Junk Food Consumption and Child Health in Khyber Pakhtunkhwa, Pakistan: A Cross-Sectional Econometric Analysis with Agricultural and Food Policy Implications

Summary

This cross-sectional study of 100 caregivers in Swabi District, Khyber Pakhtunkhwa examined associations between children's junk food consumption and reported health outcomes using Probit and Logit regression models. Frequent junk food consumption (>3 times weekly) was associated with a 27.6 percentage-point higher probability of health issues, whilst each additional junk food advertisement viewed increased this probability by 7.4 percentage points; higher parental nutritional awareness reduced it by 13.1 percentage points. The authors acknowledge the cross-sectional, single-site, convenience-sampled design limits inference to associations rather than causal effects, and propose policy interventions including school food environment reform, marketing regulation, nutrition-sensitive agriculture and nutrition education.

Regional applicability

Findings from rural/semi-rural Pakistan may have limited direct transferability to United Kingdom contexts, where food environments, regulatory frameworks and agricultural systems differ substantially. However, the associations between advertising exposure, parental nutritional knowledge and child diet-related health outcomes may be relevant to understanding health inequalities in UK populations with high ultra-processed food consumption.

Key measures

Junk food consumption frequency (categorical: more than three times weekly); advertising exposure (count); parental nutritional awareness (scale); health outcomes (binary: presence/absence of overweight/obesity, dental problems, or persistent fatigue); urban residence (binary)

Outcomes reported

The study examined associations between junk food consumption frequency, media advertising exposure, parental nutritional awareness and caregiver-reported child health outcomes (overweight/obesity, dental problems, persistent fatigue) among children aged 3–14 years. Binary logistic and probit models were used to estimate probability of health issues.

Topic tags