A Model-Based Approach to Identify Classes and Respective Cutoffs of the Brazilian Household Food Insecurity Measurement Scale.

BACKGROUND

The Brazilian Household Food Insecurity Measurement Scale (EBIA) is the main tool for assessing household food insecurity (FI) in Brazil, assisting in monitoring and improving national public policies to promote food security. Based on the sum of item scores, households have been classified into 4 levels of FI, with the use of cutoffs arising from expert discussions informed by psychometric analyses and policy considerations.

OBJECTIVES

This study aimed to identify homogeneous latent groups corresponding to levels of FI, examine whether such subgroups could be defined from discriminant cutoffs applied to the overall EBIA raw score, and compare these cutoffs against those currently used.

METHODS

A cross-sectional population-based study with a representative sample of 1105 households from a low-income metropolitan area of Rio de Janeiro was conducted. Latent class factor analysis (LCFA) models were applied to the answers to EBIA's items to identify homogeneous groups, obtaining the number of latent classes for FI measured by the scale. Based on this and a thorough classification agreement evaluation, optimal cutoffs for discriminating between different severity levels of FI were ascertained. Model-based grouping and the official EBIA classification cutoffs were also contrasted.

RESULTS

LCFA identified 4 homogeneous groups with a very high degree of class separation (entropy = 0.906), endorsing the classification of EBIA as a 4-level measure of FI. Two sets of cutoffs were identified to separate such groups according to household type: 1/2, 5/6, and 10/11 in households with children and adolescents (score range: 0-14); and 1/2, 3/4, and 5/6 in adult-only households (score range: 0-7).

CONCLUSION

Although roughly classifying EBIA as in previous studies, the current approach suggests that, in terms of raw score, households endorsing only one item of the scale would be better classified by being placed in the same stratum as those remaining negative on all items.