periods. Both methods of imputation for missing data yielded similar results; therefore only analyses that used a weight gain of 0.3 kg/mo imputed for missing data are presented. Missing dietary and physical activity data were analyzed by carrying forward baseline observations, assuming no change in dietary intake or physical activity. All analyses were conducted using SPSS 21.0 for Windows software (SPSS Inc., Chicago, IL, USA) with a P ¼ 0.05 to indicate statistically significant differences.