Association between frequency of breakfast intake before and during pregnancy and infant birth weight: the Tohoku Medical Megabank Project Birth and Three-Generation Cohort Study

Table 2 Multivariate linear regression analysis for the association between breakfast intake frequency and infant birth weight

	Model 1		Model 2		Model 3
	β	(95% CI)	β	(95% CI)	β	(95% CI)
Frequency of breakfast intake
Pre- to early pregnancy^b
Everyday	Ref		Ref		Ref
5–6 times/week	-2.1	(-21.0, 16.9)	-14.5	(-31.5, 2.5)	-12.4	(-29.4, 4.6)
3–4 times/week	-32.1	(-54.4, -9.9)	-45.0	(-65.1, -25.0)	-41.7	(-61.9, -21.6)
0–2 times/week	-26.9	(-46.6, -7.3)	-44.3	(-62.3, -26.3)	-38.2	(-56.5, -20.0)
P for trend^a	< 0.001		< 0.001		< 0.001
Early to mid- pregnancy^c
Everyday	Ref		Ref		Ref
5–6 times/week	-5.0	(-24.5, 14.5)	-12.5	(-32.2, 2.7)	-12.2	(-29.7, 5.4)
3–4 times/week	-18.3	(-42.4, 5.8)	-38.1	(-59.8, -16.4)	-33.6	(-55.4, -11.8)
0–2 times/week	-36.5	(-60.2, -12.9)	-49.0	(-70.6, -27.3)	-41.5	(-63.3, -19.6)
P for trend^a	< 0.001		< 0.001		< 0.001

CI Confidence interval, β Regression coefficients
Model 1 is crude
Model 2 is adjusted for age at delivery, pre-pregnancy BMI, parity, work status, smoking, alcohol intake, morning sickness, insomnia, child sex and gestational week
Model 3 is adjusted for covariates and folic acid supplementation, intake of energy, cereal, meat, seafood, bean, vegetable, and fruit
^aP for trend were calculated as trends across categories
^bThe coefficients of determination for Models 1, 2 and 3 were 0.0007, 0.21 and 0.21, respectively
^cThe coefficients of determination for Models 1, 2 and 3 were 0.0005, 0.21 and 0.21, respectively

ISSN: 1471-2393