Lowering Gestational Diabetes Risk by Prenatal Weight Gain Counseling

Evelyn M. Figueroa, MD; Kara Nitti, MPH; Stephen M. Sladek, MD


J Am Board Fam Med. 2020;33(2):189-197. 

In This Article



This was a pre/postintervention study in which we used a 2-year historic time period for the preintervention group (obtained by retrospective chart review 2012–2013), a 6-month clinical educational intervention on proper prenatal weight gain counseling per the 2009 IOM guidelines, and a 2-year prospective postintervention group (about 2015–2017, beginning whenever a site received their local institutional review board approval of the original protocol approved by the institutional review board at the originating site [University of Illinois, Chicago]).


Women with the following characteristics were chose for the study: a singleton pregnancy; maternal age of 15 to 45 years old at time of delivery; initiation of prenatal care by 16 weeks gestation; 6 or more prenatal visits, including 1 visit after 35 weeks; and delivery at the hospital affiliated with the practice. These characteristics were chosen to allow a reasonable opportunity for weight gain counseling during prenatal care to possibly influence maternal weight gain. Exclusion criteria were prepregnancy diabetes, daily prednisone use >21 days, or prepregnancy hypertensive disease requiring ≥2 medications.

Data Collection

Data collection included demographics (maternal age, race, ethnicity, and insurance type), parity, prepregnancy weight and BMI, as well as total weight gained. We also collected data on the following pregnancy outcomes: date of delivery, gestational age, newborn weight, and the mode of delivery (vaginal delivery, vacuum or forceps, shoulder dystocia during vaginal delivery; and primary or repeat cesarean). Data on complications of pregnancy were collected, including gestational diabetes (2-stage testing: 1-hour nonfasting 50-g oral glucose challenge test and, if ≥140 mg/dL, then 3-hour fasting 100-g oral glucose tolerance testing [Coustan criteria])[11] and hypertensive disorders of pregnancy (preeclampsia, eclampsia, and gestational hypertension).[12]


Our 6-month initial educational period involved local site training on the use of the IOM 2009 guidelines[3] (Table 1). Each site held education sessions with support staff, nurses, residents, faculty, and teaching attending physicians detailing the importance of weight gain counseling in pregnancy, and the sessions also included motivational counseling techniques.[13] The initial didactic teaching session set the expectation that weight gain should be addressed with mothers at every prenatal visit. Site principal investigators (residency faculty) were "champions" to constantly remind providers to follow this dictate and the 2009 IOM guideline throughout the entire 2-year postintervention time period. Patient handouts and brochures to help reiterate this education were provided, but their use was optional (Appendix 1). Some sites were able to provide electronic medical record weight gain counseling prompts. Six months were allowed to pass after the educational sessions before recording weight gain and pregnancy outcomes so as to give time for the new awareness to possibly influence the new group of prenatal patients entering the practice.

Data Analyses

All eligible subjects in both groups were included, as we performed an intention-to-treat analysis. A "per protocol" analysis was done only for exploratory purposes. Descriptive statistics were calculated for all variables. Group means and standard deviations are presented for continuous variables. Counts and percentages are presented for categorical variables. Comparisons were made between groups using Student's t test and Pearson's χ 2 test of proportions as applicable. Adjusted multivariate logistic regression models were constructed for the outcome of IOM guideline-recommended weight gain using backward stepwise elimination. Interactions were assessed between all covariates in the final model (SAS software, version 15.1). All statistical tests were 2-tailed and a P value of .05 was considered statistically significant in these analyses. Sensitivity analyses were performed by analyzing pre- and postintervention outcomes after omitting the data from single sites with outlying incidences of any parameter.