Factors associated with childhood malnutrition and hospitalization outcomes among children under five admitted to tertiary hospitals in Peshawar, Pakistan: a multicentre cross-sectional study

Childhood malnutrition remains a pressing public health concern in Pakistan, particularly in Khyber Pakhtunkhwa (KP). Despite decades of policy frameworks, including the National Nutrition Policy (2018) and the Alma-Ata Declaration’s principles of primary health care, implementation gaps persist, and robust multi-factorial data linking environmental, socioeconomic, and clinical factors to nutritional status and hospital-based outcomes remain scarce for this region. We conducted a multi-centre, hospital-based cross-sectional survey across three tertiary care hospitals in Peshawar, Pakistan, between March 2025 and February 2026. A total of 650 children under five years of age were enrolled. Descriptive statistics, multivariable logistic regression (with study site included as a fixed effect), and negative binomial regression were used to identify factors associated with undernutrition and to quantify associations with length of stay and complete recovery at discharge. A pilot-tested, structured questionnaire was used, and content validity was assessed by a panel of five experts. Anthropometric measurements were taken in duplicate following WHO standards. Missing birth-weight data ( \(n = 32\) ; 4.9%) were handled using multiple imputation by chained equations (20 imputations) under a missing-at-random assumption. Undernutrition was common in this hospital-based cohort: underweight 30.8%, stunting 27.5%, wasting 4.3%, and kwashiorkor 1.7%; no children were classified as overweight or obese. In the multivariable logistic regression model for composite undernutrition, unimproved sanitation (AOR = 2.66, 95% CI 1.83–3.86, \(p < 0.001\) ), urban residence (AOR = 1.58, 95% CI 1.05–2.37, \(p = 0.027\) ), food insecurity (AOR = 1.45, 95% CI 1.03–2.04, \(p = 0.034\) ), recent diarrhoea (AOR = 1.90, 95% CI 1.18–3.06, \(p = 0.009\) ), and admission to Khyber Teaching Hospital (AOR = 1.68, 95% CI 1.09–2.59, \(p = 0.019\) ) were associated with higher odds of undernutrition. Early complementary feeding was associated with lower odds (AOR = 0.63, 95% CI 0.44–0.90, \(p = 0.011\) ). High dietary diversity was not statistically associated with composite undernutrition (AOR = 0.87, 95% CI 0.56–1.35, \(p = 0.534\) ). In the negative binomial model for length of stay, underweight was associated with a higher expected length of stay (IRR = 1.63, 95% CI 1.44–1.84, \(p < 0.001\) ), whereas stunting, wasting, and kwashiorkor were not statistically associated with length of stay. Undernutrition at admission was associated with substantially reduced odds of complete recovery at discharge (AOR = 0.35, 95% CI 0.22–0.54, \(p < 0.001\) ), as was low household income (AOR = 0.51, 95% CI 0.35–0.74, \(p < 0.001\) ). In this hospital-based cohort, sanitation deficits, food insecurity, recent diarrhoeal illness, and urban residence were associated with childhood undernutrition, while early complementary feeding was associated with lower odds. Undernutrition at admission was strongly associated with reduced odds of complete recovery. These findings underscore the need for targeted WASH and nutrition-sensitive interventions, strengthened primary health care screening through Lady Health Worker programmes, and multi-sectoral policy integration. The associations observed should be interpreted within the limits of the cross-sectional, hospital-based design, and the prevalence estimates should not be generalised to all children in Peshawar or Khyber Pakhtunkhwa.