Household-level spatial modelling of child height-for-age in Northern Province, Rwanda: comparison of geographically weighted and neural network weighted regression models

Childhood stunting remains a major public health concern in low- and middle-income countries, with persistent sub-national disparities. This study examined spatially varying associations between Height-for-Age z-scores (HAZ) and selected child, maternal, and household factors, and compared Geographically Weighted Regression (GWR), Multi-scale GWR (MGWR) and Geographically Neural Network Weighted Regression (GNNWR) in their ability to capture spatial heterogeneity and non-linearity. We analysed data from a cross-sectional survey conducted in December 2021 in the Northern Province of Rwanda. The survey covered 615 households, with analyses performed for 601 children aged 1–36 months. After imputation, multicollinearity screening, and feature selection, HAZ was modelled using Ordinary Least Squares (OLS), GWR, MGWR and GNNWR. Model performance was assessed using coefficient of determination (R2), Root Mean Square Error (RMSE), Akaike Information Criterion/AIC Corrected (AIC/AICc), and Moran’s I of residuals. Overall stunting prevalence was 27.1%. Spatially varying associations were observed for child age, sex, birthweight, underweight status, selected childcare practices, maternal support, and household living conditions. GNNWR achieved the best training performance (R2 = 0.66; RMSE=0.74) followed by MGWR (R2 =0.51; RMSE=0.88) and GWR (R2 =0.43; RMSE=0.95). Validation performance declined for all models, although GNNWR retained a higher validation (R2=0.28) with negligible residual spatial autocorrelation (Moran’s I = −0.005). GNNWR provided the strongest overall fit among the compared models and highlighted local spatial variation in associations with HAZ. However, the modest validation performance indicates that the findings should be interpreted cautiously. The workflow offers a portable framework for analysing DHS-like geocoded household datasets, but broader applicability should be assessed using spatial cross-validation, and sensitivity analyses.

Amegbor PM, Zhang Z, Dalgaard R, Sabel CE, 2020. Multilevel and spatial analyses of childhood malnutrition in Uganda: Examining individual and contextual factors. Sci Rep 10:20019. DOI: https://doi.org/10.1038/s41598-020-76856-y

Amir-ud-Din R, Fawad S, Naz L, Zafar S, Kumar R, Pongpanich S, 2022. Nutritional inequalities among under-five children: A geospatial analysis of hotspots and cold spots in 73 low- and middle-income countries. Int J Equity Health 21:135. DOI: https://doi.org/10.1186/s12939-022-01733-1

Amoroso CL, Nisingizwe MP, Rouleau D, Thomson DR, Kagabo DM, Bucyana T, Drobac P, Ngabo F, 2018. Next wave of interventions to reduce under-five mortality in Rwanda: A cross-sectional analysis of demographic and health survey data. BMC Pediatr 18:27. DOI: https://doi.org/10.1186/s12887-018-0997-y

Anselin L, 1995. Local Indicators of Spatial Association—LISA. Geograph Anal 27:93–115. DOI: https://doi.org/10.1111/j.1538-4632.1995.tb00338.x

Anselin L, Bera AK, Florax R, Yoon MJ, 1996. Simple diagnostic tests for spatial dependence. Regional Sci Urban Econ 26:77–104. DOI: https://doi.org/10.1016/0166-0462(95)02111-6

Archer KJ, Kimes RV, 2008. Empirical characterization of random forest variable importance measures. Computat Stat Data Analysis 52:2249–60. DOI: https://doi.org/10.1016/j.csda.2007.08.015

Atsbeha DM, Nayga RM, Rickertsen K, 2015. Can prolonged breastfeeding duration impair child growth? Evidence from rural Ethiopia. Food Policy 53:46–53. DOI: https://doi.org/10.1016/j.foodpol.2015.03.010

Azur MJ, Stuart EA, Frangakis C, Leaf PJ, 2011. Multiple imputation by chained equations: What is it and how does it work? Int J Methods Psychiatr Res 20:40–9. DOI: https://doi.org/10.1002/mpr.329

Benjamin-Chung J, Mertens A, Colford JM Jr, Hubbard AE, van der Laan MJ, Coyle J, Sofrygin O, Cai W, Nguyen A, Pokpongkiat NN, Djajadi S, Seth A, Jilek W, Jung E, Chung EO, Rosete S, Hejazi N, Malenica I, Li H, Hafen R, Subramoney V, Häggström J, Norman T, Brown KH, Christian P, Arnold BF; Ki Child Growth Consortium, 2023. Early-childhood linear growth faltering in low- and middle-income countries. Nature 621:550–7. DOI: https://doi.org/10.1038/s41586-023-06418-5

Binagwaho A, Scott KW, Harward SH, 2016. Early childhood development in Rwanda: A policy analysis of the human rights legal framework. BMC Int Health Human Rights 16:1. DOI: https://doi.org/10.1186/s12914-016-0076-0

Biswas M, 2022. Identifying geographical heterogeneity in associations between under-five child nutritional status and its correlates across Indian districts. Spatial Demography 10:143–87. DOI: https://doi.org/10.1007/s40980-022-00104-2

Black RE, Allen LH, Bhutta ZA, Caulfield LE, De Onis M, Ezzati M, Mathers C, Rivera J, 2008. Maternal and child undernutrition: Global and regional exposures and health consequences. Lancet, 371:243–60. DOI: https://doi.org/10.1016/S0140-6736(07)61690-0

Black RE, Victora CG, Walker SP, Bhutta ZA, Christian P, De Onis M, Ezzati M, Grantham-McGregor S, Katz J, Martorell R, Uauy R, 2013. Maternal and child undernutrition and overweight in low-income and middle-income countries. Lancet 382:427–51. DOI: https://doi.org/10.1016/S0140-6736(13)60937-X

Breiman L, 2001. Random Forests. Machine Learning 45:5–32. DOI: https://doi.org/10.1023/A:1010933404324

Comber A, Brunsdon C, Charlton M, Dong G, Harris R, Lu B, Lü Y, Murakami D, Nakaya T, Wang Y, Harris P, 2023. A route map for successful applications of geographically weighted regression. Geograph Anal 55:155–78. DOI: https://doi.org/10.1111/gean.12316

Daelmans B, Manji SA, Raina N, 2021. Nurturing care for early childhood development: global perspective and guidance. Indian Pediatr 58:11–15. DOI: https://doi.org/10.1007/s13312-021-2349-5

Du Z, Wang Z, Wu S, Zhang F, Liu R, 2020. Geographically neural network weighted regression for the accurate estimation of spatial non-stationarity. Int J Geograph Informat Sci 34:1353–77. DOI: https://doi.org/10.1080/13658816.2019.1707834

Dusingizimana T, Weber JL, Ramilan T, Iversen PO, Brough L, 2021. An empirical study of factors associated with height-for-age z-scores of children aged 6−23 months in northwest Rwanda: The role of care practices related to child feeding and health. Br J Nutrition 126:1203–14. DOI: https://doi.org/10.1017/S0007114520004961

Ekholuenetale M, Okonji OC, Nzoputam CI, Edet CK, Wegbom AI, Arora A, 2023. Socioeconomic disparities in Rwanda’s under-5 population’s growth tracking and nutrition promotion: Findings from the 2019–2020 demographic and health survey. BMC Pediatrics 23:467. DOI: https://doi.org/10.1186/s12887-023-04284-8

Eryando T, Sipahutar T, Budhiharsana MP, Siregar KN, Nur Aidi M, Minarto D, Utari DM, Rahmaniati M, Hendarwan H, 2022. Spatial analysis of stunting determinants in 514 Indonesian districts/cities: Implications for intervention and setting of priority. Geospat Health 17:1055. DOI: https://doi.org/10.4081/gh.2022.1055

Fotheringham AS, Brunsdon C, Charlton M, 2003. Geographically weighted regression: The analysis of spatially varying relationships. John Wiley & Sons.

Fotheringham AS, Wong DWS, 1991. The modifiable areal unit problem in multivariate statistical analysis. Environ Planning A: Economy Space 23:1025–44. DOI: https://doi.org/10.1068/a231025

Fotheringham AS, Yang W, Kang W, 2017. Multiscale Geographically Weighted Regression (MGWR). Ann Am Assoc Geographers 107:1247–65. DOI: https://doi.org/10.1080/24694452.2017.1352480

Getis A, 2007. Reflections on spatial autocorrelation. Regional Science and Urban Economics at 35. A Retrospective/Prospective Special Issue, 37:491–6. DOI: https://doi.org/10.1016/j.regsciurbeco.2007.04.005

Hailu BA, Bogale GG, Beyene J, 2020. Spatial heterogeneity and factors influencing stunting and severe stunting among under-5 children in Ethiopia: Spatial and multilevel analysis. Sci Rep 10:16427. DOI: https://doi.org/10.1038/s41598-020-73572-5

Haque S, Price A, Mengersen K, Hu W, 2025. Evaluating the impact of the Modifiable Areal Unit Problem on ecological model inference: A case study of COVID-19 data in Queensland, Australia. Infect Dis Modelling 10:1002–19. DOI: https://doi.org/10.1016/j.idm.2025.05.003

He K, Zhang X, Ren S, Sun J, 2015. delving deep into rectifiers: surpassing human-level performance on ImageNet Classification. 2015 IEEE International Conference on Computer Vision (ICCV) 1026–1034. DOI: https://doi.org/10.1109/ICCV.2015.123

Ioffe S, Szegedy C, 2015. Batch normalization: accelerating deep network training by reducing internal covariate shift. In F. Bach & D. Blei (Eds), Proceedings of the 32nd International Conference on Machine Learning 37:448–56.

James G, Witten D, Hastie T, Tibshirani R, Taylor J, 2023. Linear regression. In An introduction to statistical learning: With applications in python (pp. 69–134). Springer. DOI: https://doi.org/10.1007/978-3-031-38747-0_3

Kalinda C, Phiri M, Simona SJ, Banda A, Wong R, Qambayot MA, Ishimwe SMC, Amberbir A, Abebe B, Gebremariam A, Nyerere JO, 2023. Understanding factors associated with rural-urban disparities of stunting among under-five children in Rwanda: A decomposition analysis approach. Mat Child Nutr 19:e13511. DOI: https://doi.org/10.1111/mcn.13511

Khan J, Mohanty SK, 2018. Spatial heterogeneity and correlates of child malnutrition in districts of India. BMC Public Health 18:1027. DOI: https://doi.org/10.1186/s12889-018-5873-z

Kinyoki DK, Osgood-Zimmerman AE, Pickering BV, Schaeffer LE, Marczak LB, Lazzar-Atwood A, Collison ML, Henry NJ, Abebe Z, Adamu AA, Adekanmbi V, Ahmadi K, Ajumobi O, Al-Eyadhy A, Al-Raddadi RM, Alahdab F, Alijanzadeh M, Alipour V, Altirkawi K, … Local Burden of Disease Child Growth Failure Collaborators. (2020). Mapping child growth failure across low- and middle-income countries. Nature 577:231–4. DOI: https://doi.org/10.1038/s41586-019-1878-8

Klein N, Kneib T, Marra G, Radice R, 2020. Chapter 5—Bayesian mixed binary-continuous copula regression with an application to childhood undernutrition. In Y. Fan, D. Nott, M. S. Smith, & J.-L. Dortet-Bernadet (Eds), Flexible Bayesian Regression Modelling Academic Press. pp. 121–152. DOI: https://doi.org/10.1016/B978-0-12-815862-3.00011-1

Kramer MS, Moodie EEM, Dahhou M, Platt RW, 2011. Breastfeeding and infant size: evidence of reverse causality. Am J Epidemiol 173(9), 978–983. DOI: https://doi.org/10.1093/aje/kwq495

Kuse KA, Debeko DD, 2023. Spatial distribution and determinants of stunting, wasting and underweight in children under-five in Ethiopia. BMC Public Health 23:641. DOI: https://doi.org/10.1186/s12889-023-15488-z

Leung Y, Mei C-L, Zhang W-X, 2000. Statistical tests for spatial nonstationarity based on the geographically weighted regression Model. Environ Planning A 32:9–32. DOI: https://doi.org/10.1068/a3162

Luo Y, Yan J, McClure S, 2021. Distribution of the environmental and socioeconomic risk factors on COVID-19 death rate across continental USA: a spatial nonlinear analysis. Environ Sci Pollut Res 28:6587–99. DOI: https://doi.org/10.1007/s11356-020-10962-2

Ma Z, Huang Z. 2023. A Bayesian implementation of the multiscale geographically weighted regression model with INLA. Ann Am Assoc Geographers 113:1501–15. DOI: https://doi.org/10.1080/24694452.2023.2187756

Marquis GS, Habicht JP, Lanata CF, Black RE, Rasmussen KM, 1997. Association of breastfeeding and stunting in Peruvian toddlers: An example of reverse causality. Int J Epidemiol 26:349–56. DOI: https://doi.org/10.1093/ije/26.2.349

McMahan LD, Sprague C, 2024. The varied perspectives of organisational effectiveness: What’s at stake for early childhood development programmes in Rwanda? Global Public Health 19:2377280. DOI: https://doi.org/10.1080/17441692.2024.2377280

Mertens A, Benjamin-Chung J, Colford JM Jr, Coyle J, van der Laan MJ, Hubbard AE, Rosete S, Malenica I, Hejazi N, Sofrygin O, Cai W, Li H, Nguyen A, Pokpongkiat NN, Djajadi S, Seth A, Jung E, Chung EO, Jilek W, Subramoney V, Hafen R, Häggström J, Norman T, Brown KH, Christian P, Arnold BF; Ki Child Growth Consortium, 2023. Causes and consequences of child growth faltering in low-resource settings. Nature 621:568–76.

Miconi D, Beeman I, Robert E, Beatson J, Ruiz-Casares M, 2018. Child supervision in low- and middle-income countries: A scoping review. Children Youth Serv Rev 89:226–42. DOI: https://doi.org/10.1016/j.childyouth.2018.04.040

Moonga G, Böse-O’Reilly S, Berger U, Harttgen K, Michelo C, Nowak D, Siebert U, Yabe J, Seiler J, 2021. Modelling chronic malnutrition in Zambia: A Bayesian distributional regression approach. PLOS ONE 16:e0255073. DOI: https://doi.org/10.1371/journal.pone.0255073

Muche A, Gezie LD, Baraki AG, Amsalu ET, 2021. Predictors of stunting among children age 6–59 months in Ethiopia using Bayesian multi-level analysis. Sci Rep 11:3759. DOI: https://doi.org/10.1038/s41598-021-82755-7

NISR [Rwanda], Ministry of Health (MOH) [Rwanda], & ICF. (2021). Rwanda Demographic and Health Survey 2019-20 Final Report. Available from: https://dhsprogram.com/pubs/pdf/FR370/FR370.pdf

Osgood-Zimmerman A, Millear AI, Stubbs RW, Shields C, Pickering BV, Earl L, Graetz N, Kinyoki DK, Ray SE, Bhatt S, Browne AJ, Burstein R, Cameron E, Casey DC, Deshpande A, Fullman N, Gething PW, Gibson H, Henry NJ, … Hay SI, 2018. Mapping child growth failure in Africa between 2000 and 2015. Nature 555:41–47. DOI: https://doi.org/10.1038/nature25760

Oshan TM, Li Z, Kang W, Wolf LJ, Fotheringham AS, 2019. mgwr: A Python Implementation of Multiscale Geographically Weighted Regression for Investigating Process Spatial Heterogeneity and Scale. ISPRS Int J Geo-Information 8:269. DOI: https://doi.org/10.3390/ijgi8060269

Perkins JM, Kim R, Krishna A, McGovern M, Aguayo VM, Subramanian SV, 2017. Understanding the association between stunting and child development in low- and middle-income countries: Next steps for research and intervention. Soc Sci Med 193:101–9. DOI: https://doi.org/10.1016/j.socscimed.2017.09.039

Prechelt L, 1998. Automatic early stopping using cross validation: Quantifying the criteria. Neural Networks, 11:761–7. DOI: https://doi.org/10.1016/S0893-6080(98)00010-0

Puri P, Khan J, Shil A, Ali M, 2020. A cross-sectional study on selected child health outcomes in India: Quantifying the spatial variations and identification of the parental risk factors. Sci Rep 10:6645. DOI: https://doi.org/10.1038/s41598-020-63210-5

Quamme SH, Iversen PO, 2022. Prevalence of child stunting in Sub-Saharan Africa and its risk factors. Clinical Nutrition Open Sci 42:49–61. DOI: https://doi.org/10.1016/j.nutos.2022.01.009

Quiñones S, Goyal A, Ahmed ZU, 2021. Geographically weighted machine learning model for untangling spatial heterogeneity of type 2 diabetes mellitus (T2D) prevalence in the USA. Sci Rep 11:6955. DOI: https://doi.org/10.1038/s41598-021-85381-5

Ruiz-Casares M, Nazif-Muñoz JI, Iwo R, Oulhote Y, 2018. Nonadult supervision of children in low- and middle-income countries: results from 61 national population-based surveys. Int J Environ Res Public Health 15:1564. DOI: https://doi.org/10.3390/ijerph15081564

Samuel VW, 2023. miceforest: Fast Imputation with Random Forests in Python. Available from: https://github.com/AnotherSamWilson/miceforest/

Seboka BT, Hailegebreal S, Mamo TT, Yehualashet DE, Gilano G, Kabthymer RH, Ewune HA, Kassa R, Debisa MA, Yawo MN, Endashaw H, Demeke AD, Tesfa GA, 2022. Spatial trends and projections of chronic malnutrition among children under 5 years of age in Ethiopia from 2011 to 2019: A geographically weighted regression analysis. J Health Population Nutr 41:28. DOI: https://doi.org/10.1186/s41043-022-00309-7

Siqi J, Yuhong W, Ling C, Xiaowen B, 2023. A novel approach to estimating urban land surface temperature by the combination of geographically weighted regression and deep neural network models. Urban Climate 47:101390. DOI: https://doi.org/10.1016/j.uclim.2022.101390

Ssentongo P, Ssentongo AE, Ba DM, Ericson JE, Na M, Gao X, Fronterre C, Chinchilli VM, Schiff SJ, 2021. Global, regional and national epidemiology and prevalence of child stunting, wasting and underweight in low-and middle-income countries, 2006–2018. Sci Rep 11:1–12. DOI: https://doi.org/10.1038/s41598-021-84302-w

Striessnig E, Bora JK, 2020. Under-five child growth and nutrition status: spatial clustering of Indian districts. Spatial Demography 8:63–84. DOI: https://doi.org/10.1007/s40980-020-00058-3

Sun Y, Jia W, Zhu W, Zhang X, Saidahemaiti S, Hu,T, Guo H, 2022. Local neural-network-weighted models for occurrence and number of down wood in natural forest ecosystem. Sci Rep 12:6375. DOI: https://doi.org/10.1038/s41598-022-10312-x

Tamir TT, Tekeba B, Mekonen EG, Zegeye AF, Gebrehana DA, 2024. Spatial heterogeneity and predictors of stunting among under five children in Mozambique: A geographically weighted regression. Front Public Health 12:1502018. DOI: https://doi.org/10.3389/fpubh.2024.1502018

UNICEF, WHO, & World Bank. 2023. Levels and trends in child malnutrition: Key findings of the 2023 Edition of the Joint Child Malnutrition Estimates. New York: UNICEF and WHO.

Utumatwishima JN, Mogren I, Elfving K, Umubyeyi A, Krantz G, 2025. Association between poor mental health in mothers and child stunting: A population-based cross-sectional study in Rwanda. BMJ Open 15:e101117. DOI: https://doi.org/10.1136/bmjopen-2025-101117

Uwiringiyimana V, Osei F, Amer S, Veldkamp A, 2022. Bayesian geostatistical modelling of stunting in Rwanda: Risk factors and spatially explicit residual stunting burden. BMC Public Health 22:159. DOI: https://doi.org/10.1186/s12889-022-12552-y

Vaivada T, Akseer N, Akseer S, Somaskandan A, Stefopulos M, Bhutta ZA, 2020. Stunting in childhood: An overview of global burden, trends, determinants, and drivers of decline. Am J Clin Nutrition 112:777S-91S. DOI: https://doi.org/10.1093/ajcn/nqaa159

Victora CG, Adair L, Fall C, Hallal PC, Martorell R, Richter L, Sachdev HS, 2008. Maternal and child undernutrition: Consequences for adult health and human capital. Lancet 371:340–57. DOI: https://doi.org/10.1016/S0140-6736(07)61692-4

WHO, 2006. WHO child growth standards: Length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: Methods and development.

WHO, UNICEF, & World Bank Group. (2018). Nurturing care for early childhood development: A framework for helping children survive and thrive to transform health and human potential.

Wu S, Wang Z, Du Z, Huang B, Zhang F. Liu R, 2021. Geographically and temporally neural network weighted regression for modeling spatiotemporal non-stationary relationships. Int J Geograph Informat Sci 35:582–608. DOI: https://doi.org/10.1080/13658816.2020.1775836

Wu Y, 2021. Can’t ridge regression perform variable selection? Technometrics 63:263–71. DOI: https://doi.org/10.1080/00401706.2020.1791254

Yi H, Li M, Dong Y, Gan Z, He L, Li X, Tao Y, Xia Z, Xia Z, Xue Y, Zhai Z, 2024. Nonlinear associations between the ratio of family income to poverty and all-cause mortality among adults in NHANES study. Sci Rep 14:12018. DOI: https://doi.org/10.1038/s41598-024-63058-z

Zhao Z, Xu Z, Hu C, Wang K, Ding X, 2024. Geographically weighted neural network considering spatial heterogeneity for landslide susceptibility mapping: A case study of Yichang City, China. CATENA, 234:107590. DOI: https://doi.org/10.1016/j.catena.2023.107590