Agreement between household food insecurity access scale and an imputed food consumption score in urban Ethiopia: implications for targeting.

backgroundThe Food Consumption Score (FCS) and Household Food Insecurity Access Scale (HFIAS) measure different domains of food security, namely dietary diversity versus perceived access constraints. In urban Ethiopia, it is unknown whether these indicators classify the same households as food insecure. This study compared FCS and HFIAS classifications among households with young children and quantified discordance between the two measures.

methodsThis secondary analysis used cross-sectional data from 542 households with children aged 6-23 months attending public health centers in Harar City, Eastern Ethiopia (July-August 2024). The original study assessed determinants of animal-source food consumption; the current comparative analysis was a pre-specified secondary objective. FCS was calculated from 7-day food frequency data using standard WFP thresholds. Missing oil and sugar values (100% missing) were imputed as 3 days/week with sensitivity analyses under alternative assumptions. HFIAS categories used continuous score cutoffs validated in Ethiopian studies. Agreement was assessed using Cohen's kappa, prevalence-adjusted bias-adjusted kappa (PABAK), and discordance analysis. Bootstrap methods calculated 95% confidence intervals. Multivariable logistic regression identified predictors of discordance.

resultsFCS classified 34.5% as acceptable, 28.6% borderline, and 36.9% poor. HFIAS classified 20.1% as food secure, 28.0% mildly, 33.0% moderately, and 18.8% severely insecure. Overall agreement was 48.5% with kappa of 0.31 (95% CI: 0.25-0.37), indicating fair agreement. PABAK was 0.42 (95% CI: 0.36-0.48). Discordance occurred in 51.5% of households: 22.9% were HFIAS-secure but FCS poor/borderline; 28.6% were HFIAS-insecure but FCS acceptable. Poorest wealth quintile (AOR = 2.43, 95% CI: 1.38-4.28) and no maternal education (AOR = 1.87, 95% CI: 1.12-3.14) were independently associated with higher odds of discordance.

conclusionsFCS and HFIAS measure different dimensions of household food security and demonstrate only fair agreement in this urban Ethiopian population. Neither measure alone provides a complete picture. The finding that more than half of households are discordantly classified highlights the importance of measuring both dietary consumption and perceived access whenever resources permit. The choice of indicator should be guided by assessment objectives and available resources.