Is obesity in the brain? Parent perceptions of brain influences on obesity

Summary

Neuroimaging studies demonstrate associations of brain structure and function with children’s eating behaviour and body weight, and the feasibility of integrating brain measures into obesity risk assessment and intervention is growing. However, little is known about lay perceptions of how the brain influences obesity. We investigated parent perceptions of brain contributions to obesity in three separate studies: 1) a study of mothers of adolescents recruited for neuroimaging research (n = 88), 2) a study of ethnically Chinese parents of 5–13 year olds participating in a parent feeding survey (n = 277), and 3) a study of parents of 3–15 year olds completing an online survey (n = 113). In general, parents believed that brain factors influence obesity, but considered them less influential than behaviours such as diet and exercise. Causal attributions for brain factors were correlated with attributions for genetic factors and biological factors (e.g., metabolism). Parents who perceived their child to be overweight or had a high concern about their child becoming overweight in the future rated brain factors as more important in determining their child’s weight and more likely to lessen their child’s ability to control their weight. Our results suggest that parents attribute obesity to the brain to a moderate degree, and that education or feedback regarding brain influences on obesity could be a promising obesity intervention component.

1 |. INTRODUCTION

Biobehavioural models of obesity susceptibility propose that individuals differ in early-appearing, biologically-influenced appetitive traits that interact with the environment to predispose some individuals to overeating and weight gain.^1,2 Recent neuroimaging studies^3–5 provide specific support for a dynamic neurobehavioral model of obesity in which brain factors may contribute to behavioural obesity risk or maintain the obese state and can be changed through intervention.^6,7 Clinical researchers are already tailoring obesity interventions to individuals’ genetically-influenced appetitive traits,⁸ and soon it may be feasible to integrate information on brain function and structure into obesity risk assessment and intervention, potentially by providing direct individualized feedback on brain factors.⁹ Meanwhile, the dissemination of scientific discovery has proliferated in the internet age, making the findings of neuroimaging studies of obesity widely available to the public.¹⁰ Despite this, we understand little about lay perceptions of the role of the brain in influencing obesity. This research gap requires attention because causal beliefs about obesity are associated with health behaviours.¹¹

Parents’ causal beliefs about obesity are of particular importance for obesity prevention, since parents are key stakeholders in their child’s weight management.¹² Parents’ beliefs about general or personalized contributions of brain factors to children’s obesity could directly influence behaviours that impact weight. For example, parents who believe that structural or functional features of their child’s brain determine their eating behaviour and obesity risk may be more, or less, likely to adopt food parenting strategies¹³ that protect against weight gain.

While beliefs about brain influences on obesity are underexplored, a substantive literature has examined responses to information about the role of genetics in obesity, with mixed findings. Some studies suggest education about genetic influences on weight increases fatalism and discourages positive behavioural changes,^14–16 others suggest genetic obesity risk testing can lessen personal blame for obesity^17,18 and increase motivation to follow a healthy diet,^19,20 and many find no effect.²¹ Fewer studies address parents’ responses to education about genetic influence on child obesity risk, but extant findings are concerning. One study found that parents receiving information about their child’s genetic risk for obesity made poorer feeding decisions for that child likely due to fatalistic beliefs.²² Another found that parents reporting genetic attributions for their child’s eating behaviour exhibited lower self-efficacy for healthy child feeding, more guilt, and chose more calories on a child feeding task.²³ Other work has suggested that parents who perceive their child as overweight (vs. healthy-weight) rate information about the impact of gene-environment interactions on their child’s obesity risk as less relevant to their family.²⁴ These findings argue for the importance of investigating beliefs about the contribution of the brain—a biological influence of greater malleability than genes. They also suggest that parents’ beliefs may vary with weight perceptions and concerns, for example in relation to perceived child weight and concern about future child weight. Such variables reflect actual child weight,²⁵ but additionally capture attitudes and emotions that may be related to beliefs about biological, including neural, influences on obesity.²⁶

We assessed parents’ beliefs about the role of brain factors compared with other factors in influencing obesity in three separate studies, using differing questions and populations. We also assessed how parent characteristics including perceived self and child weight, and concern about their child’s future weight, related to their beliefs about brain influence. We hypothesized that, across studies, parents would consider brain factors to influence obesity. We predicted that ratings for brain factors would be correlated with, but lower than, ratings for genetic factors. We expected that parents who perceived their child as overweight and reported concerns about their child’s risk of future overweight would show differential ratings of brain influence on obesity than parents who perceived their child as healthy-weight or did not report concerns about their child’s future weight.

2 |. METHODS AND RESULTS

2.1 |. STUDY 1

Study 1 assessed parents’ beliefs about the role of the brain and other factors in obesity risk in general, and perceived consequences of brain and other factors for a hypothetical individual’s responsibility for their obesity, in a sample of mothers of adolescents taking part in a neuroimaging study.

2.1.1 |. Methods

Biological mothers and their adolescent children were recruited from Baltimore, Maryland, United States for a neuroimaging study of familial obesity risk (see ^[27,28] for further details). Eligible mothers were fluent English speakers who were not currently pregnant and had no history of excessive smoking, recreational drug use, or alcohol use. N = 98 mother-child dyads were recruited and n = 88 mothers completed all study questions. The study was approved by the Johns Hopkins University School of Medicine Institutional Review Board (JHSOM IRB).

Demographic information including age, sex, race, education, and ethnicity were assessed. Mothers’ height and weight were measured via stadiometer and C-331S Total Body Composition Analyser (TANITA Corp., Tokyo) and used to calculate BMI. Mothers reported perceived current weight for themselves and child, and their level of concern about their child becoming overweight in the future. Survey questions are listed in Supplemental Materials (Appendix S1).

Mothers reported how much they agreed that 12 factors caused obesity in general (1–5 scale; 1 = Strongly disagree; 5 = Strongly agree). Factors, adapted from,²⁹ included one item to assess brain factors (‘A person’s brain’); two items to assess genetic factors (e.g., ‘A person’s genes’), one items to assess biological factors (e.g., ‘A person’s biology’); and two items to assess behavioural factors (e.g., ‘Diet or eating habits’) (See Appendix S1 for all factors). Participants were additionally presented with seven brief hypothetical scenarios in which an individual with obesity (‘Person [X]’) completes a measure showing they are high in a factor that increases the risk of obesity. Following each scenario, participants responded to the question ‘How responsible do you think Person [X] is for their obesity?’ (0–100 VAS scale; 0: ‘Not at all’; 100 = ‘Extremely’). The seven factors were: 1) behaviours; 2) characteristics; 3) traits; 4) genetic variants; 5) genetic variants that increase behaviours; 6) structural brain differences; 7) functional brain differences.

Analyses were completed in IBM SPSS Statistics (Version 27, Armonk, NY). Descriptive statistics were used to summarize demographic characteristics, measures of perceptions and concerns regarding weight, and beliefs about the brain and other influences on obesity. Perceived weight variables (perceived self-weight, perceived child weight) were recoded to form three groups (‘underweight’, ‘just right’, ‘overweight’). Weight concern responses were recoded to form three groups: no concern (‘unconcerned’), some concern (‘a little concerned’, ‘concern’) and high concern (‘fairly concerned’, ‘worried’). Based on a priori categorizations confirmed by principal components analysis, we calculated mean scores for beliefs about influences on obesity for groups of items falling into the domains of ‘behaviour’, ‘biology/metabolism’, ‘genetics’, and ‘brain’. Cronbach’s alpha was calculated to describe the agreement of items within each composite outcome.

To contrast and compare beliefs about the brain with beliefs about other domains, we conducted a series of within-subjects t-tests and paired correlations. To investigate relationships of parents’ perceived self-weight, perceived child weight and concern about future child overweight with beliefs about the brain, we used one-way ANOVA tests. We repeated the analysis of differences by concern about future child overweight controlling for perceived self-weight and perceived child weight. We also tested whether brain beliefs differed by parental education (less than college degree or college graduate or higher) using independent samples t-tests, and tested correlations between parent BMI and brain beliefs using Pearson’s r.

Tukey HSD post-hoc tests were used to probe significant differences between groups when applicable. Uncorrected p-values are presented, and p-values not surviving correction for multiple comparisons using the Benjamini-Hochberg procedure (p-FDR < 0.05) are noted.

2.1.2 |. Results

Sample characteristics, perceived self and child weight, and concern about future child overweight

Sample characteristics and response frequencies for perceived self and child weight, and concerns about future child weight are summarized in Table 1. Reflecting recruitment criteria, Study 1 contained only mothers. Around 84% of mothers had a college degree or higher. The majority described their child’s current weight as ‘just right’, but 55.7% reported at least some concern about their child becoming overweight in the future.

TABLE 1.

Sample characteristics, weight perceptions, and concern about future child overweight/obesity in studies 1–3

	Study 1 (n = 88)	Study 2 (n = 277)	Study 3 (n = 113)
Age (years)	44.14 (6.44)	43.32 (5.52)	39.67 (6.21)
BMI (kg/m 2 )	30.78 (8.45)	22.24 (3.02)	27.37 (8.58)
Sex
Male	0 (0%)	52 (18.8%)	46 (40.7%)
Female	88 (100%)	225 (81.2%)	67 (59.3%)
Race
Asian	3 (3.4%)	276 (99.7%)	17 (15.0%)
Black	20 (34.1%)	0 (0%)	16 (14.2%)
White	47 (53.7%)	0 (0%)	71 (62.8%)
More than 1 race	7 (8.0%)	1 (0.4%)	7 (6.2%)
Other	1 (1.1%)	0 (0%)	2 (1.8%)
Ethnicity
Hispanic	3 (3.4%)	0 (0%)	6 (5.3%)
Non-Hispanic	81 (92.0%)	277 (100%)	107 (94.7%)
Missing	4 (4.5%)	0 (0%)	0 (0%)
Education
Less than College Grad	14 (15.9%)	11 (4.0%)	12 (10.6%)
College Grad or Higher	74 (84.1%)	266 (96.0%)	101 (89.4%)
Child sex
Male	48 (54.5%)	141 (50.9%)	53 (46.9%)
Female	40 (45.5%)	136 (49.1%)	60 (53.1%)
Child age (years)	15.63 (1.25)	8.74 (2.19)	8.17 (3.18)
Perceived self-weight
Underweight	2 (2.3%)	19 (6.9%)	2 (1.8%)
Just right	22 (25.0%)	178 (64.3%)	62 (54.9%)
Overweight	64 (72.7%)	79 (28.6%)	49 (43.4%)
Perceived child weight
Underweight	9 (10.2%)	70 (25.3%)	16 (14.2%)
Just right	60 (68.2%)	176 (63.5%)	87 (77.0%)
Overweight	19 (21.6%)	31 (11.2%)	10 (8.8%)
Concern about future child overweight/obesity
None	39 (44.3%)	86 (31.0%)	37 (32.7%)
Some	35 (39.8%)	152 (54.9%)	50 (44.2%)
High	14 (15.9%)	36 (13.1%)	26 (23.0%)
Missing	0 (0%)	3 (1.1%)	0 (0%)

Note: Sample characteristics, perceived self and child weight, and concern about future child overweight (Studies 1 & 2)/obesity (Study 3) are presented in Table 1. Means and standard deviations [M (SD)] are given for continuous variables, and count and percent for categorical variables. Missing data is noted when applicable.

Beliefs about the brain and other influences on obesity

Figure 1 and Table 2 present descriptive statistics for responses to questions on beliefs about the brain and other influences on obesity across the three studies.

FIGURE 1.

shows the mean ratings of obesity beliefs across three samples: (A) Causal beliefs in mothers from study 1; (B) Responsibility beliefs in mothers from study 1; (C) Causal beliefs from parents from study 2; (E) Causal beliefs in parents from study 3; (F) Control beliefs in parents from study 3. Bars represent single item or composite scores across different items collected in each study. The number of items in each category is denoted in parenthesis on the x-axis bar labels. Significant differences between ratings of brain perceptions vs. biology, genetics, or behaviour are denoted by stars

TABLE 2.

Descriptive statistics for beliefs about brain and other influences on obesity

Study 1 ‘What do you think causes obesity?’
	Cronbach’s Alpha	Mean	SD	Correlation with Brain	Mean Difference from Brain
Brain	—	3.48	1.04	—	—
Biology	—	3.65	1.02	r = 0.74, p < 0.001	t = −2.2, p = 0.31
Genetics	0.88	3.81	0.98	r = 0.53, p < 0.001	t = −3.2, p = 0.002
Behaviour	0.67	4.72	0.56	r = 0.28, p = 0.008	t = −11.3, p < 0.001
Study 1 ‘Person [X] is high in [factor]. How responsible is person X for their obesity’?
	Cronbach’s Alpha	Mean	SD	Correlation with Brain	Mean Difference from Brain
Brain	0.96	42.8	22.7	—	—
Genetics	0.84	46.4	22.9	r = 0.64, p < 0.001	t = −1.9, p = 0.057
Behaviour	—	78.0	18.6	r = 0.04, p = 0.69	t = −11.5, p < 0.001
Study 2 ‘If a child is overweight, what do you think are the causes?’
	Cronbach’s Alpha	Mean	SD	Correlation with Brain	Mean Difference from Brain
Brain	—	3.57	0.87	—	—
Biology	—	4.02	0.70	r = 0.59, p < 0.001	t = −10.1, p < 0.001
Genetics	—	4.07	0.77	r = 0.34, p < 0.001	t = −8.7, p < 0.001
Behaviour	0.84	4.18	0.48	r = 0.15, p = 0.01	t = −10.8, p < 0.001
Study 3 ‘How important is this in determining your child’s future weight?’
	Cronbach’s Alpha	Mean	SD	Correlation with Brain	Mean Difference from Brain
Brain	0.96	3.10	0.91	—	—
Biology	—	3.63	0.91	r = 0.54, p < 0.001	t = −6.5, p < 0.001
Genetics	0.93	3.42	0.89	r = 0.77, p < 0.001	t = −5.7, p < 0.001
Behaviour	0.86	3.42	0.89	r = 0.68, p < 0.001	t = −4.7, p < 0.001
Study 3 ‘To what extent does this mean your child will have less control over his/her future weight?’
	Cronbach’s Alpha	Mean	SD	Correlation with Brain	Mean Difference from Brain
Brain	0.95	2.9	0.9	—	—
Genetics	0.95	3.1	0.9	r = 0.72, p < 0.001	t = −3.3, p = 0.001
Behaviour	0.87	3.0	0.9	r = 0.45, p < 0.001	t = −2.0, p = 0.044

Note: Descriptive statistics for beliefs about the brain and other influences on obesity are presented in Table 2. Internal consistency (Cronbach’s Alpha) is shown for all composite scores. Means and standard deviations (SD) are presented for each factor. Pearson correlations were used to test associations of ratings for brain factors with ratings for other factors. Paired samples t-tests were used to test for differences in mean scores between brain factors and other factors.

On average, mothers slightly agreed that ‘a person’s brain’ caused obesity, with 39.8% and 14.8% reporting that they agreed or strongly agreed. Mothers reported the highest causal attributions for obesogenic behaviours (‘diet or eating habits’, ‘lack of exercise’). Causal attributions for brain factors were correlated with biological (r = 0.74, p < 0.001), genetic (r = 0.53, p < 0.001), and behavioural (r = 0.28, p = 0.008) attributions, and were significantly lower than behavioural (t = −11.3, p < 0.001) and genetic (t = −3.2, p = 0.002) attributions.

Given the hypothetical clinical scenarios describing factors that increase the risk of obesity in ‘Person [X]’, mothers indicated that someone with structural and functional brain differences that increase obesity risk did not have full responsibility for their weight, averaging M = 42.8 SD = 22.7 on the 0–100 scale. Responsibility ratings for brain factors were correlated with ratings for genetics (r = 0.64, p < 0.001), and significantly lower than ratings for behavioural factors (t = −11.5, p < 0.001).

Individual differences in beliefs about brain influences

Associations of perceived self and child weight and concern about future child overweight with beliefs about brain influences on obesity are presented in Table 3, with plots of significant individual differences in Appendix S1.

TABLE 3.

Relationships of parents’ perceived self and child weight and concern about future child overweight/obesity with beliefs about brain influences on obesity

Perceived self-weight
Study 1: ‘What do you think causes obesity?’ [1–5]
Underweight (N = 2)	Just Right (N = 22)	Overweight (N = 64)	ANOVA
2.50 (2.12)	3.41 (0.91)	3.53 (1.05)	F(2,85) = 1.1, p = 0.34
Study 1: ‘Person [X] is high in [factor]. How responsible is person X for their obesity’? [1–100]
Underweight (N = 2)	Just Right (N = 22)	Overweight (N = 64)	ANOVA
54.0 (5.7)	48.4 (23.6)	40.4 (22.4)	F(2,84) = 1.3, p = 0.29
Study 2: ‘If a child is overweight, what do you think are the causes?’ [1–5]
Underweight (N = 19)	Just Right (N = 178)	Overweight (N = 79)	ANOVA
3.37 (0.76)	3.54 (0.91)	3.68 (0.79)	F(2, 271) = 1.2, p = 0.29
Study 3: ‘How important is this in determining your child’s future weight?’ [1–5]
Underweight (N = 2)	Just Right (N = 62)	Overweight (N = 49)	ANOVA
3.47 (1.17)	2.95 (0.89)	3.27 (0.92)	F(2,110) = 1.9, p = 0.16
Study 3: ‘To what extent does this mean your child will have less control over his/her future weight?’ [1–5]
Underweight (N = 2)	Just Right (N = 62)	Overweight (N = 49)	ANOVA
2.42 (0.0)	2.76 (0.86)	3.08 (0.82)	F(2,110) = 2.4, p = 0.09
Perceived Child Weight
Study 1 ‘What do you think causes obesity?’
Underweight (N = 9)	Just Right (N = 60)	Overweight (N = 19)	ANOVA
2.89 (1.05)	3.52 (1.07)	3.48 (1.04)	F(2,85) = 1.7, p = 0.19
Study 1: ‘Person [X] is high in [factor]. How responsible is person X for their obesity’?
Underweight (N = 9)	Just Right (N = 60)	Overweight (N = 19)	ANOVA
42.4 (20.5)	43.3 (23.2)	41.3 (23.1)	F(2,84) = 0.05, p = 0.95
Study 2: ‘If a child is overweight, what do you think are the causes?’
Underweight (N = 70)	Just Right (N = 176)	Overweight (N = 31)	ANOVA
3.60 (0.69)	2.48 (0.93)	4.06 (0.77)	F(2, 271) = 6.3, p = 0.002 O-JR: p = 0.001; O-U: p = 0.03
Study 3: ‘How important is this in determining your child’s future weight?’
Underweight (N = 16)	Just Right (N = 87)	Overweight (N = 10)	ANOVA
2.83 (1.16)	3.06 (0.84)	3.87 (0.82)	F(2,110) = 4.6, p = 0.012 O-JR: p = 0.02; O-U: p = 0.01
Study 3: ‘To what extent does this mean your child will have less control over his/her future weight?’
Underweight (N = 16)	Just Right (N = 87)	Overweight (N = 10)	ANOVA
2.58 (0.98)	2.86 (0.79)	3.71 (0.85)	F(2,110) = 6.3, p = 0.003 O-JR: p = 0.006; O-U: p = 0.002
Concern about future child overweight/obesity
Study 1: ‘What do you think causes obesity?’
None (N = 39)	Some (N = 35)	High (N = 14)	ANOVA
3.31 (1.15)	3.57 (1.01)	3.71 (0.73)	F(2,85) = 1.0, p = 0.36
Study 1: ‘Person [X] is high in [factor]. How responsible is person X for their obesity’?
None (N = 39)	Some (N = 35)	High (N = 14)	ANOVA
40.8 (22.5)	44.9 (20.9)	42.6 (28.0)	F(2,84) = 0.3, p = 0.75
Study 2: ‘If a child is overweight, what do you think are the causes?’
None (N = 86)	Some (N = 152)	High (N = 36)	ANOVA
3.42 (0.83)	3.63 (0.88)	3.69 (0.92)	F(2, 271) = 2.0, p = 0.14
Study 3: ‘How important is this in determining your child’s future weight?’
None (N = 37)	Some (N = 50)	High (N = 26)	ANOVA
3.05 (0.93)	2.84 (0.84)	3.66 (0.80)	F(2,110) = 6.4, p = 0.002 H-S: p = 0.02; H-N: p < 0.000
Study 3: ‘To what extent does this mean your child will have less control over his/her future weight?’
None (N = 37)	Some (N = 50)	High (N = 26)	ANOVA
2.82 (0.88)	2.71 (0.82)	3.35 (0.72)	F(2,110) = 5.4, p = 0.006 H-S: p = 0.005; H-N: p = 0.04

Note: Relationships between parents’ perceived self-weight, perceived child weight, and concern about future child overweight/obesity with beliefs about brain influences on obesity are presented in Table 3. One-way ANOVA was used to test for group differences in brain beliefs (significant effects in bold) and Tukey HSD post-hoc tests were to test pairwise group differences (O = overweight; JR = Just Right; U=Underweight; N=None; S=Some; H=High). Mean and standard deviation as presented as M (SD).

Mothers in the higher education group (college degree or higher) rated brain factors as more causal than the lower education group (less than a college degree) (t = −5.82, p < 0.001). Ratings for causal attributions and responsibility did not differ by parents’ perceived self-weight, perceived child weight, or concern about their child’s future weight (Table 3) and controlling for parents’ perceived self-weight or child weight did not change the null association between parents’ concern about future child overweight and their causal or responsibility ratings for brain factors (Appendix S1).

2.2 |. STUDY 2

Study 2 assessed parents’ causal beliefs about the brain in relation to children’s obesity risk, among Chinese-American mothers and fathers of younger children.

2.2.1 |. Methods

Parents were recruited from 28 Chinese language schools located in New York, New Jersey, California, Georgia, Michigan, Texas and Washington, United States for a survey study of relationships between acculturation and feeding behaviours in Chinese-American parents. Ethnically Chinese parents of 5–12 year old children were eligible. Surveys were distributed as physical copies or online via RedCap. Parents completed the survey in either Chinese or English. All questions were translated into Chinese by a bilingual researcher and comprehension was tested in 2 small pilot samples (n = 3 and n = 5). N = 297 parents were recruited, and n = 277 completed the survey. The study was approved by the JHSOM IRB.

In addition to demographic information, perceived self-weight, perceived child weight and concern about future child overweight were assessed, as for Study 1. Parents reported their height and weight, which was used to calculate BMI. Child height and weight were not assessed. Survey questions are listed in Appendix S1.

Parents reported how much they agreed that 17 factors (derived from food parenting literature and literature on parenting in ethnically Chinese populations and piloted in a small sample of Chinese American parents [n = 8]) caused a hypothetical child’s overweight status (1–5 scale; 1 = Strongly disagree; 5 = Strongly agree). Factors included one item to assess the brain (‘Differences in how the brain responds to food’); one item to assess genetics (‘Genetics/runs in the family’); one item to assess biology (e.g., ‘Biology e.g., metabolism’); and nine items to assess behaviour (e.g., ‘Diet quality/availability of healthy foods’) (All items can be found in Appendix S1).

Data analysis procedures followed those for Study 1.

2.2.2 |. Results

Sample characteristics and response frequencies for perceived self and child weight, and concern about future child overweight are summarized in Table 1.

Beliefs about brain and other influences on obesity

44% of parents agreed and 12.3% strongly agreed that ‘differences in how the brain responds to food’ caused a child to be overweight (see Table 2 and Figure 1). Behavioural factors including parenting-related causes were rated as having the most influence on a child’s overweight. For example, 50.5% of parents agreed and 44.8% strongly agreed that ‘family lifestyle’ caused a child to be overweight. Parents’ causal attributions for the brain were positively correlated with attributions for biology (r = 0.60, p < 0.001), genetics (r = 0.34, p < 0.001), and behaviour (r = 0.15, p = 0.01). Parents rated the brain as less causal than behaviour (t = −10.8, p < 0.001), genetics (t = −8.7, p < 0.001), and biology (t = −10.1, p < 0.001).

Individual differences in beliefs about brain influences

Parents’ education level, perceived self-weight, and concern about future child overweight were not related to their causal attributions for the brain (Table 3 and Appendix S1). However, parents who perceived their child as overweight rated ‘how a person’s brain responds to food’ as more causal than parents who perceived their child’s weight as underweight or just right (F[2, 274] = 6.25, p = 0.002); see Table 3 and Figure 2. Controlling for parents’ perceived self-weight or child weight did not change the null association between parents’ concern about future child overweight and their causal beliefs about the brain (Appendix S1).

FIGURE 2.

shows a selection of between group differences in obesity beliefs identified in three samples: (A) In study 2, parents’ perceived child weight was associated with their beliefs about if the brain causes overweight; (B) In study 3, parents’ concern about future child obesity was associated with their beliefs about if the brain is important in determining their child’s future weight. Significant differences between group as identified by post-hoc tests are denoted by stars

2.3 |. STUDY 3

Study 3 assessed parents’ beliefs about the role of the brain in obesity risk for their own young child using a series of personalized hypothetical clinical vignettes that described specific mechanisms of brain influence on weight. This study also assessed parents’ beliefs about how specified brain factors affected their child’s control over their future weight.

2.3.1 |. Methods

Parents from the Washington DC metro area were recruited directly from databases of individuals screened for participation in prior studies of parent feeding who had agreed to be re-contacted for future research. Parents of children 3–13 years old were eligible. To attain a 50:50 sex ratio, the sample was monitored throughout data collection and was modified to screen by sex. Parents completed a survey via SurveyMonkey. N = 118 parents participated and n = 113 contributed complete data. The study was approved by the National Human Genome Research Institute IRB.

Parents reported demographic information, perceived self-weight, perceived child weight and their level of concern about their child developing obesity in the future. Parents reported their height and weight, which was used to calculate BMI. Child height and weight was not assessed. To assess obesity beliefs, parents were presented with 13 hypothetical scenarios in which a doctor completes a test on their child and finds that they are high on a factor associated with obesity risk. For each scenario, parents reported how important they believed that result was in determining their child’s future weight, and to what extent they thought the result meant that their child would have less control over their future weight (1–7 scale; 1 = ‘not at all’; 7 = ‘completely’). Factors included six items to assess the brain (e.g., ‘child’s brain shows size linked to slower metabolism’, ‘child’s brain shows increased activity in areas linked to reward and pleasure while tasting a delicious, high-calorie milkshake’); four items to assess genetics (e.g., ‘genetic variations that increase your child’s likelihood of weight gain’); one item to assess biology (‘metabolism’); and two items to assess behaviour (‘food responsiveness’, ‘satiety responsiveness’). For all items, see Appendix S1.

Data analysis followed that for Studies 1 and 2. To allow comparison across the three samples, importance and control ratings (1–7 scale) were transformed to match the scale of ratings in Studies 1 and 2 (1–5 scale). Concern about future child obesity was recoded into three groups representing no concern (rating = 1), some concern (rating = 2–4) and high concern (rating = 5–7).

2.3.2 |. Results

Sample characteristics and response frequencies for perceived self and child weight and level of concern about future child obesity are summarized in Table 1.

Beliefs about brain and other influences on obesity

Parents rated the importance of factors relating to brain function (M = 3.08 SD = 1.04) similarly to those relating to brain structure (M = 3.11 SD = 0.90); see Table 2 and Figure 1. Parents rated ‘metabolism’ (biology) as the most important determinant of their child’s future weight (M = 3.63 SD = 0.91). Parents’ ratings of the importance of brain factors in their child’s future weight were positively correlated with ratings for genetics (r = 0.77, p < 0.001), behaviour (r = 0.68, p < 0.001), and biology (r = 0.54, p < 0.001). Parents rated brain factors as significantly less important than biology (t = −6.5, p < 0.001), genetics (t = −5.7, p < 0.001), and behaviour (t = −4.7, p < 0.001) in determining their child’s future weight.

Parents reported that of all the hypothetical obesity risk scenarios, ‘genetic variations that increased obesity’ would most strongly indicate that their child would have less control of their future weight. Parents’ control beliefs about brain factors were correlated with those for genetics (r = 0.72, p < 0.001) and behaviour (r = 0.45, p < 0.001). Parents believed that their child would have more control over their future weight given brain risk factors compared to genetic (t = −3.3, p = 0.001) and behavioural (t = −2.0, p = 0.044) risk factors.

Individual differences in beliefs about brain influences

Parents who perceived their child to have overweight (vs. just right or underweight) or expressed high (vs. some or low) concern about future child obesity gave higher ratings of the importance of brain factors in determining their child’s future weight and more strongly believed that brain factors indicated that their child would have less control over their future weight (Table 3 and Figure 2). There was no impact of parental education on importance or control ratings (Appendix S1). The effect of concern about future child obesity on importance and control ratings remained significant when controlling for parents’ perceived self-weight and perceived child weight (Appendix S1).

3 |. DISCUSSION

Causal beliefs about obesity can shape health behaviours and impact weight.¹¹ While evidence shows that people believe that obesity is influenced by behavioural, biological and environmental factors,^30–32 no previous study has specifically examined beliefs about the brain’s role in obesity. As the possibility of integrating education or individualized feedback on brain metrics into a clinical context grows more likely, and the results of neuroimaging studies continue to penetrate public discourse, it is important to understand more about how people perceive the role of the brain in influencing body weight. This is especially true for parents since parental beliefs and behaviours shape children’s eating habits.³³ Using data collected from three independent samples of parents, we demonstrate that parents moderately attribute obesity to brain factors, that brain attributions are correlated with biological and genetic attributions but perceived as less influential than proximal behaviours such as diet, and that parents’ brain attributions are related to perceptions and concerns about their child’s weight.

Across our studies, we found that parents’ beliefs about the brain’s influence on body weight were moderate in strength. In Study 1 and Study 2, over half of parents endorsed that the brain contributes to a person’s or a child’s weight. Parents in Study 3 rated brain factors as moderately important (M = 3.1; 1–5 adjusted scale) in determining their child’s future weight. The similarity of these results from three diverse samples and in relation to ‘people’, children, and ‘your child’, suggests that brain attributions for body weight are likely common across parents and moderate in strength.

We also found that ratings of causal attributions for the brain were generally positively correlated with ratings for the influence of other factors on weight across the three samples. The correlations between causal attributions for the brain and other factors are broadly consistent with other work showing that people hold multifactorial beliefs about the determinants of common health conditions (18). While differences in correlation strength should not be overinterpreted due to divergence of items between and within measures, the relative weakness of correlations of brain attributions with attributions for behaviour when operationalized narrowly (e.g., Study 1) or broadly (e.g., Study 2) was interesting. Correlations with behavioural factors were higher for Study 3, but that study assessed appetitive characteristics rather than behaviours related to energy balance. In contrast, correlations of brain attributions with genetic attributions were stronger and correlations were highest for biological factors, with the highest values in Studies 1 and 2.

We also tested whether causal attribution ratings differed between brain factors and other factors within subjects. We found that attributions for the brain were consistently lower than for behavioural factors and genetic factors, and somewhat lower than biological factors. Our findings are consistent with previous reports that obesity is most often attributed to behaviour³¹ and that behavioural attributions are more highly endorsed than genetic.^29,34,35 Combined, our results suggest that beliefs about brain influence are distinct from beliefs about other domains of influence, and that brain factors are considered less influential than proximal behavioural influences on obesity, while also being viewed somewhat similarly to other biological and genetic factors.

We additionally assessed beliefs about how the brain and other factors contribute to a person’s responsibility for their obesity (Study 1) or their child’s ability to control their future weight (Study 3). In Study 3, hypothetical feedback on genetic obesity risk was associated with least perceived control over future weight and feedback on brain risk with the most control but means were extremely similar. In contrast, in Study 1 brain and genetic domains were clearly separated from behaviours, with parents reporting that a person with behavioural risk for obesity was the most responsible for their weight. In both samples, responsibility ratings for brain factors were similar to ratings for genetic factors. These results suggest that brain factors are perceived as outside an individual’s control, which may absolve individuals from responsibility, guilt, and self-blame for elevated weight.

Two of our studies provided evidence that beliefs about brain influence on weight varied with parents’ perceptions relating to their child’s weight. Parents who perceived their child to have overweight were more likely to attribute obesity to the brain than parents who considered their child’s weight to be ‘just right’ or underweight (Studies 2 and 3), and parents with concern about their child’s future obesity risk more strongly believed that hypothetical brain risk factors would indicate that their child would have less control over their weight in the future (Study 3). Parents of children with overweight or other characteristics triggering concerns about future overweight (e.g., obesogenic appetitive traits) could be more informed about obesity causes and therefore more likely to acknowledge brain contributions. Such parents may also have witnessed neurobehavioral processes increasing obesity risk in their children or in themselves, especially if they are themselves higher weight. This seems particularly likely for Study 3, where we assessed functional brain variables (e.g., child’s brain shows increased activity in areas linked to reward and pleasure while tasting a delicious milkshake). Another possibility is that parents of children with overweight may be more motivated to attribute their child’s overweight to biological factors that are perceived as out of their control to avoid feelings of guilt. However, this does not directly explain findings in relation to concern about child weight, which persisted despite controlling for parents’ perceptions of their own and their child’s weight. It is also possible that parents who think the brain causes obesity believe that brain structure/function is inflexible, and this leads them to express greater concern for their child’s future weight. An analogous effect is found for genetic determinants of child obesity such that parents feel that genetic risk for obesity decreases the efficacy of obesity prevention.²²

While others have assessed brain factors in combination with biological/metabolic factors as collective ‘biological’ determinants of obesity (9), this is the first set of studies to focus on brain determinants. As such, our research has several limitations that should be acknowledged, and that point towards several unanswered questions meriting future investigation. First, since our goal was to test whether we observed similar patterns of results across different but complementary study designs, each sample was asked different questions to assess obesity causal beliefs, precluding direct comparison of results. Notably, while the questions all focused on similar constructs, measures included varying numbers of items to assess domains (e.g. Study 1 used 1 item to assess causal beliefs about the brain whereas Study 3 included 6 items). To facilitate qualitative comparison across studies, we combined all items relating to the brain into composite scores. However, the mean scores for items composing this score were variable (see Appendix S1) suggesting the presence of variability in parents’ perceptions of causal attributions for different facets of brain, behavioural, genetic, and biological factors that should be explored in future work. Despite this limitation, however, our results were generally consistent across the three studies supporting the reproducibility of effects. Second, since a parallel goal of the current study was to investigate generalizability of our findings across different populations, our three study groups differed in their demographic characteristics. For example, for Study 1 we recruited mothers of adolescents, two-thirds of whom had an overweight or obese BMI. As such, this sample was exclusively female and tended to be older and heavier than that for Studies 2 and 3. Also, because Study 2 recruited ethnically Chinese parents, the racial and ethnic composition of Study 2 is homogenous, while Studies 1 and 3 are comprised of primarily White, non-Hispanic parents. It is possible that cultural differences could contribute to differences in obesity beliefs and perceptions of child weight between the samples. However, we cannot determine from the current study if such differences are attributable to sample characteristics or measurement differences.

Since we provided only very brief explanations of our obesity risk factors of interest, there may be significant scope to boost people’s understanding of various brain risk factors and their contribution to body weight. Future work with an educational component could further explore questions such as whether brain structure is considered less malleable than brain function, with differential effects on perceptions of responsibility and control. Notably, we found within Study 1 that brain attributions were significantly lower among mothers with less than a college degree. Obesity disproportionately affects individuals with lower levels of education in high income countries,³⁶ and low-income parents are more likely to believe that heavy weight is a marker of child health.³⁷ Since our samples were highly educated, perceptions of the role of the brain in influencing obesity should be further explored in populations with lower socioeconomic status. One study found that individuals who identified as Black or Hispanic were more likely to endorse ‘low agency’ factors as causing obesity (e.g., ‘genetics, or a family history’) than individuals who identified as White,³⁸ suggesting that if our studies were replicated in samples higher in diversity and more representative of the broader U.S. population, brain attributions for obesity might differ.

Our findings suggesting that parents consider brain factors as unique biological influences in their causal models of obesity have several implications. First, they provide a foundation for exploring the value of providing education or personalized risk information relating to the brain in obesity prevention or treatment. The close relationships between biological factors, including the brain, suggest that combining education or feedback on brain influences with that for other biological influences may fit with individuals’ causal schemas and could have synergistic effects. More generally, science communication about the role of the brain in obesity may be able to tap into existing models³⁹ and play a role in shifting the public perception of obesity from a personal failing to a chronic disease that is caused by genetic and biological factors, in addition to behaviour,⁴⁰ thereby reducing weight stigma.

Potential negative implications of education about the role of the brain in obesity, such as defensiveness and fatalism,^24,41 should also be acknowledged. While certain features of brain structure and function can increase risk of excess weight gain, they do not inevitability lead to obesity. Brain structure and function has a reciprocal relationship with behaviour, made evident by findings that dietary interventions can alter brain structure⁴² and neurofeedback training can improve self-regulation around food.⁴³ Our findings lay the groundwork for potential future experiments and longitudinal studies to test whether beliefs about brain influences on obesity erode motivation to achieve a healthy weight or promote obesity-protective behaviours.