Strengths of this study included its randomized, blinded design,
the inclusion of NW and OW/OB women, follow-up from
early in pregnancy through 6 mo postpartum, and the use of
intervention strategies that may have practical relevance during
prenatal care in the clinical setting. A high proportion (82%) of
women who were randomly assigned completed the study
through 6 mo postpartum. A limitation of the study was selfreported
prepregnancy weight, which is a problem that exists
throughout pregnancy research. In pregnant and nonpregnant
samples, heavier women are more likely to underreport their
weight and underestimate their BMI (51). Although underreporting
may have led us to misclassify some OW/OB women as NW, this would have occurred equally in intervention and control
groups. Moreover, we were able to validate the recalled prepregnancy
weights against objectively measured weights before
pregnancy for approximately one-half of our sample. Our sample
was also self-selected, and findings may not generalize to the
population at large.