Table 1.

Cross-sectional and longitudinal associations between food insecurity and intuitive eating

Analysis	Food insecurity measure	Unadjusted					Adjusted
		b	se	t	P	f 2	B	se	t	P	f 2
Cross-sectional
Adolescence	Household	−0·08	0·03	2·41	0·02	0·004	−0·06	0·03	1·70	0·09	0·003
	Adolescent experience	−0·14	0·05	3·08	0·002	0·01	−0·10	0·05	2·27	0·02	0·01
Emerging adulthood	Emerging adult experience	−0·12	0·04	3·19	0·001	0·01	−0·13	0·04	3·43	< 0·001	0·01
Longitudinal
Adolescence to emerging adulthood	Household	−0·11	0·03	3·33	< 0·001	0·01	−0·08	0·03	2·45	0·01	0·01
	Adolescent experience	−0·09	0·05	1·94	0·045	0·002	−0·04	0·05	0·87	0·38	0·001

IE, intuitive eating.

Adjusted models included parent education, race/ethnicity, sex and age as covariates; longitudinal models were additionally adjusted for baseline IE. Food-secure participants were the reference group across analyses. The unstandardised β coefficient represents the average IE difference between food-secure and food-insecure participants (after accounting for covariates in adjusted models). For example, average IE scores, which range from 1 to 4, were 0·10 lower in adolescents experiencing food insecurity than adolescents who were food-secure. Effect sizes are presented for the individual FI variable (v. the entire model); Cohen’s conventions for interpreting f ² effect sizes are 0·02 = small, 0·15 = medium and 0·35 = large⁽⁴³⁾.