Factors Associated with Appropriate Gestational Weight Gain among Women with Obesity

Abstract

Objective

This study aimed to compare attendance of nutritional counseling, dietary composition, exercise patterns, and socioeconomic factors among obese women with inappropriate gestational weight gain (iGWG) versus appropriate GWG (aGWG).

Study Design

Medicaid-eligible women receiving prenatal care at a tertiary care center from January 2013 to December 2015 were offered individualized nutritional counseling by a registered dietitian encouraging well-balanced meals and 150 min/wk of exercise. We conducted a prospective case–control study of obese women (body mass index or BMI ≥30) with a singleton gestation with iGWG (<11 or >20 pounds) versus aGWG (11–20 pounds). Dietary intake, activity level, and socioeconomic factors were compared with Chi-square, Fisher’s exact, Student’s t-test, and Wilcoxon Rank Sum tests as indicated, and odds ratios with 95% confidence intervals were calculated. Multivariate regression analysis for significant variables was performed. A subgroup analysis of women with BMI ≥40 was planned.

Results

A total of 401 women were analyzed: 78% (n = 313) with iGWG and 22% (n = 88) with aGWG. Demographics were similar between groups. Women with iGWG less frequently reported physician reinforcement of counseling and reported more physical inactivity and unemployment; there were no differences in caloric intake or macronutrient profile between groups. Multivariate regression identified physician reinforcement and employment as independent predictors of aGWG. Among women with BMI ≥40 (n = 133), those with iGWG (78%) were less likely to attend counseling, report physician reinforcement of counseling, and have adequate caloric and protein intake when compared with those with aGWG (22%). Activity level and socioeconomic factors were not different between groups.

Conclusion

Physician reinforcement of nutritional counseling, greater activity level, and employment are associated with aGWG in women with BMI ≥30, while individualized professional nutritional counseling and dietary modifications were further associated with aGWG in women with BMI ≥40. Thus, greater focus should be placed on enhancing exposure to counseling and altering nutritional and exercise choices to optimize aGWG.

Obesity in the United States is a growing epidemic which disproportionally affects Hispanic and non-Hispanic black women,¹ and is a major risk factor for multiple medical complications and even all-cause mortality.^2–5 Obesity predisposes pregnant women to numerous maternal complications, including gestational diabetes, hypertensive disorders, prolonged labor, and cesarean delivery, as well as neonatal complications, including stillbirth, shoulder dystocia, low APGAR scores, and low umbilical artery pH.^6–11 The American College of Obstetricians and Gynecologists (ACOG) recommends that prepregnancy or initial prenatal visit body mass index (BMI) be used to provide counseling guided by the Institute of Medicine (IOM) recommendations for gestational weight gain (GWG), e.g., 11 to 20 pounds for all obese women given limited data on GWG per obesity class.^12,13 GWG above the IOM guidelines is associated with several poor perinatal and maternal outcomes including low APGAR scores, seizures, hypoglycemia, fetal macrosomia, large for gestational age birth weight infants, gestational diabetes, hypertensive disorders, cesarean delivery, and postpartum weight retention,^14–18 whereas GWG under the guidelines is associated with small for gestational age birth weight infants.¹⁹ Obese women are particularly impacted by disparities in GWG, with over 70% of obese women having excessive GWG.¹⁶

While several studies have investigated risk factors for abnormal GWG, fewer studies have focused on interventions to remediate it. Weight management strategies for pregnant women generally center around dietary control, exercise, and behavioral modification. However, little guidance is provided by ACOG to physicians on how these interventions should be enacted. A meta-analysis including 7,278 women compared dietary, physical activity, and combined strategies for weight management during pregnancy, and demonstrated that women receiving any intervention significantly reduced their GWG compared with no intervention, with the greatest reduction found among dietary interventions.²⁰ Another recent meta-analysis of 11,444 women analyzed dietary interventions of low glycemic load or total caloric restriction, combination dietary and exercise interventions, and exercise-only interventions to limit excessive GWG, and found that any intervention reduced the risk of excessive GWG by 20%, with the largest reduction occurring in supervised dietary and exercise interventions.²¹

Despite the increasing prevalence of obesity, especially class III obesity, few studies have examined the impact of lifestyle interventions on GWG in this high-risk population. The ability to achieve appropriate GWG (aGWG) may be limited by socioeconomic status via restriction of food options and patient support systems, which are important to the effectiveness of any lifestyle intervention program. Thus, we sought to investigate whether engagement in a nutritional and lifestyle counseling intervention, dietary composition, exercise level, and socioeconomic factors differed among obese women with inappropriate, e.g., excessive or inappropriate gestational weight gain (iGWG) versus those with aGWG.

Materials and Methods

We performed a prospective case–control study of women receiving prenatal care at the University of Alabama at Birmingham from January 1, 2013, to December 31, 2015 iGWG (<11 or >20 pounds) versus with aGWG (11–20 pounds). Institutional Review Board approval was obtained prior to the initiation of this study (IRB-130417003).

Women at our predominantly government-insured clinics entering prenatal care in the first or second trimester were offered participation in a Centers for Medicare and Medicaid Services sponsored “Strong Start for Mothers and Newborns” program. All women screened for participation were required to have a Medicaid number and be Medicaid eligible in the state of Alabama, a determination made by family size and income; for example, in 2013, a pregnant woman qualified for Medicaid if her family income was at or below 133% of the federal poverty level (available at: medicaid.alabama.gov).²² This Strong Start initiative included education on healthy lifestyle choices, and screening for depression and substance abuse which prompted referrals to a social worker or to treatment facilities, respectively. All Strong Start participants were provided access to video and/or written instructions on topics including nutrition, exercise, the importance of regular prenatal visits, the benefits of breastfeeding, normal pregnancyrelated emotions, and substance use avoidance. Additionally, women who were obese (BMI ≥30 kg/m²) were offered individualized nutritional counseling (NC) by a single registered dietitian (M.D.). Only English-speaking women were offered counseling as the clinic was unable to reliably provide language interpretation services throughout all sessions.

Nutritional counseling encouraged well-balanced meals using the Department of Agriculture’s MyPlate guide (accessible at ChooseMyPlate.gov) (►Supplementary Material 1, available in the online version). Counseling focused on daily breakfast, small frequent meals, portion control, lean proteins, whole grains, low-fat dairy, and limitation of sugary beverages. Additional recommendations were made for swapping unhealthy for healthy choices in an “eat this not that” format, i.e., “fried chicken” versus “grilled, baked, or broiled chicken” (►Supplementary Material 2, available in the online version), as well as for 150 min/wk of low impact exercise. Take-home printed handouts regarding meal planning, including a sample meal plan (►Supplementary Material 3, available in the online version), and resistance training exercises (►Supplementary Material 4, available in the online version) were provided to participants, with additional links to educational videos on these topics.

Women enrolled in Strong Start with a singleton pregnancy, delivery estimated gestational age (EGA) >23^0/7 weeks, and a BMI ≥30 at initial prenatal visit were included for analysis. Women were excluded if they had a multifetal gestation, delivered at a previable EGA, had BMI <30 at initial exam, presented to care after 29^0/7 weeks EGA, or had a prior gastric bypass or sleeve gastrectomy.

The initial height and weight used to calculate BMI ≥as the primary screening for inclusion was obtained by trained research nurses. Additionally, maternal demographics, comorbidities, and mode of delivery were ascertained by trained research staff if the patient was eligible for counseling. Subsequently, individualized chart review by two trained reviewers (G.D.C., M.L.C.) was performed regarding singleton versus multifetal gestation, delivery EGA, surgical history, and weight at delivery to allow for calculation of GWG. Standardized abstraction forms were utilized.

Socioeconomic characteristics, e.g., partner involvement, neighborhood safety, were assessed on standardized surveys completed by patients at program intake. Macronutrient composition, caloric intake, and activity level were recorded in conjunction with the dietitian through verbal recall of a patient’s typical week, given historical experience and anticipated low compliance with food diary recordings. Caloric needs, usually 1,800 to 2,200 kilocalories, were estimated and discussed by the dietitian (M.D.) at the counseling appointment. Recalled intake was determined to be excessive or inadequate using the Department of Agriculture MyPlate recommendations, SuperTracker analysis, and Dietary Reference Intakes for pregnancy.²³ Further, during the counseling session, the dietitian inquired whether the patient recalled her primary provider, e.g., physician or nurse practitioner, previously addressing GWG recommendations and the importance of aGWG regarding the impact on perinatal outcomes.

The primary exposure in this analysis was attendance of NC, with the primary outcome of iGWG versus aGWG. Secondary exposures included activity level and multiple nutritional factors, including the patient’s perception of her primary provider’s reinforcement for NC, caloric intake, and measures of dietary composition. Additional secondary exposures included socioeconomic characteristics, i.e., maternal education, maternal and partner employment, perceived partner and familial support, and housing. A subgroup analysis of women with BMI ≥40 was planned. Further analyses evaluating the primary outcome across three levels of inadequate GWG (<11 pounds), aGWG, and excessive GWG (>20 pounds) groups were conducted in the entire cohort. Sensitivity analyses excluding those entering care after 14^0/7 weeks EGA or delivering <37^0/7 weeks EGA were planned, given a truncated time frame for GWG calculation, as well as analyses excluding women with pregestational diabetes.

Maternal demographics, primary exposures, and secondary exposures were compared between those with iGWG and aGWG using Chi-square tests and Fisher’s exact tests as appropriate for categorical variables. Student’s t-tests and Wilcoxon Rank Sum tests, as appropriate, were used to compare continuous variables. Odds ratios (OR) with 95% confidence intervals (CI) were calculated using aGWG as the reference group. A multivariable logistic regression model that included significant covariates identified in bivariate analyses of maternal demographics was performed to identify independent predictors in women with BMI ≥30. All analyses were performed using SAS Version 9.4 (SAS Institute, Inc., Cary, NC). Exposures were evaluated at a 0.05 level of significance.

Results

Of the 1646 women who participated in the Strong Start initiative, 401 met final inclusion criteria. Specific indications for inclusion and exclusion are detailed in ►Fig. 1.

Fig. 1.

Flow diagram depicting women identified for inclusion in final analysis. EMR, electronic medical record; BMI, body mass index.

Of the 401 women with BMI ≥30, 78% (n = 313) had iGWG and 22% (n = 88) had aGWG. Maternal demographics were similar between those with iGWG and aGWG (►Table 1). Notably, the majority of women were non-Hispanic Black race/ethnicity, unmarried, and had government-assisted insurance.

Table 1.

Maternal demographics of obese women with appropriate versus inappropriate gestational weight gain

	Appropriate GWG (n = 88)	Inappropriate GWG (n = 313)	p-Value
Maternal demographics for BMI ≥30
Age (y)	26.6 (5.5)	26.1 (5.6)	0.39
Parity	1 (0–6)	1 (0–7)	0.26
BMI at initial exam (kg/m2)	37.9 (30.1–86.6)	37.1 (30.0–70.4)	0.54
Gestational age at initial exam (wk)	9.7 (4.7–26.6)	9.9 (1.9–62.0)	0.73
Gestational age at delivery (wk)	38.0 (36.7–39.6)	38.9 (37.0–39.7)	0.09
Race/Ethnicity
White, non-Hispanic	10 (11.4)	55 (17.6)	0.23
Black, non-Hispanic	75 (85.2)	252 (80.8)
Hispanic	3 (3.4)	5(1.6)
Marital status
Married	13 (14.8)	45 (14.4)	0.93
Other	75 (85.2)	268 (85.6)
Payer status
Private	7 (8.0)	19 (6.1)	0.29
Government-assisted	80 (90.9)	293 (93.6)
No insurance	1 (1.1)	1 (0.3)
Tobacco use	12 (14.5)	57 (19.1)	0.33
Pregestational diabetes	13 (14.8)	33 (10.6)	0.28
Chronic hypertension	26 (29.6)	71 (22.8)	0.19
Prior cesarean section	24 (27.3)	74 (23.8)	0.50
Maternal demographics for BMI ≥40
Age (y)	28.0 (5.1)	26.3 (5.4)	0.13
Parity	1 (0–4)	1 (0–7)	0.12
BMI at initial exam (kg/m2)	44.4 (40.0–86.6)	45.3 (40.0–70.4)	0.89
Gestational age at initial exam (wk)	9.2 (5.3–24.9)	9.3 (1.9–25.7)	0.99
Gestational age at delivery (wk)	37.9 (36.6–38.9)	38.6 (37.1–39.4)	0.039
Race/Ethnicity
White, non-Hispanic	5(16.7)	14 (13.7)	0.86
Black, non-Hispanic	25 (83.3)	86 (84.3)
Hispanic	0 (0.0)	2 (2.0)
Marital status
Married	6 (20.0)	15 (14.6)	0.47
Other	24 (80.0)	88 (85.4)
Payer status
Private	3 (10.0)	5 (4.9)	0.38
Government-assisted	27 (90.0)	98 (95.2)
Tobacco use	3 (10.0)	17(17.7)	0.56
Pregestational diabetes	13(43.3))	33 (32.0)	0.54
Chronic hypertension	13 (43.3)	34 (33.3)	0.31
Prior cesarean section	5(16.7)	30 (29.4)	0.16

Abbreviations: BMI, body mass index; GWG, gestational weight gain.

Note: Data presented as n (%) or median (range) as appropriate.

When considering women with BMI ≥30, there was no significant difference between those with iGWG versus aGWG with respect to attendance of NC (75 vs. 72%, p = 0.51, ►Table 2). However, women with iGWG were less likely to report that their primary medical provider had discussed GWG recommendations during their pregnancy (22 vs. 38%, p = 0.008,OR0.45[0.25–0.82]).There was no difference in self-reported caloric intake or dietary composition with respect to meals, fat, protein, fruits, vegetables, and sweetened beverages. Women with iGWG were more than twice as likely to report inactivity (25 vs. 13%, p = 0.041, OR 2.27 [1.02–5.04]). With respect to social factors, women with iGWG and aGWG had similar education levels, financial and social support, and neighborhood and housing characteristics. However, women with iGWG were less likely to be employed (37 vs. 51%, p = 0.018, OR 0.57 [0.35–0.91]). The multivariable model identified perceived physician reinforcement of NC (OR 0.44 [0.24–0.80]) and employment (OR 0.50 [0.28–0.89]) as independent predictors of iGWG for those with BMI ≥30.

Table 2.

Nutritional and socioeconomic exposures among obese women (BMI ≥30) with appropriate versus inappropriate gestational weight gain

	Appropriate GWG (n = 88)	Inappropriate GWG (n = 313)	p-Value (OR [95% CI] for significant factors)
Nutritional exposures
Attendance of nutritional sessions	63 (72)	235 (75)	0.51
Perceived physician reinforcement of NC	24 (38)	51 (22)	0.008 (OR 0.45 [0.25–0.82])
Physical inactivity	8(13)	58 (25)	0.041 (OR 2.27 [1.02–5.04])
Excessive caloric intake (>2,400 kcal)	22 (37)	56 (27)	0.13
Inadequate caloric intakea	12 (20)	57 (27)	0.27
Baseline dietary composition
Number of meals	3 (2–3)	3(1–22)	0.66
Number of snacks	2 (0–4)	2 (0–4)	0.90
Excessive sweetened Beverage intakeb	40 (67)	153 (73)	0.38
Excessive fat intakeb	30 (50)	81 (39)	0.12
Inadequate fruit intakeb	27 (46)	89 (42)	0.62
Inadequate vegetable intakeb	47 (80)	179 (85)	0.34
Inadequate protein intakeb	8(13)	50 (24)	0.08
Inadequate fat intakeb	1 (2)	10(5)	0.46
Socioeconomic exposures
Planned pregnancy	19 (22)	46 (15)	0.11
Education (y)	12 (9–18)	12 (7–17)	0.33
Employment	45 (51)	116 (37)	0.018 (OR 0.57 [0.35–0.91])
Employment of father of baby	51 (73)	206 (77)	0.52
Financial support by father of baby	59 (82)	214 (79)	0.58
Emotional support by family	82 (93)	288 (93)	0.85
Unsafe neighborhood	6(7)	23(7)	0.85
Housing			0.77
House	33 (38)	111 (36)	0.77
Apartment	25 (28)	75 (24)	0.42
With relatives	13(15)	61 (20)	0.30
Public housing	12(14)	42 (14)	0.98
Mobile home	3(3)	17(5)	0.58
Other	2(2)	1 (1)	0.12

Abbreviations: BMI, body mass index; CI, confidence interval; GWG, gestational weight gain; NC, nutritional counseling; OR, odds ratio.

Note: Data presented as n (%) or median (range) as appropriate; aGWG is the reference group.

^aIndividualized dietary caloric needs as determined by registered dietician.

^bExcessive or inadequate intake was determined using the United States Department of Agriculture MyPlate recommendations, SuperTracker analysis, and Dietary Reference Intakes for pregnancy.

Of the 133 women with BMI ≥40, 78% (n = 103) had iGWG and 22% (n = 30) had aGWG (►Table 1). Maternal demographics were again similar between those with iGWG and aGWG, with the exception of EGA at delivery (p = 0.039). Although the median EGA at delivery for those with aGWG was earlier than those with iGWG, both median EGA were at term (►Table 1).

Among women BMI ≥40, those with iGWG versus aGWG were less likely to attend a single session of NC (78 vs. 97%, p = 0.017, OR 0.12 [0.02–0.93]) or report physician reinforcement of counseling provided by the dietitian (24 vs. 48%, p = 0.014, OR 0.33 [0.14–0.81], ►Table 3). Those with iGWG were also less likely to report excessive caloric intake (19 vs. 38%, p = 0.041, OR 0.37 [0.14–0.98]), and more likely to report inadequate caloric intake (41 vs. 17%, p = 0.024, OR 3.31 [1.13–9.70]) and inadequate protein intake (32 vs. 10%, p = 0.023, OR 4.15 [1.14–15.15]). Among those with iGWG, there was a higher rate of physical inactivity; however, this association was not statistically significant (37 vs. 17%, p = 0.054). There were no differences between groups with respect to numerous socioeconomic factors (►Table 3). Multivariable regression models to identify independent associations were limited by small sample size.

Table 3.

Nutritional and socioeconomic exposures among class III obese women (BMI ≥ 40) with appropriate versus inappropriate gestational weight gain

	Appropriate GWG (n = 30)	Inappropriate GWG (n = 103)	p-Value (OR [95% CI] for significant factors)
Nutritional exposures
Attendance of nutritional sessions	29 (97)	80 (78)	0.017 (OR 0.12 [0.02–0.93])
Perceived physician reinforcement of NC	14 (48)	19 (24)	0.014 (OR 0.33 [0.14–0.81])
Physical inactivity	5(17)	29 (37)	0.054
Excessive caloric intake (>2,400 kcal)	11 (38)	13 (19)	0.041 (OR 0.37 [0.14–0.98])
Inadequate caloric intakea	5(17)	29 (41)	0.024 (OR 3.31 [1.13–9.70])
Baseline dietary composition
Number of meals	3 (2–3)	3(1–3)	0.11
Number of snacks	2 (0–3)	1 (0–4)	0.70
Excessive sweetened Beverage intakeb	19 (66)	49 (69)	0.73
Excessive fat intakeb	15(52)	23 (33)	0.08
Inadequate fruit intakeb	16(55)	34 (48)	0.51
Inadequate vegetable intakeb	21 (72)	61 (86)	0.11
Inadequate protein intakeb	3(10)	23 (32)	0.023 (OR 4.15 [1.14–15.15])
Inadequate fat intakeb	1 (3)	5(7)	0.67
Socioeconomic exposures
Planned pregnancy	6(20)	13 (13)	0.37
Education (y)	12 (10–18)	12 (8–16)	0.70
Employment	15(50)	3(36)	0.16
Employment of father of baby	22 (81)	68 (76)	0.58
Financial support by father of baby	21 (78)	70 (79)	0.92
Emotional support by family	28 (93)	91 (88)	0.74
Unsafe neighborhood	2(7)	10 (10)	>0.99
Housing			0.55
House	13(43)	37 (36)	0.46
Apartment	10 (33)	23 (22)	0.22
With relatives	3(10)	18(17)	0.40
Public housing	3(10)	18(17)	0.40
Mobile home	1 (3)	7(7)	0.68

Abbreviations: BMI, body mass index; CI, confidence interval; GWG, gestational weight gain; NC, nutritional counseling; OR, odds ratio.

Note: Data presented as n (%) or median (range) as appropriate; aGWG is the reference group.

^aIndividualized dietary caloric needs as determined by registered dietician.

^bExcessive or inadequate intake was determined using the United States Department of Agriculture MyPlate recommendations, SuperTracker analysis, and Dietary Reference Intakes for pregnancy.

We conducted an additional stratified analysis, categorizing women with iGWG into inadequate GWG and excessive GWG. In our entire cohort, 138 had inadequate GWG, 88 aGWG, and 175 excessive GWG. Sixty-three (71.6%) of women with aGWG attended NC compared with 103 (74.6%) with inadequate and 132 (75.4%) with excessive GWG. Consistent with the primary analysis, there was no significant association with attendance of NC among these three groups (p = 0.79). For those with BMI ≥40, 54 had inadequate GWG, 30 aGWG, and 49 excessive GWG (n = 49). In this analysis, the 29 (96.7%) women with aGWG attended NC, which was significantly higher than the 40 (74.1%) with inadequate and 40 (81.6%) with excessive GWG (p = 0.036).

A sensitivity analysis excluding those with pregestational diabetes (n = 46) was consistent with results for the primary outcome in women with BMI ≥30, with no significant difference in attendance of NC (74.6 vs. 72.0%, p = 0.64); however, for women with BMI ≥40, attendance of counseling remained more infrequent among those with iGWG (76.9 vs. 96.0%, p = 0.041). Additionally, a sensitivity analysis excluding patients entering care after 14 ^0/7 weeks EGA included 183 women: 79% (n = 145) with iGWG and 21% (n = 38) with aGWG. The results demonstrated no difference in attendance of NC between those with iGWG versus aGWG. Similarly, another sensitivity analysis of 308 women excluding preterm deliveries (EGA <37^0/7 weeks) also demonstrated no difference in attendance of counseling, consistent with the primary analysis (results not shown).

Discussion

Our analysis of obese patients who participated in the Strong Start program of lifestyle management demonstrated that women who reported that their physician reinforced the concepts reviewed in NC by a dietitian, had greater physical activity levels, and were employed were more likely to have aGWG. Specifically, regression analysis identified physician reinforcement of counseling and employment as independent predictors of aGWG. This data suggests that it is crucial for physicians to reinforce the importance of nutrition and exercise at every prenatal visit. When evaluating women with BMI ≥40, those who participated in individualized NC and had increased activity levels were more likely to have aGWG, whereas those who reported inadequate caloric or protein intake were less likely to have aGWG. There was no difference in a variety of socioeconomic factors, including education level, familial and partner support, and housing situation in those women.

Few studies to date have evaluated interventions to limit GWG in women with obesity, as many studies specifically exclude this population.^20,21 A recent meta-analysis investigating lifestyle interventions to limit GWG demonstrated a reduced GWG among obese women, however, did not demonstrate this association among overweight or class III obese women, which the authors note may be attributable to significant heterogeneity between study interventions.²⁴ Among 14 studies assessing interventions performed during pregnancy, only four analyzed obese patients and only two used the current IOM guidelines for GWG. Haakstad and Bø performed a randomized controlled trial of 105 obese patients evaluating the impact of a structured 12-week exercise program on GWG and demonstrated a significant increase in aGWG for those completing the program, although noted a 21% drop out rate.²⁵ The parent Strong Start program, although not a randomized controlled trial, is a significantly larger study evaluating the impact of a lifestyle intervention for women with obesity (n = 401) and with class III obesity (n = 133). Additionally, it uses the current IOM guidelines to evaluate a population which has been infrequently studied and provides specifics of a uniformly executed program by a dietitian trained in counseling during pregnancy.

Our data support the efficacy of directed in-person NC for this high-risk population of women with BMI ≥40. Patients with BMI ≥40 with iGWG were more likely to report inadequate caloric intake, which may be due to recall bias. However, it may suggest that women with financial constraints may have variant dietary patterns, e.g., choosing small meals of high-calorie or nutrient-poor food. Although there was no significant difference in the reported number of meals per day among women with iGWG versus aGWG, those with inadequate and aGWG had fewer meals per day than those with excessive GWG (p = 0.007, data not shown). We were unable to assess location of housing in proximity to grocery stores, e.g., women living in food deserts, which may be an important element underlying many other socioeconomic and dietary composition variables. Additionally, employment was the only socioeconomic factor associated with aGWG in initial analyses, which may reflect our incredibly low-resource population. Although not a modifiable risk factor, employment can be a screening tool for providers to identify patients at higher risk of iGWG.

As this analysis demonstrates the effectiveness of NC in women with class III obesity, it suggests the need for prospective studies and randomized trials to analyze this association. Additionally, future studies should be performed to assess the impact of NC in women with BMI ≥40 on maternal and fetal outcomes, as these patients are already at high risk of poor perinatal outcomes. Furthermore, this data lay the groundwork to consider expanding NC by a dietitian to a group counseling model. Group prenatal care has been shown to besuccessful in improving pregnancy outcomes in certain high risk populations, including gestational diabetes.²⁶ Thus far, data are mixed regarding the efficacy of group prenatal care for the prevention of excessive GWG,²⁷ and there is a paucity of data on group prenatal care specifically for obese women.

The strengths of our study include a large sample size of patients prospectively enrolled in this analysis, all of whom sought prenatal care and delivered at a single tertiary care center. Nutritional counseling was performed in a one-on-one setting by a single dietitian specialized in counseling in pregnancy in our clinical practice (M.D.) and allowed for standardization of the patient experience. Trained research personnel assisting in the Strong Start program enrollment and data collection, and subsequent individual detailed chart review by two trained data reviewers, ensure accuracy in this work. Our study is unique in that it focuses on both women with BMI ≥30 and ≥40, the latter of which is an understudied study population.

Limitations of our study include the possibility of selection bias, that patients who opted to participate in counseling may be inherently different than those who refused counseling. Additionally, patient-reported food composition is subject to recall bias, and as such may be an inaccurate or incomplete representation of their true intake. There was also limited ability of the registered dietitian to meet with patients sequentially to assess interval progress in dietary alteration, insofar as there may have been an even greater impact of counseling if a patient had more counseling sessions; however, this was not provided in the program nor could it have been assessed in this analysis. Our sensitivity analyses assess the impact of including those women whose GWG calculation may be inaccurate if they entered prenatal care late or had a preterm delivery, however, these analyses demonstrate that our results are not materially altered. Finally, our results may not be generalizable to all demographics of women, as we were unable to enroll nonEnglish speaking patients in individualized counseling.

Among this cohort of low socioeconomic status women with BMI ≥30 at a single tertiary care center, perceived physician reinforcement of NC by a registered dietitian, greater physical activity level, and employment were associated with aGWG. Among women with BMI ≥40, attendance of individualized professional NC was associated with aGWG, whereas physical inactivity, inadequate protein intake, and inadequate caloric intake were associated with iGWG. Further studies should assess these associations in randomized controlled trials as well as explore the possibility of group NC for these high-risk women.

Key Points.

• Physician reinforcement of nutritional counseling by a dietitian is crucial for obese women.

• Physical inactivity and unemployment are associated with inappropriate gestational weight gain.

• Nutritional counseling is associated with appropriate gestational weight gain in women with BMI ≥40.