Introduction

Excess or inadequate gestational weight gain (GWG) has serious health consequences for pregnant women and their offspring. Women who gain excessive weight during pregnancy are more likely to have a caesarean section delivery;[1] pre-eclampsia;[1] and excessive weight retention after delivery.[2] Inadequate gestational gain also has adverse consequences with birth outcomes such as increased risk of preterm birth and small for gestational age.[3] Furthermore, gestational diabetes (GDM), a common pregnancy complication has been found to increase the risk of perinatal complications,[4] most notably fetal overgrowth (i.e. macrosomia)[5, 6] and the associated birth injuries due to the large size of the fetus, progression to type 2 diabetes in the mothers, and obesity and type 2 diabetes in the offspring.[7–9] The potentially synergistic adverse effects of excess or inadequate GWG and GDM makes understanding factors associated with optimal GWG in this high risk population particularly important.

Existing studies suggest that maternal pregnancy stress may influence perinatal outcomes including GWG by disrupting the hypothalamic pituitary axis (HPA), resulting in elevation of cortisol levels and promotion of intake of nutrient-dense “comfort” food and central fat distribution.[10–12] Despite the biological plausibility of this framework, previous observational studies have failed to demonstrate an association between psychosocial stress during pregnancy and the risk of excess or inadequate GWG among women with uncomplicated pregnancies.[13] However, no previous study has examined this association among women with gestational diabetes (GDM). Psychosocial stress is a potentially modifiable, yet understudied potential risk factor that is prevalent among pregnant women,[14, 15] particularly those diagnosed with pregnancy complications such as GDM.[16, 17]

We conducted a cross-sectional analysis to explore whether there is an association between perceived stress during pregnancy and GWG, classified in accordance with the 2009 IOM recommendations for GWG,[18] among women diagnosed with GDM, using a large multiethnic cohort delivering at Kaiser Permanente Northern California (KPNC).

Materials and methods

This study is a secondary cross-sectional analysis of baseline data collected for the “Gestational Diabetes’s Effects on Moms (GEM)” trial at KPNC (Fig 1). Details of the GEM trial are described elsewhere.[19] Briefly, GEM was a cluster randomized clinical trial of the comparative effectiveness of diabetes prevention strategies for women with GDM. A telephonic lifestyle intervention was implemented after the delivery. The study was approved by the Kaiser Permanente institutional review board. The human subjects committee of the Kaiser Foundation Research Institute waived the requirement for individual informed consent for the intervention component, given the pragmatic, cluster randomized design.[19] All intervention and data assessment materials for the GEM study have been approved by the human subjects committee of the Kaiser Foundation Research Institute; the trial is also registered at clinicaltrials.gov (NCT01344278).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Overview of the pregnancy component of the GEM cluster randomized trial.

https://doi.org/10.1371/journal.pone.0174290.g001

From March 2011 to March 2012, all women with a diagnosis of GDM according to the Carpenter and Coustan criteria, as recommended by ACOG during the study period,[20] who were 18 years of age or older were identified in KPNC’s electronic health record (EHR). Women were contacted soon after the diagnosis of GDM and invited to participate in the baseline GEM survey, in English or Spanish. Among all 2,280 eligible women in the 44 medical facilities who were sent study surveys,[19] 1,783 (78.2%) completed the baseline questionnaire. Survey responders did not differ from non-responders except for being less likely to be of non-Hispanic white origin.[19]

Results

Among the 1,783 women who completed the baseline survey, we excluded those who did not fill in the stress questionnaire (n = 408) or were missing data for GWG (n = 1). We also excluded women whose pregravid BMI was below 18.5 (n = 21) since this group was too small to include in the sub-group analysis. Our final sample included 1,353 women with GDM who delivered a term singleton (≥ 37 weeks’ gestation). Those who were included in the analyses were not substantially different from those who were not included in the analyses with regard to demographic characteristics or obesity status. The PSS10 was completed at a median of 32.1 weeks gestation (IQR 29.0–34.6). Overall, the cohort had a median total GWG of 9.6 kg (IQR 6.2–13.6); 28.3% of the cohort gained below the IOM recommendation, 34.6% met the recommendation, and 37.1% exceeded the recommendation.

Baseline characteristics by stress levels

Table 1 presents the baseline demographic and lifestyle characteristics, and obesity status of study participants, stratified by the quartile categories of PSS10 score. The ranges for each quartile were 6–10, 13–15, 16–19, and 21–25 for Q1, Q2, Q3, and Q4, respectively. There were no significant differences in demographic characteristics across the quartiles of stress level, except that women with higher education were more likely to be in lower stress categories (Q1 or Q2) than their less educated counterparts (p = 0.02). Total caloric intake, percent of calories from fat and physical activity levels were similar between the groups.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics by perceived stress scale (PSS) score; Gestational Diabetes Effects’ on Mom (GEM) study.

https://doi.org/10.1371/journal.pone.0174290.t001

GWG by stress levels and pregravid obesity

Table 2 shows proportions of women who had inadequate, adequate and excessive GWG categorized by stress levels and pregravid BMI. Generally, higher proportion of women in the highest quartile of stress score tended to exceed the IOM recommendation for total GWG compared to those in the lowest stress category, but the associations were significant only among those with pregravid BMI<25 (p = 0.02).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Total GWG by stress level and pregravid BMI; Gestational Diabetes Effects’ on Mom (GEM) study.

https://doi.org/10.1371/journal.pone.0174290.t002

Association between perceived stress and excess GWG by pregravid BMI

Table 3 presents the association between PSS10 score and the risk of exceeding the IOM recommendation for total GWG categorized by pregravid BMI. Among women with a normal pregravid BMI, a high level (Q4) of stress was significantly associated with over twice the risk of exceeding IOM recommendations compared with women with low stress (Q1) [Rate Ratio (RR) = 2.16, 95% CI 1.45–3.21], after adjusting for maternal age and race/ethnicity. There was a significant test for trend across the stress categories for excessive GWG (p = 0.02). High stress level was also associated with the risk of inadequate GWG [RR = 1.39 95%CI (1.01–1.91)], though test for trend was not statistically significant. On the other hand, there was no association between stress level and the risk of excess or inadequate GWG among women who were overweight or obese prior to their pregnancy.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Adjusted rate ratios from log multinomial regression estimating the association between perceived stress level and exceeding or being below the IOM recommendation for total GWG; Gestational Diabetes Effects’ on Mom study.

https://doi.org/10.1371/journal.pone.0174290.t003

Discussion

Our study suggests that there may be an association between high perceived stress during pregnancy and excess or inadequate GWG among women diagnosed with GDM who had a healthy BMI prior to the index pregnancy. The association between high stress and excess GWG persisted when we examined maternal pregnancy weight gain (defined as total pregnancy weight gain minus infant birthweight) suggesting maternal stress was impacting maternal weight independent of fetal growth. This study, although not designed to investigate temporal relationship, suggests that there may be complex interplay between perceived stress and gestational weight gain among women with GDM.

Reports from the White House Task Force on Childhood Obesity[26] and the U.S Surgeon General[27] recommend promoting effective perinatal interventions and research on health factors affecting weight gain during pregnancy in diverse populations in an effort to interrupt the intergenerational cycle of obesity. Furthermore, the 2009 IOM report highlights the need to understand the role of psychosocial factors in optimizing GWG.[18] Women with pregnancy complications such as GDM are more likely to have a higher level of psychological stress compared to women with a healthy pregnancy.[16, 17, 28] However, the association between stress and GWG among women with GDM has been understudied.

There is some evidence supporting a relationship between psychosocial stress and/or depression and weight gain from the overall population of pregnant women. A study by Picone et al. examined psychological stress as a factor in GWG and pregnancy outcome in a small study and found a correlation between higher stress scores and lower GWG, independent of nutrient or caloric intake.[29] The authors concluded that perhaps stress did not directly affect food intake, but rather that it may have impacted the utilization of calories and nutrients from the foods consumed to support pregnancy. The 2009 IOM report, “Weight Gain During Pregnancy: Reexamining the Guidelines,” reviewed the existing literature on depression and gestational weight gain and concluded most studies suggested that both low and high GWG may be a sign of depression during pregnancy.[30] These data from general pregnant populations suggest that different effects of stress and depression on GWG can occur depending on how an individual responds to stress. The IOM report highlighted the need for more research on the role of stress in gestational weight gain.[8, 30] The association between psychosocial stress and risk of exceeding or gaining below the GWG recommendations was observed only among women with a normal pregravid BMI. For the present study, we speculate that the effect modification might have been observed because overweight or obese women are already at higher risk of exceeding the recommended GWG compared to women with a normal pregravid BMI;[31, 32] therefore, perceived stress had no additional effect on excess weight gain.

Further, some women’s stress levels might have been affected by the diagnosis of GDM. Women with the diagnosis of GDM are often put on a strict diet to optimize weight gain, therefore some women may have overcompensated the dietary regimen resulting in less overall weight gain. Among women with higher BMI, there may be many genetic, psychological, and/or environmental risk factors explaining why they exceed the recommended GWG, masking the effect of psychosocial stress on GWG and resulting in the observed effect modification by pregravid BMI.

There have been numerous behavioral intervention studies conducted in an effort to optimize GWG and reduce the risk of maternal pregnancy complications, though the results are mixed.[33, 34] In a review by Skouteris, et al., the authors suggested a need for the consideration of psychological factors relevant to pregnancy, in addition to health behaviors, including healthy eating and increased physical activity.[34] Pregnancy is a time of considerable psychological and physiological stress for many women.[14, 15] The presence of a diagnosed pregnancy complication such as GDM further escalates the level of stress and anxiety,[16, 17, 28] making stress reduction a potential candidate for future prevention strategies.

There are several limitations in this study. First, this study only included women with GDM and their stress level was assessed a few weeks after their GDM diagnosis. Since we did not assess their perceived stress prior to the GDM diagnosis, we are unable to assess whether stress increased the risk of GDM among these women or GDM diagnosis affected the level of stress, and how that impacted gestational weight gain up to the time of diagnosis. A previous study by Silveira et al. reported that an increase in stress level between early to mid-pregnancy substantially increased the odds of developing GDM among Hispanic women, compared to those with no change or a reduction in stress.[35] Although we were unable to assess the temporal relationship between stress during the early stage of pregnancy and the risk of developing GDM or excess/inadequate GWG, our study extended previous findings by suggesting that there may be associations between stress around the time of GDM diagnosis and increased risk of excess or inadequate total GWG. Second, although we did not observe mediating role of diet and physical activity, it is possible that the measures were not accurate, resulting in lack of observation. More rigorous method of diet and exercise measures are needed to better understand the role of these variables in the association between stress and weight gain during pregnancy. Third, we were unable to examine whether maternal stress had an effect on glycemic control and whether that may have impacted GWG; future studies are needed to clarify the inter-relationship between stress, glycemic control and GWG. However, we tried to assess confounding by including as many potential factors that were available as possible in the model, including income and education; none of these variables explained the observed associations (data not shown). Fourth, although we did not observe interaction between education, race/ethnicity and income, it is possible that the study did not have enough power to detect the effect modification. Lastly, there may be different types of stress (e.g., HPA of SAM responses) that we were unable to distinguish in the present study because PSS10 only measures perceived stress. If more evidence is found on the association between stress and GWG, further study examining how different types of stress affect weight gain may be warranted.

In conclusion, psychosocial stress may be an independent risk factor for gaining outside of the IOM’s recommended GWG range among women with GDM who had a normal BMI prior to the pregnancy. Future longitudinal studies of pregnant women to evaluate the interplay of stress, measured at multiple times during pregnancy, weight gain, and changes in blood glucose levels are warranted. If an association exists, future intervention studies may include stress reduction programs in conjunction with modifying lifestyle to improve the effectiveness of interventions to optimize GWG, thereby preventing future intergenerational transmission of obesity and diabetes.

Acknowledgments

We would like to thank GEM participants and research staff. Authors have no conflict of interest to disclose.