Maternal perceived stress, household disorder, eating behaviors and adiposity of women and their children

Abstract

Background:

Stress is associated with physiological and behavioral adaptations that increase the risk for obesity and related diseases in adults and children. Mechanisms linking stress to chronic disease are diverse and not fully elucidated, but research suggests stress may impact eating behaviors and increase food intake and thereby, risk for obesity.

Objective:

The purpose of this study was to test the hypotheses that women’s perceived stress and household disorder are associated with more uncontrolled and emotional eating among women, more food responsiveness and emotional overeating among their children, and greater adiposity in both women and their children.

Methods:

Women (n = 86) completed the Perceived Stress Scale, Confusion, Hubbub and Order Scale, Three Factor Eating Questionnaire, and Child Eating Behavior Questionnaire. Total body fat (%) was measured via dual-energy X-ray absorptiometry. Linear regression models evaluated associations of perceived stress and household disorder with eating behaviors and adiposity of women and their children (4–10 years old).

Results:

In a sample of predominantly non-Hispanic Black women (84.9%, n = 73), more perceived stress and household disorder were associated with more uncontrolled and emotional eating (p < 0.05). Women’s perceived stress was not associated with their children’s eating behaviors; however, household disorder was positively associated with children’s food responsiveness and emotional overeating (p < 0.05). Perceived stress and household disorder were not associated with adiposity of women or their children.

Conclusions:

These findings suggest household disorder may be a factor for home-based interventions to consider when addressing eating behaviors among families with children.

1. Introduction

Stress is defined as a state in which an individual’s homeostasis is challenged, or perceived to be challenged, by physical or psychological stimuli, called stressors (Chrousos, 2009). In response to stressors, the body activates physiological and behavioral processes to re-establish homeostasis (Godoy et al., 2018). These stress responses are fundamental to survival; however, in the context of repetitive or chronic stressors across the lifespan, they can have deleterious impacts on health and behavior (Chrousos, 2009; McEwen, 1998; McEwen & Stellar, 1993; Schneiderman et al., 2005).

In adults, chronic stress is a risk factor for obesity and related diseases such as type 2 diabetes and cardiovascular disease (Kelly & Ismail, 2015; Steptoe & Kivimaki, 2013). Similarly, children who experience chronic stress are at greater risk for obesity and future cardiometabolic disease (Jakubowski et al., 2018; Miller & Lumeng, 2018). The mechanisms underlying the association of chronic stress and disease are not fully understood but may involve physiological and behavioral factors. Repeated or prolonged exposure to hormones released during the body’s physiological stress response, such as cortisol, is detrimental to cardiometabolic health (Chrousos, 2009; Hackett & Steptoe, 2017; Tank & Lee Wong, 2015). Previous research has also shown chronic stress is adversely related to behaviors associated with health and body weight such as physical activity, sleep, and eating (Cohen et al., 1997; Gardani et al., 2022; Stults-Kolehmainen & Sinha, 2014).

Regarding eating specifically, research in adults and children shows that stress is associated with increased food intake, especially of highly-palatable, energy-dense foods (Hill et al., 2022; Hill et al., 2018). Perceived stress in adults, which reflects an individual’s perception of the stress in their lives, is associated with less control of eating and more eating in response to negative emotions (Joseph et al., 2018; Richardson et al., 2015; Sims et al., 2008). It is not clear, however, if women’s perceived stress is associated with comparable eating behaviors in their children, such as food responsiveness (i.e., wanting food, having food in mouth) and emotional overeating. Additionally, little is known about whether exposure to household disorder, wherein the home environment is characterized by noise, disorganization, and a lack of predictability and routine, is associated with eating behaviors of women and children. Prior research shows an association of household disorder with stress and risk factors for cardiometabolic disease such as blood pressure and body weight so it is plausible that household disorder may be associated with eating behaviors (Marsh et al., 2020; Martin et al., 2023).

The objective of the study was to evaluate whether women’s perceived stress and household disorder are associated with eating behaviors and adiposity of women and their children. Among women, we hypothesized higher perceived stress and more household disorder would be associated with more uncontrolled eating and emotional eating. Among children, we hypothesized higher maternal perceived stress and more household disorder would be associated with more food responsiveness and more emotional overeating. Much of the previous work demonstrating an association between stress and obesity risk has used outcomes of body weight and body mass index (BMI), however the accuracy of BMI as a proxy measure of adiposity and predictor of cardiometabolic disease risk is limited and may not be consistent across races (Dugas et al., 2011; Fairchild et al., 2018; Gomez-Ambrosi et al., 2012; Gujral et al., 2017). Therefore, we used a more rigorous assessment of adiposity and hypothesized that women’s perceived stress and household disorder would be positively associated with percent total body fat of women and their children independent of eating behaviors.

2. Methods

2.1. Participants

This is a secondary analysis of observational, cross-sectional data collected from a cohort of mother-child dyads enrolled in a study investigating health during pregnancy and long-term outcomes of women and their children 4 to 10 years after pregnancy (Martin et al., 2022). Mother-child dyads were eligible to participate in the parent study if the index pregnancy was a singleton gestation of ≥ 36 weeks, the women were between 20 and 36 years old at delivery, and the child was between 4 and 10 years old at enrollment without any medical condition that could impact growth or metabolic health, or developmental disability that would preclude completion of study procedures. As part of the parent study, a subgroup of participants completed additional phenotyping including evaluation of body composition and eating behaviors; the data from this subset of mother-child dyads was used for this analysis. The Institutional Review Board of Human Use at the University of Alabama at Birmingham approved all procedures for the parent study (record number IRB-170518007). All mothers provided written consent and children ≥ 7 years provided written assent.

2.2. Procedure

Participants were recruited from June 2017 to April 2019. In the parent study, mother-child dyads were recruited such that dyads filled groups characterized by maternal BMI at entry to prenatal care and gestational diabetes mellitus (GDM) status during the index pregnancy: maternal BMI < 25 kg/m² without GDM, maternal BMI ≥ 25 kg/m² without GDM, and maternal BMI ≥ 25 kg/m² with GDM. Women and their children reported to the research clinic for 2 study visits. During the first study visit, a trained research assistant collected anthropometrics and verbally administered surveys to assess demographics, household food security, and household disorder. A licensed nurse practitioner performed a brief physical examination of children aged ≥ 7 years to assess pubertal stage. Mother-child dyads returned within 4 weeks to complete the second study visit, however, given the children were school-aged, some return visits were delayed until a weekend visit could be scheduled. During the second study visit, a trained research assistant verbally administered validated questionnaires to women to evaluate their perceived stress, eating behaviors, and their child’s eating behaviors. The women and children also completed a whole-body dual-energy x-ray absorptiometry (DXA) scan to evaluate body composition.

2.3. Measures

Perceived stress –

Women’s perceived stress was evaluated using the 10-item Perceived Stress Scale (PSS) (Cohen et al., 1983). The PSS is a widely used and validated survey that asks individuals to report how often they found their lives unpredictable, uncontrollable, and overloaded in the preceding month (Lee, 2012). Women responded on a scale from 1 (never) to 5 (very often) (Cronbach’s α = 0.87). Scores range from 10 to 50; higher scores indicate more perceived stress.

Household disorder –

Household disorder was measured using the Confusion, Hubbub and Order Scale (CHAOS), a 15-item survey that has been used among families with children to assess an individual’s perception of characteristics of their home environment such as noise, disorganization, and a lack of predictability and routine (Dumas et al., 2005; Matheny, 1995). On a scale of 1 (very much like your own home) to 4 (not at all like your own home) women indicated how much the statements described their own home (Cronbach’s α = 0.85). The range of scores is 15 to 60, higher scores indicate more household disorder.

Women’s eating behaviors –

Women’s eating behaviors were assessed using the 18-item Three-Factor Eating Questionnaire (TFEQ-R) (Karlsson et al., 2000). The TFEQ-R has been validated and used in the general population to evaluate the following cognitive and behavioral components of eating: cognitive restraint, uncontrolled eating, and emotional eating, of which, the current study used uncontrolled eating and emotional eating (de Lauzon et al., 2004). Uncontrolled eating was evaluated with 9 questions of the TFEQ-R and is characterized by an individual’s experience of eating large quantities of food and feeling little control while eating (Cronbach’s α = 0.78). Emotional eating was evaluated with 3 questions of the TFEQ-R and is defined by an individual’s experience of eating while feeling anxious, sad, or lonely (Cronbach’s α = 0.83). On a scale of 1 (definitely true) to 4 (definitely false) women indicated the extent to which statements describing eating behaviors described their own behavior. Scores for uncontrolled eating range from 9 to 36, and for emotional eating from 3 to 12; higher scores indicate less uncontrolled eating and emotional eating.

Children’s eating behaviors –

Women completed the Child Eating Behavior Questionnaire (CEBQ) to evaluate their child’s eating behaviors (Wardle et al., 2001). The CEBQ has been used among preschool and school-aged children and uses 35 items to assess how frequently a child displays habitual eating behaviors across 8 constructs, 2 of which are food responsiveness and emotional overeating (Domoff et al., 2015; Sleddens et al., 2008). Food responsiveness was evaluated with 5 items of the CEBQ and is a measure of a child’s tendency to want something to eat and/or to have food in their mouth (Cronbach’s α = 0.81). Emotional overeating was assessed with 4 items and is characterized by a child’s tendency to eat more in different emotional states (Cronbach’s α = 0.86). Women answered each question on a scale of 0 (never) to 4 (always). Subscale scores were calculated using the average score of the items within each construct, yielding a possible range from 0 to 4; higher scores indicate more food responsiveness and emotional overeating.

Body composition –

DXA (GE-Lunar Radiation Corp., Madison, WI) assessed total body fat mass and lean soft tissue mass, and calculated percent total body fat for women and children. The manufacturer software (enCORE Version 15[SP2]) selected pediatric settings and equations for children given they were < 20 years old.

2.4. Covariates

Self-reported race was dichotomized as Black or White, which included Hispanic and non-Hispanic ethnicities. Women reported their highest education level, which was grouped as: less than high school, some high school, high school graduate, some college, college graduate, and graduate degree. Women’s GDM status was self-reported with verification from prenatal care records, if available_. Child sex assigned at birth was reported by women. Child age was calculated as the number of years between the child’s date of birth and date of the first study visit. The U.S. Department of Agriculture Adult Food Security Survey Module was used to assess food security in the home (Bickel et al., 2000). Due to limited distribution across the score range, household food security status was dichotomized as food secure or food insecure following methods used previously (Lumeng et al., 2014; So et al., 2024) scores on the food security screener that deviated from high food security were categorized as food insecure.

2.5. Statistical Analyses

Descriptive statistics were calculated to summarize participant demographics and characteristics. Descriptive statistics are displayed as means and standard deviations for continuous variables, and frequencies followed by percentages for categorical variables.

Linear regression analyses were performed to evaluate the influence of women’s perceived stress and household disorder on outcome variables of interest among women (uncontrolled eating, emotional eating, and percent total body fat) and their children (food responsiveness, emotional overeating, and percent total body fat). Perceived stress, household disorder, and all outcome variables were treated as continuous variables in the models. Missing values were treated by complete case analysis, in which missing values were excluded from the analysis. Outliers were identified using visual inspection of residuals from regression models; residuals ± 3 standard deviations from the mean were eliminated.

Non-parametric Spearman correlations and Mann-Whitney U tests were used to evaluate the associations of outcome variables with the following covariates, which were selected based on prior literature: women’s education, race, household food security status, GDM status during pregnancy, child sex assigned at birth and child age (Dunton et al., 2017; Eagleton et al., 2022; Martin-Biggers et al., 2018; Shapiro et al., 2017). GDM status was not associated with outcome variables, and therefore was not included in final models. Models for the women were adjusted for women’s education, race, and household food security status, whereas models for the children were adjusted for child sex, age, and household food security status.

G*Power version 3.1.9.6 was used to determine the minimum sample size required to test the study hypotheses (Faul et al., 2009). The required sample size to achieve 80% power for detecting a medium effect (f² = 0.15) (Cohen, 1992), with α = 0.05 was n = 43. Thus, the obtained sample of n = 86 is more than adequate the test the study hypotheses. Statistical significance was α = 0.05 for all tests. All analyses were performed using Analytics Software & Solutions (SAS 9.4, SAS Institute Inc., Cary, NC).

3. Results

Women and children who completed all procedures in the parent study were considered for this secondary analysis. As shown in the participant flow chart in Figure 1, mother-child dyads were excluded due to either incomplete data (n = 11) or not meeting inclusion criteria for the current study (n = 1), yielding a final sample of n = 86 mother-child dyads for this analysis.

Figure 1.

Flow chart depicting the number of participants who completed the parent study to those who were included in the current study.

Sample characteristics are shown in Table 1. The average age of women was 34.63 ± 5.38 years. Most women were non-Hispanic Black (84.9%, n = 73), and the remainder were non-Hispanic White (14.0%, n = 12) and Hispanic White (1.1%, n = 1). Women’s highest level of education ranged from less than high school (7.0%, n = 6), to high school and some college (59.3%, n = 51), followed by undergraduate and graduate degrees (33.7%, n = 29). Children in this sample were on average 7.45 ± 2.15 years old, and just over half of the children were assigned female at birth (54.7%, n = 47). Of the represented households, 40.8% (n = 41) reported food insecurity, while the remainder were food secure. The average number of days between study visits was 58.56 ± 43.29 days.

Table 1.

Sample characteristics (n = 86)

Women	Mean + SD Unless noted otherwise
Age, years	34.63 ± 5.38
Non-Hispanic, African American or Black race, n (%)	73 (84.9%)
Marital status, single, n (%)	51 (59.3%)
Education, n (%)
Less than high school	1 (1.2%)
Some high school	5 (5.8%)
High school graduate	32 (37.2%)
Some college	19 (22.1%)
College graduate	16 (18.6%)
Graduate degree	13 (15.1%)
Household income, n (%)
< $25,000	43 (50.0%)
$25,000 - $49,999	22 (25.6%)
$50,000 - $99,999	11 (12.8)
$100,000+	10 (11.6%)
Food secure, n (%)a	45 (59.2%)
Perceived stress, PSS score	24.52 ± 7.54
Household disorder, CHAOS score	26.30 ± 8.46
Eating behaviors, TFEQ-R score
Uncontrolled eating	27.48 ± 4.77
Emotional eating	9.47 ± 2.32
BMI, kg/m2	33.35 ± 9.15
Percent total body fat	41.24 ± 8.89
Children
Age, years	7.45 ± 2.15
Female at birth, n (%)	47 (54.7%)
Tanner stage 1, n (%)b	69 (81.2%)
Eating behaviors, CEBQ score
Food responsiveness	1.77 ± 1.01
Emotional overeating	1.22 ± 0.92
BMIz	0.76 ± 1.23
Percent total body fat	27.58 ± 9.50

Note:

^aData from n = 76; missing participant response.

^bData from n = 85; missing participant response.

Abbreviations: BMI, body mass index; BMIz, body mass index z-score; CEBQ, Child Eating Behavior Questionnaire; CHAOS, Confusion Hubbub and Order Scale; PSS, Perceived Stress Scale. TFEQ-R, Three-Factor Eating Questionnaire Revised.

Figure 2 displays the associations of perceived stress and household disorder with women’s eating behaviors. Perceived stress was significantly associated with women’s eating behaviors in the unadjusted and adjusted models. Specifically, women who experienced higher perceived stress reported more uncontrolled eating (b = −0.24, p < 0.01) and more emotional eating (b = −0.11, p < 0.01), independent of women’s education, race, and household food security. Household disorder was significantly associated with more uncontrolled eating (b = −13.36, p < 0.01) and more emotional eating (b = −4.40, p = 0.02) in unadjusted models (Table S1). However, after adjusting for women’s education, race, and household food security, household disorder remained a significant predictor of women’s uncontrolled eating (b = −11.21, p < 0.01), but not emotional eating (b = −4.32, p = 0.09).

Figure 2.

Associations of (A and B) women’s perceived stress and (C and D) household disorder with women’s eating behaviors. Models adjusted for women’s education, race, and household food security status. Data shown in figure represent the transformed data used in analyses. CHAOS, Confusion Hubbub and Order Scale; PSS, Perceived Stress Scale; TFEQ-R, Three-Factor Eating Questionnaire Revised.

Maternal perceived stress was not associated with children’s food responsiveness or emotional overeating (b = 0.01 and 0.02, respectively, p > 0.05; Table S2). However, as shown in Figure 3, household disorder was associated with children’s eating behaviors such that more household disorder was predictive of more food responsiveness (b = 2.51, p = 0.03) and more emotional overeating (b = 1.80, p = 0.02), independent of child’s sex, age, and household food security.

Figure 3.

Associations of household disorder with children’s (A) food responsiveness and (B) emotional overeating. Models adjusted for child sex, age, and household food security status. Data shown in figure represent the transformed data used in analyses. CEBQ, Child Eating Behavior Questionnaire; CHAOS, Confusion Hubbub and Order Scale.

Among women and their children, women’s perceived stress and household disorder were not significantly associated with percent total body fat (p > 0.05; Table S3). Additional models (not shown) evaluated the association of household disorder with child adiposity within each sex and pubertal stage, however no significant relationships were observed.

4. Discussion

This study’s objective was to evaluate the association of women’s perceived stress and household disorder with their uncontrolled eating and emotional eating, and comparable eating behaviors among their children including food responsiveness and emotional overeating. Consistent with our hypotheses, women’s perceived stress and household disorder were positively associated with their uncontrolled and emotional eating. Among children, household disorder, but not maternal perceived stress, was positively associated with child food responsiveness and emotional overeating. Finally, contrary to what was hypothesized, neither women’s perceived stress nor household disorder were associated with adiposity in women or children. Results of this study suggest that although maternal perceived stress and household disorder are associated with eating behaviors that increase food intake, they may not be directly related to weight-related outcomes of women and children, although effects on adiposity may become apparent as children age.

Consistent with previous studies, women who reported higher perceived stress engaged in more uncontrolled eating and emotional eating (Groesz et al., 2012; Pickett et al., 2020; Pickett, 2018; Sims et al., 2008). Prior work related to household disorder and eating behaviors in adults is limited, with one study to date evaluating these associations among women (Martin-Biggers et al., 2018). As such, this study was the first to show that women with school-aged children who live in households with more noise and commotion, and less routine and stability engage in more uncontrolled eating and emotional eating. The observed associations between stressors of the home environment and women’s eating behaviors may be consequent, in part, to physiological and behavioral processes. For example, cortisol, a glucocorticoid released by the hypothalamic-pituitary adrenal axis in response to stress, has been associated with increased levels of appetite-stimulating hormones such as ghrelin and neuropeptide Y, and reduced sensitivity to appetite-suppressing hormones including leptin and insulin (Kuckuck et al., 2023). Alternately, stress may reduce executive function, which is a collection of cognitive processes responsible for planning, thinking ahead, and goal-directed behaviors (Shields et al., 2016). In turn, lower executive function has previously been linked to engagement in more disinhibited eating (O’Neill et al., 2020). It has also been proposed that the eating behaviors and patterns of food consumption associated with states of negative affect (i.e., stress) may serve as coping mechanisms that individuals use to manage their mood and regulate their emotions (Garg et al., 2007). When considered together, these findings provide physiological and psychological explanations for why women who report more perceived stress and household disorder report engaging in more uncontrolled and emotional eating.

Children’s eating behaviors were related to household disorder but not to maternal perceived stress. Specifically, children of women who reported more disorder in the home environment displayed more food responsiveness and emotional overeating. Previous work has suggested there may be a positive relationship between parental stress and their child’s food responsiveness and emotional overeating (Varghese et al., 2023), however, the absence of a relationship between women’s perceived stress and eating behaviors of their children has been reported as well (Dunton et al., 2017). Regarding household disorder and children’s eating behavior, the results of the current study extend our understanding of this relationship. Previous work by Buchanan and colleagues reported a positive relationship between household disorder and child food responsiveness among children aged 5 to 7 years old, however, the study was limited by its use of the 4-item CHAOS rather than the full 15-item version (Buchanan et al., 2021). Another study found that food responsiveness and emotional overeating of children aged 18 to 24 months old were positively associated with household disorder, particularly if mothers were not attentive and responsive to children’s needs during mealtimes (Saltzman et al., 2019). Further investigation is needed to identify factors that may influence the associations among maternal perceived stress, household disorder, and children’s eating behaviors together with effective interventions to improve eating behaviors.

No relationships were observed between women’s perceived stress or household disorder and adiposity in either women or their children. This was an unexpected finding as it was inconsistent with results from prior studies among adults which demonstrated associations between either perceived stress or household disorder with greater body weight and adiposity (Hruska et al., 2020; O’Neill et al., 2020). The results from this study were consistent, however, with a prior study of non-Hispanic Black women, which reported that despite an observed relationship between women’s perceived stress and emotional eating, perceived stress was not related to women’s adiposity, indicated by BMI (Pickett, 2018). Inconsistent results in the literature and the current study, may be explained, in part, by differences in racial diversity, with studies reporting associations of perceived stress and household disorder with adult adiposity having predominantly non-Hispanic White participants (83.6% and 70.7%, respectively), whereas our study and that from Pickett et al., included predominantly non-Hispanic Black women. It is also possible that associations of maternal perceived stress and household disorder with child adiposity may emerge later in life as there is longitudinal evidence to suggest that stress experienced during early childhood and adolescence is associated with increases in obesogenic eating behaviors over time, and is predictive of BMI and ratings of overall physical health in adulthood (Farrell et al., 2017; Greenfield & Marks, 2009; Miller et al., 2018).

The findings of the study should be interpreted in the context of its strengths and limitations. Strengths include the evaluation of eating behaviors in women and their children, as opposed to evaluating women and children separately, and the use of DXA to assess body composition rather than dependence on anthropometric measures (i.e., BMI, waist circumference). Although the assessments for stress, household disorder and eating behavior were subject to women’s perception and self-report, the surveys used in this study have been shown to be valid and reliable methods to evaluate these constructs in adults and children (de Lauzon et al., 2004; Domoff et al., 2015; Dumas et al., 2005; Karlsson et al., 2000; Lee, 2012; Matheny, 1995; Pérez-Fuentes et al., 2019; Sleddens et al., 2008). Due to the cross-sectional observational design of this analysis, results demonstrate correlational relationships and causation cannot be inferred. Finally, as mentioned previously, this study was conducted in a relatively homogeneous sample of predominantly non-Hispanic Black mother-child dyads, which may limit generalizability.

5. Conclusion

In conclusion, our findings extend prior studies by showing that in addition to the association of perceived stress and women’s eating behaviors, household disorder is also associated with eating behaviors. Further, children’s eating behaviors (at ages 4–10 years) were related to household disorder but not to maternal perceived stress. These findings suggest that household disorder, characterized by noise, commotion, and a lack of predictability and routine in the home, may be a factor to consider during the development and evaluation of home-based interventions designed to address obesity risk and eating behaviors among families with children. Although it was unexpected that women’s perceived stress and household disorder were not associated with adiposity of women and their children, the cross-sectional nature of this study may have precluded the ability to detect such an association. Future research evaluating the relationship between household disorder and eating should include other caregivers and adults in the home environment to expand our understanding of these associations beyond mother-child dyads. Additionally, more research is needed to better understand the physiological and psychological processes that underlie the associations between stressors in the home environment and eating behaviors in adults and children.

Highlights.

• Stress may be associated with eating behaviors that increase risk for obesity.

• Stressors in the home environment are associated with women’s eating behaviors.

• Household disorder is adversely related to children’s eating behaviors.

• Interventions to promote healthy eating should consider role of household disorder.