Associations Among Parents’ Internalized Weight Bias, Negative Child-focused Body Talk, and Feeding Behaviors

Abstract

Objective:

Parents’ negative body talk about children (negative weight/shape comments) and internalized weight bias (application of negative weight-based stereotypes to oneself) are associated with children’s maladaptive eating behaviors, but mechanisms are poorly understood. Conceptually, parental behaviors and attitudes may translate to implicit and explicit concerns about their child’s weight and influence parents’ feeding practices. These associations are underexplored in the literature.

Method:

242 parents (59.4% mothers) completed a one-time assessment that included measures of internalized weight bias, negative body talk, and feeding practices. Parents also completed assessments about one of their children, of any weight status (age range 5-15 years; 40% daughters).

Results:

Parents’ internalized weight bias was positively associated with concern about their child’s weight and restriction of their child’s diet. Parents’ negative body talk towards their child was positively associated with concern about their child’s weight and shape, restriction of their child’s diet, and monitoring of their child’s diet. Relative to internalized weight-bias, negative body talk had a stronger correlation with parents’ concern about child weight and monitoring of child’s diet. Correlations did not vary by child gender or weight category.

Discussion:

Parents’ internalized weight bias and negative body talk about their child were associated with their feeding practices across child gender and weight categories, with correlations in the small to medium range. Thus, children of all genders and weights may be vulnerable to enacted weight bias. Future studies should examine whether addressing parents’ internalized weight bias and communication in family-focused interventions improves feeding practices and child well-being.

Introduction

Weight stigma, which refers to the negative societal evaluation of people living in larger bodies, has increased over time (Latner & Stunkard, 2003). Weight stigma can manifest as unfavorable attitudes towards one’s own body weight/shape, applying society’s attitude to one’s own body (i.e., “weight-bias internalization”), and/or negative comments about weight consistent with society’s attitude, directed at oneself or others (i.e., “negative body talk”). Although there is evidence that negative stereotypes about higher weight individuals have been internalized in general adolescent populations (Puhl et al., 2009), youth with a higher body mass index (BMI; kg/m²) are more likely to experience weight-based teasing than their non-overweight peers (Brixal et al.,2012). Negative body talk and weight-bias have been associated with numerous negative mental health outcomes, even those unrelated to body image and eating behaviors (e.g., depression; Shannon & Mills, 2015). Parents have a critical role in shaping the behaviors, attitudes, and values of their offspring, particularly during childhood and adolescence, developmental periods marked by normative changes in body shape/weight, increased autonomy regarding eating- and weight-related behaviors, and increasing understanding of complex constructs. Thus, a better understanding of the relationship between parents’ cognitions and behaviors around shape and weight, and their children’s health outcomes, is warranted.

Parents are highly influential in the management of their children’s eating behaviors and in their children’s attitudes about weight (Arroyo et al., 2013), and are one source of weight-based stigma (Lydecker et al., 2018). Parents influence their children’s weight-related attitudes and eating behaviors both directly by providing guidelines and information, and indirectly through modeling attitudes and behaviors that children observe (Khandpur et al., 2014; Mazzeo et al., 2005). Indeed, high levels of familial concern with weight are associated with eating disorders and other mental health outcomes in children (Kluck, 2010). Negative body talk and parental encouragement to diet are associated with disordered eating and increased weight (Balantekin et al.,2014; Neumark-Sztainer et al.,2010). Overall, while most parents strive to optimize their children’s physical and mental health (Bryant-Waugh et al., 2007), their attempts to control and oversee their child’s diet can be associated with negative outcomes such as poor emotional regulation and disordered eating (Birch, 2006).

To this end, the present study examined how parental weight bias internalization and negative body talk relate to parents’ feeding practices. We hypothesized that higher parental internalized weight bias and more frequent negative body talk would be associated with more restrictive feeding practices, greater concern about their child’s weight, and more monitoring of their child’s diet. Additionally, we explored whether the relationships between weight bias internalization, negative body talk, and parent feeding practices differed by child gender or weight status.

2. Methods

2.1. Procedures

Community-based adults (aged ≥21y) with at least one child (aged 5-15y) who lived with them over half of the time (n=242) were recruited via Mechanical Turk to complete online assessments regarding parenting, bullying, eating, and weight. Participants recruited from Mechanical Turk yield data with reliability and validity comparable to other forms of internet-based surveys and samples have greater age and geographic diversity (Buhrmester et al., 2011). To ensure valid responding, we included multiple validity checks, including tests of attention and using multiple response formats, as recommended for best practices (Behrend et al., 2011). Parents were asked to have one child in mind when completing the surveys even if they had multiple children.

Electronic consent was obtained from participants before completing surveys. This study was approved by the university’s institutional review board.

2.2. Measures

2.2.1. Body Mass Index (BMI)

Parents reported their child’s age, height, and weight, which were used to calculate age- and sex-specific BMI z-scores and percentiles. We created weight status categories using Centers for Disease Control (CDC) growth charts that considers children’s age and sex (Kuczmarski et al., 2000); these calculations (CDC, 2023) flag “biologically implausible values,” which we did not include in our data as part of the data-cleaning protocol. Parents also self-reported their own height and weight, which were used to compute parent BMI (kg/m²).

2.2.2. Fat Talk Questionnaire (FTO)

The FTQ is a 16-item self-report measure that assesses negative talk regarding weight and body image. The original measure was developed to target young adult women and their negative talk about their own bodies (Royal et al, 2013); however, the scale in the current study was adapted (Lydecker et al., 2018) to evaluate how parents speak to their child about their child’s appearance and weight (e.g., “I comment that they have gained weight”). Response options ranged from 1=never through 5=always. Greater scores indicate higher occurrences of negative body talk directed at children from parents. Items yielded an internally consistent total score, α=.98.

2.2.3. Weight-Bias Internalization Scale (WBIS)

The WBIS is a 11-item self-report questionnaire examining the degree to which individuals internalize harmful weight-related stereotypes and biases. The current study used the modified version, which uses language inclusive of all weight categories (Pearl et al., 2014). Higher scores on the WBIS indicate greater levels of weight-bias internalization. Items yielded an internally consistent total score in the current study, α=.94.

2.2.4. Child Feeding Questionnaire (CFQ)

The CFQ assesses parental feeding practices and attitudes (Birch et. al, 2001). We used the scoring structure identified by Anderson and colleagues, which generates five subscales (perceived responsibility, concerns about child weight, restriction, pressure to eat, monitoring; Anderson et al., 2005; Lydecker et al., 2017). The CFQ has been validated for children ages 5-15y (Berger et al., 2016; Birch et al., 2001; Kaur et al., 2006). For the present study, we included three subscales that have been shown to relate to child weight status and disordered eating: concern about child weight (reflecting the extent to which the parent is concerned about their child’s risk of elevated weight), restriction (reflecting how much parents restrict the types of foods their child eat) and monitoring (reflecting how much parents watch what their children eat). Greater scores reflect greater levels of each construct. Alphas for the subscales used in the current study ranged from .79 to .93.

2.3. Statistical Analyses

SPSS and SAS were used for statistical analyses. We first examined correlations between WBIS and CFQ scores, as well as (separately) FTQ and CFQ scores. Next, we applied Fisher’s R-to-Z transformations to compare whether the magnitude of the correlations differed between stigma variables (WBIS or FTQ) and then by child gender (sons vs. daughters) and by child weight category (BMI at or above vs. below the 85^th percentile for age and sex).

Results

3.1. Parent and Child Characteristics

Parent and child demographic characteristics are summarized overall, as well as by child gender and child weight status, in Table 1. Parents included mothers (n=143; 59.1%), fathers (n=96; 39.7%), and those identifying as another gender (n=3; 1.2%). Parents self-identified their race as American Indian (n=1; 0.4%), Asian (n=24; 9.9%), Black (n=26; 10.7%), White (n=178; 73.6%), or Multiracial (n=6; 2.5%); 5.0% self-identified their ethnicity as Hispanic (n=12). On average, parents were 38.9 years of age (SD=8.9). Children had a mean age of 11.8 years (SD=2.9) and 59.1% were sons (n=143). Two parents (0.8%) opted not to report the gender of their children. Child BMI percentiles were across the weight spectrum (M percentile=68.0%; SD=32.0).

Table 1.

Demographic characteristics of parents and children

		Child Gender		Child Weight Status
	Full sample	Sons	Daughters	BMI <85th %le	BMI ≥85th %le
	N=242	n=143	n=97	n=129	n=107
Parent gender, n (%)
Female	143 (59.1)	74 (51.7)	67 (69.1)	77 (59.7)	62 (57.9)
Male	96 (39.7)	67 (46.9)	29 (29.9)	50 (38.8)	44 (41.1)
Another gender	3 (1.2)	2 (1.4)	1 (1.0)	2 (1.6)	1 (0.9)
Race, n (%)
American Indian	1 (0.4)	1(0.7)	0 (0.0)	1 (0.8)	0 (0.0)
Asian	24 (9.9)	14 (9.8)	10 (10.3)	13 (10.1)	10 (9.3)
Black	26 (10.7)	18 (12.6)	8 (8.2)	16 (12.4)	10 (9.3)
Multiracial	6 (2.5)	3 (2.1)	3 (3.1)	5 (3.9)	1 (0.9)
White	178 (73.6)	103 (72.0)	73 (75.3)	93 (72.1)	81 (75.7)
Not specified	7 (2.9)	4 (2.8)	3 (3.1)	1 (0.8)	5 (4.7)
Ethnicity, n (%)
Hispanic/ Latinx	12 (5.0)	6 (4.2)	6 (6.2)	4 (3.1)	7 (6.5)
Not Hispanic/ Latinx	230 (95.0)	137 (95.8)	91 (93.8)	125 (96.9)	100 (93.5)
Education, n (%)
High school or less	34 (14.0)	16 (11.2)	18 (18.6)	15 (11.6)	18 (16.8)
Some college	59 (24.4)	36 (25.2)	22 (22.7)	32 (24.8)	25 (23.4)
College degree	147 (60.7)	90 (62.9)	56 (57.7)	81 (62.8)	64 (59.8)
Not specified	2 (0.8)	1 (0.7)	1 (1.0)	1 (0.8)	0 (0.0)
Household, n (%)
Single parent	55 (22.7)	28 (19.6)	27 (27.8)	36 (27.9)	18 (16.8)
Dual parent	175 (72.3)	108 (75.5)	66 (68.0)	86 (66.7)	86 (80.4)
Other	11 (4.5)	6 (4.2)	4 (4.1)	7 (5.4)	3 (2.8)
Not specified	1 (0.4)	1 (0.7)	0 (0.0)	0 (0.0)	0 (0.0)
Child gender, n (%)
Female	97 (40.1)			60(46.5)	36 (33.6)
Male	143 (59.1)			69 (53.5)	71 (66.4)
Another gender	0 (0.0)			0 (0.0)	0 (0.0)
Not specified	2 (0.8)			0 (0.0)	0 (0.0)
Child age, M±SD	11.8±2.9	11.8±2.8	11.7±3.0	12.3±2.6	11.1±3.0
Parent age, M±SD	38.9±8.9	39.5±9.0	38.1±8.7	39.9±8.8	37.97±9.0
Child BMI %le, M±SD	68.0±32.0	70.0.±33.3	65.2±30.1	45.6±27.4	95.1±4.0

Note. BMI %le=Body mass index percentile (age- and sex-normed). Child BMI %le was not available for 6 participants.

Table 2 describes correlations between WBIS and FTQ scores and parenting variables overall, by child gender, and by child weight category. In the overall study sample, WBIS scores were positively associated with CFQ concern (r=.26, p<.001) and restriction (r=.16, p=.03), but not with CFQ monitoring (r=.04, p=.54). FTQ scores were also positively associated with CFQ concern (r=.59, p<.001), restriction (r=.23, p=.002), and monitoring (r=.29, p<.001). WBIS and FTQ scores were also positively associated with each other (r=.29, p<.001).

Table 2.

Correlations between bias variables and parent feeding variables as well as child gender and child weight status.

		Child Gender			Child Weight Category
	Overall r	Sons r	Daughters r	Z	BMI%le <85 r	BMI%le ≥85 r	Z
WBIS
Concern about child weight	0.26***	0.25**	0.27*	−0.13	−0.23*	0.27*	−0.29
Restriction	0.16*	0.25**	−0.01	1.80	0.10	0.23*	−0.89
Monitoring	0.04	0.09	−0.04	0.92	0.05	0.07	−0.13
FTQ
Concern about child weight	0.59***	0.56***	0.63***	−0.66	0.67***	0.50***	1.71
Restriction	0.23**	0.18	0.28*	−0.71	0.26**	0.15	0.76
Monitoring	0.29***	0.29**	0.27*	0.20	0.36***	0.19	1.25

Note. N=242.

^*p<.05;

^**p<.01;

^***p <.001.

Omitted asterisks denote non-significant correlations. None of the z-scores comparing child gender or child weight category was significant.

WBIS=weight bias internalization scale; FTQ=fat talk questionnaire. BMI%le= BMI percentile (age/sex normed)

Overall, FTQ scores were more strongly positively associated with CFQ concern (z= −3.98, p<.001) and monitoring (z= −2.45, p=.01) than were WBIS scores. However, the magnitude of the correlation between FTQ scores and CFQ restriction did not differ significantly from the magnitude of the WBIS correlation with that same subscale (z= −0.72, p=.47).

The magnitude of correlations between WBIS and FTQ scores (separately), and CFQ feeding practices, did not differ significantly between parents of sons and daughters (see Table 2; all ps≥.07), or between parents of children with or without elevated BMI (see Table 2; all ps≥.09).

Discussion

The current study examined whether parental weight bias internalization and negative body talk was associated with parent feeding practices. Higher levels of parent weight bias internalization and negative body talk both were positively associated with parent concern about their child’s weight, as well as restrictive feeding patterns. Of these, negative body talk had a stronger positive association than internalized weight bias with parent feeding practices. Previous research has found that negative body talk directed at children was associated with disordered eating patterns in children (Lydecker et. al., 2018); however, this is the first study, to our knowledge, to examine internalized weight bias and parent feeding practices. The pattern of associations showing that internalized weight bias and negative body talk both were positively associated with concern about child weight and restriction of the child’s diet suggests that parents may be acting on their concerns about their child’s weight by trying to help their child lose weight. However, restrictive feeding patterns have been associated with an increased risk for disordered eating and weight gain (Ventura, 2008; Balantekin et al., 2014; Birch et al., 2001). Taken together, these data highlight the importance of including parents in interventions, rather than solely focusing on the child, when addressing weight or eating behavior. Future research should test this hypothesis.

We also found that correlations were not different across child gender and weight categories. This suggests that prevention work might target all families rather than specific child subgroups. Future research with longitudinal and controlled designs are essential to clarify optimal prevention strategies.

This study had several limitations. Given the cross-sectional study design, we cannot infer causal mechanisms. Future research could explore whether interventions addressing parental weight bias and negative body talk could help to prevent or minimize children’s maladaptive eating behaviors and body image concerns. An additional limitation is that data were collected via self-report assessments and participants were recruited on the Internet, which could increase risks of sampling and/or measurement bias. Alternatively, these methods could also enhance the likelihood that participants responded truthfully to survey items given the associated anonymity. Additionally, parent report of child height and weight is less accurate than measured height and weight; parents tend to under-report both height and weight (Chai et al, 2019) but parent-report and measured height and weight are strongly correlated (Brault, et al., 2015). Additionally, we recognize that the weight categorizations defined by the CDC growth charts are arbitrary and may not accurately reflect health. Whenever feasible, measured adiposity should be used in future studies. Another limitation is that the age range of parents’ children was large. While there is support for the CFQ within our age range (5-15y), CFQ constructs may differ for adolescents relative to younger children, given that parents may not have as much awareness of or authority over their older children’s eating. Further research is needed to examine child age and development as potential moderators. Likewise, parent characteristics, such as gender and weight status, should be examined as potential moderators. Finally, we did not assess children directly. Children’s view of their parents’ feeding practices, as well as children’s own body image concerns and eating concerns, could add meaningfully to our understanding of the association of parental bias and negative body talk with parenting.

Strengths of the current study included the relatively diverse sample, inclusion of both mothers and fathers, and use of well-validated questionnaires assessing constructs of interest (Khandpur et al., 2014). A key next step to build on our findings is longitudinal research that explores whether parents’ feeding practices mediate associations between weight stigma and children’s development of eating disorders.

Taken together, our findings suggest that parents’ internalized weight bias and negative body talk are related to their feeding practices. Future research should examine whether addressing parents’ attitudes and behaviors about weight in family-focused interventions improves restrictive feeding patterns. Future research should also incorporate children’s perspectives on their own shape- and weight-related cognitions and behaviors, as well as how parent characteristics (e.g., gender, BMI) influence relationships among the variables of interest, to better understand associations between parent constructs and childhood health outcomes. Weight bias and weight-related teasing are public health concerns that negatively impact children’s and families’ functioning (Lydecker et al., 2020). Prevention and intervention efforts targeting parents and children is needed to reduce adverse mental health outcomes and improve well-being.

Highlights:

• Parent negative body talk and weight bias were associated with feeding practices

• Negative body talk was more strongly correlated than weight bias with feeding practices

• Results did not differ by child weight status or child gender