Parenting Stress, Child Weight-Related Behaviors, and Child Weight Status

Abstract

Background:

There has been limited examination of the association between parenting stress and child weight-related behaviors. We aimed to determine whether parenting stress is associated with child weight-related behaviors, including physical activity, screen time, diet, sedentary time, and eating in the absence of hunger (EAH). Secondarily, we assessed association between parenting stress and child weight status.

Methods:

Mother-child dyads (N = 291) enrolled in the Newborn Epigenetic STudy (NEST), a longitudinal cohort study, completed surveys to describe parenting stress, and child diet. Children participated in the EAH task and wore accelerometers to assess sedentary time and physical activity. Child weight status was assessed using measured height and weight. Outcomes and exposures were examined using generalized linear models and restricted cubic splines as appropriate based on linear lack-of-fit test.

Results:

Child sedentary time and vegetable consumption were inversely associated with parenting stress (Total Stress B = −0.78; 95% confidence interval [CI]: −1.35 to −0.20; p = 0.017; and Total Stress adjusted odds ratio [aOR] = 0.98; 95% CI: 0.99 to 1.00; p = 0.022, respectively). Child screen time was directly associated with parenting stress (Total Stress = aOR 1.01; 95% CI: 1.00–1.02; p = 0.032). Fast-food intake was nonlinearly associated with parenting stress. There was no evidence of association between parenting stress and child EAH, physical activity, or weight status. Associations between parenting stress and child weight-related behaviors were not moderated by race or family structure.

Conclusions:

Parenting stress was associated with important child weight-related behaviors but not weight status. Management of parenting stress may represent a reasonable adjunct to family-based behavioral interventions.

Introduction

Parenting stress is psychosocial distress resulting from the demands and responsibilities of parenting.¹ Parenting stress is associated with child psychosocial distress as well as distress in the family unit.¹ Children of parents with higher levels of parenting stress may be at higher risk of developing child obesity; however, findings vary across studies.² The mechanisms by which parenting stress may influence child obesity risk include parental inability to effectively enforce household rules, such as limitations around food or screen time, to minimize parent-child conflict.^3–5 Alternatively, children may exhibit unhealthy behavioral responses to parent-child conflict or emulate unhealthy parental responses to stress.⁶

There has been limited examination of the influence of parenting stress on child weight-related behaviors.⁷ The majority of existing studies rely on parent-reported physical activity and assess limited dimensions of parenting stress.² In addition, few studies have examined the association between parenting stress and child eating behaviors, such as eating in the absence of hunger (EAH), an eating behavior associated with both psychosocial stress, and obesity among children.^8,9 Finally, few studies have examined whether associations between parenting stress and child weight-related behaviors are the same across racial groups and family structures.^7,10 Prior studies have shown that parent psychosocial stress may have a particularly negative impact on health behaviors among children of color and children living in single-parent homes.^10,11 Further elucidation of the association between parenting stress and child weight-related behaviors is needed as such findings could inform family-based interventions that include parenting stress management as an adjunct to behavior-based approaches.¹²

The primary objective of our study was to determine whether parenting stress was associated with key child behaviors associated with child obesity (child diet, EAH, physical activity, sedentary time, and screen time).^8,13 As a secondary objective, we examined whether these associations were moderated by parent race and family structure. Finally, we assessed association between parenting stress and child weight status. Overall, we hypothesized that children of parents with higher levels of parenting stress would be more likely to engage in weight-related behaviors, which may lead to obesity, including less physical activity, more sedentary time, and higher frequency of calorically dense, low nutrient foods.

Methods

Participants and Procedures

Study participants were part of the Newborn Epigenetic STudy (NEST), a longitudinal cohort assembled to examine the effects of prenatal and early childhood exposures on epigenetic and developmental outcomes.¹⁴ Starting in 2016, women with children between 4 and 10 years from the NEST study were recruited for a follow-up study. Eligibility criteria required women participants to speak English and children to have been a singleton birth. Spanish-speaking mothers were excluded in the follow-up study because several measures were unavailable in Spanish at the time. Women provided consent and children provided verbal assent. All women identified themselves as the mother of the child. The present study is a cross-sectional analysis of 291 mothers and children who were enrolled in the NEST follow-up study from 2016 to 2019. Compared to the overall NEST sample, the sample for the present study included a higher percentage of women who were college graduates (39.9% vs. 33.2%) and who identified as black (60.8% vs. 43.6%).The NEST study and follow-up study were approved by the Duke University Institutional Review Board.

Demographic Data

Demographic data collected from enrolled mothers included race/ethnicity, age, marital status, and educational attainment. Mothers were asked to describe their financial security using a single item, “How would you describe your household's financial situation right now?” Response choices were as follows: “(1) After paying the bills, you still have enough money for special things that you want; (2) You have enough money to pay the bills, but little spare money to buy extra or special things; (3) You have money to pay the bills, but only because you have to cut back on things; and (4) You are having difficulty paying the bills, no matter what you do.” Financial security was operationalized as a 4-level categorical variable.

Parenting Stress

Mothers completed the Parenting Stress Index Short Form, 4th edition (PSI).¹⁵ The PSI is composed of 36 items and provides a total score and scores for 3 subscales: Parental Distress, Parent-Child Dysfunctional Interaction, and Difficult Child. The PSI is validated in multiple populations, including U.S. black and Latinx parents.¹⁶ The normal range for scores is within the 16th to 84th percentiles.¹⁵ Mothers within our cohort had a mean Total Stress score of 58.0 (standard deviation [SD] 43.9) (27th percentile). Parenting stress subscale mean scores were 17.6 (SD 16.7) (28th percentile) for Parent-Child Dysfunctional Interaction, 20.0 for Difficult Child (SD 14.1) (18th percentile), and 20.5 (SD 14.7) (27th percentile) for Parental Distress. Raw scores were used for analysis. The full distribution of scores is included as Supplementary Data.

Moderate to Vigorous Physical Activity and Sedentary Time

Accelerometers (ActiGraph) were used to directly measure children's daily physical activity counts and to summarize moderate-vigorous activity (MVPA) and sedentary time. Children were fitted for the device during the study visit and asked to wear the accelerometers for 7 days, including 2 weekend days, for at least 8 continuous waking hours each day. Written instructions and verbal instructions provided by study staff emphasized wearing the devices as much as possible. Mothers received daily texts to encourage compliance. After 7 days, accelerometer data were downloaded by study staff to ensure that there were at least 3 weekdays and 1 weekend day recorded. The Freedson cut points were used to measure the amount of time children spent in MVPA and sedentary time over the course of the week.¹⁷ For each child, a day was selected at random and minutes spent in sedentary activity, and minutes spent in MVPA were calculated based on this index day.¹⁸

Screen Time

Mothers answered questions regarding their child's screen time. Specifically, mothers were asked the following: “Think back over the past 7 days. On a typical day, how many hours of screen time does your child get? Screen time includes time watching TV, playing on the computer, or playing video games.” Response options included “None,” “1 hour or less,” “2 hours,” “3 hours,” “4 hours,” “5 hours,” “6 hours or more,” and “I don't know.” Screen time was dichotomized as 0–2 hours per day versus 3 or more hours per day corresponding to American Academy of Pediatrics screen time recommendations for school-aged children.¹⁹

Diet

Mothers answered questions regarding their child's typical daily consumption of a healthy food (vegetables) and calorie dense, nutrient poor foods (fast-food, desserts, and French fries/chips) using a food frequency questionnaire widely used in North Carolina public health programs.^20,21 Response options to describe vegetable, dessert and fry/chip frequency were as follows: “None,” “More than none, but less than 1 serving per day,” “1 serving per day,” “2 servings per day,” or “3 or more servings per day.” Vegetable consumption was dichotomized as less than 2 servings per day or 2 or more servings per day based on age-based dietary guidelines.²² French fries/chip and dessert frequency were dichotomized as less than every day versus 1 or more servings per day. Examples of vegetable serving sizes were provided, as were definitions of chips (i.e., potato chips, corn chips, Cheetos, and so on). Mothers were also asked the number of times per week that their child usually ate from a fast-food restaurant. Response options were as follows: “None,” “More than never, but less than once a week,” “1 time per week, “2 times per week,” “3 times per week,” “4 times per week,” and “5 or more times per week.” Examples of fast-food restaurants were provided. Fast-food frequency was dichotomized as less than once per week versus 1 or more times per week.

Eating in the Absence of Hunger

The EAH test has been validated in samples of young children and adolescents.⁸ Mothers were asked to have their children refrain from eating for two hours before the task and received a reminder the day before the study visit. At their appointment, children and their mothers were provided with a meal, and children were given up to 30 minutes to eat until they felt full. Hunger before the meal and fullness following the meal were verified using a 1 (very hungry) to 5 (not at all hungry) Likert scale via a developmentally appropriate self-report measure of satiety.²³ On average, children rated their satiety as 2.06 (SD = 0.93) before the meal and 4.69 (SD = 0.57) after the meal. After the meal, the child was taken to a room by themselves with age-appropriate toys and containers of preweighed portions of five snack foods (e.g., M&Ms, CheezIts) that the child had identified preference for. Children were told that they could play with any of the toys and eat as much or as little of the snacks as they liked, and the examiner would check on them in 10 minutes. Following the task, the amounts of remaining snacks were weighed and total calories consumed were calculated.

Anthropometric Data

Mothers' and children's height and weight were measured and BMI was calculated using National Health and Nutrition Examination Survey (NHANES) anthropometry procedures.²⁴ Child weight status was defined as underweight (BMI <5th percentile); healthy weight (BMI ≥5 and <85th percentile); overweight (BMI ≥85th percentile and <95th percentile), or obese (BMI ≥95th percentile).²⁵ Analyses of child weight status and parenting stress compare healthy weight versus overweight and healthy weight versus obesity. Children with underweight were excluded from analyses of parenting stress and child weight status.

Statistical Analysis

Descriptive statistics are given for the sample under investigation. Continuous variables are presented as mean (SD) and categorical variables as number (percentage). A bivariate regression model was built to explore the unadjusted association between PSI scores and their unadjusted association with child weight-related behaviors and child weight status. A generalized linear model was utilized for each combination of exposure and outcome. An underlying assumption of generalized linear models is that continuous exposures have a linear association with the outcome. This assumption was assessed for all continuous exposures when fitting each model. The linearity assumption was assessed by performing a lack-of-fit test comparing a linear fit with a nonlinear fit based on a restricted cubic spline with 3 knots at observed percentiles (10th, 50th, and 90th) of the exposure. If evidence of a nonlinear association was found (linearity test p-value <0.05), a piecewise linear spline transformation with a single knot, that adequately accounts for the nonlinearity, was applied to the covariate before including it in the final association model to enhance interpretability.²⁶ The nonlinear association between the exposure and the outcome is modeled as a linear association between the exposure and outcome for exposure values less than the knot and another linear association for exposure values after the knot. Odds ratios and confidence intervals (CIs) are reported for each part of the linear spline. Associations between parenting stress and all of the outcomes (child weight-related behaviors and child weight status) were adjusted for mother's BMI, child age, mother's education level, mother's race/ethnicity, and financial security.

To describe moderation between parenting stress, race, and family structure, a model was built, including an interaction term for child weight-related behaviors. Using a lack of fit test, we compared unadjusted models with and without the interaction term. If the lack of fit test was significant (p-value <0.05), we concluded that the corresponding variable moderated the effect between parenting stress and child weight-related behaviors. The false discovery rate was used to correct for multiple testing.

Results

There were 291 mother-child dyads included in our study (Table 1). Approximately half of children were female, with a mean age of 7.4 years. Fourteen percent of children had overweight and 23% had obesity. The mean age of enrolled mothers was 35.8% and 59.9% of mothers reported having a partner or spouse. Nearly two-thirds of mothers identified as black while one-third identified as non-Hispanic white. Three percent of mothers identified as Hispanic and 5% as other race. Mothers were diverse in their educational attainment and financial security (Table 1).

Table 1.

Demographic Characteristics of Mothers and Enrolled Children (N = 291 Mother-Child Dyads)

Children (N = 291)
Female, N (%)	151 (51.9)
Child age, mean (SD)	7.4 (2.0)
Overweight,a N (%)	41 (14.1)
Obese,b N (%)	68 (23.4)
Healthy weight,c N (%)	168 (57.8)
Mothers (N = 291)
Mom BMI, mean (SD)	32.6 (9.7)
Mother's age, mean (SD)	35.8 (5.8)
Partner or spouse, N (%)	173 (59.9)
Race, N (%)
Black	177 (60.8)
Hispanic	8 (2.7)
Other	13 (4.5)
White	93 (32.0)
Maternal education, N (%)
Less than high school completion	40 (14.1)
High school completion or GED completion	69 (24.4)
Completed some college	61 (21.6)
College graduate	113 (39.9)
Financial security (%)
After paying the bills, you still have enough money for special things that you want	97 (33.3)
You have enough money to pay the bills, but little spare money to buy extra or special things	116 (39.9)
You have money to pay the bills, but only because you have to cut back on things	49 (16.8)
You are having difficulty paying the bills, no matter what you do	29 (10.0)
Parenting Stress Index Short Form Score, mean (SD)/percentiled
Parental Distress	20.5 (14.7)/27
Parent-Child Dysfunctional Interaction	17.6 (16.7)/28
Difficult child	20.0 (14.1)/18
Total stress	58.0 (43.9)/27

^a BMI percentile <95th and >85th.

^b BMI percentile >95th.

^c BMI percentile >5th and <85th.

^d Percentiles per Abidin.¹⁵

GED, General Educational Development; SD, standard deviation.

Adjusted Associations between Parenting Stress and Child Diet

Higher parenting stress was associated with lower odds of children meeting vegetable intake guidelines using Total Stress and across all subscales of the PSI (Total Stress aOR = 0.99; 95% CI: 0.99–1.00; p = 0.022) (Table 2). Parenting stress was not associated with dessert or French fries/chips consumption.

Table 2.

Adjusted Associations between Parenting Stress and Dietary Behaviors (N = 291 Mother-Child Dyads)

Child weight-related behaviors	Exposure	Odds ratio (95% CI)		Corrected p-valuea
Linear Models
Vegetables	Parental Distress	0.98 (0.96–1.00)		0.022
	Parent-Child Dysfunctional Interaction	0.98 (0.97–1.00)		0.029
	Difficult Child	0.98 (0.96–0.99)		0.022
	Total Stress	0.99 (0.99–1.00)		0.022
Desserts	Parental Distress	0.99 (0.98–1.01)		0.874
	Parent-Child Dysfunctional Interaction	1.00 (0.98–1.01)		0.874
	Difficult Child	1.00 (0.99–1.02)		0.874
	Total Stress	1.00 (1.00–1.01)		0.874
French fries/chips	Parental Distress	1.01 (0.99–1.03)		0.280
	Parent-Child Dysfunctional Interaction	1.01 (0.99–1.03)		0.283
	Difficult Child	1.01 (0.99–1.03)		0.265
	Total Stress	1.00 (1.00–1.01)		0.280

Nonlinear Models		Levels (based on apex of curve)
Fast food	Parental Distress	≤25	1.06 (1.02–1.11)	0.032
		>25	0.95 (0.91,0.99)
	Parent-Child Dysfunctional Interaction	≤25	1.06 (1.01–1.11)	0.032
		>25	0.94 (0.88–0.99)
	Difficult Child	≤25	1.06 (1.02–1.11)	0.009
		>25	0.94 (0.90–0.99)
	Total Stress	≤55	1.03 (1.01–1.05)	0.009
		>55	0.99 (0.97–1.00)

Bolded p values are the ones which are statistically significant (p < .05).

Adjusted for maternal BMI, child age, maternal education, maternal race, financial security; Logistic regression used, estimates are odds ratios.

^a p Values are corrected for multiple testing within each outcome by using the False Discovery Rate.

CI, confidence interval.

There was nonlinear association between parenting stress and fast-food frequency (Chi-square nonlinear lack-of-fit association 11.02; p = 0.004) (Fig. 1). Mothers with median Total Stress score (50th percentile or Total Stress score of 44) had children with higher odds of frequent fast-food consumption than mothers at the 5th or 90th percentile for Total Stress score. Similar associations were found between fast-food frequency and the 3 PSI subscales (Table 2).

Figure 1.

Association of parenting stress and fast-food frequency (N = 291 mother-child dyads).

Adjusted Associations between Parenting Stress, Child Activity Behaviors, and EAH

Child screen time was directly associated with the Total Stress score as well as the Parent-Child Dysfunctional Interaction and Difficult Child subscales (Total Stress aOR 1.01; 95% CI: 1.00–1.02; p = 0.032) (Table 3). Child screen time was not associated with the Parental Distress subscale (aOR 1.02; 95% CI: 1.00–1.04; p = 0.095). Child sedentary activity was inversely associated with mothers' Total Stress score as well as each of the PSI subscores (Total Stress B = −0.78; 95% CI: −1.35 to −0.20; p = 0.017). Neither child MVPA nor EAH was associated with parenting stress (Total Stress B = 0.06; 95% CI: −0.14 to 0.03; p = 0.288 for child MVPA and Total Stress B = −0.22; 95% CI: −0.52 to 0.07; p = 0.265 for EAH).

Table 3.

Adjusted Associations between Parenting Stress, Activity Behaviors, Eating in the Absence of Hunger (N = 291 Mother-Child Dyads)

Child weight-related behaviors	Exposure	Odds ratio or estimate (95% CI)	Corrected p-valuea
Screen time	Parental Distress	1.02 (1.00 to 1.04)	0.095
	Parent-Child Dysfunctional Interaction	1.02 (1.00 to 1.04)	0.035
	Difficult Child	1.03 (1.01 to 1.06)	0.012
	Total Stress	1.01 (1.00 to 1.02)	0.032
Sedentary	Parental Distress	−2.00 (−3.71 to −0.29)	0.025
	Parent-Child Dysfunctional Interaction	−1.73 (−3.24 to −0.23)	0.025
	Difficult Child	−2.89 (−4.65 to −1.13)	0.006
	Total Stress	−0.78 (−1.35 to −0.20)	0.017
Eating in the absence of hunger	Parental Distress	−1.01 (−1.88 to −0.14)	0.094
	Parent-Child Dysfunctional Interaction	−0.38 (−1.14 to 0.39)	0.336
	Difficult Child	−0.52 (−1.42 to 0.38)	0.336
	Total Stress	−0.22 (−0.52 to 0.07)	0.265
Total MVPA	Parental Distress	−0.16 (−0.42 to 0.09)	0.288
	Parent-Child Dysfunctional Interaction	−0.18 (−0.40 to 0.05)	0.288
	Difficult Child	−0.13 (−0.39 to 0.14)	0.362
	Total Stress	−0.06 (−0.14 to 0.03)	0.288

Adjusted for maternal BMI, child age, maternal education, maternal race, financial security; Logistic or linear Regression used, estimates are usual estimates from a linear model.

^a p Values are corrected for multiple testing within each outcome by using the False Discovery Rate.

MVPA, moderate to vigorous physical activity.

Moderation by Race and Family Structure and Association with Child Weight Status

There was some evidence that race moderated the association between parenting stress and child MVPA (Table 4). Before correction for multiple testing, higher levels of Parental Distress and Total Stress were associated with lower levels of MVPA in black children only (total stress B = −0.12; 95% CI: −0.25 to 0.00; p = 0.046 and Parental Distress B = −0.37; 95% CI: −0.73 to −0.02; p = 0.030). Following correction for multiple testing, there was no evidence of significant moderation (corrected p = 0.075 for Parental Distress and p = 0.085 for Total Stress). There was neither an evidence of moderation by race across any of the other outcomes nor an evidence of moderation of the association between parenting stress and child weight-related behaviors by family structure (mothers who reported a partner/spouse vs. those who did not). Parenting stress was not associated with child overweight (total stress aOR 1.00; 95% CI: 0.99–1.01; p = 0.958) or child obesity (total stress aOR 1.00; 95% CI: 1.00–1.01; p = 0.292) (Table 5).

Table 4.

Moderation of the Association of Parenting Stress and Child Moderate to Vigorous Physical Activity by Race and Family Structure

Outcome	Exposure	Estimate (95% CI)	p	Corrected p-valuea
Race
Total MVPA	Parental Distress
	Parental Distress | White	0.24 (−0.28 to 0.77)	0.030	0.075
	Parental Distress | Black	−0.37 (−0.73 to −0.02)
	Parent-Child Dysfunctional Interaction
	Parent-Child Dysfunctional Interaction | White	0.13 (−0.37 to 0.62)	0.085	0.085
	Parent-Child Dysfunctional Interaction | Black	−0.32 (−0.64 to −0.01)
	Difficult Child
	Difficult Child | White	0.22 (−0.33 to 0.77)	0.056	0.075
	Difficult Child | Black	−0.35 (−0.72 to 0.03)
	Total Stress
	Total Stress | White	0.07 (−0.11 to 0.25)	0.046	0.075
	Total Stress | Black	−0.12 (−0.25 to 0.00)
Family structure
Total MVPA	Parental Distress
	Parental Distress | Partner/Spouse	0.02 (−0.34 to 0.37)	0.092	0.092
	Parental Distress | No Partner/Spouse	−0.41 (−0.86 to 0.04)
	Parent-Child Dysfunctional Interaction
	Parent-Child Dysfunctional Interaction | Partner/Spouse	−0.01 (−0.33 to 0.31)	0.068	0.092
	Parent-Child Dysfunctional Interaction | No Partner/Spouse	−0.42 (−0.81 to −0.03)
	Difficult Child
	Difficult Child | Partner/Spouse	0.03 (−0.35 to 0.40)	0.092	0.092
	Difficult Child | No Partner/Spouse	−0.43 (−0.90 to 0.05)
	Total Stress
	Total Stress | Partner/Spouse	0.00 (−0.12 to 0.12)	0.069	0.092
	Total Stress | No Partner/Spouse	−0.15 (−0.31 to 0.00)

Linear regression used, estimates are usual estimates from a linear model.

^a p Value for lack of fit test corrected for multiple testing using False Discovery Rate.

Table 5.

Adjusted Associations between Parenting Stress and Child Weight Status (N = 277 Mother-Child Dyads)

Child weight status	Exposure	Odds ratio (95% CI)	Corrected p-valuea
Overweightb	Parental Distress	0.99 (0.96–1.02)	0.860
	Parent-Child Dysfunctional Interaction	1.01 (0.98–1.03)	0.860
	Difficult Child	1.00 (0.97–1.03)	0.860
	Total Stress	1.00 (0.99–1.01)	0.860
Obesityc	Parental Distress	1.01 (0.99–1.04)	0.292
	Parent-Child Dysfunctional Interaction	1.02 (0.99–1.04)	0.292
	Difficult Child	1.01 (0.99–1.04)	0.326
	Total Stress	1.00 (1.00–1.01)	0.292

Logistic regression^, model, adjusted for maternal BMI, child age, maternal education, maternal race, financial security.

^a p Value for lack of fit test corrected for multiple testing using False Discovery Rate.

^b Adjusted odds of overweight (BMI percentile <95 percentile and ≥85% percentile) compared to healthy weight (BMI ≥5 and <85th percentile).

^c Adjusted odds of obesity (BMI percentile ≥95%) compared to healthy weight (BMI ≥5 and <85th percentile).

Discussion

Children of parents with higher parenting stress had more screen time, less sedentary time, and less vegetable consumption compared to children of parents with lower parenting stress. In models corrected for multiple testing, the association between parenting stress and child weight-related behaviors was similar across children of different races and family structures. Finally, parenting stress was not associated with child overweight or obesity.

Prior studies have examined parenting stress, child sedentary behavior, and child screen time.^4,10,27,28 However, there is significant methodological heterogeneity. Similar to our study, two of these found a direct association between parenting stress and child screen time or television time.^27,28 Notably, a study by Walton et al., utilizing the Parental Distress subscale of the PSI-Short Form, did not find an association between parenting stress and child television time, however, the authors did find that parents with higher levels of parental distress were less likely to set limits regarding television time, providing a potential mechanism for our findings.⁴ We found an inverse association between parenting stress and sedentary time, contrary to a prior study which found no association.¹⁰ This may reflect increased parenting stress, which can occur when parenting a child with a more active temperament.^29–31 Notably, the relationship between child activity levels and child temperament is complex and a more active child temperament may not lead to higher levels of physical activity.^29,31 Further study using comparable measures of stress and activity as well as child temperament are needed to better understand the association between child activity levels and parenting stress.

Few studies have assessed the association between parenting stress and child diet.² Given previously described associations between stress and emotional eating, we anticipated that parenting stress might be associated with greater child intake of calorically dense foods (desserts, French fry/chips, and fast food).³² However, we only found evidence of an association between parenting stress and less frequent child vegetable consumption. High parenting stress has been linked with higher levels of picky eating in children, which could result in reduced intake of vegetables.³ The potential link between higher parenting stress and lower child intake of healthy foods warrants further study using a more comprehensive, longitudinal examination of dietary quality to determine whether parenting stress has a clinically relevant impact on child diet.

We examined whether there was effect moderation of the association between parenting stress and child weight-related behaviors by race and family structure. Effect moderation of the association between parenting stress and child MVPA by family structure was previously noted by Maher et al. revealing an inverse association between parenting stress and MVPA only in children of single-parent families.¹⁰ By contrast, we found no evidence of moderation by family structure. This may be because Maher et al. compared multigenerational and dual-parent families to single-parent families. In our examination of moderation by race, black children appeared to have increased vulnerability to low MVPA in high parenting stress contexts, before correction for multiple testing. Although there was no apparent association after correcting for multiple testing, the impact of parenting stress on weight-related behaviors in black families and single-parent families warrants further study given their risk for child obesity and higher levels of parenting stress.^33,34 Mixed method studies of parenting stress, which focus on these key populations, may reveal aspects of parenting stress that are important for reducing obesity disparities.

A unique feature of our study is our examination of linear and nonlinear associations between parenting stress and child weight-related behaviors. Nonlinear associations between maternal psychosocial stress and perinatal and child outcomes have been previously described.^35,36 Parents with the median level of parenting stress reported the highest frequency of child's fast food consumption. While difficult to interpret fully, this may indicate that parenting stress plays a minor or inconsistent role in child's fast-food consumption.³⁷ Importantly, parent psychosocial stress (i.e., from any stressor, not exclusively related to the parent role) has been found to be directly associated with child's fast-food consumption and thus may represent a more promising intervention target than parenting stress alone.³⁷

Just under half of studies assessing associations between parenting stress and child weight status have found a direct association.² Overall, our study shows that parenting stress is associated with some child weight-related behaviors, but not child overweight or obesity risk. For family-based obesity interventions, parenting stress management may be best viewed as an adjunct to current strategies and could be helpful when targeting behaviors such as screen time and vegetable consumption. It is also possible that parents who report high levels of parenting stress may be more successful with child behavior change if parenting stress is addressed.

Our study has several limitations. Diet and screen time measures are self-reported and thus vulnerable to recall bias. Parents may not be aware of their child's dietary intake and screen time while in school or childcare. We also acknowledge that we measured the parenting stress of only one primary caregiver, the child's mother. Due to the demographics of our sample, we were only able to assess racial moderation for white and black mothers. Examination of other racial and ethnic groups is important given child obesity disparities. Finally, because this is a cross-sectional study, we are unable to ascribe causation to our findings.

Strengths of our study include the use of accelerometry to measure child activity and examination of multiple dimensions of parenting stress. Notably, our results indicate that the total score and subscales of the PSI similarly capture aspects of parenting stress that influence child weight-related behaviors. Finally, we are among the first to test the association between parenting stress and child EAH.

Conclusions

Higher levels of parenting stress were associated with more child screen time, less sedentary time, and lower frequency of vegetable consumption. These findings invite further investigation into the mechanisms of these associations and longitudinal studies to determine causation. Cumulatively, our findings support incorporating strategies for managing parenting stress as an adjunct to family-based obesity interventions.

Authors' Contributions

All authors made substantial contributions to the conception or design of the study, analysis, or interpretation of the analysis. M.J.W. drafted the initial draft of the article and revised it. J.C.S., E.M.P., A.J.K., S.H.K., B.F.F., B.N., C.R., and R.L.M. provided critical revisions. All authors approve the version to be published and agree to be accountable for all aspects of the work.

Disclaimer

The content is solely the responsibility of the authors and does not necessarily represent the official views of the funders. Further, USEPA does not endorse the purchase of any commercial products or services mentioned in the publication.

Funding Information

This work was supported by National Center for Advancing Translational Sciences of the National Institutes of Health under Award Number 1KL2TR002554 (M.J.W.), the National Institute of Environmental Health Sciences R01ES016772, R21ES014947, P30ES011961 pilot project, P01ES022831, the US Environmental Protection Agency RD-83543701, the National Institute of Diabetes and Digestive and Kidney Diseases (R01DK085173), and the Eunice Kennedy Shriver National Institute of Child Health and Human Development R01HD08448.

Author Disclosure Statement

No competing financial interests exist.

Supplementary Material

Supplementary Data