Table 2.

Prevalence ratios (PRs) for skipping breakfast by socioeconomic status^a

	Reference	Category	Main effect			Interaction
			PR	95% CI	p	PR	95% CI	p
Young adults (n = 2,179)
Socioeconomic status
EA	Higher	Lower	1.21	1.09–1.36	0.001	1.20	1.04–1.39	0.014
SFS	Moderate	Poor	1.15	1.03–1.27	0.009	1.14	1.01–1.29	0.035
EA × SFSb								0.796
Middle-aged adults (n = 2,992)
Socioeconomic status
EA	Higher	Lower	1.42	1.27–1.60	< 0.001	1.44	1.25–1.66	< 0.001
SFS	Moderate	Poor	1.10	0.97–1.25	0.123	1.12	0.95–1.33	0.182
EA × SFSb								0.743
Older adults (n = 3,184)
Socioeconomic status
EA	Higher	Lower	1.02	1.00–1.05	0.095	1.29	1.04–1.60	0.020
SFS	Moderate	Poor	1.04	1.00–1.07	0.032	1.61	1.18–2.19	0.002
EA × SFSb								0.042

Abbreviations: PR, prevalence ratio; CI, confidence interval; EA, educational attainment; SFS, subjective financial status

Bold values denote significant at α = 0.025 (0.05/2) by Bonferroni’s correction adjustment

ª Poisson regression model with robust variance. Dependent variable: skipping breakfast; Independent variable: socioeconomic status;

Covariates: sex, age, living arrangement, marital status, current work status, number of comorbidities, K6 and IADL (only for older adults)

^b Overall test of interaction