Table 2.

Models of longitudinal prediction of serum homocysteine and vitamin B-12 and milk B-12 trajectories based on the presence of depressive symptoms at the 3^rd trimester of pregnancy, Rio de Janeiro, Brazil.

Unadjusted model	Log serum homocysteine (mg/ L) (n = 90)		Log serum B-12 (pmol/L) (n = 90)		Log milk B-12 (pmol/L) (n = 94)
	β (SE)a	P	β (SE)a	P	β (SE)a	P
Main effects
With vs. without depressive symptoms at 3rd trimesterb	0.013 (0.066)	0.839	0.136 (0.112)	0.228	0.150 (0.182)	0.412
Postpartum days	−0.001 (0.000)	0.046	0.000 (0.001)	0.008	−0.003 (0.002)	0.133
Interaction term Postpartum days # with vs. without depressive symptoms at 3rd trimester	−0.001 (0.001)	0.232	−0.000 (0.000)	0.775	−0.002 (0.004)	0.594
Adjusted model	β (SE)a	P c	β (SE)a	P c	β (SE)a	P d
Main effects
With vs. without depressive symptoms at 3rd trimesterb	0.020 (0.068)	0.771	0.192 (0.116)	0.098	0.140 (0.179)	0.435
Postpartum days	−0.001 (0.001)	0.053	0.002 (0.001)	0.006	−0.002 (0.002)	0.332
Interaction term Postpartum days # With vs. without depressive symptoms at 3rd trimester	−0.001 (0.001)	0.194	−0.001 (0.002)	0.445	−0.003 (0.004)	0.455

^aLongitudinal mixed-effect model with beta coefficients (β) and standard error (SE), p-value refers to maximum likelihood estimator.

^bDepressive symptoms were assessed by the Edinburgh Postnatal Depression Scale (EPDS), yes (EPDS ≥11) and no (EPDS <11).

^cModels were adjusted for maternal age, education, pre-pregnancy BMI, B-vitamin (including folate) supplement intake during pregnancy, and serum folate concentration at the 3^rd trimester of pregnancy.

^dModel was adjusted for pre-pregnancy BMI, B-vitamin (including folate) supplement intake during pregnancy, and serum B-12 concentration at the 3^rd trimester of pregnancy. # refers to the interaction term.