Peer-reviewed article
Multilevel modeling and mediation analysis of acute respiratory infections among under five children in Bangladesh: evidence from multiple indicator cluster survey.
Shohel Mahmud, Razia Sultana, Md Sakawat Hossain, Morzina Akhter, Mohammad Emran Hossain ·
Summary
Written by AI, not yet AICS's view · claude-haiku-4-5-20251001This study analyzed acute respiratory infections (ARI) among 24,686 children under five years old in Bangladesh using nationally representative survey data from 2019. The overall prevalence of ARI was 7.54%, with higher rates among boys and younger children. The research found that unsafe cooking environments (separate or outdoor kitchens) and malnutrition (stunting) were risk factors, while better wealth was protective. Kitchen location and stunting together explained nearly half of the wealth-related differences in ARI risk.
Key points
- ARI prevalence was 7.54% among under-five children in Bangladesh
- Boys had higher ARI rates (56.3%) than girls (43.7%)
- Children aged 24–59 months had 39% lower ARI risk compared to younger children
- Separate or outdoor kitchens increased ARI risk by 22-28%
- Kitchen location and stunting mediated 30.5% and 14.7% respectively of the wealth-ARI relationship
Limitations
The study is cross-sectional and cannot establish causal relationships, only associations. The analysis relies on reported symptoms of ARI rather than clinical diagnosis, which may introduce recall bias or misclassification.
Abstract
BACKGROUND: Acute respiratory infection (ARI) is a major health problem in Bangladesh but limited evidence exists on its prevalence, determinants, and mediating factors among under-five children using nationally representative data. METHODS: Using data from the 2019 Multiple Indicator Cluster Survey (MICS), this study analyzed ARI prevalence among 24,686 children aged 0–59 months (under five years). Multilevel logistic regression was employed to identify ARI determinants while accounting for the hierarchical data structure. Furthermore, mediation analysis was conducted to evaluate the mediating roles of maternal education, stunting, and kitchen location in the association between the wealth index and ARI. RESULTS: The prevalence of ARI was 7.54%. The mean age of the children was 2.02 years (SD: 1.42 years), and among respondents 51.5% were male and 48.5% were female. The prevalence of ARI was higher among male (56.3%) children than female (43.7%) children. Chi-square tests showed that gender, age, parental age, wealth index, kitchen location, and stunting were significantly associated with ARI (p < 0.05). Multilevel logistic regression revealed that females (AOR: 0.79, 95% CI: 0.72–0.88) and older children aged 24–59 months (AOR: 0.61, 95% CI: 0.55–0.68) had significantly lower ARI risk. Conversely, separate (AOR: 1.28) or outdoor kitchens (AOR: 1.22) and stunting (AOR: 1.19) were associated with increased risk. Compared to Barishal, the ARI risk was lower in Dhaka (AOR: 0.54) but higher in Mymensingh (AOR: 1.87). Mediation analysis showed that kitchen location (30.5%) and stunting (14.7%) mediated the relationship between the wealth index and ARI prevalence. CONCLUSION: In addition to treatment, nutrition improvement, safe cooking environment, and regional planning are essential. Promoting and strengthening health education could further help in reducing ARI. These interventions are also aligned with the Sustainable Development Goals (SDG 3: good health and well-being), and could contribute to achieving national and global health targets.