Household Crowding and Malaria Infection Among Under-Five Children in Nigeria: A Design-Based and Multilevel Analysis of the 2021 Malaria Indicator Survey
Faculty Mentor
Dr. Jing Kersey
Location
Russell Union Ballroom
Type of Research
On-going
Session Format
Poster Presentation
College
Jiann-Ping Hsu College of Public Health
Department
Biostatistics, Epidemiology, and Environmental Health Sciences
Abstract
Background Household crowding may increase malaria exposure through shared sleeping spaces and housing vulnerability; however, nationally representative evidence remains limited. Most malaria studies using Demographic and Health Survey (DHS)-type data rely solely on design-based models without explicitly assessing contextual heterogeneity. We examined the association between household crowding and laboratory-confirmed malaria infection among under-five children in Nigeria, comparing design-based and multilevel modeling approaches.
Methods We analyzed data from 1,849 under-five children with valid rapid diagnostic test (RDT) results from the 2021 Nigeria Malaria Indicator Survey. Household overcrowding was defined as persons per sleeping room and categorized as none (≤2), moderate (>2–< 5), intermediate (5–6), and extreme (>6). Primary analyses used survey-weighted logistic regression incorporating sampling strata, primary sampling units, and weights to obtain nationally representative estimates. Covariates were selected a priori and included child age and sex, bed net use, household wealth, residence, sanitation, water source, and housing materials. As a robustness analysis, multilevel logistic regression with a random intercept for cluster was conducted to model intra-cluster correlation.
Results The weighted prevalence of RDT-confirmed malaria was 36.5%. In adjusted survey-weighted models, overcrowding was not significantly associated with malaria infection (overall p=0.48). Children in extremely overcrowded households had elevated but imprecise odds (aOR 2.94, 95% CI 0.64–13.51). Strong socioeconomic gradients were observed: children in the poorest households had markedly higher odds compared to the richest (aOR 8.24, 95% CI 3.92–17.33), and rural residence was independently associated with malaria (aOR 1.71, 95% CI 1.21–2.42). Multilevel models yielded consistent estimates and demonstrated substantial between-cluster heterogeneity (intraclass correlation coefficient ≈26%).
Conclusions Household crowding was not independently associated with malaria infection after adjustment. Substantial cluster-level heterogeneity and strong socioeconomic gradients suggest that structural and contextual factors, rather than occupancy density alone, drive malaria risk.
Program Description
.
Start Date
4-23-2026 2:00 PM
End Date
4-23-2026 4:00 PM
Recommended Citation
Asifat, Olamide; Das, Keya; Adenusi, Adedeji; Adebile, Tolulope; Alliu, Ibrahim; and Kersey, Jing, "Household Crowding and Malaria Infection Among Under-Five Children in Nigeria: A Design-Based and Multilevel Analysis of the 2021 Malaria Indicator Survey" (2026). GS4 Student Scholars Symposium. 202.
https://digitalcommons.georgiasouthern.edu/research_symposium/2026/2026/202
Household Crowding and Malaria Infection Among Under-Five Children in Nigeria: A Design-Based and Multilevel Analysis of the 2021 Malaria Indicator Survey
Russell Union Ballroom
Background Household crowding may increase malaria exposure through shared sleeping spaces and housing vulnerability; however, nationally representative evidence remains limited. Most malaria studies using Demographic and Health Survey (DHS)-type data rely solely on design-based models without explicitly assessing contextual heterogeneity. We examined the association between household crowding and laboratory-confirmed malaria infection among under-five children in Nigeria, comparing design-based and multilevel modeling approaches.
Methods We analyzed data from 1,849 under-five children with valid rapid diagnostic test (RDT) results from the 2021 Nigeria Malaria Indicator Survey. Household overcrowding was defined as persons per sleeping room and categorized as none (≤2), moderate (>2–< 5), intermediate (5–6), and extreme (>6). Primary analyses used survey-weighted logistic regression incorporating sampling strata, primary sampling units, and weights to obtain nationally representative estimates. Covariates were selected a priori and included child age and sex, bed net use, household wealth, residence, sanitation, water source, and housing materials. As a robustness analysis, multilevel logistic regression with a random intercept for cluster was conducted to model intra-cluster correlation.
Results The weighted prevalence of RDT-confirmed malaria was 36.5%. In adjusted survey-weighted models, overcrowding was not significantly associated with malaria infection (overall p=0.48). Children in extremely overcrowded households had elevated but imprecise odds (aOR 2.94, 95% CI 0.64–13.51). Strong socioeconomic gradients were observed: children in the poorest households had markedly higher odds compared to the richest (aOR 8.24, 95% CI 3.92–17.33), and rural residence was independently associated with malaria (aOR 1.71, 95% CI 1.21–2.42). Multilevel models yielded consistent estimates and demonstrated substantial between-cluster heterogeneity (intraclass correlation coefficient ≈26%).
Conclusions Household crowding was not independently associated with malaria infection after adjustment. Substantial cluster-level heterogeneity and strong socioeconomic gradients suggest that structural and contextual factors, rather than occupancy density alone, drive malaria risk.
