Repeated-measures analysis of variance was used for secondary
aims that assessed the effects of treatment group and BMI
category on total weight gain and postpartum weight loss, with
adjustment for the same covariates. Logistic regression analyses
with similar covariates were also used to compare the intervention
with standard care on differential delivery complications
and maternal and fetal outcomes. Both R (version 2.11.1;
Palo Alto, CA) and SPSS (PASW version 18.0.1; IBM, Somers,
NY) statistical packages were used.