Can teenagers reliably report food insecurity using a six-question survey?

Details

The study evaluated the reliability and validity of a six-item food security scale when adolescents completed it themselves. It used a cross-sectional questionnaire survey that included the six-item food security measure, socio-economic variables and a food-frequency questionnaire. Participants were 1,903 students aged approximately 16 years from a representative sample of 29 schools in Trinidad.

Affirmative responses ranged from 514 (27%) for the 'balanced meal' item to 128 (7%) for the 'skipped or cut meals often' item and 141 (7%) for the 'hungry' item. Item-score correlations ranged from 0.444 to 0.580. Cronbach's alpha was 0.77.

Relative item severities estimated using the Rasch model ranged from -1.622 (standard error 0.043) for the 'balanced meal' item to 1.103 (standard error 0.068) for the 'skipped or cut meals often' item and 0.944 (standard error 0.062) for the 'hungry' item. The 'hungry' item had a slightly lower relative severity in boys than in girls.

Food insecurity was associated with household overcrowding: the adjusted odds ratio comparing the highest and lowest quartiles was 2.61 (95% confidence interval 1.75 to 3.91). It was also associated with lack of pipe-borne water in the home, low paternal education or paternal unemployment. After adjustment for socio-economic variables, food insecurity was associated with less frequent consumption of fruit (0.75, 0.60 to 0.94) or fish (0.72, 0.58 to 0.88), but more frequent consumption of biscuits or cakes (1.47, 1.02 to 2.11).

The researchers concluded that the food security scale provides a valid, reliable measure in adolescents, although young people report being hungry but not eating relatively more frequently than adults. They also concluded that food-insecure adolescents have low socio-economic position and may eat less healthy diets.