Assessing the Impact of Agricultural Droughts on Child Undernutrition in Varying Crop-Growing Periods and Agricultural Land Use: Analysis across 33 Sub-Saharan African Countries

Assessing the Impact of Agricultural Droughts on Child Undernutrition in Varying Crop-Growing Periods and Agricultural Land Use: Analysis across 33 Sub-Saharan African Countries | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Assessing the Impact of Agricultural Droughts on Child Undernutrition in Varying Crop-Growing Periods and Agricultural Land Use: Analysis across 33 Sub-Saharan African Countries Anna Dimitrova, Alexander Gershunov, Raya Muttarak, Francois Rerolle, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7765713/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Countries in sub-Saharan Africa (SSA) face persistent challenges in addressing child malnutrition, with approximately one-third of children under the age of five in the region experiencing chronic undernourishment. Previous studies have demonstrated that agricultural droughts contribute to child undernutrition; however, they have not adequately accounted for variations in agricultural land specialization and corresponding crop growing periods. In this study, we aim to address these heterogeneities by identifying vulnerable communities and specific exposure periods. We utilize anthropometric data for over 300,000 children ages 1 to 5 collected in nationally representative household surveys in 33 countries in SSA over three decades (1990-2022) and employ a novel approach that integrates detailed crop calendar information (crop planting and harvesting dates) with high-resolution agricultural and climatological datasets to detect droughts during specific crop-growing periods. The Standardized Precipitation Evapotranspiration Index (SPEI), a multi-scalar drought index, is used to measure the intensity and spatial distribution of droughts. We explore community-level heterogeneities determined by agricultural land specialization (crop farming or pastoralism) and the types of crops cultivated. Additionally, socioeconomic factors underlying vulnerability are explored. Our analysis reveals a significant association between agricultural droughts and the risk of child undernutrition among communities where agricultural land-use practices are dominated by crop farming. Specifically, droughts occurring during the growing seasons for grains and oilseeds exhibit a strong association with child undernutrition. Furthermore, residing in a rural area, being employed in agriculture, and belonging to a lower socioeconomic class are found to amplify the risk of undernutrition related to agricultural droughts among these communities. The findings presented in this study call for urgent action to improve drought monitoring and response in SSA where the risks to child health posed by global warming are considerable. Under climate change, the severity and frequency of extreme weather and climate events, including droughts, are projected to increase. This will place millions of children at risk of malnutrition unless timely action plans are taken to improve food security in the region. Earth and environmental sciences/Environmental social sciences/Climate-change impacts/Environmental health Earth and environmental sciences/Climate sciences/Climate change/Climate-change impacts/Environmental health drought agriculture undernutrition food security sub-Saharan Africa Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Climate conditions and stable precipitation regimes are critical for livelihoods in sub-Saharan Africa (SSA) given the high reliance on subsistence farming in the region 1 . Smallholder farmers constitute over half of the population in SSA and agriculture accounted for 17% of the region's GDP in 2022 2 . Hence, the loss of crop and livestock production due to precipitation variability poses a major threat to food security and economic well-being in the region 3 – 5 . The increased incidence of agricultural droughts due to climate change and its potential impact on child health is particularly concerning. Agricultural droughts, as opposed to meteorological droughts, can be driven by warmer temperatures in addition to precipitation deficits. Therefore, anthropogenic climate change can exacerbate agricultural drought by enhancing water demand directly via warming 6 as well as via precipitation regime change 7 , 8 . Droughts can impact food security through multiple channels, directly by reducing crop yields and the amount of food available to households, and indirectly by eroding the economic resources of smallholder farmers and agricultural laborers and causing spikes and volatility in food prices 9 , 10 . Droughts can also aggravate underlying socioeconomic conditions that impact food security, such as economic instability, conflict, and forced displacement 11 , 12 . A growing body of literature shows that climate variability and extreme climate events, including droughts, are associated with an increased risk of experiencing food insecurity and malnutrition among households in low- and middle-income countries (LMICs) 13 – 15 . In fact, malnutrition has been recognized as one of the main health risks posed by global warming 16 , 17 . Young children are at heightened risk of undernutrition in households experiencing food insecurity. Children need sufficient, nutritious, and safe foods and not meeting these nutritional needs can seriously affect their long-term health and well-being 18 . Besides being one of the main causes of death for children under 5 years of age 19 , 20 , undernutrition has been associated with loss of physical growth potential, diminished neurodevelopmental and cognitive function, and learning challenges in the classroom 21 , 22 . Children who are too short for their age, an indication of chronic undernutrition known as stunting, were found to have an increased risk of chronic health conditions and poorer socioeconomic status in adulthood 23 , 24 . Malnourished children are also more susceptible to infectious diseases such as diarrhea and acute respiratory infections because of their compromised immunity 20 , which has been associated with poor intake and absorption of nutrients and further growth retardation 25 . The societal implications of child undernutrition can be substantial as there is evidence that poor development due to stunting can be transmitted to the subsequent generation of children 26 , which can perpetuate poverty and undermine the economic growth of nations where child undernutrition is widespread 27 . Countries in SSA have some of the highest levels of child undernutrition in the world, with 31.5% of children under 5 in the region being stunted in 2022 28 . Despite some reduction in the prevalence of child stunting in SSA over the past decade, dropping 36.2% in 2012 28 , progress has been slow, possibly due to the long-lasting effects of the COVID-19 pandemic 29 , recurrent harvest failures due to the impacts of climate change on the frequency of agricultural droughts 3 , 4 , and other compounding factors 30 . While droughts are known to pose a serious threat to food security 3 – 5 , the precise impact of these climate shocks on child health and the underlying mechanisms remain understudied. A recent systematic review 31 of 27 studies conducted in 20 different LMICs found that there is a dearth of rigorous research and a small number of observational studies that link droughts to undernutrition in children. While some studies found a positive association, no conclusive evidence was established in the systematic review due to multiple reasons, including small sample sizes, high risk of bias, and the lack of comparable estimates across studies 31 . One recent study combining data from 53 countries found an association between precipitation shocks and child stunting 15 , however, the factors driving this association remained unclear. Overall, the existing literature provides little understanding of which periods and types of droughts are most critical for child nutrition and fails to consider the importance of agricultural practices, such as land allocation and types of crops cultivated. Yet, such information targeted in space and time is critical for developing targeted intervention strategies. With the recent development of spatially resolved agricultural 32 , 33 and climate 34 datasets that integrate satellite-derived information with traditional on-the-ground data sources, this paper is able to delve deeper into these questions. Another important gap in the literature concerns the identification of population groups most vulnerable to the impacts of drought. Such investigation can benefit the development of targeted intervention strategies. Identifying households that are most in need of assistance will be critical for allocating the limited resources available to LMICs most effectively. We address some of these gaps in the literature by combining spatially disaggregated climate and agricultural land-use information with nationally representative household survey data from 33 sub-Saharan African countries to investigate the relationship between precipitation shocks during key agricultural periods and child stunting. We focus on SSA because of the region’s high rates of child undernutrition and the growing impacts of climate change on food production systems 35 . We compile anthropometric data (height-for-age) for 380,653 children aged 1 to 5 from 105 household surveys collected between 1990 and 2022. Using high spatial resolution data on the geographical distribution of crop areas for 15 main crops 33 , detailed crop calendar information for each crop 36 , and climatological data, we were able to capture climate conditions during key crop-growing periods. We calculated the Standardized Precipitation and Evapotranspiration Index (SPEI) from the climatological data and linked it with the household survey data to investigate the non-linear relationship between child stunting and sub-annual variations in climate conditions. Agricultural droughts are driven by the compounded effects of increased evaporation and transpiration caused by higher temperatures and precipitation deficits 6 – 8 . Hence, SPEI is well-suited for capturing agricultural droughts since it incorporates information on both temperature-related evapotranspiration and precipitation changes 37 . We focused on exposure to precipitation shocks during infancy since most growth faltering is shown to occur prior to the first 23 months of life 38 , 39 but we also investigated the impact of agricultural drought experienced at later ages. We additionally examine heterogeneities between communities characterized by predominant land-use practices (crop-farming, pastoralism, mixture of both, and non-agricultural activities) since the direction and magnitude of the association may vary depending on the community’s dependence on the natural environment and their ability to adapt to environmental changes 35 , 40 . For example, there is evidence that live-stock herders may relocate to greener areas or sell livestock to smooth their consumption during dry periods 41 , while crop farmers may be more constrained in their response to an agricultural drought 42 . Among communities whose livelihoods do not directly depend on agriculture, agricultural droughts may still impact child nutrition via food import disruptions and food price spikes, however, this has been understudied. Finally, we investigate whether the association between droughts and child stunting is modified by households’ socioeconomic characteristics such as relative wealth, education, and type of occupation. Little is known about the complex interaction of factors contributing to households’ experience of food insecurity following a climate-related shock and how this affects children’s nutrition 9 . Households’ ability to sell assets or diversify income may influence how well they can smooth consumption in such adverse situations. Parental education may also have a protective effect during droughts. Research indicates that maternal education is associated with behaviors that improve child nutrition, such as improved dietary quality and other care practices 43 . Differences in childhood undernutrition by sex and birth order have also been identified in the literature, possibly due to the uneven allocation of scarce resources, including food, within the household, among other factors. However, it is unclear whether these factors play a role when households experience an external shock such as a drought. In what follows, we attempt to shed light on the link between agricultural drought and child undernutrition in SSA, investigating the interactive role of environmental and socioeconomic factors that may heighten population vulnerability to such climatic shocks. Results Our dataset consists of 380,653 children ages 1 to 5 from 33 countries in SSA for whom complete information was available on anthropometric measurements, geo-location, and individual and household characteristics. Considering the agricultural land distribution, cropland areas dominate in western Africa (Fig. 1 b), particularly areas characterized by sub-humid climate (Supplementary Fig. 4), as well as in Uganda and the tropical highlands of Rwanda and Burundi. Pastures extend across the continent (Fig. 1 a) and are dominant in eastern Africa (e.g., Mozambique, Tanzania, and Zambia) and southern Africa (e.g., Lesotho, Eswatini, South Africa, and Namibia). Grains are the most common type of crops grown across the survey sites (Fig. 1 c), while roots are commonly grown in the humid and sub-humid areas of central Africa. Based on the share of land dedicated to crop-farming and pastures in the area of the survey sites (Fig. 1 and Methods), we split our sample into four groups: cropland-dominated areas (24% of our sample), mixed cropland and pasture areas (22%), pasture-dominated areas (31%), and areas where agricultural land accounts for less than 20% of the total land area (24% of the sample) (see Supplementary Fig. S2 ). The prevalence of child stunting is slightly higher in mixed areas and cropland-dominated areas (44% and 43% of children, respectively) as compared to the rest (about 40% of children). In Fig. 2 we show the results of our main model where we estimate the association between localized SPEI during specific agricultural periods and child stunting adjusting for potential confounders (see Methods). In all models described below, we focus specifically on exposure to a climatic shock during the child’s infancy period. Considering SPEI during the main-crop growing period, we find a reverse J-shaped relationship with child stunting (Fig. 2 a), with a sharp and statistically significant increase in the risk of stunting during relatively dry periods (negative SPEI values) and a more shallow and not statistically significant increase in the risk of stunting during abnormally wet periods (positive SPEI values). Using weighted SPEI values, which combine multiple crop growth periods (see Methods), we find a more pronounced reverse J-shaped relationship with the risk of stunting (Fig. 2 b). Examining the association for specific crop-growing periods, we find strong associations between stunting and relatively dry conditions during the grains and oilseeds growth periods and more shallow and non-significant associations for relatively wet conditions (Fig. 2 c). We find non-significant associations between stunting and SPEI during the growth periods for roots and pulses (Fig. 2 c). We construct indicators for drought and extreme precipitation, based on standard SPEI thresholds proposed in the literature 44 , and conduct a stratified sample analysis where we distinguish by agricultural land specialization. For the rest of the analysis, we use the weighted SPEI measure aggregated over all crop-growing seasons since it showed a stronger association with child stunting than the main crop-growing season SPEI. In the full sample, we find that an agricultural drought (SPEI<-1) increases the risk of child stunting by 1.07 (95% CIs: 1.03 to 1.11; Fig. 3 ). Among the cropland-dominated areas, this risk is more pronounced (aOR = 1.18 with 95% CIs: 1.07–1.29), while in the mixed areas it is comparable (aOR = 1.08 with 95% CIs: 1.02–1.15). In the pasture-dominated areas and the areas where less than 20% of the land is used for agriculture, we do not find an association between agricultural drought and child stunting. Extreme precipitation (SPIE ≥ 1) does not seem to pose a risk in any of the four agricultural groups (Fig. 3 ). We additionally estimate the association with different categories of droughts – from mild to severe (SI Appendix 1, Table S6). We find that the risk of stunting increases progressively with the severity of the drought event. We further explore how the association changes across different cropland and pasture distributions (SI Appendix 1, Table S7). We find the strongest associations between agricultural drought and child stunting in areas where the cropland area is 20% or more of the total land area and the pasture area is under 20% of the total land area. The higher the share of land used for pastures and the lower the share of land used for croplands, the weaker the association becomes. We also explore whether irrigation infrastructure may diminish the impact of agricultural droughts on child stunting. Generally, about 1 percent of the agricultural area is equipped for irrigation across the 33 countries included in the analysis. We find some evidence that irrigation infrastructure diminishes the impact of droughts on child stunting in mixed areas but not in the cropland-dominated areas (SI Appendix 1, Table S8). It must be noted that we do not have information on irrigation practices but only on whether the area is equipped for irrigation. We also do not know what type of irrigation system is available, whether it is functional and its capacity. We additionally explore whether individual children’s and household’s sociodemographic factors modify the associations between agricultural droughts and child stunting. We use a Wald test for heterogeneity to identify statistically significant heterogeneity in the effect estimates across population subgroups. We carry out this analysis for the sample of children residing in cropland-dominated areas and mixed areas, and for the full sample of children (Fig. 4 ). We do not find evidence of effect modification by birth order and sex. The Wald test did not show statistically significant differences by urban-rural place of residence and by mother’s level of education either, even though the effect estimates were stronger among children residing in rural areas as compared to urban areas, and among children whose mothers had no formal education as compared to children whose mothers had completed at least primary education. We find some evidence that the type of occupation of the household head modifies the association between agricultural droughts and child stunting in mixed cropland and pasture areas. Specifically, we observe a positive and statistically significant association between agricultural drought and stunting among children whose households are employed in the agricultural sector (aOR = 1.13 with 95% CIs: 1.05–1.22; Fig. 4 e), whereas no association is found among children whose households are employed in non-agricultural sectors (aOR = 1.001 with 95% CIs: 0.93–1.08). The Wald test shows that the pairwise difference is statistically significant ( p -value = 0.01). Within the cropland-dominated areas, we do not find evidence of effect modification by occupational group. We also find that household wealth modifies the association between agricultural droughts and child stunting for children in the cropland-dominated areas and for the full sample of children. Among children in the poorer and middle-wealth households in the cropland-dominated areas, we see a positive and statistically significant association between agricultural droughts and child stunting (aOR = 1.27 with 95% CIs: 1.16–1.43 for the poorer group and aOR = 1.20 with 95% CIs: 1.08–1.33 for the middle-wealth group, Fig. 4 f), whereas this association is close to null among children from wealthier households (aOR = 1.04 with 95% CIs: 0.92–1.19). The Wald test shows that these differences in the effect estimates are highly statistically significant ( p -value = 0.003 for the pairwise difference between the poorer and wealthier groups and p -value = 0.02 for the pairwise difference between the middle-wealth and wealthier groups). We find that the impact of agricultural droughts experienced during infancy can be lasting. We see that children exposed to an agricultural drought during infancy display a higher risk of stunting at ages 1, 2, and 3 as compared to children of the same age groups who were not exposed to a drought early in life (Table 1 ) and the magnitude of the association remains about the same. Among children measured at age 4, we do not find a statistically significant association between exposure to an agricultural drought during infancy and the risk of stunting. We also find that agricultural droughts experienced at later ages affect children’s risk of stunting, especially within a few years after exposure. For example, among children who were measured at age 4, we see that an agricultural drought experienced at age 3 increases the risk of stunting by 1.13 (95% CIs: 1.04–1.24) and a drought exposure at age 2 increases the risk of stunting by 1.09 (95% Cis: 1.00-1.18). Table 1 Associations between child stunting and drought (wtd SPEI<-1) measured at different ages of exposure and stratified by child’s age at measurement. All estimates are adjusted for potential confounders (see Methods). Age at measurement: 1 Age at measurement: 2 Age at measurement: 3 Age at measurement: 4 aOR [95% CIs] aOR [95% CIs] aOR [95% CIs] aOR [95% CIs] Drought: infancy 1.09 [1.03–1.15] 1.10 [1.02–1.18] 1.09 [1.00-1.20] 0.97 [0.90–1.04] Drought: age 1 1.09 [1.02–1.16] 1.10 [1.03–1.18] 1.03 [0.94–1.13] Drought: age 2 1.16 [1.08–1.24] 1.09 [1.00-1.18] Drought: age 3 1.13 [1.04–1.24] Individual & household-level covariates Yes Yes Yes Yes Geographical covariates Yes Yes Yes Yes Obs. 104,725 98,893 90,431 86,533 Discussion We conduct the largest study to date linking agricultural droughts to child undernutrition using data from 33 countries in sub-Saharan Africa collected over three decades (1990–2022). By explicitly considering multiple location-specific features such as agricultural land distribution, types of crops grown, and detailed crop calendar information, we were able to identify precise dose-response relationships that were not identifiable previously due to the lack of harmonized and geographically disaggregated agricultural and climatological data. We find compelling evidence that agricultural droughts are associated with an increased risk of child stunting in sub-Saharan Africa. The effect is most notable in places where a high fraction of the land is used for crop cultivation, implying that crop-farming is an important source of livelihood there, and less pronounced in areas where a high fraction of the land is used for pastures. These findings are in line with previous research, which shows high sensitivity of crop production to climate variability in Africa 5 , whereas the impact on livestock productivity was found to be less significant. A recent study that looked at the impact of climate variability on birth weight in Kenya and Mali also documented weaker associations among pastoral communities as compared to crop-farming communities 42 . Temporary migration to greener areas, reducing livestock numbers, conserving feed resources and shifting to browse livestock species have all been documented as common coping strategies employed by pastoralists in response to droughts 41 . However, these communities may still be vulnerable to longer-lasting, recurrent, and widespread droughts which can erode their resources over time 45 . Pastoral communities usually occupy dryland areas that are less productive and are often economically marginalized. Future research should look more closely at how these communities are affected by climate change. Likewise, we did not find an association between child stunting and agricultural droughts in areas where a relatively low fraction of the land area is used for agriculture. In such areas, populations may be less dependent on agricultural resources as compared to other sources of livelihood or may be more adaptable to variable climate conditions. Such populations may still be impacted by food import disruptions due to drought or conflict further away, which should be investigated in future research. Overall, we find that droughts during the growing seasons for grains and oilseeds are most strongly associated with child stunting. Grains, such as maize, millet, rice, sorghum, and wheat, are staple crops in SSA and contribute to most of the calorie availability there. Apart from reducing agricultural production 4 , climate change has been shown to impact the micronutrient content of food crops 46 , which may have important consequences for child nutrition. Even though certain crops, such as certain oilseeds (e.g., cotton and soybean) are sold for profit (so called “cash crops”), decline in their yields may still impact food security via income shocks. Apart from the implications for food availability and agricultural income, droughts can impact child health through infectious illnesses 47 , 48 . The presence of chronic infectious illnesses in children, especially during critical periods in early childhood, has been linked to hindered growth and development 49 . This is particularly the case in areas where access to clean water and sanitation infrastructure is limited. We do not find an association between heavy precipitation and child growth, even though such an association has been documented in previous research 15 . A reason for that could be our focus on precipitation shocks during the main agricultural periods rather than annual precipitation variability. Poverty seems to be an important factor exacerbating the impact of agricultural droughts on child growth. We find the strongest effect among the poorest strata in cropland-dominated areas, whereas the association is close to null among the wealthiest strata. Smallholder and subsistence farmers are among the poorest groups in less developed countries and they are also among the most vulnerable to the effects of climate change since their livelihoods directly depend on natural resources 35 . Drought-related crop loss and decreased agricultural income are likely to pressure poor households to reduce food consumption and cut back on non-food expenses, such as healthcare expenditure, compromising children’s health. Wealthier households may be less reliant on farming activities or may be able to smooth consumption by using savings or selling assets. When droughts persist, however, household assets may get depleted and no longer provide a buffer against agricultural shocks. For example, droughts have been shown to affect land-ownership rates in Africa 50 , implying a shift towards wage labor and rural-urban migration. Such developments should be considered when designing effective intervention strategies that protect vulnerable populations. Finally, our findings suggest that the impact of agricultural droughts experienced early in life can persist. Children who experienced a drought during infancy have a higher risk of stunting at ages 1 to 3 and the magnitude of this association remains relatively constant. For children measured at age 4, we do not find an association between drought experienced in infancy and stunting, which could be explained by “catch-up” growth 51 and/or selective survival of the healthier children. Identifying the exact mechanism nevertheless would require panel data that follows children’s growth trajectories over their lives. For children measured at aged 4, we find an association between stunting and agricultural drought experienced in more recent periods (ages 2 and 3). Our study has certain limitations that must be acknowledged. The survey data used in the analysis is cross-sectional and does not allow us to determine the longitudinal effects of agricultural droughts on child growth. Furthermore, there are multiple pathways, including food availability, poverty, and infectious diseases, which we are not able to disentangle with the available data. It would be particularly important to explore the contribution of multiple mechanisms on such relationships by capitalizing on datasets that collected data on child growth, nutrition as well as variables capturing such potential pathways, such as the World Bank Living Standards Measurement Study – Integrated Surveys on Agriculture (LSMS- ISA). Another limitation is that we use information on the distribution of land for croplands and pastures as well as specific crop types that is fixed around the year 2000. To the best of our knowledge, no longitudinal data are available at the spatial resolution and for the variety of crops used in this study. We also do not consider changes in the crop calendars, potential crop diversification and other adaptation measures, which may be employed in response to the impacts of climate change. We are also not able to capture land desertification 52 . We do not distinguish between subsistence agriculture and industrial crop cultivation, which may have very different implications for food security in SSA 53 . Furthermore, even though we consider whether the area is equipped for irrigation, we are not able to distinguish between irrigated and non-irrigated farming systems, which may have important implications for drought resistance. Also, we are not able to determine the agricultural specializations at the household level and to distinguish between food croppers, who are usually subsistence farmers and may be more sensitive to agricultural shocks, and cash croppers. All of these limitations should be considered in future research. In conclusion, we find a strong association between agricultural droughts and child undernutrition in SSA using data from 33 countries spanning three decades. We show that vulnerability to the impact of drought can be influenced by factors at the community level, such as agricultural land allocation and crop specialization, as well as at the household level, such as relative wealth and occupation of the household head. In this study, we focus specifically on agricultural droughts, i.e., droughts that affect crop production. These can be driven by region-specific changes in precipitation regimes 7 , 8 and warming temperatures 6 caused by anthropogenic climate change as well as natural processes. Other types of droughts may also impact population health via changes to the ecological system and socioeconomic processes and should be addressed in future research. Moreover, the interaction of local climate conditions with global crises such as the COVID-19 pandemic 29 and the war in Ukraine 60 (a major wheat exporter to countries in SSA) may further undermine food security and child nutrition in the region, which should be studied in the future. Considering the additional challenges posed by climate change on achieving global nutrition goals, our study provides insights that can be used for targeting food aid programs and other interventions that strengthen food security in a changing climate. Methods Nutrition and covariate data We obtained anthropometric data on children under 5 years of age from the DHS program 54 . DHS surveys are conducted via a two-stage procedure, which ensures the representativeness of the samples at both the national and subnational levels. We included all surveys that contained child anthropometric information and global positioning system (GPS) coordinates of the primary sampling units (PSUs). PSUs are enumeration areas used in DHS which usually represent villages in rural areas and city blocks in urban areas. To maintain the confidentiality of survey participants, DHS displaces the PSU coordinates by 2km in urban areas, 5km in rural areas, and an extra 10km for 5% of all clusters. In total, we used information from 105 surveys collected across 33 countries between 1990 and 2022 (SI Appendix, Table S1 ). While DHS surveys are repeatedly collected in certain countries, the surveys are cross-sectional as new households are interviewed in every survey round. The surveys are a rich source of information on fertility, reproductive behavior, mortality, health, and wellbeing in LMICs. Women in the selected households who are of reproductive age (15–49 years) are interviewed in-depth to gather details on their past pregnancies and the health of their children born in the 5 years preceding the survey. We infer children’s nutritional status from their height, which is measured by trained fieldworkers during the interviews. In particular, we calculated height-for-age z-scores (HAZ) for children under 5 years of age and based on the scores constructed a binary indicator for child stunting, which takes the value 1 if a child’s HAZ score is more than 2 standard deviations below the World Health Organization’s (WHO) growth standard median for children of the same age-group 55 and 0 otherwise. Observations were dropped if the HAZ scores were implausible (below − 6 and above 6). Low HAZ values reflect cumulative effects of sustained undernutrition and infections experienced since birth and even during pregnancy 56 . Our final sample consists 380,653 children aged 1 to 5 for whom complete anthropometric and sociodemographic information was available. In addition to children’s anthropometric measurements, we retrieve information about the demographic and socioeconomic characteristics of the children and their households: children’s sex and birth order, mother’s highest achieved level of education (none, primary, secondary and higher), type of occupation of the household head (agricultural or non-agricultural), place of residence (urban or rural) and household wealth group. We used this contextual information to explore potential heterogeneities in the effect size among population subgroups. Following standard DHS procedures 57 , we use principle component analysis to construct an index of household wealth by combining information about household ownership of select assets (radio, fridge, television, and car), building characteristics (material of the floor), and the availability of electricity and improved water and sanitation facilities on the premise. Children of women that are not permanent residents at the place of interview are excluded from the analysis to reduce the risk of misclassification. Agricultural data We use global gridded dataset of cropland and pastures available at a high spatial resolution (~ 10km) 32 to determine the predominant land use practice around each PSU (within 10km radius). The data is representative for the year 2000 and provides greater accuracy than other data sources since it combines agricultural inventory data with satellite-derived land-cover datasets. We use the cropland and pasture data to distinguish between four types of areas: cropland-dominated (areas where over 20% of the land is cropland and the share of land used for growing crops is at least twice as high as the share of land used for pasture), pasture-dominated (areas where over 20% of the land is dedicated to pastures and the share of land dedicated to pastures is at least twice as high as the share of land dedicated to crops), mixed cropland and pasture areas (areas which do not fall in either of the above categories but where the combined share of land dedicated to crops and pastures is at least 20% of the total land area), and areas where less than 20% of the total land area is used for agriculture. Fig. S2 in SI Appendix 1 shows how the above groups are spatially distributed. We additionally retrieve information about the area harvested for main crop grown in SSA 33 (SI Appendix 2). The data are available globally at ~ 10km spatial resolution and indicate multiple crops grown in each location. Crop planting and harvesting dates are additionally obtained 36 and used to determine the start and end day and the duration of the growing season for each crop. The growing season is defined as the period between the start of planting and the start of harvesting. Our final dataset includes information about 15 main crops grown in SSA (SI Appendix 2), which, for the purpose of our analysis, we group into: grains (barley, maize, millet, rice, sorghum, and wheat), oilseeds (cotton, groundnut, soybean, sunflower, and pulses), pulses, and roots (cassava, potato, sweet potato, and yam). PSUs for which no land area is used to grow crops are excluded from the analysis. Climate data We use the SPEI as an indicator of agricultural droughts. The SPEI measures the variation in hydrological conditions during a certain period in relation to the area-specific long-term norm and it can take both negative values, indicating abnormally dry conditions, and positive values, indicating abnormally wet conditions. The SPEI is preferred over other drought indices since it can be computed over different temporal scales and considers how both temperature and precipitation contribute to the development of droughts 58 . We generate SPEI values using the global gridded precipitation and potential evapotranspiration data provided by the Climatic Research Unit (CRU) at the University of East Anglia (time series 4) at monthly temporal resolution and 0.5° spatial resolution and covering the period 1901- 2022 59 . The R package “SPEI” is used to compute the monthly SPEI values based on the input CRU data at different time scales considering the duration of each crop-growing season in every PSU location. The seasonal SPEI values for all crops grown in specific PSU are combined into a single index by computing a weighted SPEI value where the weights correspond to the share of land dedicated to each crop. Similarly, weighted SPEI values are generated for the crop-growing seasons of specific crop group (grains, roots, oilseeds, and pulses). Estimation strategy We model the effect of droughts experienced during the period of infancy (the first year of life) on the child’s risk of stunting. We run a non-linear logistic regression, which takes the following form: $$\:{y}_{i,j,r}=\alpha\:+{{\beta\:}\varvec{X}}_{i,j,r}+\:\delta\:{SPEI}_{j}+{A}_{r}+{\text{ϵ}}_{i,j,a}$$ Where \(\:{y}_{i,a}\) is the nutrition status of child i interviewed in PSU j in administrative area (e.g., district or province) r . \(\:{\varvec{X}}_{i}\) is a vector of individual, household, and geographic factors, including sex of the child, their age in months, bird order, size at birth (smaller than average, average, or larger than average), age of the mother at giving birth, mother’s level of education, occupation of the household head (agricultural or non-agricultural), household wealth group, urban or rural place of residence. Birth quarter and interview quarter are also included as covariates to account for seasonal effects in the child’s risk of stunting. We also include indicators for the main crop and the length of the main crop-growing season in the area. We additionally control for the share of land equipped for irrigation (see Fig. S3 in SI Appendix 1) and agroecological zone (Fig. S4 in SI appendix 1). \(\:{\beta\:}\) is a vector of corresponding slope parameters for each covariate. A is an intercept for the administrative area The most up-to-date subnational boundaries were retrieved from the DHS spatial data repository ( https://spatialdata.dhsprogram.com/home/ ). The subnational boundaries were harmonized in the case when changes occurred between survey rounds. \(\:{SPEI}_{i,j,r}\) captures the drought conditions in each PSU site. We use a cubic spline transformation of the SPEI value with 4 equally spaced knots to allow for non-linearity in the relationship with stunting. The errors are clustered at the administrative level. References You L, Ringler C, Wood-Sichra U, et al. What is the irrigation potential for Africa? A combined biophysical and socioeconomic approach. Food Policy . 2011;36(6):770-782. doi:10.1016/J.FOODPOL.2011.09.001 World Bank. World Development Indicators. Published online 2024. Accessed January 30, 2024. http://data.worldbank.org/data-catalog/world-development-indicators. Sultan B, Defrance D, Iizumi T. Evidence of crop production losses in West Africa due to historical global warming in two crop models. Sci Reports 2019 91 . 2019;9(1):1-15. doi:10.1038/s41598-019-49167-0 Ortiz-Bobea A, Ault TR, Carrillo CM, Chambers RG, Lobell DB. Anthropogenic climate change has slowed global agricultural productivity growth. Nat Clim Chang 2021 114 . 2021;11(4):306-312. doi:10.1038/s41558-021-01000-1 Stige LC, Stave J, Chan KS, et al. The effect of climate variation on agro-pastoral production in Africa. Proc Natl Acad Sci U S A . 2006;103(9):3049-3053. doi:10.1073/pnas.0600057103 Williams AP, Cook ER, Smerdon JE, et al. Large contribution from anthropogenic warming to an emerging North American megadrought. Science (80- ) . 2020;368(6488):314-318. doi:10.1126/SCIENCE.AAZ9600/SUPPL_FILE/AAZ9600_WILLIAMS_SM.PDF Polade SD, Pierce DW, Cayan DR, Gershunov A, Dettinger MD. The key role of dry days in changing regional climate and precipitation regimes. Sci Rep . 2014;4. doi:10.1038/srep04364 Polade SD, Gershunov A, Cayan DR, Dettinger MD, Pierce DW. Precipitation in a warming world: Assessing projected hydro-climate changes in California and other Mediterranean climate regions. Sci Reports 2017 71 . 2017;7(1):1-10. doi:10.1038/s41598-017-11285-y Hadley K, Wheat S, Rogers HH, et al. Mechanisms underlying food insecurity in the aftermath of climate-related shocks: a systematic review. Lancet Planet Heal . 2023;7(3):e242-e250. doi:10.1016/S2542-5196(23)00003-7 FAO. The Impact of Disasters and Crises on Agriculture and Food Security .; 2018. Accessed January 27, 2020. www.fao.org/publications Abel GJ, Brottrager M, Crespo Cuaresma J, Muttarak R. Climate, conflict and forced migration. Glob Environ Chang . 2019;54:239-249. doi:10.1016/J.GLOENVCHA.2018.12.003 Von Uexkull N, Croicu M, Fjelde H, Buhaug H. Civil conflict sensitivity to growing-season drought. Proc Natl Acad Sci U S A . 2016;113(44):12391-12396. doi:10.1073/PNAS.1607542113/SUPPL_FILE/PNAS.201607542SI.PDF Niles MT, Salerno JD. A cross-country analysis of climate shocks and smallholder food insecurity. PLoS One . 2018;13(2):e0192928. doi:10.1371/JOURNAL.PONE.0192928 Baker RE, Anttila-Hughes J. Characterizing the contribution of high temperatures to child undernourishment in Sub-Saharan Africa. Sci Rep . 2020;10(1). doi:10.1038/s41598-020-74942-9 Cooper MW, Brown ME, Hochrainer-Stigler S, et al. Mapping the effects of drought on child stunting. Proc Natl Acad Sci . Published online August 12, 2019:201905228. doi:10.1073/pnas.1905228116 IPCC. Climate Change 2022: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Climate Change 2021: The Physical Science Basis. ; 2022. doi:10.1017/9781009325844.Front Hales S, Kovats S, Simon Lloyd DC-L. Quantitative Risk Assessment of the Effects of Climate Change on Selected Causes of Death, 2030s and 2050s .; 2014. doi:ISBN 978 92 4 150769 1 Matrins VJB, Toledo Florêncio TMM, Grillo LP, et al. Long-lasting effects of undernutrition. Int J Environ Res Public Health . 2011;8(6):1817-1846. doi:10.3390/ijerph8061817 Black RE, Victora CG, Walker SP, et al. Maternal and child undernutrition and overweight in low-income and middle-income countries. Lancet . 2013;382(9890):427-451. doi:10.1016/S0140-6736(13)60937-X Arthur SS, Nyide B, Soura AB, Kahn K, Weston M, Sankoh O. Tackling malnutrition: a systematic review of 15-year research evidence from INDEPTH health and demographic surveillance systems. doi:10.3402/gha.v8.28298 Victora CG, Adair L, Fall C, et al. Maternal and child undernutrition: consequences for adult health and human capital. Lancet . 2008;371(9609):340-357. doi:10.1016/S0140-6736(07)61692-4 Poveda NE, Hartwig FP, Victora CG, et al. Patterns of Growth in Childhood in Relation to Adult Schooling Attainment and Intelligence Quotient in 6 Birth Cohorts in Low- and Middle-Income Countries: Evidence from the Consortium of Health-Oriented Research in Transitioning Societies (COHORTS). J Nutr . 2021;151(8):2342-2352. doi:10.1093/JN/NXAB096 Prendergast AJ, Humphrey JH. The stunting syndrome in developing countries. Paediatr Int Child Health . 2014;34(4):250-265. doi:10.1179/2046905514Y.0000000158 McGovern ME, Krishna A, Aguayo VM, Subramanian S V. A review of the evidence linking child stunting to economic outcomes. Int J Epidemiol . 2017;46(4):1171-1191. doi:10.1093/ije/dyx017 Walson JL, Berkley JA. The impact of malnutrition on childhood infections. Curr Opin Infect Dis . 2018;31(3):231-236. doi:10.1097/QCO.0000000000000448 Walker SP, Chang SM, Wright A, Osmond C, Grantham-McGregor SM. Early Childhood Stunting Is Associated with Lower Developmental Levels in the Subsequent Generation of Children. J Nutr . 2015;145(4):823-828. doi:10.3945/JN.114.200261 Heltberg R. Malnutrition, poverty, and economic growth. Health Econ . 2009;18 Suppl 1(SUPPL. 1). doi:10.1002/HEC.1462 UNICEF, WHO, World Bank. Level and Trend in Child Malnutrition: UNICEF/WHO/World Bank Group Joint Child Malnutrition Estimates: Key Findings of the 2023 Edition . UNICEF and WHO; 2023. Accessed January 21, 2024. https://www.who.int/publications/i/item/9789240073791 Kansiime MK, Tambo JA, Mugambi I, Bundi M, Kara A, Owuor C. COVID-19 implications on household income and food security in Kenya and Uganda: Findings from a rapid assessment. World Dev . 2021;137. doi:10.1016/j.worlddev.2020.105199 FAO, IFAD, UNICEF, WFP, WHO. The State of Food Security and Nutrition in the World 2023. Urbanization, Agrifood Systems Transformation and Healthy Diets across the Rural–Urban Continuum . FAO; 2023. doi:10.4060/CC3017EN Belesova K, Agabiirwe CN, Zou M, Phalkey R, Wilkinson P. Drought exposure as a risk factor for child undernutrition in low- and middle-income countries: A systematic review and assessment of empirical evidence. Environ Int . 2019;131. doi:10.1016/j.envint.2019.104973 Ramankutty N, Evan AT, Monfreda C, Foley JA. Farming the planet: 1. Geographic distribution of global agricultural lands in the year 2000. Global Biogeochem Cycles . 2008;22(1):1003. doi:10.1029/2007GB002952 Monfreda C, Ramankutty N, Foley JA. Farming the planet: 2. Geographic distribution of crop areas, yields, physiological types, and net primary production in the year 2000. Global Biogeochem Cycles . 2008;22(1). doi:10.1029/2007GB002947 Vicente-Serrano SM, Beguería S, López-Moreno JI. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index. J Clim . 2010;23(7):1696-1718. doi:10.1175/2009JCLI2909.1 Morton JF. The impact of climate change on smallholder and subsistence agriculture. Proc Natl Acad Sci U S A . 2007;104(50):19680-19685. doi:10.1073/pnas.0701855104 Sacks WJ, Deryng D, Foley JA, Ramankutty N. Crop planting dates: an analysis of global patterns. Glob Ecol Biogeogr . 2010;19(5):607-620. doi:10.1111/J.1466-8238.2010.00551.X Vicente-Serrano SM, Beguería S, López-Moreno JI, Angulo M, Kenawy AE. A New Global 0.5° Gridded Dataset (1901–2006) of a Multiscalar Drought Index: Comparison with Current Drought Index Datasets Based on the Palmer Drought Severity Index. J Hydrometeorol . 2010;11(4):1033-1043. doi:10.1175/2010JHM1224.1 Alderman H, Headey D. The timing of growth faltering has important implications for observational analyses of the underlying determinants of nutrition outcomes. PLoS One . 2018;13(4):e0195904. doi:10.1371/JOURNAL.PONE.0195904 Victora CG, De Onis M, Hallal PC, Blössner M, Shrimpton R. Worldwide timing of growth faltering: Revisiting implications for interventions. Pediatrics . 2010;125(3). doi:10.1542/peds.2009-1519 Hoffmann R, Wiederkehr C, Dimitrova A, Hermans K. Agricultural livelihoods, adaptation, and environmental migration in sub-Saharan drylands: a meta-analytical review. Environ Res Lett . 2022;17(8):083003. doi:10.1088/1748-9326/AC7D65 Habte M, Eshetu M, Maryo M, Andualem D, Legesse A. Effects of climate variability on livestock productivity and pastoralists perception: The case of drought resilience in Southeastern Ethiopia. Vet Anim Sci . 2022;16. doi:10.1016/J.VAS.2022.100240 Bakhtsiyarava M, Grace K, Nawrotzki RJ. Climate, Birth Weight, and Agricultural Livelihoods in Kenya and Mali. Am J Public Health . 2018;108(S2):S144-S150. doi:10.2105/AJPH.2017.304128 Alderman H, Headey D. How Important is Parental Education for Child Nutrition? World Dev . 2017;94:448-464. Beguería S, Vicente-Serrano SM, Angulo-Martínez M. A Multiscalar Global Drought Dataset: The SPEIbase: A New Gridded Product for the Analysis of Drought Variability and Impacts. Bull Am Meteorol Soc . 2010;91(10):1351-1356. doi:10.1175/2010BAMS2988.1 Inman EN, Hobbs RJ, Tsvuura Z. No safety net in the face of climate change: The case of pastoralists in Kunene Region, Namibia. PLoS One . 2020;15(9). doi:10.1371/JOURNAL.PONE.0238982 Semba RD, Askari S, Gibson S, Bloem MW, Kraemer K. The Potential Impact of Climate Change on the Micronutrient-Rich Food Supply. Adv Nutr . 2022;13(1):80. doi:10.1093/ADVANCES/NMAB104 Wang P, Asare E, Pitzer VE, Dubrow R, Chen K. Associations between long-term drought and diarrhea among children under five in low- and middle-income countries. Nat Commun 2022 131 . 2022;13(1):1-10. doi:10.1038/s41467-022-31291-7 Dimitrova A, McElroy S, Levy M, Gershunov A, Benmarhnia T. Precipitation variability and risk of infectious disease in children under 5 years for 32 countries: a global analysis using Demographic and Health Survey data. Lancet Planet Heal . 2022;6(2):e147-e155. doi:10.1016/S2542-5196(21)00325-9 Schaible UE, Kaufmann SHE. Malnutrition and Infection: Complex Mechanisms and Global Impacts. PLOS Med . 2007;4(5):e115. doi:10.1371/JOURNAL.PMED.0040115 Linke AM, Tollefsen AF. Environmental stress and agricultural landownership in Africa. Glob Environ Chang . 2021;67:102237. doi:10.1016/J.GLOENVCHA.2021.102237 Desmond C, Casale D. Catch-up growth in stunted children: Definitions and predictors. Schooling CM, ed. PLoS One . 2017;12(12):e0189135. doi:10.1371/journal.pone.0189135 United Nations Economic and Social Council. Africa Review Report on Drought and Desertification .; 2007. doi:E/ECA/ACSD/5/3 Jarzebski MP, Ahmed A, Boafo YA, et al. Food security impacts of industrial crop production in sub-Saharan Africa: a systematic review of the impact mechanisms. Food Secur . 2020;12(1):105-135. doi:10.1007/S12571-019-00988-X/TABLES/3 ICF. The DHS Program website [various datasets]. Funded by USAID. Accessed January 17, 2024. http://www.dhsprogram.com World Health Organiozation. Nutrition Landscape Information System (NLiS) Country Profile Indicators: Interpretation Guide .; 2019. de Onis M. Child Growth and Development. In: Nutrition and Health in a Developing World . Springer International Publishing; 2017:119-141. doi:10.1007/978-3-319-43739-2_6 Rutstein SO. Steps to Constructing the New DHS Wealth Index .; 2015. Vicente-Serrano SM, Beguería S, López-Moreno JI, Angulo M, El Kenawy A. A New Global 0.5° Gridded Dataset (1901–2006) of a Multiscalar Drought Index: Comparison with Current Drought Index Datasets Based on the Palmer Drought Severity Index. J Hydrometeorol . 2010;11(4):1033-1043. Harris I, Osborn TJ, Jones P, Lister D. Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. Sci Data 2020 71 . 2020;7(1):1-18. doi:10.1038/s41597-020-0453-3 Mottaleb KA, Kruseman G, Snapp S. Potential impacts of Ukraine-Russia armed conflict on global wheat food security: A quantitative exploration. Glob Food Sec . 2022;35:100659. doi:10.1016/J.GFS.2022.100659 Additional Declarations There is NO Competing Interest. Supplementary Files SIAppendix1.docx Appendix 1 SIAppendix2.docx Appendix 2 Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7765713","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":530237223,"identity":"b99b51e5-ac41-41b5-8043-8fcebebbeb73","order_by":0,"name":"Anna Dimitrova","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzElEQVRIiWNgGAWjYFCCxAaGBDCD+QCIAOIEYrSAFbElEKslAaaIx4A4Lfztya0bHv64J2dw/Mw3iR9/rBn42XMM8GqROPOw7UZCQrGxwZncbZK9bekMkj1v8GthuJEI0pKQOLMhd5s0Y8NhBoMbBGyRh2vpf/NMmuHPYQZ7QloMYFr6JXLYpBnYgLZIENBiCPZLWoIxv8QzY0ugX3gkzjwrwKtF7nj6s5s/bBLk2PiTH94AhpgcMAw34NWCAXhIUz4KRsEoGAWjACsAAM41St4rDTbmAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0003-2715-9276","institution":"Scripps Institution of Oceanography, University of California San Diego","correspondingAuthor":true,"prefix":"","firstName":"Anna","middleName":"","lastName":"Dimitrova","suffix":""},{"id":530237224,"identity":"3d48132e-78e9-40f6-9732-e2853095d788","order_by":1,"name":"Alexander Gershunov","email":"","orcid":"https://orcid.org/0000-0002-6598-6638","institution":"University of California San Diego","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Gershunov","suffix":""},{"id":530237225,"identity":"2574a737-a318-45ce-a6fd-cba76f5f654d","order_by":2,"name":"Raya Muttarak","email":"","orcid":"https://orcid.org/0000-0003-0627-4451","institution":"University of Bologna","correspondingAuthor":false,"prefix":"","firstName":"Raya","middleName":"","lastName":"Muttarak","suffix":""},{"id":530237226,"identity":"df34e856-c0c6-41cc-9c8d-94e093da5a89","order_by":3,"name":"Francois Rerolle","email":"","orcid":"","institution":"University of California, San Francisco","correspondingAuthor":false,"prefix":"","firstName":"Francois","middleName":"","lastName":"Rerolle","suffix":""},{"id":530237227,"identity":"5b058200-e4dc-4f9c-8d15-00629614a1aa","order_by":4,"name":"Tarik Benmarhnia","email":"","orcid":"","institution":"University of California San Diego","correspondingAuthor":false,"prefix":"","firstName":"Tarik","middleName":"","lastName":"Benmarhnia","suffix":""}],"badges":[],"createdAt":"2025-10-02 10:30:44","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7765713/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7765713/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":93779924,"identity":"f5806c2e-f5c5-4fb5-af14-c45bf32da501","added_by":"auto","created_at":"2025-10-17 12:55:07","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":605197,"visible":true,"origin":"","legend":"","description":"","filename":"AgriculturaldroughtsandchildundernutritioninSSAtb.docx","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/00c4219bd34152e9f9912a33.docx"},{"id":93778458,"identity":"0cf4b8fb-e6b1-43e4-9311-cd89a6878671","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7635,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS2579077.json","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/7f06bae17cee7d2a48e16543.json"},{"id":93779925,"identity":"ba179276-5158-435f-8de0-fbd9a098b047","added_by":"auto","created_at":"2025-10-17 12:55:07","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":851053,"visible":true,"origin":"","legend":"","description":"","filename":"SIAppendix1.docx","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/31cddbdd115b3ecf7b83b8b2.docx"},{"id":93778477,"identity":"f45ebc07-ccff-4501-ac91-b55eea5b3297","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3638247,"visible":true,"origin":"","legend":"","description":"","filename":"SIAppendix2.docx","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/858365a550f8c2a1de3d4c2d.docx"},{"id":93778469,"identity":"f5cd8c4d-eb13-451c-bc91-d6e2b8f69e55","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"xml","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":135659,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25790770enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/ceb8f560593a17a0497b1ae2.xml"},{"id":93779926,"identity":"3e11f560-792b-4e20-95e5-871ef0791e62","added_by":"auto","created_at":"2025-10-17 12:55:07","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":234036,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/655fd7f1dce73b0f0883583d.png"},{"id":93778466,"identity":"a3aaee2f-ab50-48d1-9f75-0acc7e768bdf","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":63426,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/4350819f9c2fd2fcb94c0ebf.png"},{"id":93778472,"identity":"803d344f-4966-435f-adff-748823a156e2","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":52999,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/d094622a7b09f335d52725f3.png"},{"id":93779927,"identity":"af770b16-3f3e-480b-9fd8-53b1841a51a0","added_by":"auto","created_at":"2025-10-17 12:55:07","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":117617,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/454a9ba6bf67a7f4c3caf793.png"},{"id":93778470,"identity":"3dccbb83-5326-4194-bf19-b8c229c5d263","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":58850,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/f08c4499f949ba467c04a8a3.png"},{"id":93778463,"identity":"d339f1de-0d84-45dc-8d8c-a5953c22f3a3","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":26922,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/a22c611978e5fa881a22ae64.png"},{"id":93778467,"identity":"6fa8b59e-b339-4652-9610-48efc03a2b7d","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":32699,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/f4bbad50ccf037434b1f5b40.png"},{"id":93778474,"identity":"a9473a24-5760-43b5-b1e7-c1fe2c04ac30","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":41750,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/45afeda4cee5cd3e94924e43.png"},{"id":93778476,"identity":"01418642-c4bf-4a9e-8edc-05d18b9aa45b","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"xml","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":134136,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25790770structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/2c741b483a56f696b67e192c.xml"},{"id":93779928,"identity":"05ad2717-a24b-4d25-8262-8e60800503d8","added_by":"auto","created_at":"2025-10-17 12:55:07","extension":"html","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145821,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/9c912c664c1cac9289ed7dbf.html"},{"id":93778457,"identity":"00b9f2cd-7948-41ab-95b8-9ad310394630","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":122083,"visible":true,"origin":"","legend":"\u003cp\u003eShare of land dedicated to (a) pastures and (b) cropland. (c) Main crop per PSU determined based on the share of land dedicated to each major crop grown in the area.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/3c2cec9334ebb521f1b00099.jpg"},{"id":93778462,"identity":"baec609f-22c8-4482-8495-4b347d476493","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":108358,"visible":true,"origin":"","legend":"\u003cp\u003eAssociations between crop-growing season SPEI during child’s infancy and child stunting. In (a) SPEI is measured over the growing season for the main crop identified in the area. In (b) and (c) weighted SPEI is constructed by combining the SPEI values over multiple crop-growing seasons, where the weights are determined based on the share of land dedicated to each crop. In (b) all crop-growing seasons in the area are considered and in (c) the crop-growing seasons for specific crop groups are considered. Samples consist of children ages 1 to 5. All estimates are adjusted for potential confounders (see Methods).\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/9939006138910183ecc238dc.jpg"},{"id":93778459,"identity":"f455f966-d4e9-40d1-a2c3-27876c3c572a","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":86354,"visible":true,"origin":"","legend":"\u003cp\u003eAssociations between child stunting and drought (wtd SPEI\u0026lt;-1) and extreme precipitation (wtd SPEI≥1) during the child’s infancy period for the full sample and stratified by the agricultural land use practices. Samples consist of children ages 1 to 5. All estimates are adjusted for potential confounders (see Methods). Detailed results are available in SI Appendix 1, Tables S5.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/599732101722fa05adfa5609.jpg"},{"id":93778460,"identity":"f141a915-f67c-4423-b543-300656d36ab0","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":189183,"visible":true,"origin":"","legend":"\u003cp\u003eAssociations between drought (wtd SPEI\u0026lt;-1) during child’s infancy and child stunting by individual and household characteristics and stratified by agricultural land use. Samples consist of children ages 1 to 5. All estimates are adjusted for potential confounders (see Methods). Detailed results are available in SI Appendix 1, Table S9. Wald test is used to determine statistically significant pairwise differences in effect estimates between effect modifier categories. § and þ indicate pairwise differences that are statistically significant at 0.1 significance level or higher.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/e89037708fa321a2dd0c90b6.jpg"},{"id":93780297,"identity":"e9d2762e-02f9-4c4b-9e8b-4a71a6be102b","added_by":"auto","created_at":"2025-10-17 13:03:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1067506,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/77f87b72-2801-46a1-9b56-1975641dff6e.pdf"},{"id":93778461,"identity":"498962c4-2ca9-414e-8f75-d3060af86b0b","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":851053,"visible":true,"origin":"","legend":"Appendix 1","description":"","filename":"SIAppendix1.docx","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/0eb79ce6d7bd39fe869caeb1.docx"},{"id":93778471,"identity":"d7b3723c-005a-4958-b5f7-3b78e04525a6","added_by":"auto","created_at":"2025-10-17 12:47:07","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":3638247,"visible":true,"origin":"","legend":"Appendix 2","description":"","filename":"SIAppendix2.docx","url":"https://assets-eu.researchsquare.com/files/rs-7765713/v1/ebf0a17127e83f320a91b9bd.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Assessing the Impact of Agricultural Droughts on Child Undernutrition in Varying Crop-Growing Periods and Agricultural Land Use: Analysis across 33 Sub-Saharan African Countries","fulltext":[{"header":"Introduction","content":"\u003cp\u003eClimate conditions and stable precipitation regimes are critical for livelihoods in sub-Saharan Africa (SSA) given the high reliance on subsistence farming in the region\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Smallholder farmers constitute over half of the population in SSA and agriculture accounted for 17% of the region's GDP in 2022\u003csup\u003e2\u003c/sup\u003e. Hence, the loss of crop and livestock production due to precipitation variability poses a major threat to food security and economic well-being in the region\u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. The increased incidence of agricultural droughts due to climate change and its potential impact on child health is particularly concerning. Agricultural droughts, as opposed to meteorological droughts, can be driven by warmer temperatures in addition to precipitation deficits. Therefore, anthropogenic climate change can exacerbate agricultural drought by enhancing water demand directly via warming\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e as well as via precipitation regime change\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eDroughts can impact food security through multiple channels, directly by reducing crop yields and the amount of food available to households, and indirectly by eroding the economic resources of smallholder farmers and agricultural laborers and causing spikes and volatility in food prices\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. Droughts can also aggravate underlying socioeconomic conditions that impact food security, such as economic instability, conflict, and forced displacement\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. A growing body of literature shows that climate variability and extreme climate events, including droughts, are associated with an increased risk of experiencing food insecurity and malnutrition among households in low- and middle-income countries (LMICs)\u003csup\u003e\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. In fact, malnutrition has been recognized as one of the main health risks posed by global warming\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e,\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e. Young children are at heightened risk of undernutrition in households experiencing food insecurity. Children need sufficient, nutritious, and safe foods and not meeting these nutritional needs can seriously affect their long-term health and well-being\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e. Besides being one of the main causes of death for children under 5 years of age\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e,\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, undernutrition has been associated with loss of physical growth potential, diminished neurodevelopmental and cognitive function, and learning challenges in the classroom\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Children who are too short for their age, an indication of chronic undernutrition known as stunting, were found to have an increased risk of chronic health conditions and poorer socioeconomic status in adulthood\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e,\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. Malnourished children are also more susceptible to infectious diseases such as diarrhea and acute respiratory infections because of their compromised immunity\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, which has been associated with poor intake and absorption of nutrients and further growth retardation\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. The societal implications of child undernutrition can be substantial as there is evidence that poor development due to stunting can be transmitted to the subsequent generation of children\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e, which can perpetuate poverty and undermine the economic growth of nations where child undernutrition is widespread\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eCountries in SSA have some of the highest levels of child undernutrition in the world, with 31.5% of children under 5 in the region being stunted in 2022\u003csup\u003e28\u003c/sup\u003e. Despite some reduction in the prevalence of child stunting in SSA over the past decade, dropping 36.2% in 2012\u003csup\u003e28\u003c/sup\u003e, progress has been slow, possibly due to the long-lasting effects of the COVID-19 pandemic\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, recurrent harvest failures due to the impacts of climate change on the frequency of agricultural droughts\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, and other compounding factors\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. While droughts are known to pose a serious threat to food security\u003csup\u003e\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, the precise impact of these climate shocks on child health and the underlying mechanisms remain understudied. A recent systematic review\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e of 27 studies conducted in 20 different LMICs found that there is a dearth of rigorous research and a small number of observational studies that link droughts to undernutrition in children. While some studies found a positive association, no conclusive evidence was established in the systematic review due to multiple reasons, including small sample sizes, high risk of bias, and the lack of comparable estimates across studies\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e. One recent study combining data from 53 countries found an association between precipitation shocks and child stunting\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, however, the factors driving this association remained unclear. Overall, the existing literature provides little understanding of which periods and types of droughts are most critical for child nutrition and fails to consider the importance of agricultural practices, such as land allocation and types of crops cultivated. Yet, such information targeted in space and time is critical for developing targeted intervention strategies. With the recent development of spatially resolved agricultural\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e and climate\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e datasets that integrate satellite-derived information with traditional on-the-ground data sources, this paper is able to delve deeper into these questions. Another important gap in the literature concerns the identification of population groups most vulnerable to the impacts of drought. Such investigation can benefit the development of targeted intervention strategies. Identifying households that are most in need of assistance will be critical for allocating the limited resources available to LMICs most effectively.\u003c/p\u003e\u003cp\u003eWe address some of these gaps in the literature by combining spatially disaggregated climate and agricultural land-use information with nationally representative household survey data from 33 sub-Saharan African countries to investigate the relationship between precipitation shocks during key agricultural periods and child stunting. We focus on SSA because of the region\u0026rsquo;s high rates of child undernutrition and the growing impacts of climate change on food production systems\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. We compile anthropometric data (height-for-age) for 380,653 children aged 1 to 5 from 105 household surveys collected between 1990 and 2022. Using high spatial resolution data on the geographical distribution of crop areas for 15 main crops\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e, detailed crop calendar information for each crop\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e, and climatological data, we were able to capture climate conditions during key crop-growing periods. We calculated the Standardized Precipitation and Evapotranspiration Index (SPEI) from the climatological data and linked it with the household survey data to investigate the non-linear relationship between child stunting and sub-annual variations in climate conditions. Agricultural droughts are driven by the compounded effects of increased evaporation and transpiration caused by higher temperatures and precipitation deficits\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Hence, SPEI is well-suited for capturing agricultural droughts since it incorporates information on both temperature-related evapotranspiration and precipitation changes\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eWe focused on exposure to precipitation shocks during infancy since most growth faltering is shown to occur prior to the first 23 months of life\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e,\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e but we also investigated the impact of agricultural drought experienced at later ages. We additionally examine heterogeneities between communities characterized by predominant land-use practices (crop-farming, pastoralism, mixture of both, and non-agricultural activities) since the direction and magnitude of the association may vary depending on the community\u0026rsquo;s dependence on the natural environment and their ability to adapt to environmental changes\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e,\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. For example, there is evidence that live-stock herders may relocate to greener areas or sell livestock to smooth their consumption during dry periods\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e, while crop farmers may be more constrained in their response to an agricultural drought\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. Among communities whose livelihoods do not directly depend on agriculture, agricultural droughts may still impact child nutrition via food import disruptions and food price spikes, however, this has been understudied.\u003c/p\u003e\u003cp\u003eFinally, we investigate whether the association between droughts and child stunting is modified by households\u0026rsquo; socioeconomic characteristics such as relative wealth, education, and type of occupation. Little is known about the complex interaction of factors contributing to households\u0026rsquo; experience of food insecurity following a climate-related shock and how this affects children\u0026rsquo;s nutrition\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Households\u0026rsquo; ability to sell assets or diversify income may influence how well they can smooth consumption in such adverse situations. Parental education may also have a protective effect during droughts. Research indicates that maternal education is associated with behaviors that improve child nutrition, such as improved dietary quality and other care practices\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. Differences in childhood undernutrition by sex and birth order have also been identified in the literature, possibly due to the uneven allocation of scarce resources, including food, within the household, among other factors. However, it is unclear whether these factors play a role when households experience an external shock such as a drought. In what follows, we attempt to shed light on the link between agricultural drought and child undernutrition in SSA, investigating the interactive role of environmental and socioeconomic factors that may heighten population vulnerability to such climatic shocks.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eOur dataset consists of 380,653 children ages 1 to 5 from 33 countries in SSA for whom complete information was available on anthropometric measurements, geo-location, and individual and household characteristics. Considering the agricultural land distribution, cropland areas dominate in western Africa (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eb), particularly areas characterized by sub-humid climate (Supplementary Fig.\u0026nbsp;4), as well as in Uganda and the tropical highlands of Rwanda and Burundi. Pastures extend across the continent (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea) and are dominant in eastern Africa (e.g., Mozambique, Tanzania, and Zambia) and southern Africa (e.g., Lesotho, Eswatini, South Africa, and Namibia). Grains are the most common type of crops grown across the survey sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ec), while roots are commonly grown in the humid and sub-humid areas of central Africa. Based on the share of land dedicated to crop-farming and pastures in the area of the survey sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Methods), we split our sample into four groups: cropland-dominated areas (24% of our sample), mixed cropland and pasture areas (22%), pasture-dominated areas (31%), and areas where agricultural land accounts for less than 20% of the total land area (24% of the sample) (see Supplementary Fig.\u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). The prevalence of child stunting is slightly higher in mixed areas and cropland-dominated areas (44% and 43% of children, respectively) as compared to the rest (about 40% of children).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e we show the results of our main model where we estimate the association between localized SPEI during specific agricultural periods and child stunting adjusting for potential confounders (see Methods). In all models described below, we focus specifically on exposure to a climatic shock during the child\u0026rsquo;s infancy period. Considering SPEI during the main-crop growing period, we find a reverse J-shaped relationship with child stunting (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea), with a sharp and statistically significant increase in the risk of stunting during relatively dry periods (negative SPEI values) and a more shallow and not statistically significant increase in the risk of stunting during abnormally wet periods (positive SPEI values). Using weighted SPEI values, which combine multiple crop growth periods (see Methods), we find a more pronounced reverse J-shaped relationship with the risk of stunting (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). Examining the association for specific crop-growing periods, we find strong associations between stunting and relatively dry conditions during the grains and oilseeds growth periods and more shallow and non-significant associations for relatively wet conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec). We find non-significant associations between stunting and SPEI during the growth periods for roots and pulses (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe construct indicators for drought and extreme precipitation, based on standard SPEI thresholds proposed in the literature\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e, and conduct a stratified sample analysis where we distinguish by agricultural land specialization. For the rest of the analysis, we use the weighted SPEI measure aggregated over all crop-growing seasons since it showed a stronger association with child stunting than the main crop-growing season SPEI. In the full sample, we find that an agricultural drought (SPEI\u0026lt;-1) increases the risk of child stunting by 1.07 (95% CIs: 1.03 to 1.11; Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Among the cropland-dominated areas, this risk is more pronounced (aOR\u0026thinsp;=\u0026thinsp;1.18 with 95% CIs: 1.07\u0026ndash;1.29), while in the mixed areas it is comparable (aOR\u0026thinsp;=\u0026thinsp;1.08 with 95% CIs: 1.02\u0026ndash;1.15). In the pasture-dominated areas and the areas where less than 20% of the land is used for agriculture, we do not find an association between agricultural drought and child stunting. Extreme precipitation (SPIE\u0026thinsp;\u0026ge;\u0026thinsp;1) does not seem to pose a risk in any of the four agricultural groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWe additionally estimate the association with different categories of droughts \u0026ndash; from mild to severe (SI Appendix 1, Table S6). We find that the risk of stunting increases progressively with the severity of the drought event. We further explore how the association changes across different cropland and pasture distributions (SI Appendix 1, Table S7). We find the strongest associations between agricultural drought and child stunting in areas where the cropland area is 20% or more of the total land area and the pasture area is under 20% of the total land area. The higher the share of land used for pastures and the lower the share of land used for croplands, the weaker the association becomes. We also explore whether irrigation infrastructure may diminish the impact of agricultural droughts on child stunting. Generally, about 1 percent of the agricultural area is equipped for irrigation across the 33 countries included in the analysis. We find some evidence that irrigation infrastructure diminishes the impact of droughts on child stunting in mixed areas but not in the cropland-dominated areas (SI Appendix 1, Table S8). It must be noted that we do not have information on irrigation practices but only on whether the area is equipped for irrigation. We also do not know what type of irrigation system is available, whether it is functional and its capacity.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe additionally explore whether individual children\u0026rsquo;s and household\u0026rsquo;s sociodemographic factors modify the associations between agricultural droughts and child stunting. We use a Wald test for heterogeneity to identify statistically significant heterogeneity in the effect estimates across population subgroups. We carry out this analysis for the sample of children residing in cropland-dominated areas and mixed areas, and for the full sample of children (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). We do not find evidence of effect modification by birth order and sex. The Wald test did not show statistically significant differences by urban-rural place of residence and by mother\u0026rsquo;s level of education either, even though the effect estimates were stronger among children residing in rural areas as compared to urban areas, and among children whose mothers had no formal education as compared to children whose mothers had completed at least primary education.\u003c/p\u003e\u003cp\u003eWe find some evidence that the type of occupation of the household head modifies the association between agricultural droughts and child stunting in mixed cropland and pasture areas. Specifically, we observe a positive and statistically significant association between agricultural drought and stunting among children whose households are employed in the agricultural sector (aOR\u0026thinsp;=\u0026thinsp;1.13 with 95% CIs: 1.05\u0026ndash;1.22; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ee), whereas no association is found among children whose households are employed in non-agricultural sectors (aOR\u0026thinsp;=\u0026thinsp;1.001 with 95% CIs: 0.93\u0026ndash;1.08). The Wald test shows that the pairwise difference is statistically significant (\u003cem\u003ep\u003c/em\u003e-value\u0026thinsp;=\u0026thinsp;0.01). Within the cropland-dominated areas, we do not find evidence of effect modification by occupational group.\u003c/p\u003e\u003cp\u003eWe also find that household wealth modifies the association between agricultural droughts and child stunting for children in the cropland-dominated areas and for the full sample of children. Among children in the poorer and middle-wealth households in the cropland-dominated areas, we see a positive and statistically significant association between agricultural droughts and child stunting (aOR\u0026thinsp;=\u0026thinsp;1.27 with 95% CIs: 1.16\u0026ndash;1.43 for the poorer group and aOR\u0026thinsp;=\u0026thinsp;1.20 with 95% CIs: 1.08\u0026ndash;1.33 for the middle-wealth group, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003ef), whereas this association is close to null among children from wealthier households (aOR\u0026thinsp;=\u0026thinsp;1.04 with 95% CIs: 0.92\u0026ndash;1.19). The Wald test shows that these differences in the effect estimates are highly statistically significant (\u003cem\u003ep\u003c/em\u003e-value\u0026thinsp;=\u0026thinsp;0.003 for the pairwise difference between the poorer and wealthier groups and \u003cem\u003ep\u003c/em\u003e-value\u0026thinsp;=\u0026thinsp;0.02 for the pairwise difference between the middle-wealth and wealthier groups).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe find that the impact of agricultural droughts experienced during infancy can be lasting. We see that children exposed to an agricultural drought during infancy display a higher risk of stunting at ages 1, 2, and 3 as compared to children of the same age groups who were not exposed to a drought early in life (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) and the magnitude of the association remains about the same. Among children measured at age 4, we do not find a statistically significant association between exposure to an agricultural drought during infancy and the risk of stunting. We also find that agricultural droughts experienced at later ages affect children\u0026rsquo;s risk of stunting, especially within a few years after exposure. For example, among children who were measured at age 4, we see that an agricultural drought experienced at age 3 increases the risk of stunting by 1.13 (95% CIs: 1.04\u0026ndash;1.24) and a drought exposure at age 2 increases the risk of stunting by 1.09 (95% Cis: 1.00-1.18).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eAssociations between child stunting and drought (wtd SPEI\u0026lt;-1) measured at different ages of exposure and stratified by child\u0026rsquo;s age at measurement. All estimates are adjusted for potential confounders (see Methods).\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAge at measurement: 1\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAge at measurement: 2\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAge at measurement: 3\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eAge at measurement: 4\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eaOR [95% CIs]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eaOR [95% CIs]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eaOR [95% CIs]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eaOR [95% CIs]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDrought: infancy\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.09 [1.03\u0026ndash;1.15]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.10 [1.02\u0026ndash;1.18]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.09 [1.00-1.20]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.97 [0.90\u0026ndash;1.04]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDrought: age 1\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.09 [1.02\u0026ndash;1.16]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.10 [1.03\u0026ndash;1.18]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.03 [0.94\u0026ndash;1.13]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDrought: age 2\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.16 [1.08\u0026ndash;1.24]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.09 [1.00-1.18]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eDrought: age 3\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e1.13 [1.04\u0026ndash;1.24]\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIndividual \u0026amp; household-level covariates\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGeographical covariates\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eObs.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e104,725\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e98,893\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e90,431\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e86,533\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eWe conduct the largest study to date linking agricultural droughts to child undernutrition using data from 33 countries in sub-Saharan Africa collected over three decades (1990–2022). By explicitly considering multiple location-specific features such as agricultural land distribution, types of crops grown, and detailed crop calendar information, we were able to identify precise dose-response relationships that were not identifiable previously due to the lack of harmonized and geographically disaggregated agricultural and climatological data.\u003c/p\u003e\u003cp\u003eWe find compelling evidence that agricultural droughts are associated with an increased risk of child stunting in sub-Saharan Africa. The effect is most notable in places where a high fraction of the land is used for crop cultivation, implying that crop-farming is an important source of livelihood there, and less pronounced in areas where a high fraction of the land is used for pastures. These findings are in line with previous research, which shows high sensitivity of crop production to climate variability in Africa\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e, whereas the impact on livestock productivity was found to be less significant. A recent study that looked at the impact of climate variability on birth weight in Kenya and Mali also documented weaker associations among pastoral communities as compared to crop-farming communities\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. Temporary migration to greener areas, reducing livestock numbers, conserving feed resources and shifting to browse livestock species have all been documented as common coping strategies employed by pastoralists in response to droughts\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. However, these communities may still be vulnerable to longer-lasting, recurrent, and widespread droughts which can erode their resources over time\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e\u003c/sup\u003e. Pastoral communities usually occupy dryland areas that are less productive and are often economically marginalized. Future research should look more closely at how these communities are affected by climate change. Likewise, we did not find an association between child stunting and agricultural droughts in areas where a relatively low fraction of the land area is used for agriculture. In such areas, populations may be less dependent on agricultural resources as compared to other sources of livelihood or may be more adaptable to variable climate conditions. Such populations may still be impacted by food import disruptions due to drought or conflict further away, which should be investigated in future research.\u003c/p\u003e\u003cp\u003eOverall, we find that droughts during the growing seasons for grains and oilseeds are most strongly associated with child stunting. Grains, such as maize, millet, rice, sorghum, and wheat, are staple crops in SSA and contribute to most of the calorie availability there. Apart from reducing agricultural production\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e, climate change has been shown to impact the micronutrient content of food crops\u003csup\u003e\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e, which may have important consequences for child nutrition. Even though certain crops, such as certain oilseeds (e.g., cotton and soybean) are sold for profit (so called “cash crops”), decline in their yields may still impact food security via income shocks. Apart from the implications for food availability and agricultural income, droughts can impact child health through infectious illnesses\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e,\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e\u003c/sup\u003e. The presence of chronic infectious illnesses in children, especially during critical periods in early childhood, has been linked to hindered growth and development\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e\u003c/sup\u003e. This is particularly the case in areas where access to clean water and sanitation infrastructure is limited.\u003c/p\u003e\u003cp\u003eWe do not find an association between heavy precipitation and child growth, even though such an association has been documented in previous research\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. A reason for that could be our focus on precipitation shocks during the main agricultural periods rather than annual precipitation variability.\u003c/p\u003e\u003cp\u003ePoverty seems to be an important factor exacerbating the impact of agricultural droughts on child growth. We find the strongest effect among the poorest strata in cropland-dominated areas, whereas the association is close to null among the wealthiest strata. Smallholder and subsistence farmers are among the poorest groups in less developed countries and they are also among the most vulnerable to the effects of climate change since their livelihoods directly depend on natural resources\u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Drought-related crop loss and decreased agricultural income are likely to pressure poor households to reduce food consumption and cut back on non-food expenses, such as healthcare expenditure, compromising children’s health. Wealthier households may be less reliant on farming activities or may be able to smooth consumption by using savings or selling assets. When droughts persist, however, household assets may get depleted and no longer provide a buffer against agricultural shocks. For example, droughts have been shown to affect land-ownership rates in Africa\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e, implying a shift towards wage labor and rural-urban migration. Such developments should be considered when designing effective intervention strategies that protect vulnerable populations.\u003c/p\u003e\u003cp\u003eFinally, our findings suggest that the impact of agricultural droughts experienced early in life can persist. Children who experienced a drought during infancy have a higher risk of stunting at ages 1 to 3 and the magnitude of this association remains relatively constant. For children measured at age 4, we do not find an association between drought experienced in infancy and stunting, which could be explained by “catch-up” growth\u003csup\u003e\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e\u003c/sup\u003e and/or selective survival of the healthier children. Identifying the exact mechanism nevertheless would require panel data that follows children’s growth trajectories over their lives. For children measured at aged 4, we find an association between stunting and agricultural drought experienced in more recent periods (ages 2 and 3).\u003c/p\u003e\u003cp\u003eOur study has certain limitations that must be acknowledged. The survey data used in the analysis is cross-sectional and does not allow us to determine the longitudinal effects of agricultural droughts on child growth. Furthermore, there are multiple pathways, including food availability, poverty, and infectious diseases, which we are not able to disentangle with the available data. It would be particularly important to explore the contribution of multiple mechanisms on such relationships by capitalizing on datasets that collected data on child growth, nutrition as well as variables capturing such potential pathways, such as the World Bank Living Standards Measurement Study – Integrated Surveys on Agriculture (LSMS- ISA).\u003c/p\u003e\u003cp\u003eAnother limitation is that we use information on the distribution of land for croplands and pastures as well as specific crop types that is fixed around the year 2000. To the best of our knowledge, no longitudinal data are available at the spatial resolution and for the variety of crops used in this study. We also do not consider changes in the crop calendars, potential crop diversification and other adaptation measures, which may be employed in response to the impacts of climate change. We are also not able to capture land desertification\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e. We do not distinguish between subsistence agriculture and industrial crop cultivation, which may have very different implications for food security in SSA\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eFurthermore, even though we consider whether the area is equipped for irrigation, we are not able to distinguish between irrigated and non-irrigated farming systems, which may have important implications for drought resistance. Also, we are not able to determine the agricultural specializations at the household level and to distinguish between food croppers, who are usually subsistence farmers and may be more sensitive to agricultural shocks, and cash croppers. All of these limitations should be considered in future research.\u003c/p\u003e\u003cp\u003eIn conclusion, we find a strong association between agricultural droughts and child undernutrition in SSA using data from 33 countries spanning three decades. We show that vulnerability to the impact of drought can be influenced by factors at the community level, such as agricultural land allocation and crop specialization, as well as at the household level, such as relative wealth and occupation of the household head. In this study, we focus specifically on agricultural droughts, i.e., droughts that affect crop production. These can be driven by region-specific changes in precipitation regimes\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e,\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e and warming temperatures\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e caused by anthropogenic climate change as well as natural processes. Other types of droughts may also impact population health via changes to the ecological system and socioeconomic processes and should be addressed in future research. Moreover, the interaction of local climate conditions with global crises such as the COVID-19 pandemic\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e and the war in Ukraine\u003csup\u003e\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u003c/sup\u003e (a major wheat exporter to countries in SSA) may further undermine food security and child nutrition in the region, which should be studied in the future. Considering the additional challenges posed by climate change on achieving global nutrition goals, our study provides insights that can be used for targeting food aid programs and other interventions that strengthen food security in a changing climate.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eNutrition and covariate data\u003c/p\u003e\u003cp\u003eWe obtained anthropometric data on children under 5 years of age from the DHS program\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e. DHS surveys are conducted via a two-stage procedure, which ensures the representativeness of the samples at both the national and subnational levels. We included all surveys that contained child anthropometric information and global positioning system (GPS) coordinates of the primary sampling units (PSUs). PSUs are enumeration areas used in DHS which usually represent villages in rural areas and city blocks in urban areas. To maintain the confidentiality of survey participants, DHS displaces the PSU coordinates by 2km in urban areas, 5km in rural areas, and an extra 10km for 5% of all clusters. In total, we used information from 105 surveys collected across 33 countries between 1990 and 2022 (SI Appendix, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). While DHS surveys are repeatedly collected in certain countries, the surveys are cross-sectional as new households are interviewed in every survey round. The surveys are a rich source of information on fertility, reproductive behavior, mortality, health, and wellbeing in LMICs. Women in the selected households who are of reproductive age (15–49 years) are interviewed in-depth to gather details on their past pregnancies and the health of their children born in the 5 years preceding the survey. We infer children’s nutritional status from their height, which is measured by trained fieldworkers during the interviews. In particular, we calculated height-for-age z-scores (HAZ) for children under 5 years of age and based on the scores constructed a binary indicator for child stunting, which takes the value 1 if a child’s HAZ score is more than 2 standard deviations below the World Health Organization’s (WHO) growth standard median for children of the same age-group\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e and 0 otherwise. Observations were dropped if the HAZ scores were implausible (below − 6 and above 6). Low HAZ values reflect cumulative effects of sustained undernutrition and infections experienced since birth and even during pregnancy\u003csup\u003e\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e\u003c/sup\u003e. Our final sample consists 380,653 children aged 1 to 5 for whom complete anthropometric and sociodemographic information was available.\u003c/p\u003e\u003cp\u003eIn addition to children’s anthropometric measurements, we retrieve information about the demographic and socioeconomic characteristics of the children and their households: children’s sex and birth order, mother’s highest achieved level of education (none, primary, secondary and higher), type of occupation of the household head (agricultural or non-agricultural), place of residence (urban or rural) and household wealth group. We used this contextual information to explore potential heterogeneities in the effect size among population subgroups. Following standard DHS procedures\u003csup\u003e\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e\u003c/sup\u003e, we use principle component analysis to construct an index of household wealth by combining information about household ownership of select assets (radio, fridge, television, and car), building characteristics (material of the floor), and the availability of electricity and improved water and sanitation facilities on the premise. Children of women that are not permanent residents at the place of interview are excluded from the analysis to reduce the risk of misclassification.\u003c/p\u003e\u003cp\u003eAgricultural data\u003c/p\u003e\u003cp\u003eWe use global gridded dataset of cropland and pastures available at a high spatial resolution (~ 10km)\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e to determine the predominant land use practice around each PSU (within 10km radius). The data is representative for the year 2000 and provides greater accuracy than other data sources since it combines agricultural inventory data with satellite-derived land-cover datasets. We use the cropland and pasture data to distinguish between four types of areas: cropland-dominated (areas where over 20% of the land is cropland and the share of land used for growing crops is at least twice as high as the share of land used for pasture), pasture-dominated (areas where over 20% of the land is dedicated to pastures and the share of land dedicated to pastures is at least twice as high as the share of land dedicated to crops), mixed cropland and pasture areas (areas which do not fall in either of the above categories but where the combined share of land dedicated to crops and pastures is at least 20% of the total land area), and areas where less than 20% of the total land area is used for agriculture. Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e in SI Appendix 1 shows how the above groups are spatially distributed.\u003c/p\u003e\u003cp\u003eWe additionally retrieve information about the area harvested for main crop grown in SSA\u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e (SI Appendix 2). The data are available globally at ~ 10km spatial resolution and indicate multiple crops grown in each location. Crop planting and harvesting dates are additionally obtained\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e and used to determine the start and end day and the duration of the growing season for each crop. The growing season is defined as the period between the start of planting and the start of harvesting. Our final dataset includes information about 15 main crops grown in SSA (SI Appendix 2), which, for the purpose of our analysis, we group into: grains (barley, maize, millet, rice, sorghum, and wheat), oilseeds (cotton, groundnut, soybean, sunflower, and pulses), pulses, and roots (cassava, potato, sweet potato, and yam). PSUs for which no land area is used to grow crops are excluded from the analysis.\u003c/p\u003e\u003cp\u003eClimate data\u003c/p\u003e\u003cp\u003eWe use the SPEI as an indicator of agricultural droughts. The SPEI measures the variation in hydrological conditions during a certain period in relation to the area-specific long-term norm and it can take both negative values, indicating abnormally dry conditions, and positive values, indicating abnormally wet conditions. The SPEI is preferred over other drought indices since it can be computed over different temporal scales and considers how both temperature and precipitation contribute to the development of droughts\u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. We generate SPEI values using the global gridded precipitation and potential evapotranspiration data provided by the Climatic Research Unit (CRU) at the University of East Anglia (time series 4) at monthly temporal resolution and 0.5° spatial resolution and covering the period 1901- 2022\u003csup\u003e59\u003c/sup\u003e. The R package “SPEI” is used to compute the monthly SPEI values based on the input CRU data at different time scales considering the duration of each crop-growing season in every PSU location. The seasonal SPEI values for all crops grown in specific PSU are combined into a single index by computing a weighted SPEI value where the weights correspond to the share of land dedicated to each crop. Similarly, weighted SPEI values are generated for the crop-growing seasons of specific crop group (grains, roots, oilseeds, and pulses).\u003c/p\u003e\u003cp\u003eEstimation strategy\u003c/p\u003e\u003cp\u003eWe model the effect of droughts experienced during the period of infancy (the first year of life) on the child’s risk of stunting. We run a non-linear logistic regression, which takes the following form:\u003c/p\u003e\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:{y}_{i,j,r}=\\alpha\\:+{{\\beta\\:}\\varvec{X}}_{i,j,r}+\\:\\delta\\:{SPEI}_{j}+{A}_{r}+{\\text{ϵ}}_{i,j,a}$$\u003c/div\u003e\u003c/div\u003e\u003cp\u003eWhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{y}_{i,a}\\)\u003c/span\u003e\u003c/span\u003e is the nutrition status of child \u003cem\u003ei\u003c/em\u003e interviewed in PSU \u003cem\u003ej\u003c/em\u003e in administrative area (e.g., district or province) \u003cem\u003er\u003c/em\u003e. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{X}}_{i}\\)\u003c/span\u003e\u003c/span\u003eis a vector of individual, household, and geographic factors, including sex of the child, their age in months, bird order, size at birth (smaller than average, average, or larger than average), age of the mother at giving birth, mother’s level of education, occupation of the household head (agricultural or non-agricultural), household wealth group, urban or rural place of residence. Birth quarter and interview quarter are also included as covariates to account for seasonal effects in the child’s risk of stunting. We also include indicators for the main crop and the length of the main crop-growing season in the area. We additionally control for the share of land equipped for irrigation (see Fig. S3 in SI Appendix 1) and agroecological zone (Fig. S4 in SI appendix 1). \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\beta\\:}\\)\u003c/span\u003e\u003c/span\u003e is a vector of corresponding slope parameters for each covariate. \u003cem\u003eA\u003c/em\u003e is an intercept for the administrative area The most up-to-date subnational boundaries were retrieved from the DHS spatial data repository (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://spatialdata.dhsprogram.com/home/\u003c/span\u003e\u003cspan address=\"https://spatialdata.dhsprogram.com/home/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The subnational boundaries were harmonized in the case when changes occurred between survey rounds. \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{SPEI}_{i,j,r}\\)\u003c/span\u003e\u003c/span\u003e captures the drought conditions in each PSU site. We use a cubic spline transformation of the SPEI value with 4 equally spaced knots to allow for non-linearity in the relationship with stunting. The errors are clustered at the administrative level.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eYou L, Ringler C, Wood-Sichra U, et al. What is the irrigation potential for Africa? A combined biophysical and socioeconomic approach. \u003cem\u003eFood Policy\u003c/em\u003e. 2011;36(6):770-782. doi:10.1016/J.FOODPOL.2011.09.001\u003c/li\u003e\n\u003cli\u003eWorld Bank. World Development Indicators. Published online 2024. Accessed January 30, 2024. http://data.worldbank.org/data-catalog/world-development-indicators.\u003c/li\u003e\n\u003cli\u003eSultan B, Defrance D, Iizumi T. Evidence of crop production losses in West Africa due to historical global warming in two crop models. \u003cem\u003eSci Reports 2019 91\u003c/em\u003e. 2019;9(1):1-15. doi:10.1038/s41598-019-49167-0\u003c/li\u003e\n\u003cli\u003eOrtiz-Bobea A, Ault TR, Carrillo CM, Chambers RG, Lobell DB. Anthropogenic climate change has slowed global agricultural productivity growth. \u003cem\u003eNat Clim Chang 2021 114\u003c/em\u003e. 2021;11(4):306-312. doi:10.1038/s41558-021-01000-1\u003c/li\u003e\n\u003cli\u003eStige LC, Stave J, Chan KS, et al. The effect of climate variation on agro-pastoral production in Africa. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e. 2006;103(9):3049-3053. doi:10.1073/pnas.0600057103\u003c/li\u003e\n\u003cli\u003eWilliams AP, Cook ER, Smerdon JE, et al. Large contribution from anthropogenic warming to an emerging North American megadrought. \u003cem\u003eScience (80- )\u003c/em\u003e. 2020;368(6488):314-318. doi:10.1126/SCIENCE.AAZ9600/SUPPL_FILE/AAZ9600_WILLIAMS_SM.PDF\u003c/li\u003e\n\u003cli\u003ePolade SD, Pierce DW, Cayan DR, Gershunov A, Dettinger MD. The key role of dry days in changing regional climate and precipitation regimes. \u003cem\u003eSci Rep\u003c/em\u003e. 2014;4. doi:10.1038/srep04364\u003c/li\u003e\n\u003cli\u003ePolade SD, Gershunov A, Cayan DR, Dettinger MD, Pierce DW. Precipitation in a warming world: Assessing projected hydro-climate changes in California and other Mediterranean climate regions. \u003cem\u003eSci Reports 2017 71\u003c/em\u003e. 2017;7(1):1-10. doi:10.1038/s41598-017-11285-y\u003c/li\u003e\n\u003cli\u003eHadley K, Wheat S, Rogers HH, et al. Mechanisms underlying food insecurity in the aftermath of climate-related shocks: a systematic review. \u003cem\u003eLancet Planet Heal\u003c/em\u003e. 2023;7(3):e242-e250. doi:10.1016/S2542-5196(23)00003-7\u003c/li\u003e\n\u003cli\u003eFAO. \u003cem\u003eThe Impact of Disasters and Crises on Agriculture and Food Security\u003c/em\u003e.; 2018. Accessed January 27, 2020. www.fao.org/publications\u003c/li\u003e\n\u003cli\u003eAbel GJ, Brottrager M, Crespo Cuaresma J, Muttarak R. Climate, conflict and forced migration. \u003cem\u003eGlob Environ Chang\u003c/em\u003e. 2019;54:239-249. doi:10.1016/J.GLOENVCHA.2018.12.003\u003c/li\u003e\n\u003cli\u003eVon Uexkull N, Croicu M, Fjelde H, Buhaug H. Civil conflict sensitivity to growing-season drought. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e. 2016;113(44):12391-12396. doi:10.1073/PNAS.1607542113/SUPPL_FILE/PNAS.201607542SI.PDF\u003c/li\u003e\n\u003cli\u003eNiles MT, Salerno JD. A cross-country analysis of climate shocks and smallholder food insecurity. \u003cem\u003ePLoS One\u003c/em\u003e. 2018;13(2):e0192928. doi:10.1371/JOURNAL.PONE.0192928\u003c/li\u003e\n\u003cli\u003eBaker RE, Anttila-Hughes J. Characterizing the contribution of high temperatures to child undernourishment in Sub-Saharan Africa. \u003cem\u003eSci Rep\u003c/em\u003e. 2020;10(1). doi:10.1038/s41598-020-74942-9\u003c/li\u003e\n\u003cli\u003eCooper MW, Brown ME, Hochrainer-Stigler S, et al. Mapping the effects of drought on child stunting. \u003cem\u003eProc Natl Acad Sci\u003c/em\u003e. Published online August 12, 2019:201905228. doi:10.1073/pnas.1905228116\u003c/li\u003e\n\u003cli\u003eIPCC. \u003cem\u003eClimate Change 2022: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Climate Change 2021: The Physical Science Basis.\u003c/em\u003e; 2022. doi:10.1017/9781009325844.Front\u003c/li\u003e\n\u003cli\u003eHales S, Kovats S, Simon Lloyd DC-L. \u003cem\u003eQuantitative Risk Assessment of the Effects of Climate Change on Selected Causes of Death, 2030s and 2050s\u003c/em\u003e.; 2014. doi:ISBN 978 92 4 150769 1\u003c/li\u003e\n\u003cli\u003eMatrins VJB, Toledo Flor\u0026ecirc;ncio TMM, Grillo LP, et al. Long-lasting effects of undernutrition. \u003cem\u003eInt J Environ Res Public Health\u003c/em\u003e. 2011;8(6):1817-1846. doi:10.3390/ijerph8061817\u003c/li\u003e\n\u003cli\u003eBlack RE, Victora CG, Walker SP, et al. Maternal and child undernutrition and overweight in low-income and middle-income countries. \u003cem\u003eLancet\u003c/em\u003e. 2013;382(9890):427-451. doi:10.1016/S0140-6736(13)60937-X\u003c/li\u003e\n\u003cli\u003eArthur SS, Nyide B, Soura AB, Kahn K, Weston M, Sankoh O. Tackling malnutrition: a systematic review of 15-year research evidence from INDEPTH health and demographic surveillance systems. doi:10.3402/gha.v8.28298\u003c/li\u003e\n\u003cli\u003eVictora CG, Adair L, Fall C, et al. Maternal and child undernutrition: consequences for adult health and human capital. \u003cem\u003eLancet\u003c/em\u003e. 2008;371(9609):340-357. doi:10.1016/S0140-6736(07)61692-4\u003c/li\u003e\n\u003cli\u003ePoveda NE, Hartwig FP, Victora CG, et al. Patterns of Growth in Childhood in Relation to Adult Schooling Attainment and Intelligence Quotient in 6 Birth Cohorts in Low- and Middle-Income Countries: Evidence from the Consortium of Health-Oriented Research in Transitioning Societies (COHORTS). \u003cem\u003eJ Nutr\u003c/em\u003e. 2021;151(8):2342-2352. doi:10.1093/JN/NXAB096\u003c/li\u003e\n\u003cli\u003ePrendergast AJ, Humphrey JH. The stunting syndrome in developing countries. \u003cem\u003ePaediatr Int Child Health\u003c/em\u003e. 2014;34(4):250-265. doi:10.1179/2046905514Y.0000000158\u003c/li\u003e\n\u003cli\u003eMcGovern ME, Krishna A, Aguayo VM, Subramanian S V. A review of the evidence linking child stunting to economic outcomes. \u003cem\u003eInt J Epidemiol\u003c/em\u003e. 2017;46(4):1171-1191. doi:10.1093/ije/dyx017\u003c/li\u003e\n\u003cli\u003eWalson JL, Berkley JA. The impact of malnutrition on childhood infections. \u003cem\u003eCurr Opin Infect Dis\u003c/em\u003e. 2018;31(3):231-236. doi:10.1097/QCO.0000000000000448\u003c/li\u003e\n\u003cli\u003eWalker SP, Chang SM, Wright A, Osmond C, Grantham-McGregor SM. Early Childhood Stunting Is Associated with Lower Developmental Levels in the Subsequent Generation of Children. \u003cem\u003eJ Nutr\u003c/em\u003e. 2015;145(4):823-828. doi:10.3945/JN.114.200261\u003c/li\u003e\n\u003cli\u003eHeltberg R. Malnutrition, poverty, and economic growth. \u003cem\u003eHealth Econ\u003c/em\u003e. 2009;18 Suppl 1(SUPPL. 1). doi:10.1002/HEC.1462\u003c/li\u003e\n\u003cli\u003eUNICEF, WHO, World Bank. \u003cem\u003eLevel and Trend in Child Malnutrition: UNICEF/WHO/World Bank Group Joint Child Malnutrition Estimates: Key Findings of the 2023 Edition\u003c/em\u003e. UNICEF and WHO; 2023. Accessed January 21, 2024. https://www.who.int/publications/i/item/9789240073791\u003c/li\u003e\n\u003cli\u003eKansiime MK, Tambo JA, Mugambi I, Bundi M, Kara A, Owuor C. COVID-19 implications on household income and food security in Kenya and Uganda: Findings from a rapid assessment. \u003cem\u003eWorld Dev\u003c/em\u003e. 2021;137. doi:10.1016/j.worlddev.2020.105199\u003c/li\u003e\n\u003cli\u003eFAO, IFAD, UNICEF, WFP, WHO. \u003cem\u003eThe State of Food Security and Nutrition in the World 2023. Urbanization, Agrifood Systems Transformation and Healthy Diets across the Rural\u0026ndash;Urban Continuum\u003c/em\u003e. FAO; 2023. doi:10.4060/CC3017EN\u003c/li\u003e\n\u003cli\u003eBelesova K, Agabiirwe CN, Zou M, Phalkey R, Wilkinson P. Drought exposure as a risk factor for child undernutrition in low- and middle-income countries: A systematic review and assessment of empirical evidence. \u003cem\u003eEnviron Int\u003c/em\u003e. 2019;131. doi:10.1016/j.envint.2019.104973\u003c/li\u003e\n\u003cli\u003eRamankutty N, Evan AT, Monfreda C, Foley JA. Farming the planet: 1. Geographic distribution of global agricultural lands in the year 2000. \u003cem\u003eGlobal Biogeochem Cycles\u003c/em\u003e. 2008;22(1):1003. doi:10.1029/2007GB002952\u003c/li\u003e\n\u003cli\u003eMonfreda C, Ramankutty N, Foley JA. Farming the planet: 2. Geographic distribution of crop areas, yields, physiological types, and net primary production in the year 2000. \u003cem\u003eGlobal Biogeochem Cycles\u003c/em\u003e. 2008;22(1). doi:10.1029/2007GB002947\u003c/li\u003e\n\u003cli\u003eVicente-Serrano SM, Beguer\u0026iacute;a S, L\u0026oacute;pez-Moreno JI. A multiscalar drought index sensitive to global warming: The standardized precipitation evapotranspiration index. \u003cem\u003eJ Clim\u003c/em\u003e. 2010;23(7):1696-1718. doi:10.1175/2009JCLI2909.1\u003c/li\u003e\n\u003cli\u003eMorton JF. The impact of climate change on smallholder and subsistence agriculture. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e. 2007;104(50):19680-19685. doi:10.1073/pnas.0701855104\u003c/li\u003e\n\u003cli\u003eSacks WJ, Deryng D, Foley JA, Ramankutty N. Crop planting dates: an analysis of global patterns. \u003cem\u003eGlob Ecol Biogeogr\u003c/em\u003e. 2010;19(5):607-620. doi:10.1111/J.1466-8238.2010.00551.X\u003c/li\u003e\n\u003cli\u003eVicente-Serrano SM, Beguer\u0026iacute;a S, L\u0026oacute;pez-Moreno JI, Angulo M, Kenawy AE. A New Global 0.5\u0026deg; Gridded Dataset (1901\u0026ndash;2006) of a Multiscalar Drought Index: Comparison with Current Drought Index Datasets Based on the Palmer Drought Severity Index. \u003cem\u003eJ Hydrometeorol\u003c/em\u003e. 2010;11(4):1033-1043. doi:10.1175/2010JHM1224.1\u003c/li\u003e\n\u003cli\u003eAlderman H, Headey D. The timing of growth faltering has important implications for observational analyses of the underlying determinants of nutrition outcomes. \u003cem\u003ePLoS One\u003c/em\u003e. 2018;13(4):e0195904. doi:10.1371/JOURNAL.PONE.0195904\u003c/li\u003e\n\u003cli\u003eVictora CG, De Onis M, Hallal PC, Bl\u0026ouml;ssner M, Shrimpton R. Worldwide timing of growth faltering: Revisiting implications for interventions. \u003cem\u003ePediatrics\u003c/em\u003e. 2010;125(3). doi:10.1542/peds.2009-1519\u003c/li\u003e\n\u003cli\u003eHoffmann R, Wiederkehr C, Dimitrova A, Hermans K. Agricultural livelihoods, adaptation, and environmental migration in sub-Saharan drylands: a meta-analytical review. \u003cem\u003eEnviron Res Lett\u003c/em\u003e. 2022;17(8):083003. doi:10.1088/1748-9326/AC7D65\u003c/li\u003e\n\u003cli\u003eHabte M, Eshetu M, Maryo M, Andualem D, Legesse A. Effects of climate variability on livestock productivity and pastoralists perception: The case of drought resilience in Southeastern Ethiopia. \u003cem\u003eVet Anim Sci\u003c/em\u003e. 2022;16. doi:10.1016/J.VAS.2022.100240\u003c/li\u003e\n\u003cli\u003eBakhtsiyarava M, Grace K, Nawrotzki RJ. Climate, Birth Weight, and Agricultural Livelihoods in Kenya and Mali. \u003cem\u003eAm J Public Health\u003c/em\u003e. 2018;108(S2):S144-S150. doi:10.2105/AJPH.2017.304128\u003c/li\u003e\n\u003cli\u003eAlderman H, Headey D. How Important is Parental Education for Child Nutrition? \u003cem\u003eWorld Dev\u003c/em\u003e. 2017;94:448-464.\u003c/li\u003e\n\u003cli\u003eBeguer\u0026iacute;a S, Vicente-Serrano SM, Angulo-Mart\u0026iacute;nez M. A Multiscalar Global Drought Dataset: The SPEIbase: A New Gridded Product for the Analysis of Drought Variability and Impacts. \u003cem\u003eBull Am Meteorol Soc\u003c/em\u003e. 2010;91(10):1351-1356. doi:10.1175/2010BAMS2988.1\u003c/li\u003e\n\u003cli\u003eInman EN, Hobbs RJ, Tsvuura Z. No safety net in the face of climate change: The case of pastoralists in Kunene Region, Namibia. \u003cem\u003ePLoS One\u003c/em\u003e. 2020;15(9). doi:10.1371/JOURNAL.PONE.0238982\u003c/li\u003e\n\u003cli\u003eSemba RD, Askari S, Gibson S, Bloem MW, Kraemer K. The Potential Impact of Climate Change on the Micronutrient-Rich Food Supply. \u003cem\u003eAdv Nutr\u003c/em\u003e. 2022;13(1):80. doi:10.1093/ADVANCES/NMAB104\u003c/li\u003e\n\u003cli\u003eWang P, Asare E, Pitzer VE, Dubrow R, Chen K. Associations between long-term drought and diarrhea among children under five in low- and middle-income countries. \u003cem\u003eNat Commun 2022 131\u003c/em\u003e. 2022;13(1):1-10. doi:10.1038/s41467-022-31291-7\u003c/li\u003e\n\u003cli\u003eDimitrova A, McElroy S, Levy M, Gershunov A, Benmarhnia T. Precipitation variability and risk of infectious disease in children under 5 years for 32 countries: a global analysis using Demographic and Health Survey data. \u003cem\u003eLancet Planet Heal\u003c/em\u003e. 2022;6(2):e147-e155. doi:10.1016/S2542-5196(21)00325-9\u003c/li\u003e\n\u003cli\u003eSchaible UE, Kaufmann SHE. Malnutrition and Infection: Complex Mechanisms and Global Impacts. \u003cem\u003ePLOS Med\u003c/em\u003e. 2007;4(5):e115. doi:10.1371/JOURNAL.PMED.0040115\u003c/li\u003e\n\u003cli\u003eLinke AM, Tollefsen AF. Environmental stress and agricultural landownership in Africa. \u003cem\u003eGlob Environ Chang\u003c/em\u003e. 2021;67:102237. doi:10.1016/J.GLOENVCHA.2021.102237\u003c/li\u003e\n\u003cli\u003eDesmond C, Casale D. Catch-up growth in stunted children: Definitions and predictors. Schooling CM, ed. \u003cem\u003ePLoS One\u003c/em\u003e. 2017;12(12):e0189135. doi:10.1371/journal.pone.0189135\u003c/li\u003e\n\u003cli\u003eUnited Nations Economic and Social Council. \u003cem\u003eAfrica Review Report on Drought and Desertification\u003c/em\u003e.; 2007. doi:E/ECA/ACSD/5/3\u003c/li\u003e\n\u003cli\u003eJarzebski MP, Ahmed A, Boafo YA, et al. Food security impacts of industrial crop production in sub-Saharan Africa: a systematic review of the impact mechanisms. \u003cem\u003eFood Secur\u003c/em\u003e. 2020;12(1):105-135. doi:10.1007/S12571-019-00988-X/TABLES/3\u003c/li\u003e\n\u003cli\u003eICF. The DHS Program website [various datasets]. Funded by USAID. Accessed January 17, 2024. http://www.dhsprogram.com\u003c/li\u003e\n\u003cli\u003eWorld Health Organiozation. \u003cem\u003eNutrition Landscape Information System (NLiS) Country Profile Indicators: Interpretation Guide\u003c/em\u003e.; 2019.\u003c/li\u003e\n\u003cli\u003ede Onis M. Child Growth and Development. In: \u003cem\u003eNutrition and Health in a Developing World\u003c/em\u003e. Springer International Publishing; 2017:119-141. doi:10.1007/978-3-319-43739-2_6\u003c/li\u003e\n\u003cli\u003eRutstein SO. \u003cem\u003eSteps to Constructing the New DHS Wealth Index\u003c/em\u003e.; 2015.\u003c/li\u003e\n\u003cli\u003eVicente-Serrano SM, Beguer\u0026iacute;a S, L\u0026oacute;pez-Moreno JI, Angulo M, El Kenawy A. A New Global 0.5\u0026deg; Gridded Dataset (1901\u0026ndash;2006) of a Multiscalar Drought Index: Comparison with Current Drought Index Datasets Based on the Palmer Drought Severity Index. \u003cem\u003eJ Hydrometeorol\u003c/em\u003e. 2010;11(4):1033-1043.\u003c/li\u003e\n\u003cli\u003eHarris I, Osborn TJ, Jones P, Lister D. Version 4 of the CRU TS monthly high-resolution gridded multivariate climate dataset. \u003cem\u003eSci Data 2020 71\u003c/em\u003e. 2020;7(1):1-18. doi:10.1038/s41597-020-0453-3\u003c/li\u003e\n\u003cli\u003eMottaleb KA, Kruseman G, Snapp S. Potential impacts of Ukraine-Russia armed conflict on global wheat food security: A quantitative exploration. \u003cem\u003eGlob Food Sec\u003c/em\u003e. 2022;35:100659. doi:10.1016/J.GFS.2022.100659\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"drought, agriculture, undernutrition, food security, sub-Saharan Africa","lastPublishedDoi":"10.21203/rs.3.rs-7765713/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7765713/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Countries in sub-Saharan Africa (SSA) face persistent challenges in addressing child malnutrition, with approximately one-third of children under the age of five in the region experiencing chronic undernourishment. Previous studies have demonstrated that agricultural droughts contribute to child undernutrition; however, they have not adequately accounted for variations in agricultural land specialization and corresponding crop growing periods. In this study, we aim to address these heterogeneities by identifying vulnerable communities and specific exposure periods. We utilize anthropometric data for over 300,000 children ages 1 to 5 collected in nationally representative household surveys in 33 countries in SSA over three decades (1990-2022) and employ a novel approach that integrates detailed crop calendar information (crop planting and harvesting dates) with high-resolution agricultural and climatological datasets to detect droughts during specific crop-growing periods. The Standardized Precipitation Evapotranspiration Index (SPEI), a multi-scalar drought index, is used to measure the intensity and spatial distribution of droughts. We explore community-level heterogeneities determined by agricultural land specialization (crop farming or pastoralism) and the types of crops cultivated. Additionally, socioeconomic factors underlying vulnerability are explored. Our analysis reveals a significant association between agricultural droughts and the risk of child undernutrition among communities where agricultural land-use practices are dominated by crop farming. Specifically, droughts occurring during the growing seasons for grains and oilseeds exhibit a strong association with child undernutrition. Furthermore, residing in a rural area, being employed in agriculture, and belonging to a lower socioeconomic class are found to amplify the risk of undernutrition related to agricultural droughts among these communities. The findings presented in this study call for urgent action to improve drought monitoring and response in SSA where the risks to child health posed by global warming are considerable. Under climate change, the severity and frequency of extreme weather and climate events, including droughts, are projected to increase. This will place millions of children at risk of malnutrition unless timely action plans are taken to improve food security in the region.","manuscriptTitle":"Assessing the Impact of Agricultural Droughts on Child Undernutrition in Varying Crop-Growing Periods and Agricultural Land Use: Analysis across 33 Sub-Saharan African Countries","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-17 12:47:02","doi":"10.21203/rs.3.rs-7765713/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"a343866f-a9f1-450f-8c1a-eb833bbdf6cd","owner":[],"postedDate":"October 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":56366919,"name":"Earth and environmental sciences/Environmental social sciences/Climate-change impacts/Environmental health"},{"id":56366920,"name":"Earth and environmental sciences/Climate sciences/Climate change/Climate-change impacts/Environmental health"}],"tags":[],"updatedAt":"2025-10-17T12:47:02+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-17 12:47:02","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7765713","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7765713","identity":"rs-7765713","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}