral0.183 [0.096 - 0.271]0.00096.4% Subgroup analysis - underweight The pooled prevalence of underweight among children aged ≤ 5 years was 20.5% (95% CI: 14.7%–26.3%), reflecting a substantial burden of malnutrition in this vulnerable population. The analysis also revealed very high heterogeneity (p < 0.001; I² = 95.5%), indicating significant variability among studies, likely due to differences in study settings, measurement techniques, or population characteristics. Afar had the highest pooled prevalence of underweight at 24.8% (95% CI: 21.8%–27.7%), indicating a serious nutritional challenge. Notably, no heterogeneity was detected, implying consistent findings across studies. Benishangul-Gumuz followed with a prevalence of 21.6% (95% CI: 10.8%–32.5%), with high heterogeneity (I² = 94.6%), reflecting differences between included studies. Oromia reported 20.0% (95% CI: 3.5%–36.4%) but exhibited very high heterogeneity (I² = 98.2%), suggesting substantial variation in underweight prevalence within the region. SNNPR had the lowest prevalence at 15.3% (95% CI: 12.6%–18.1%), with no reported heterogeneity, indicating consistent findings. The between-region variation was statistically significant (p < 0.001), confirming meaningful differences in underweight prevalence across geographic locations. Sample category 641 showed a lower prevalence of 17.2% (95% CI: 9.6%–24.8%) but still exhibited very high heterogeneity (I² = 96.0%). However, no statistically significant difference was observed between the two sample categories (p = 0.228), indicating that sample-related characteristics did not meaningfully influence variations in underweight prevalence. In the subgroup analysis for underweight, the pooled prevalence was 22.7% (95% CI: 16.4%–29.0%) in pastoral areas and 18.3% (95% CI: 9.6%–27.1%) in semi-pastoral areas, with substantial heterogeneity in both groups (I² = 90.6% and 96.4%, respectively; p = 0.000). Publication bias analysis Funnel plot The funnel plots shown in supplementary file 2 assess potential publication bias in the meta-analysis of the pooled prevalence of malnutrition indicators (a) stunting, (b) wasting, and (c) underweight among children in the pastoral and semi pastoral regions of Ethiopia. Visual inspection of the plots shows that the distribution of studies is not symmetrical around the central line for all three indicators, suggesting evidence of publication bias.(Fig. 5). Funnel Plots Assessing Publication Bias for (A) Wasting, (B) Stunting, and (C) Underweight Eagers regression Egger’s test detected significant small-study effects across all malnutrition indicators among Ethiopian children. For stunting, the bias coefficient was 106.95 (p = 0.014) and the slope − 3.77 (p = 0.001), indicating potential publication bias. In wasting, both the bias (119.18; p = 0.035) and slope (–3.11; p = 0.000) were significant, suggesting smaller studies may be inflating prevalence estimates. Similarly, underweight showed evidence of bias (bias = 95.07, p = 0.046; slope = − 3.16, p = 0.003), reinforcing concerns of small-study effects in the meta-analysis. Table 5. Table 5Egger’s test for Small-Study effectsNutritional StatusBias CoefficientP-value (Bias)SlopeP-value (Slope) Stunting 106.950.014−3.770.001 Wasting 119.180.035−3.110.000 Underweight 95.070.046−3.160.003 Egger’s test for Small-Study effects Trim and fill analysis To account for this bias, a Trim and Fill analysis was conducted to adjust the pooled prevalence estimates. The revised figures demonstrated a downward shift across all indicators: Stunting was initially reported at 39.3% (95% CI: 31.4%–47.2%), but after adjustment, the estimate dropped to 31.1% (95% CI: 23.0%–39.1%), with five imputed studies contributing to this correction. Wasting originally had a prevalence of 13.0% (95% CI: 9.0%–17.0%), but the adjusted figure fell to 8.5% (95% CI: 4.1%–12.9%), incorporating three imputed studies. Underweight saw a modest decline from 20.5% (95% CI: 14.7%–26.3%) to 16.5% (95% CI: 10.5%–22.5%), with two imputed studies involved in the adjustment (Table 6). Table 6Trim and fill analysisDatasetOriginal Pooled Effect (Random)Adjusted Pooled Effect (Random)Number of StudiesNumber of Filled (Imputed) Studies Stunting 0.393 (0.314, 0.472)0.311 (0.230, 0.391)135 Wasting 0.130 (0.090, 0.170)0.085 (0.041, 0.129)83 underweight 0.205 (0.147, 0.263)0.165 (0.105, 0.225)62 Trim and fill analysis Sensitivity analysis This figure presents a leave-one-out sensitivity analysis conducted to examine the influence of individual studies on the pooled prevalence estimates of malnutrition indicators in children from pastoral and semi pastoral regions of Ethiopia. Each line represents the overall pooled estimate after omitting one study at a time. Figure 6. Leave-One-Out Sensitivity Analysis for Pooled Prevalence of (A) Stunting, (B) Wasting, and (C) Underweight the analyses, for wasting, for underweight and stunting the pooled estimate remained stable across all iterations, with no single study significantly altering the overall result. The narrow fluctuation in effect sizes and confidence intervals confirms the reliability and stability of the meta-analysis results across the three indicators. Discussion This systematic review and meta-analysis, including 13 studies with a total of 7,769 children from pastoral and semi-pastoral regions of Ethiopia, reveals alarmingly high levels of undernutrition. The pooled prevalence estimates were 39.3% for stunting (95% CI: 31.6%–47.1%), 13% for wasting (95% CI: 9%–17%), and 20% for underweight (95% CI: 15%–25%). These results indicate a serious public health concern, with prevalence rates exceeding both the national averages reported in the 2019 Ethiopian Demographic and Health Survey (37% stunting, 7% wasting) [24] and global estimates reported by WHO in 2022 (22% stunting) [25]. The stunting prevalence reported here is markedly higher than that observed in other countries. For instance, meta-analyses in North Africa and Iran found pooled stunting prevalence’s of 23.5% [26] and 12.6% [27], respectively. Similarly, an earlier Ethiopian meta-analysis reported a lower prevalence of 23.1% [28]. However, our findings are in line with more recent Ethiopian studies. SA Abdilahi et al. [29]. reported a prevalence of 43%, while A. Kassaw et al. [30]. found a stunting rate of 35.0%, underscoring the ongoing burden of chronic malnutrition in Ethiopia. Adongo et al. also reported a high stunting prevalence of 38.1% at pastoral region of kenya, which is consistent with our findings [31]. The pooled prevalence of wasting in this review was 13%, indicating a significant burden of acute malnutrition. This finding aligns with studies conducted in Iran, which reported a wasting prevalence of 12.5% [32] among children aged 6–14 years, and with a prior Ethiopian meta-analysis that found 10.9% among children under two. However, our result is lower than other Ethiopian studies, such as the 22% [28] reported by Demissie et al. and 17.7% reported by Assemie et al. [33]. On the pastoral region of kenya Adongo et al. [31]. found a similar wasting prevalence of 18.5%. Conversely, it is higher than the 7.9% [26]wasting prevalence observed in Iranian children under five. These variations may reflect differences in study populations, age groups, seasonal effects, or local socio-economic and environmental conditions. A study by Okidi et al. found that wasting prevalence varied with agroecology, and household expenses were associated with a reduced likelihood of wasting in both pastoral and agricultural zones [34]. Regarding underweight, our pooled estimate of 20% also reflects a substantial nutritional burden. This finding is consistent with the 27.5% prevalence reported in another Ethiopian study [35] but higher than that of Assemie et al., who reported 18.2% [33]. Comparable underweight prevalence of 23.0% noted by Adongo et al. [31]. It also exceeds the 12.9% prevalence found among Iranian children [26]. The observed variation may be due to factors such as regional disparities, ongoing drought and conflict, differences in food security, child feeding practices, maternal education, and household income. Okidi et al. found that being a male child was associated with a higher likelihood of being underweight in pastoral agroecologies additionally, cereal storage in both sacks and granaries was linked to a reduced likelihood of underweight in the pastoral zone [34]. The substantial heterogeneity observed across studies (I² > 95% for all three outcomes) suggests that contextual factors play a major role in influencing nutritional outcomes in these communities. Children in Ethiopia’s pastoralist regions face unique challenges, including limited access to health services, poor infrastructure, recurrent droughts, food insecurity, and social marginalization, all of which likely contribute to the high prevalence of undernutrition. Conclusion and recommendation Undernutrition remains a critical public health issue among children in Ethiopia’s pastoral and semi-pastoral areas, with stunting and underweight affecting nearly two in five and one in five children, respectively. These rates surpass both national and global averages, underscoring the urgent need for region-specific nutrition interventions and policies that address the unique vulnerabilities of these communities. Strength of the study Nutrition programs must be context-specific, reflecting the unique agroecological and cultural realities of these communities. Strengthening community-based services, including regular growth monitoring and timely treatment of malnutrition, is essential to reach remote and mobile populations. Efforts should also address underlying determinants such as food security, access to safe water, maternal education, and healthcare. Multi-sectoral collaboration between health, agriculture, education, and social protection sectors, coupled with ongoing nutrition surveillance and operational research, will be critical to guide and evaluate interventions. Limitations of the study Limitations Despite its contributions, this systematic review and meta-analysis has limitations. Including substantial heterogeneity across the included studies, and evidence of small-study effects was detected. To address homogeneity we perform subgroup analysis and to address publication bias we performed a trim-and-fill. Supplementary Information Supplementary Material 1 Acknowledgments Guarantors All authors. Authors’ contributions Zekarias Markos (ZM) conceived and designed the study, coordinated the review process, performed the data analysis, and drafted the manuscript. Ayenew Kassegn (AK) contributed to the development of the search strategy, study selection, and data extraction. Yegle Asrat (YA) participated in quality assessment of the included studies, interpretation of results, and manuscript revision. Esubalew Muluneh (EM) contributed to statistical analysis, data interpretation, and critically reviewed the final manuscript for intellectual content. All authors read and approved the final version of the manuscript. Funding No financial aid. Data availability All data generated or analyzed during this study are included in this published article(Table 1) and its supplementary information files. The datasets used for the meta-analysis were extracted from previously published studies, which are cited in the reference list. If additional information is required, it can be made available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate Not Applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All data generated or analyzed during this study are included in this published article(Table 1) and its supplementary information files. The datasets used for the meta-analysis were extracted from previously published studies, which are cited in the reference list. If additional information is required, it can be made available from the corresponding author upon reasonable request.