Gestational diabetes mellitus (GDM), defined as glucose intolerance with onset or first recognition during pregnancy, is one of the most common complications in pregnancy, representing about 19% of pregnancies annually, using the new International Association of the Diabetes and Pregnancy Study Groups (IADPSG) criteria [1]. Being overweight or obese before pregnancy, as well as excessive gestational weight gain (GWG), are leading risk factors for developing GDM [2, 3]. Women with GDM compared to women with normoglycemic pregnancies are seven times more likely to develop type 2 diabetes (T2D) later in life [4]. Postpartum weight change appears to influence future risk of T2D in women with a history of GDM. In more than 600 Hispanic women with GDM, a postpartum weight gain of 4.5 kg was independently associated with a twofold increase in the risk of developing diabetes [5]. In a prospective cohort study with up to 18 years of follow-up among 1695 primarily Caucasian GDM women, each 5 kg increment of weight gain after GDM was associated with a 27% higher risk of developing T2D [6]. Pregnancy weight retention can vary greatly, with 20% of women in general having substantial postpartum weight retention (SPPWR), defined as ≥5 kg at 1 year postpartum [7], and women with GWG above recommendations retaining the most weight [8].

Identifying dietary predictors of SPPWR, particularly in GDM women, would focus behavioral intervention efforts for weight reduction. A large national retrospective study showed that women with GDM compared to their counterparts without GDM were significantly less likely to meet the five or more servings of fruit and vegetable recommendations [9]. Recently, there have been a handful of randomized controlled trials targeting postpartum weight reduction in GDM women [10]. A lifestyle intervention for 397 postpartum women (45% African American) who were overweight/obese reported that lower junk food at six weeks postpartum was associated with greater weight loss at 12, 18 and 24 months postpartum [11]. In the KAN-DO postpartum nutrition intervention study with 308 women who were overweight/obese [12], only energy intake at baseline (2–7 months postpartum) was a significant negative predictor of weight change at 12–17 month follow-up. A cluster-randomized controlled trial of 2280 women with GDM randomized to usual care or lifestyle (diet + physical activity) perinatal intervention found that the intervention group retained less weight at 6 months postpartum than the usual care group, but the weight retention was markedly attenuated at 12 months postpartum [13]. All of the above intervention studies are fairly non-specific and target a variety of healthful dietary and/or physical activity behaviors.

The purpose of this study was to evaluate if the intake of specific food and beverages at 6–9 weeks postpartum predicts SPPWR at 1-year in women after GDM pregnancy. We hypothesized that women who consume high amounts of high fat foods (such as processed meats, butter and fried foods), regular soda, sweet grains, and low amounts of dietary fiber, low-fat dairy, lean meats, whole grains, fruit and vegetables at 6–9 weeks postpartum will have higher odds of SPPWR at 1-year postpartum.

Eating fried foods was associated with higher odds of SPPWR in unadjusted and adjusted models. Consuming ≥5 servings of fried food a week (n = 32) versus none (n = 295) was associated with a threefold higher odds of SPPWR (OR = 3.72, 95% CI:1.74–7.96, P < 0.001), and consuming 2–4 servings per week of fried foods versus none was associated with a two-fold greater odds of SPPWR (OR = 1.99, 95% CI:1.30–3.03, P = 0.002). These results for fried foods remained highly significant after adjusting for other food and beverage intake and all a priori covariates (OR = 3.38, 95% CI:3.38–1.36, P = 0.009 for intake of ≥ 5 serv/wk., and OR = 1.80, 95% CI:1.8–1.08, P = 0.02 for intake of 2–4 serv/wk. versus none).

This study found that higher intakes of soda, and fried food intake during early postpartum periods were independently associated with substantial postpartum weight retention. However, processed meats and egg intake showed no independent association after controlling for other food and beverage intake. To our knowledge, this is the first study to show direct associations for high intakes of fried food and soda intake with SPPWR in a high-risk GDM population.

Fried food intake at baseline was linked to three-fold higher odds of SPPWR and this impact remained highly significant after controlling for soda intake and other confounding variables. National data shows that approximately 31% of adults consume fast food on any given day and fast food intake is associated with an increase in daily energy intake of 190 kcals per day and over 10 additional grams of fat [25]. In a cross-sectional study of 226 women with GDM by Zehle et al., 26% of the women consumed fried food at least twice per week [26]. Similarly, we found that 27% of our sample consumed fried food at least twice per week. Numerous adult studies have consistently found a link between fast food intake and obesity, [27–29] however, few studies have examined the relationship of fast food and/or fried foods on postpartum weight retention. In the Active Mothers Postpartum study, which included 450 overweight or obese White and African American postpartum women, lower junk food intake (which included French fries, fast food, chips, and soda and sugar sweetened beverages) at baseline (6 weeks postpartum) was linked to greater weight loss at 12-, 18, and 24-months postpartum [11]. These findings along with our findings, suggests that targeting the reduction of fried foods early in postpartum may have profound impacts on reducing PPWR.

Strong evidence exists to support that sugar sweetened beverage (SSB) intake, including soda intake, is a leading culprit in both pediatric and adult obesity [30, 31]. National data in 2009–2010 found that US adults consumed an estimated average 151 kcal/day of SSB, with regular soda and fruit drinks representing the leading sources of SSB energy intake [32]. National data showed that 26% of adults consumed regular soda or fruit drinks or both at least once daily and consumption was highest among adults between 18 and 34 years of age and in minority adults [33]. Fewer studies have examined SSB intake during pregnancy or postpartum periods. In a cross-sectional study with 115 lactating Mexican women living in Mexico City, 23% of women drank approximately 2.5 servings of SSB per day, in the first three months postpartum [34]. In the current study, reported soda intake was significantly lower than the above study and the national average, where only 23% of GDM women drank soda 2–4+ servings per week. Regardless of the lower intake, soda intake during early postpartum periods was linked to increased rates of SPPWR. This data suggests that even low to moderate soda intake (i.e., 2–4+ servings per week) can hinder postpartum weight loss.

However, the link between soda consumption and SPPWR was attenuated somewhat after controlling for processed meats, and egg consumption. While research consistently shows that increased fast food intake is linked to increases in SSB intake [25, 29], our findings suggest that fried food intake compared to soda intake has a more profound impact on SPPWR. One possible explanation for this finding is that these GDM women were often counseled by their healthcare team to limit sugar intake, but fried food intake may not have been addressed.

In the current study, processed meat intake and egg intake were not associated with odds of SPPWR after accounting for covariates, fried foods and regular soda. It is possible that some processed meats were also counted as fried foods, for example bacon and sausage, but fried foods in total had a stronger impact on weight gain. Thus, the fried foods intake may overwhelm any impact of processed meats on SPPWR. Another explanation is that subjects who consumed fried foods, also consumed higher quantities of processed foods and lower quantities of eggs, but fried foods had a stronger effect on PPWR than the other high fat foods.

Recently a number of RCT have been conducted with GDM women targeting postpartum weight reduction. A feasibility RCT conducted by Ferrara et al. [35] with 197 women with GDM found no significant differences in postpartum weight loss between those women receiving an intensive telephone-based lifestyle intervention, which targeted increases in moderate intensity physical activity, exclusive breastfeeding for six months, and reductions in caloric and fat intake, compared to the control group. However, the intervention appeared to be more effective among women who did not exceed the guidelines for GWG. A cluster-RCT conducted by Ferrara et al. [13] including 2280 women with GDM found that a similar telephone based lifestyle intervention compared to usual care group resulted in approximately one kilogram less weight retention at 6 months postpartum, but the weight retention was attenuated at 12 months postpartum. A 12-week lifestyle intervention conducted by Bertz et al. [36] with 61 Swedish women who are overweight/obese, found that women who decreased their energy intake, primarily by decreasing carbohydrates, sucrose and fat intake at 12 weeks and sustained these reductions at 1-year postpartum had significant reductions in weight loss at both 12 weeks and 1-year. Another RCT with 75 women with GDM found that a 1-year web-based lifestyle intervention that began at 6 weeks postpartum decreased postpartum weight retention at 1-year [37]. To date, the interventions targeting postpartum weight reduction in GDM women have primarily used lifestyle interventions that target multiple dietary and physical activity behaviors.

Evidence remains unclear on whether intervention targeting multiple behavior changes vs. single behavior change is more effective at weight loss. A recent review of 22 meta-analyses studies, showed that interventions targeting a single health behavior change (i.e., nutrition vs. physical activity) were more effective at improving the targeted behavior, while multiple health behavior interventions (i.e., nutrition + physical activity) resulted in greater weight loss [38]. However, this review classified the single health behavior interventions as either nutrition or physical activity behaviors, which ultimately still includes multiple behavior targets within each category. Few studies have examined whether or not changes in a single dietary behavior could have more weight loss effects compared to changing multiple dietary behaviors. More long-term interventions targeting single or few dietary behavior changes are warranted to fully understand the long-term benefits of such an approach.

Focusing on one dietary component, such as decreasing fried food intake, is a very concise and easy message to focus on, and might be easier for postpartum women with GDM to understand and achieve. Individuals asked to make multiple behavior changes report feeling confused and overwhelmed [39] and this may push them to quit and return to their unhealthy lifestyles. Particularly, because new mothers are facing a variety of stressors, including sleep deprivation, hormonal changes, and the demands of caring for a newborn [40]. In addition, body dissatisfaction and depression increases significantly during postpartum periods [41, 42]. Thus, a simpler and more succinct intervention to follow, such as decreasing soda intake, may be easier and more attainable for new mothers.

Contrary to our hypothesis, low intakes of whole grains, fruit and vegetables were not linked to SPPWR. In a large national sample, women with self-reported GDM who lived with their children compared to their counterparts without GDM were significantly less likely to meet the five or more servings of fruit and vegetable recommendations [9]. In another cross-sectional study of 226 women with GDM, only 5% consumed the five or more servings a day of vegetables and 44% consumed two or more servings a day of fruit. Higher intakes of fruit, vegetables, and whole grains have been shown to reduce risk of obesity and associated chronic diseases in adults [43–45], however, few studies have examined how these specific foods are related to postpartum weight retention. One retrospective study with 444 Swedish women showed no difference in fruit and vegetable intake during pregnancy between women with a history of normal pregnancy compared to those with a history of GDM [46]. A possible explanation for our null findings with fruit and vegetable intake, is that the PrimeScreen survey was not designed or validated to evaluate the combined fruit and vegetable intake, and it is possible that the sum of vegetables and/or fruits would have been linked to reductions in PPWR. In addition, energy, saturated fat, animal fat and fiber intake were not associated with SPPWR, and this also might be because the PrimeScreen is less accurate in assessing energy and nutrient intake.

There are some other limitations to consider for the current study. One limitation is that the caffeine questionnaire did not include questions for other SSBs, such as juice, fruit drinks and sports drinks, which are also known to be large contributors to added sugar intake in the American diet. Another limitation is the rather small sample size of subjects consuming fried food ≥5 servings a week at baseline (n = 32), however, at least 23% of the population (n = 208) consumed fried food 2–4 times per week at baseline and this group compared to the never group also had SPPWR. There was also a large percent of women (48%) in this cohort that reported not consuming SSBs at baseline. This is likely because clinicians treating these GDM women advised them not to consume SSB intake to control their blood sugars during pregnancy, and this abstinence may have continued into early postpartum periods. Another limitation is that dietary intake may have changed from baseline to 1-year postpartum. Lastly, this was not an intervention study, so causality could not directly be addressed. However, the precise longitudinal measures obtained at baseline and 1-year postpartum allowed for models to be run assessing which dietary variables predicted postpartum weight retention.

The study evaluated dietary intake and risk of weight retention among women with GDM who are seven times more likely to develop T2D later in life. Although postpartum weight loss may lower future risk of T2D [4, 5], few studies have evaluated postpartum weight retention among women with GDM, or evaluated the link between specific food and beverage intake and PPWR. Given the limited evidence on specific dietary behaviors linked to postpartum weight retention in women with GDM, these findings may be important to reduction in future diabetes risk. This study suggests that dietary modifications focused on fried food and soda reduction after delivery may lower risk of SPPWR at 1-year postpartum, particularly for women with excessive gestational weight gain. This is a very simple and straightforward message for healthcare providers to convey to women that may be an effective strategy for diabetes prevention in postpartum women with GDM.