Abstract
Introduction: Accumulating evidence suggests that preconception epigenetic changes elevate the risk for obesity throughout the lifespan. Little is known about how parents may react to learning about parent-child epigenetic transmission of obesity risk. Further, it is unclear how trust in science may moderate these responses. Methods: We compared risk perceptions, behavioral intentions, perceived control, and information-focused ratings of 322 parents with high weight status who were randomized to read an article about the role of preconception epigenetics in intergenerational obesity risk transmission, versus three comparators that focused on genetics, family environment, or an unrelated topic. Results: Parents had largely similar reactions to the epigenetics, genetics, and family environment articles in terms of perceived credibility, relevance, and threat response, but the epigenetics article failed to produce the elevated cognitive (F(3, 310) = 3.027, p = 0.030) and affective/intuitive (F(3, 310) = 3.05, p = 0.029) risk perceptions observed in response to the genetics and family environment articles compared to control. Science trust moderated individual reactions to the epigenetics concepts, such that those with low science trust exhibited lower attentiveness to the epigenetics article (F(4, 249) = 2.92, p = 0.022), and groups with low, medium, or high science trust exhibited distinct reaction profiles in terms of affective/intuitive risk perception (F(6, 310) = 2.40, p = 0.028). Conclusion: An audience’s trust in science should be considered when tailoring messages about the role of epigenetics in conveying obesity risk from parent to child.
Introduction
On April 28, 2021, The New York Times published an article titled “Parents’ Diet and Exercise Habits, Even Before Birth, May Contribute to Child’s Well-Being” [1]. The article described findings from a study of intergenerational murine epigenetics, stating that “rodent parents-to-be that fatten on a greasy diet before mating produce offspring with sky-high later risks for metabolic problems” [1]. The article concluded that physical activity should be encouraged for expecting parents, especially mothers. Reader comments further suggested that the article had immediate health implications for parents and prospective parents. The present study examines cognitive and affective reactions of parents to such media messages about the role of epigenetics in intergenerational obesity risk transmission compared to other causal attributions using standardized newspaper-style articles in a randomized experiment.
Parental weight status, including prior to or during pregnancy, has a well-documented association with high weight status for offspring in childhood and adulthood [2‒4]. Evidence suggests that epigenetic changes play a role in conveying this risk [5‒9]. Intergenerational transmission of epigenetic markers can confer elevated risk for obesity among individuals with at least one parent at high weight status at the time of conception [10‒14].
Epigenetics, as a concept, is beginning to enter the sphere of public knowledge. Media organizations play a pivotal role in communicating the role of epigenetics in health and disease development to the general public [15], and epigenetics communication being “inaccurate, exaggerated, premature, or misleading” has been raised as a major concern among researchers [16]. Little is known about how parents react to learning about the role of preconception epigenetic changes they pass down in their children’s obesity risk. Given the potential for cognitive and affective reactions to child risk messages to influence parental health behavior [17], it is crucial to understand whether and how parents respond to epigenetics information.
Effective communication about epigenetics may result in opportunities to motivate behavior change during this window [18]. Early research suggests that increased knowledge about developmental origins of health and disease concepts is associated with better diet quality among childless young adults [19]. However, exposure to out-of-context or incomplete communication about epigenetics could pose risks for parents and those considering parenthood. Prospective parents and pregnant people are groups particularly vulnerable to misleading health advice [20].
Preconception epigenetics, similar to genetics, may be perceived as relatively unchangeable, especially for parents considering epigenetic changes that were transmitted to a child at conception. Beliefs that this risk is unalterable may result in feelings of helplessness, lack of perceived control, or fatalism about a child’s obesity risk. If parents adopt such beliefs, it may result in suppression of positive health behavior, discounting of risk information, and elicitation of self-blame and guilt [21], particularly for parents who had high weight status when their child was conceived [22]. Evidence suggests that some parents with high weight status do feel guilty for “passing down” genetically oriented risk to their child [23]. Another risk of epigenetic information exposure is increasing societal-level blame of individuals portrayed as responsible for passing along obesity risk. For example, individuals with high weight status due to factors associated with socioeconomic status may experience heightened blame for their children’s weight while drawing attention away from causative structural factors [24‒26]. This is particularly likely when epigenetics messages are presented without consideration for the larger societal forces underlying weight status [27, 28]. On the other hand, learning about epigenetics may help distribute the responsibility for child risk and well-being more equally between biological parents, rather than putting undue blame upon mothers for children’s obesity risk [10, 29]. However, although epigenetic changes from both biological parents appear to contribute to risk, negligent communication approaches may imply that biological mothers or pregnant parents alone are responsible for transmitting epigenetic risk [25, 30].
Receptivity to epigenetics concepts may be particularly affected by one’s trust in science. The Risk Information Seeking and Processing Model posits that individuals’ perceptions of information sources has important influence on how risk messages are mentally processed [31]. In this case, trust in the science that gives rise to epigenetics-oriented risk messages is likely to be relevant given the topic’s novel nature. Indeed, epigenetics concepts are less familiar to the public than genetics concepts [32, 33], can be technical, may seem to contradict scientific understanding of genetics as unmodifiable, and are often described in shorthand ways that can cause confusion [32, 34, 35]. These factors can result in increased perceived uncertainty and potentially resistance to epigenetics concepts depending upon beliefs about the scientific enterprise that produced the information [36]. Science trust, therefore, is likely to moderate receptivity to epigenetics messages as information about epigenetics begins to become increasingly prevalent in the public sphere.
Existing literature about parental engagement with and responses to epigenetics concepts is sparse. In a recent study of parent knowledge on epigenetics concepts, 57.3% of parents from a highly educated sample indicated they had heard of epigenetics previously. Of these individuals, 14.1% described the intergenerational transmission of epigenetic effects as part of their broader epigenetics definition [32]. Bombak et al. [25] reported that lower-income women with high weight status largely did not endorse epigenetics as a causal factor in intergenerational obesity risk transmission concepts. One study indicated that exposure to educational information about epigenetics caused parents to put higher importance on both mothers’ and fathers’ preconception lifestyle for child health and elicited feelings of guilt and negativity, particularly among parents with high weight status [22].
This study explores parental reactions to messages about preconception epigenetic transmission of obesity risk as compared to more familiar causes, including genetics and the family environment. We further examine the moderating effect of science trust on these reactions. In this study, parents with self-reported overweight read newspaper-style articles explaining that the children of parents with overweight are at elevated risk for high weight status in the future matched by parent gender (Table 1). Conditions varied by the mechanism through which the risk was said to be transmitted: epigenetics, genetics, or family environment. We selected genetic and family environment-based risk factors as comparators against which to evaluate epigenetic concepts because they also play a role in generational transmission of risk for obesity, are more familiar to the public, and have differing similarities to epigenetics concepts as highlighted in hypotheses:
Excerpts from articles by experimental condition
| Section of article | Epigenetics | Genetics | Family environment |
|---|---|---|---|
| Explanation | Epigenetics is defined as changes in the way genes (DNA) are expressed. Genes are instruction manuals for the body, and are the directions for building all the proteins that make the body function. They are inherited from biological parents. While genes stay the same throughout a person’s lifetime, the way genes are read and expressed in the body changes based on one’s behavior, environmental exposures, and other factors. Just like a switch turns a light on and off, epigenetics is the process that turns genes on and off. These changes in how genes are read and expressed underlie an individual’s risk for obesity. Epigenetic changes in a [father or mother] can be passed down to [his or her] child, influencing the child’s change of developing obesity in his or her lifetime. For example, a [father or mother] who eats an unhealthy diet can pass down epigenetic changes that influence [his or her] child’s weight | Genes are instruction manuals for the body, and are the directions for building all the proteins that make the body function. They are inherited from biological parents. Differences in genes underlie an individual’s risk for obesity. Weight-related genes in a [father OR mother] can thus be passed down to [his OR her] child, influencing the child’s chances of developing obesity in his or her lifetime. For example, genes that influence a [father’s OR mother’s] metabolism can be passed down and influence [his OR her] child’s weight | The family environment is defined as habits, behaviors, and beliefs that parents set. [Fathers OR Mothers] are especially important in establishing the family environment. This environment sets in motion behaviors that children continue throughout their lives and affects their risk of becoming obese |
| Quote | “I was shocked to learn that my eating behaviors even before I had Taylor influence Taylor’s weight in this way,” said Mr. Bryson | “I was shocked to learn that my genes influence Taylor’s weight in this way,” said Mr. Bryson | “I was shocked to learn that my behaviors influence Taylor’s weight in this way,” said Mr. Bryson |
| Studies | These studies offer clues to a big question in the field of obesity research: Why can’t other known factors like the environment or genetics explain why children with heavy parents grow up to be overweight? This recent research shows that the answer is epigenetics | These studies offer clues into a big question in the field of obesity: Why do people gain different amounts of weight while eating the same number of calories? This recent research shows that the answer is genetics | These studies offer clues into a big question in the field of obesity: How do unhealthy habits that parents establish persist once parents are no longer around? This recent research shows that the answer is the early family environment |
| Conclusions | These studies show how the consequences of parents behaviors are passed down to children, affecting children’s gene expression, which explains a child’s risk for developing obesity | These studies show how genes passed down from [fathers OR mothers] can form the basis of a child’s risk for obesity | The studies show how weight-related behaviors like food preference and consumption of fatty foods are taught by parents (rather than being due to innate preferences and cravings), which explains a child’s risk for obesity |
| Section of article | Epigenetics | Genetics | Family environment |
|---|---|---|---|
| Explanation | Epigenetics is defined as changes in the way genes (DNA) are expressed. Genes are instruction manuals for the body, and are the directions for building all the proteins that make the body function. They are inherited from biological parents. While genes stay the same throughout a person’s lifetime, the way genes are read and expressed in the body changes based on one’s behavior, environmental exposures, and other factors. Just like a switch turns a light on and off, epigenetics is the process that turns genes on and off. These changes in how genes are read and expressed underlie an individual’s risk for obesity. Epigenetic changes in a [father or mother] can be passed down to [his or her] child, influencing the child’s change of developing obesity in his or her lifetime. For example, a [father or mother] who eats an unhealthy diet can pass down epigenetic changes that influence [his or her] child’s weight | Genes are instruction manuals for the body, and are the directions for building all the proteins that make the body function. They are inherited from biological parents. Differences in genes underlie an individual’s risk for obesity. Weight-related genes in a [father OR mother] can thus be passed down to [his OR her] child, influencing the child’s chances of developing obesity in his or her lifetime. For example, genes that influence a [father’s OR mother’s] metabolism can be passed down and influence [his OR her] child’s weight | The family environment is defined as habits, behaviors, and beliefs that parents set. [Fathers OR Mothers] are especially important in establishing the family environment. This environment sets in motion behaviors that children continue throughout their lives and affects their risk of becoming obese |
| Quote | “I was shocked to learn that my eating behaviors even before I had Taylor influence Taylor’s weight in this way,” said Mr. Bryson | “I was shocked to learn that my genes influence Taylor’s weight in this way,” said Mr. Bryson | “I was shocked to learn that my behaviors influence Taylor’s weight in this way,” said Mr. Bryson |
| Studies | These studies offer clues to a big question in the field of obesity research: Why can’t other known factors like the environment or genetics explain why children with heavy parents grow up to be overweight? This recent research shows that the answer is epigenetics | These studies offer clues into a big question in the field of obesity: Why do people gain different amounts of weight while eating the same number of calories? This recent research shows that the answer is genetics | These studies offer clues into a big question in the field of obesity: How do unhealthy habits that parents establish persist once parents are no longer around? This recent research shows that the answer is the early family environment |
| Conclusions | These studies show how the consequences of parents behaviors are passed down to children, affecting children’s gene expression, which explains a child’s risk for developing obesity | These studies show how genes passed down from [fathers OR mothers] can form the basis of a child’s risk for obesity | The studies show how weight-related behaviors like food preference and consumption of fatty foods are taught by parents (rather than being due to innate preferences and cravings), which explains a child’s risk for obesity |
H1: Epigenetics messages will result in reduced perceived control, increased cognitive risk perceptions of one’s child’s obesity risk, and increased intent to change children’s health behaviors in the future (similar to effects seen for genetics messages).
H2: Epigenetics messages will produce increased threat response and increased affective/intuitive risk perception of one’s child’s obesity risk (similar to effects seen for family environment messages).
H3: Trust in science will moderate parental reactions to the epigenetics article. (a) Those with low trust in science will exhibit reduced attentiveness to the article, perceived relevance, and perceived credibility of the article, perceived risk (cognitive and affective/intuitive), perceived control, and intent to change behavior in response to the epigenetics messages compared to the genetics and family environment messages and (b) those with higher trust in science will have increased attentiveness, relevance, credibility, risk perception (cognitive and affective/intuitive), perceived control over their child’s weight, and intent to change their child’s health behaviors after reading the epigenetic message versus the genetics or family environment messages.
Methods
Data for this manuscript were generated within a larger study that assessed parental responses to messages about the effect of parent weight status on children’s obesity risk with attributions to different causal contexts [37]. The study included a sample of adult parents with self-reported overweight or obesity who had at least one biological child 3–13 years of age. Participants were recruited on Amazon Mechanical Turk, a data crowdsourcing platform shown to replicate findings from research in nationally representative samples in the USA [38, 39] through the Cloud Research platform. We excluded 94 of 416 qualifying participants due to incomplete data on measures and other data quality concerns (see online suppl. material at https://doi.org/10.1159/000543627 for further details), which yielded a sample of 322 for analysis. Data collection was completed in 2019 and the study was IRB approved.
This study is a 4 (article type, manipulated) × 3 (trust in science category, measured) factorial design with obesity-causing mechanisms described in article content and self-report trust in science categorized into three levels as described below. Participants included in the current analysis were randomized to read one of four articles: (a) the article on epigenetic causal factors in children’s obesity risk; (b) a comparator article on genetic factors; (c) a comparator article on family environment risk factors; or (d) a comparator scientific article about ocean exploration (unrelated to obesity). Participants completed a short battery of pre-exposure measures including an assessment of trust in science (shown in Table 2). Participants then read a newspaper-style article consistent with their assigned condition. Participants completed a post-exposure questionnaire with open-ended and standardized measures that evaluated their response to the message, including potential implications for themselves and their child (shown in Table 2).
Study measures
| Construct and reliability | Items | Response scale |
|---|---|---|
| Trust in Science α = 0.80 (Brossard and Nisbet, 2007) | Average of | |
| “Scientists know what is good for the public” | ||
| “It is important for scientists to get research done even if they displease people by doing it” | ||
| “Scientists should do what they think is best, even if they have to persuade people that it is right” | ||
| “Scientists should make decisions about research that is done on obesity” | ||
| Attentiveness to article [40‒44] | “I paid a great deal of attention when reading this article” | 1 (strongly disagree) – 7 (strongly agree) |
| Threat response (de novo) | “I felt threatened when reading this article” | 1 (strongly disagree) – 7 (strongly agree) |
| Perceived article relevance [41] | Average of | |
| “I felt this article was relevant to me personally” | 1 (strongly disagree) – 7 (strongly agree) | |
| “I felt this article was relevant to my family” | 1 (strongly disagree) – 7 (strongly agree) | |
| Perceived article credibility α = 0.90 (Meyer, 1988) | Average of | |
| Trust | 1 (cannot be trusted) – 5 (can be trusted) | |
| Accuracy | 1 (is inaccurate) – 5 (is accurate) | |
| Fairness | 1 (is unfair) – 5 (is fair) | |
| Objectivity | 1 (does not tell the whole story) – 5 (tells the whole story) | |
| Bias | 1 (is biased) – 5 (is unbiased) | |
| Intent to change parental feeding behavior [43] | Average of | 1 (strongly disagree) – 7 (strongly agree) |
| “I intend to make changes to my child’s diet in the next 6 months” | ||
| “I intend to make changes such that my child gets more exercise in the next 6 months” | ||
| Cognitive risk perception r(320) = 0.78, p < 0.0001 (Ferrer et al., 2016) | Average of | |
| “How likely do you think it is that your child will be obese as an adult?” | 1 (certain not to happen) - 7 (certain to happen) | |
| “Compared to the average child, what do you think are your child’s chances of developing obesity?” | 1 (much lower than average) - 7 (much higher than average) | |
| Affective/intuitive risk perception r(320) = 0.83, p < 0.0001 (Ferrer et al., 2016) | Average of | |
| “How vulnerable do you feel your child is to developing obesity?” | 1 (not vulnerable at all) - 7 (very vulnerable) | |
| “How worried are you that your child will be obese as an adult?” | 1 (not at all worried) - 7 (very worried) | |
| Perceived control of child’s obesity risk (de novo) | “How much do you think there are things you can do to control your child’s risk for developing obesity?” | 1 (not at all) – 7 (completely) |
| Construct and reliability | Items | Response scale |
|---|---|---|
| Trust in Science α = 0.80 (Brossard and Nisbet, 2007) | Average of | |
| “Scientists know what is good for the public” | ||
| “It is important for scientists to get research done even if they displease people by doing it” | ||
| “Scientists should do what they think is best, even if they have to persuade people that it is right” | ||
| “Scientists should make decisions about research that is done on obesity” | ||
| Attentiveness to article [40‒44] | “I paid a great deal of attention when reading this article” | 1 (strongly disagree) – 7 (strongly agree) |
| Threat response (de novo) | “I felt threatened when reading this article” | 1 (strongly disagree) – 7 (strongly agree) |
| Perceived article relevance [41] | Average of | |
| “I felt this article was relevant to me personally” | 1 (strongly disagree) – 7 (strongly agree) | |
| “I felt this article was relevant to my family” | 1 (strongly disagree) – 7 (strongly agree) | |
| Perceived article credibility α = 0.90 (Meyer, 1988) | Average of | |
| Trust | 1 (cannot be trusted) – 5 (can be trusted) | |
| Accuracy | 1 (is inaccurate) – 5 (is accurate) | |
| Fairness | 1 (is unfair) – 5 (is fair) | |
| Objectivity | 1 (does not tell the whole story) – 5 (tells the whole story) | |
| Bias | 1 (is biased) – 5 (is unbiased) | |
| Intent to change parental feeding behavior [43] | Average of | 1 (strongly disagree) – 7 (strongly agree) |
| “I intend to make changes to my child’s diet in the next 6 months” | ||
| “I intend to make changes such that my child gets more exercise in the next 6 months” | ||
| Cognitive risk perception r(320) = 0.78, p < 0.0001 (Ferrer et al., 2016) | Average of | |
| “How likely do you think it is that your child will be obese as an adult?” | 1 (certain not to happen) - 7 (certain to happen) | |
| “Compared to the average child, what do you think are your child’s chances of developing obesity?” | 1 (much lower than average) - 7 (much higher than average) | |
| Affective/intuitive risk perception r(320) = 0.83, p < 0.0001 (Ferrer et al., 2016) | Average of | |
| “How vulnerable do you feel your child is to developing obesity?” | 1 (not vulnerable at all) - 7 (very vulnerable) | |
| “How worried are you that your child will be obese as an adult?” | 1 (not at all worried) - 7 (very worried) | |
| Perceived control of child’s obesity risk (de novo) | “How much do you think there are things you can do to control your child’s risk for developing obesity?” | 1 (not at all) – 7 (completely) |
Cronbach’s alpha (α) is reported for constructs with 3+ items, and correlation (r) with p value (p) is reported for constructs with 2 items.
The article in each experimental condition (i.e., epigenetics, genetics, family environment) referenced a parent (matched to participant gender) and child visiting their pediatrician. The doctor in the article described the increased obesity risk that a parent’s high weight status can pose to their child. This elevated risk was then attributed to a specific cause. The epigenetics condition explained that a parent’s weight status before conception causes epigenetic changes, which are inherited by their children and can affect their weight status. This was compared with articles describing the potential for parents to pass down weight-related genes to their children (genetics condition) or parents building habits in their children based on the environment they create (family environment condition). The unrelated science article described a record-breaking dive by a fictional ocean explorer; subject matter was unrelated to obesity or parenting. All articles were matched in terms of reading difficulty and word count.
Measures
Measures included article judgments, perceptions related to their child’s obesity risk, and behavioral intentions (shown in Table 2). Parents completed these measures in reference to one of their children if they had more than 1 biological child 3–13 years of age.
Analyses
For the trust in science measure, participants were grouped into low (below 25th percentile; score 0–3.25), medium (middle 50th percentile; score 3.25–4.25), and high (above 75th percentile; score 4.25–5) trust in science (shown in Table 2). Sensitivity analyses were also performed with the trust in science scores grouped based on 1 standard deviation above or below the mean. All effects with this categorization remained significant. We report groupings based on percentiles here because of small cell sizes that result from standard deviation-based categorization. We conducted two-way ANOVAs to investigate the main effects and interaction between article type and trust in science. For responses relating directly to the article (e.g., reported attentiveness to article), the epigenetics condition was compared to the genetics and family environment conditions only. Due to the inherently reduced personal relevance of the unrelated, control article, for the outcomes related to health behavior and not directly related article response (e.g., perceived control over child’s weight status), a group that read a control article unrelated to obesity was included in the analyses as a no-information comparator. We conducted targeted post hoc comparisons for interaction effects if the interaction term was significant.
Results
Sample characteristics by article type and trust in science appear in Table 3 and descriptives of study outcomes appear in Table 4. No differences were observed between randomly assigned article condition groups by demographic variables (shown in Table 3). Differences were observed between science trust levels by parent gender, child age, and child gender. We ran each of the following models with and without these three demographic variables which did not change the pattern of results. We report models without demographics as covariates (shown in Table 5).
Demographics by article type condition and science trust group
| Total (n = 322) | Article type | p value | Science trust level | p value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| epigenetics (n = 84) | genetics comparator (n = 93) | family environment comparator (n = 81) | unrelated comparator (n = 64) | low (n = 84) | medium (n = 135) | high (n = 103) | ||||
| Parent gender, n (%)a | ||||||||||
| Male | 107 (33.2) | 28 (33.3) | 29 (31.2) | 32 (39.5) | 18 (28.1) | NS | 22 (26.2) | 55 (40.7) | 30 (29.1) | 0.048b |
| Female | 215 (66.8) | 56 (66.7) | 64 (68.8) | 49 (60.5) | 46 (71.9) | NS | 62 (73.8) | 80 (59.3) | 73 (70.9) | |
| Parent age, years | ||||||||||
| Mean (SD) | 36.7 (7.3) | 37.6 (6.8) | 37.0 (7.9) | 35.9 (7.0) | 35.9 (7.6) | NS | 37.7 (±7.2) | 36.3 (±7.1) | 36.4 (±7.7) | NS |
| Parent education, n (%) | ||||||||||
| No college | 34 (10.6) | 10 (11.9) | 10 (10.8) | 9 (11.1) | 5 (7.8) | NS | 9 (10.7) | 13 (9.6) | 12 (11.7) | NS |
| College + | 288 (89.4) | 74 (88.1) | 83 (89.2) | 72 (88.9) | 59 (92.2) | NS | 75 (89.3) | 122 (90.4) | 91 (88.3) | NS |
| Parent race/ethnicity, n (%) | ||||||||||
| White | 270 (83.9) | 69 (82.1) | 81 (87.1) | 65 (80.2) | 55 (7.8) | NS | 71 (84.5) | 113 (83.7%) | 86 (83.5) | NS |
| Latinx | 24 (7.4) | 6 (7.1) | 6 (6.5) | 9 (11.1) | 3 (4.7) | NS | 6 (7.1) | 10 (7.4) | 8 (7.8) | NS |
| Black | 11 (3.4) | 3 (3.6) | 2 (2.2) | 3 (3.7) | 3 (4.7) | NS | 3 (3.6) | 5 (3.7) | 3 (2.9) | NS |
| Other | 17 (5.3%) | 6 (7.1%) | 4 (4.3%) | 4 (4.9%) | 3 (4.7%) | NS | 4 (4.8%) | 2 (1.5%) | 6 (5.9%) | NS |
| Parent weight status, n (%) | ||||||||||
| Overweight | 278 (86.0) | 77 (91.7) | 80 (86.0) | 67 (82.7) | 54 (84.4) | NS | 71 (84.5) | 123 (91.1) | 84 (81.6) | NS |
| Very overweight | 44 (13.7) | 7 (8.3) | 13 (14.0) | 14 (17.3) | 10 (15.6) | NS | 13 (15.5) | 12 (8.9) | 9 (18.4) | NS |
| Parent BMI | ||||||||||
| Mean (SD) | 32.8 (6.9) | 31.7 (5.0) | 32.1 (5.4) | 34.4 (8.6) | 33.1 (8.2) | NS | 33.8 (±7.7) | 32.0 (±5.7) | 32.9 (±7.6) | NS |
| Child age, years | ||||||||||
| Mean (SD) | 6.6 (2.7) | 6.3 (2.5) | 6.9 (2.8) | 6.2 (2.5) | 7.2 (2.8) | NS | 7.2 (±2.7) | 6.4 (±2.8) | 6.5 (±2.4) | NS |
| Child gender, n (%) | ||||||||||
| Male | 162 (50.3) | 44 (52.4) | 48 (51.6) | 37 (45.7) | 33 (51.6) | NS | 53 (63.1) | 69 (51.1) | 40 (38.8) | 0.004c |
| Female | 160 (49.7) | 40 (47.6) | 45 (48.4) | 44 (54.3) | 31 (48.4) | NS | 31 (36.9) | 66 (48.9) | 63 (61.2) | |
| Child perceived weight, n (%) | ||||||||||
| Overweight | 59 (18.3) | 14 (16.7) | 20 (20.5) | 12 (14.8) | 13 (20.3) | NS | 13 (15.5) | 26 (19.3) | 20 (19.4) | NS |
| Not overweight | 263 (81.7) | 70 (83.3) | 73 (78.5) | 69 (85.2) | 51 (79.7) | NS | 71 (84.5) | 109 (80.7) | 83 (80.6) | NS |
| Science trust | ||||||||||
| Mean (SD) | 3.76 (0.74) | 3.80 (0.62) | 3.79 (0.66) | 3.72 (0.83) | 3.72 (0.86) | NS | - | |||
| Low, n (%) | 84 (26.1) | 15 (17.9) | 23 (24.7) | 27 (33.3) | 19 (29.7) | NS | ||||
| Medium, n (%) | 135 (41.9) | 47 (55.9) | 40 (43.0) | 25 (30.9) | 23 (35.9) | NS | ||||
| High, n (%) | 103 (32.0) | 22 (26.2) | 30 (32.3) | 29 (35.8) | 22 (34.4) | NS | ||||
| Total (n = 322) | Article type | p value | Science trust level | p value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
| epigenetics (n = 84) | genetics comparator (n = 93) | family environment comparator (n = 81) | unrelated comparator (n = 64) | low (n = 84) | medium (n = 135) | high (n = 103) | ||||
| Parent gender, n (%)a | ||||||||||
| Male | 107 (33.2) | 28 (33.3) | 29 (31.2) | 32 (39.5) | 18 (28.1) | NS | 22 (26.2) | 55 (40.7) | 30 (29.1) | 0.048b |
| Female | 215 (66.8) | 56 (66.7) | 64 (68.8) | 49 (60.5) | 46 (71.9) | NS | 62 (73.8) | 80 (59.3) | 73 (70.9) | |
| Parent age, years | ||||||||||
| Mean (SD) | 36.7 (7.3) | 37.6 (6.8) | 37.0 (7.9) | 35.9 (7.0) | 35.9 (7.6) | NS | 37.7 (±7.2) | 36.3 (±7.1) | 36.4 (±7.7) | NS |
| Parent education, n (%) | ||||||||||
| No college | 34 (10.6) | 10 (11.9) | 10 (10.8) | 9 (11.1) | 5 (7.8) | NS | 9 (10.7) | 13 (9.6) | 12 (11.7) | NS |
| College + | 288 (89.4) | 74 (88.1) | 83 (89.2) | 72 (88.9) | 59 (92.2) | NS | 75 (89.3) | 122 (90.4) | 91 (88.3) | NS |
| Parent race/ethnicity, n (%) | ||||||||||
| White | 270 (83.9) | 69 (82.1) | 81 (87.1) | 65 (80.2) | 55 (7.8) | NS | 71 (84.5) | 113 (83.7%) | 86 (83.5) | NS |
| Latinx | 24 (7.4) | 6 (7.1) | 6 (6.5) | 9 (11.1) | 3 (4.7) | NS | 6 (7.1) | 10 (7.4) | 8 (7.8) | NS |
| Black | 11 (3.4) | 3 (3.6) | 2 (2.2) | 3 (3.7) | 3 (4.7) | NS | 3 (3.6) | 5 (3.7) | 3 (2.9) | NS |
| Other | 17 (5.3%) | 6 (7.1%) | 4 (4.3%) | 4 (4.9%) | 3 (4.7%) | NS | 4 (4.8%) | 2 (1.5%) | 6 (5.9%) | NS |
| Parent weight status, n (%) | ||||||||||
| Overweight | 278 (86.0) | 77 (91.7) | 80 (86.0) | 67 (82.7) | 54 (84.4) | NS | 71 (84.5) | 123 (91.1) | 84 (81.6) | NS |
| Very overweight | 44 (13.7) | 7 (8.3) | 13 (14.0) | 14 (17.3) | 10 (15.6) | NS | 13 (15.5) | 12 (8.9) | 9 (18.4) | NS |
| Parent BMI | ||||||||||
| Mean (SD) | 32.8 (6.9) | 31.7 (5.0) | 32.1 (5.4) | 34.4 (8.6) | 33.1 (8.2) | NS | 33.8 (±7.7) | 32.0 (±5.7) | 32.9 (±7.6) | NS |
| Child age, years | ||||||||||
| Mean (SD) | 6.6 (2.7) | 6.3 (2.5) | 6.9 (2.8) | 6.2 (2.5) | 7.2 (2.8) | NS | 7.2 (±2.7) | 6.4 (±2.8) | 6.5 (±2.4) | NS |
| Child gender, n (%) | ||||||||||
| Male | 162 (50.3) | 44 (52.4) | 48 (51.6) | 37 (45.7) | 33 (51.6) | NS | 53 (63.1) | 69 (51.1) | 40 (38.8) | 0.004c |
| Female | 160 (49.7) | 40 (47.6) | 45 (48.4) | 44 (54.3) | 31 (48.4) | NS | 31 (36.9) | 66 (48.9) | 63 (61.2) | |
| Child perceived weight, n (%) | ||||||||||
| Overweight | 59 (18.3) | 14 (16.7) | 20 (20.5) | 12 (14.8) | 13 (20.3) | NS | 13 (15.5) | 26 (19.3) | 20 (19.4) | NS |
| Not overweight | 263 (81.7) | 70 (83.3) | 73 (78.5) | 69 (85.2) | 51 (79.7) | NS | 71 (84.5) | 109 (80.7) | 83 (80.6) | NS |
| Science trust | ||||||||||
| Mean (SD) | 3.76 (0.74) | 3.80 (0.62) | 3.79 (0.66) | 3.72 (0.83) | 3.72 (0.86) | NS | - | |||
| Low, n (%) | 84 (26.1) | 15 (17.9) | 23 (24.7) | 27 (33.3) | 19 (29.7) | NS | ||||
| Medium, n (%) | 135 (41.9) | 47 (55.9) | 40 (43.0) | 25 (30.9) | 23 (35.9) | NS | ||||
| High, n (%) | 103 (32.0) | 22 (26.2) | 30 (32.3) | 29 (35.8) | 22 (34.4) | NS | ||||
p reflects significance test for chi square for categorical variables and anova/t test for continuous variables.
NS, not significant.
aSurvey question used the term gender, but survey responses incorrectly labeled options typically used for sex, male, and female.
bPost hoc analyses applied with Bonferroni correction reveal no significant differences.
cPost hoc analyses applied with Bonferroni correction reveal significant differences in female and males in the low science trust groups and in the high science trust group.
Descriptives of study outcomes by article type and science trust
| Overall | Article type | Science trust | ||||||
|---|---|---|---|---|---|---|---|---|
| epigenetics | genetics comparator | family environment comparator | unrelated comparator1 | low | medium | high | ||
| Attention to article | 6.21 (1.02) | 6.24 (1.04) | 6.18 (1.04) | 6.21 (0.98) | - | 6.05 (1.12) | 6.19 (1.06) | 6.37 (0.84) |
| Article credibility | 3.64 (0.94) | 3.46 (1.00) | 3.69 (0.88) | 3.76 (0.92) | - | 3.35 (0.99)A | 3.66 (0.87)B | 3.83 (0.94)B |
| Threat response to article | 2.55 (1.79) | 2.42 (1.81) | 2.51 (1.67) | 2.74 (1.91) | - | 2.52 (1.71) | 2.61 (1.73) | 2.48 (1.95) |
| Article relevance | 5.36 (1.54) | 5.10 (1.58) | 5.41 (1.59) | 5.57 (1.42) | - | 5.05 (1.67) | 5.37 (1.41) | 5.59 (1.58) |
| Intent to change parental feeding behavior | 4.81 (1.62) | 4.86 (1.59) | 4.72 (1.65) | 5.00 (1.60) | 4.65 (1.68) | 4.65 (1.73) | 4.81 (1.51) | 4.95 (1.68) |
| Perceived control | 5.52 (1.04) | 5.48 (1.04)A | 5.27 (1.07)AC | 5.62 (0.92)D | 5.81 (1.07)B | 5.38 (1.19) | 5.47 (1.04) | 5.70 (0.86) |
| Cognitive risk perception | 3.67 (1.32) | 3.49 (1.21)A | 3.91 (1.25)BC | 3.81 (1.39)C | 3.37 (1.43)D | 3.67 (1.29) | 3.67 (1.29) | 3.67 (1.41) |
| Affective/intuitive risk perception | 4.02 (1.66) | 3.82 (1.58)A | 4.32 (1.65)BC | 4.20 (1.69)C | 3.63 (1.68)D | 4.04 (1.53) | 3.97 (1.61) | 4.07 (1.84) |
| Overall | Article type | Science trust | ||||||
|---|---|---|---|---|---|---|---|---|
| epigenetics | genetics comparator | family environment comparator | unrelated comparator1 | low | medium | high | ||
| Attention to article | 6.21 (1.02) | 6.24 (1.04) | 6.18 (1.04) | 6.21 (0.98) | - | 6.05 (1.12) | 6.19 (1.06) | 6.37 (0.84) |
| Article credibility | 3.64 (0.94) | 3.46 (1.00) | 3.69 (0.88) | 3.76 (0.92) | - | 3.35 (0.99)A | 3.66 (0.87)B | 3.83 (0.94)B |
| Threat response to article | 2.55 (1.79) | 2.42 (1.81) | 2.51 (1.67) | 2.74 (1.91) | - | 2.52 (1.71) | 2.61 (1.73) | 2.48 (1.95) |
| Article relevance | 5.36 (1.54) | 5.10 (1.58) | 5.41 (1.59) | 5.57 (1.42) | - | 5.05 (1.67) | 5.37 (1.41) | 5.59 (1.58) |
| Intent to change parental feeding behavior | 4.81 (1.62) | 4.86 (1.59) | 4.72 (1.65) | 5.00 (1.60) | 4.65 (1.68) | 4.65 (1.73) | 4.81 (1.51) | 4.95 (1.68) |
| Perceived control | 5.52 (1.04) | 5.48 (1.04)A | 5.27 (1.07)AC | 5.62 (0.92)D | 5.81 (1.07)B | 5.38 (1.19) | 5.47 (1.04) | 5.70 (0.86) |
| Cognitive risk perception | 3.67 (1.32) | 3.49 (1.21)A | 3.91 (1.25)BC | 3.81 (1.39)C | 3.37 (1.43)D | 3.67 (1.29) | 3.67 (1.29) | 3.67 (1.41) |
| Affective/intuitive risk perception | 4.02 (1.66) | 3.82 (1.58)A | 4.32 (1.65)BC | 4.20 (1.69)C | 3.63 (1.68)D | 4.04 (1.53) | 3.97 (1.61) | 4.07 (1.84) |
Cells represent means and standard deviation. Bolded values reperesent tests that demonstrated statistical significance. Superscript A/B or C/D represents significantly different groups.
1Unrelated comparator article included in analyses only for health behavior-related outcomes, as article was entirely irrelevant to intergenerational epigenetic risk transmission.
Effects of article type and trust in science on study outcomes
| Article type | Trust in science | Interaction | ||||
|---|---|---|---|---|---|---|
| F (df) | p value | F (df) | p value | F (df) | p value | |
| Attention to article1 | 0.07 (2, 249) | 0.935 | 1.94 (2, 249) | 0.147 | 2.92 (4, 249) | 0.022 |
| Article credibility1 | 2.49 (2, 249) | 0.085 | 5.51 (2, 249) | 0.005 | 1.76 (4, 249) | 0.138 |
| Threat response to article1 | 0.73 (2, 249) | 0.743 | 0.95 (2, 249) | 0.743 | 2.02 (4, 249) | 0.093 |
| Article relevance1 | 2.13 (2, 249) | 0.121 | 2.59 (2, 249) | 0.077 | 1.15 (4, 249) | 0.330 |
| Intent to change parental feeding behavior | 0.70 (3, 310) | 0.551 | 0.81 (2, 310) | 0.447 | 1.08 (6, 310) | 0.372 |
| Perceived control | 3.91 (3, 310) | 0.009 | 2.54 (2, 310) | 0.081 | 0.34 (6, 310) | 0.915 |
| Cognitive risk perception | 3.03 (3, 310) | 0.030 | 0.01 (2, 310) | 0.984 | 1.52 (6, 310) | 0.171 |
| Affective/intuitive risk perception | 3.05 (3, 310) | 0.029 | 0.07 (2, 310) | 0.934 | 2.40 (6, 310) | 0.028 |
| Article type | Trust in science | Interaction | ||||
|---|---|---|---|---|---|---|
| F (df) | p value | F (df) | p value | F (df) | p value | |
| Attention to article1 | 0.07 (2, 249) | 0.935 | 1.94 (2, 249) | 0.147 | 2.92 (4, 249) | 0.022 |
| Article credibility1 | 2.49 (2, 249) | 0.085 | 5.51 (2, 249) | 0.005 | 1.76 (4, 249) | 0.138 |
| Threat response to article1 | 0.73 (2, 249) | 0.743 | 0.95 (2, 249) | 0.743 | 2.02 (4, 249) | 0.093 |
| Article relevance1 | 2.13 (2, 249) | 0.121 | 2.59 (2, 249) | 0.077 | 1.15 (4, 249) | 0.330 |
| Intent to change parental feeding behavior | 0.70 (3, 310) | 0.551 | 0.81 (2, 310) | 0.447 | 1.08 (6, 310) | 0.372 |
| Perceived control | 3.91 (3, 310) | 0.009 | 2.54 (2, 310) | 0.081 | 0.34 (6, 310) | 0.915 |
| Cognitive risk perception | 3.03 (3, 310) | 0.030 | 0.01 (2, 310) | 0.984 | 1.52 (6, 310) | 0.171 |
| Affective/intuitive risk perception | 3.05 (3, 310) | 0.029 | 0.07 (2, 310) | 0.934 | 2.40 (6, 310) | 0.028 |
F statistic (F) reflects the overall differences in variances in study outcomes between study groups; df, degrees of freedom for anova test, p = significance level.
1Unrelated comparator article included in analyses only for health behavior-related outcomes.
Article Judgments
There was a significant interaction by article type and science trust (shown in Fig. 1). Consistent with hypotheses, those with low trust in science paid significantly less attention to the epigenetics article (M = 2.72, SD = 0.98) than the genetics article (M = 3.46, SD = 0.93). Among participants who read the epigenetics article, those with low trust in science reported paying significantly less attention (M = 2.72, SD = 0.98) than those in the medium (M = 3.56, SD = 0.89) or high (M = 3.74, SD = 1.05) science trust groups.
A main effect of science trust was observed for perceived article credibility. Consistent with hypotheses, those with low trust in science rated articles as significantly less credible (M = 3.35, SD = 0.99) than those in the medium (M = 3.66, SD = 0.87) or high (M = 3.83, SD = 0.94) science trust groups. Inconsistent with hypotheses, there were no differences observed in threat response or in perceived article relevance.
Health Behavior-Related Outcomes
Inconsistent with hypotheses, there were no significant main effects or interaction for intent to change parental feeding behavior in the next 6 months. There was a significant main effect by condition observed in perceived control over child’s weight. Consistent with hypotheses, in the epigenetics condition, perceived control was significantly lower (M = 5.48, SD = 1.04) than those in the unrelated control condition (M = 5.81, SD = 1.07), but not significantly different from the family environment condition (M = 5.62, SD = 0.92). Similarly, perceived control in the genetics condition was significantly lower (M = 5.27, SD = 1.07) than in the unrelated control condition and the family environment condition. Inconsistent with hypotheses, no significant main or moderation effects of science trust were observed.
Risk Perception
Cognitive risk perception differed significantly by article type. Inconsistent with hypotheses, cognitive perceptions of risk in response to the epigenetics article (M = 3.49, SD = 1.21) were not significantly different from the unrelated article (M = 3.37, SD = 1.43). However, the genetics (M = 3.91, SD = 1.25) and family environment (M = 3.81, SD = 1.39) articles were associated with increased risk perception over the unrelated article. Further, cognitive risk perception in the epigenetics condition was significantly lower than in the genetics condition. Inconsistent with hypotheses, we did not observe a main effect of science trust or any interaction.
Similar main effects were observed for affective/intuitive risk perception (shown in Fig. 2). Inconsistent with hypotheses, the epigenetics article did not elicit significantly elevated risk perceptions (M = 3.82, SD = 1.58) compared to the unrelated article (M = 3.63, SD = 1.68), whereas the genetics (M = 4.32, SD = 1.65) and family environment (M = 4.20, SD = 1.69) articles were associated with higher levels of risk perception than the unrelated article. The epigenetics article also elicited significantly lower affective/intuitive risk perception than the genetics condition. These main effects are qualified by a significant moderation effect by trust in science. Among those with medium science trust, the genetics (M = 4.45, SD = 1.63) and family environment (M = 4.64, SD = 1.35) articles produced significantly elevated affective risk perception compared to both the unrelated (M = 3.63, SD = 1.63) and epigenetics (M = 3.37, SD = 1.54) articles. Consistent with hypotheses, among those with high science trust, the epigenetics (M = 4.61, SD = 1.65) and genetics (M = 4.30, SD = 1.76) articles produced elevated risk perception compared to the unrelated article (M = 3.27, SD = 1.72), while the family environment (M = 4.03, SD = 2.01) article showed no significant difference from the unrelated article. Finally, inconsistent with hypotheses, among those with low trust in science, there were no differences in risk perception by condition. In addition, within the group that read the epigenetics article, those with medium science trust (M = 3.37, SD = 1.54) had significantly lower affective/intuitive risk perception than those with high science trust (M = 4.61, SD = 1.65). There were no differences by science trust within the genetics, family environment, or unrelated article conditions.
Discussion
Parents who read a newspaper-style article about preconception epigenetic mechanisms of obesity risk transmission to their children exhibited a range of reactions that, at times, mirrored reactions to genetic and family environment-based risk information; however, patterns differed substantially from those originally hypothesized. We predicted that epigenetic messages might boost perceived health risk for one’s offspring, in which case, future work might explore using such messaging to elicit health behavior change among individuals contemplating parenthood. However, we found that the epigenetics article produced levels of cognitive and affective/intuitive risk perception equivalent to the unrelated article, and lower levels of risk perception than the genetic and family environment articles (which are longstanding and frequently used frames for providing risk information). That is, epigenetics information did not increase parents’ perceptions of their child’s obesity risk despite the fact that the numeric and other risk information associated with epigenetic mechanisms was the same as the risk level shown for the genetic and family environment conditions.
Across science trust groups, parents were equally accepting of the epigenetics article as they were of other obesity risk-oriented articles such that they reported high levels of relevance and low levels of threat response. Those in the moderate to high science trust groups agreed with the notion that the article was relevant and credible. However, those in the low science trust group reported all articles as less credible and paid less attention to the epigenetics article than those in the medium or high science trust groups.
Changing health-risk perception with messaging is an important route toward spurring health behavior change in many domains. However, as stated, the epigenetics message, unlike its family environment and genetic counterparts, did not influence perceived risk. Research has suggested interventions are more effective in increasing risk perception and related outcomes when people are able to envision and understand how mechanisms of risk relate to mechanisms of behavioral response [17]. This may partly explain why learning about risk conferred by the novel and complex pathway of epigenetics did not result in heightened risk perception. It is, however, apparent from study results that the lack of increase in risk perception was not universally due to discounting or derogation of the message as this did not occur generally in the epigenetics condition.
In contrast, the epigenetics article did result in reduced perceived control over children’s future obesity status compared to the control comparator, although this effect was not as extreme as that produced by the genetics article. This is perhaps an area to carefully consider when crafting epigenetics messages, as perceived control has been identified as a crucial pathway underlying parental health behavior for their children [17]. The present study focused a description of on intergenerational transmission of epigenetic changes at conception, and as the participants were all current parents these changes were likely viewed as unmodifiable. However, many epigenetic markers are affected by environmental factors, and research is ongoing regarding the degree to which intergenerationally inherited epigenetic changes may change throughout the offspring’s lifetime. Epigenetics as a concept has the benefit of including within it a specific mechanism for risk reduction via enacting health behaviors. As such, preconception epigenetics messages might have a different effect on those who have not yet conceived a child or in the case where messages are accompanied by information about the dynamic nature of epigenetic changes. This is a key area for future research.
Because of the unfamiliar and somewhat unintuitive nature of epigenetics concepts and the role of information source judgments in risk processing, we predicted that science trust would be an important moderator of reactions to messages about epigenetics-based risk. Those with low science trust did demonstrate lower perceived credibility of the articles overall. Interestingly, this was not specific to the epigenetics article. The novelty of the epigenetics article may have magnified this effect, however, as those with low science trust who read the epigenetics article reported lower attentiveness to the article than any other group. This suggests that although readers’ trust in science may be especially important in predicting engagement with unfamiliar risk information like epigenetics, trust in science may be an individual difference to attend to when communicating about risk across the spectrum of causal pathways regardless of novelty.
Beyond engagement, we also examined how trust in science would influence the application of risk messages to one’s own child. Here, effects of science trust proved to be more complex than expected, with some null results, but also different considerations across trust groups rather than a linear effect of trust level. With respect to affective/intuitive risk, the epigenetics article resulted in increased risk perception only in the high science trust group. In the medium science trust group, it failed to produce the increased risk perception observed in response to the genetics and family environment articles, while among those with low science trust, none of the articles appeared to provoke increased affective/intuitive risk perception. While affective/intuitive risk perception is only one of multiple outcomes evaluated here, it is an important outcome as it is a robust predictor of preventive health behavior [45, 46]. This variable pattern of responses by science trust group suggests that such trust may not act in a linear way, and those who endorse low versus medium versus high-trust responses may have categorically distinct approaches to evaluating scientific arguments in the media which should be explored further. Overall, the combination of decreased attentiveness, decreased perceived control, and lack of risk perception change in response to the epigenetics messages is worth attending to. Our results support the consideration of audience trust in science when developing health messages, particularly for novel messages like epigenetics. Trust is considered a crucial topic in the broader genetics literature for engagement in research endeavors, biobanking, and genetic testing and treatment decisions [47‒50], so it is sensible that trust was an important construct across study conditions in the current work despite the novel context. This is the first study to our knowledge to investigate the role of trust in science on acceptance and interpretation of risk information stemming from epigenetics and thus is likely to be a fruitful area for research in the future, particularly since public trust in science in the U.S. is variable and has continued to decline since the COVID-19 pandemic [51]. In particular, our findings suggest potential utility to evaluating whether varying aspects of genomics concepts (e.g., complexity, unfamiliarity) might differentially elicit health message discounting depending upon individuals’ general orientation toward trust versus distrust of science.
The current study had several limitations and strengths. This study was focused on information about preconception epigenetic changes, a small subset of the information parents may encounter (e.g., changes in epigenetic markers over time). Epigenetics concepts as discussed here were likely perceived as more immutable than a broader overview of epigenetic processes. Although the experimental articles were written in plain language so as to be understandable by participants, there was no measure that confirmed the participants’ accurate understanding of epigenetics concepts. The use of Amazon Mechanical Turk, despite some limitations in representativeness, has good ecological validity for the spread of online information that we were attempting to achieve. We were able to collect data from fathers as well as mothers which has been a challenge in past research involving parents [52, 53]. Mothers were overrepresented in our sample relative to fathers. However, evidence suggests that despite fathers’ increasing responsibility for child feeding, mothers are more frequently considered the primary caregiver responsible for mealtimes [54]. Individuals with high educational status were overrepresented in this sample. This is a major limitation given the known association between educational status and decreased trust in science. Also, most parents in the sample did not perceive their children to be overweight. It is conceivable that individuals whose children are currently perceived as at high weight status may have more pronounced reactions to the type of information presented in this study. Finally, data were collected in 2019 and many discoveries about epigenetics have been made in the time since. However, the concept of epigenetics has not had a precipitating event to increase its visibility and does not seem to have grown substantially in public awareness. For this reason, the processes reported here are likely relevant at the present time.
Conclusion
Peak motivational potency of epigenetic information is likely to occur as a result of increased risk perception when appropriate while mitigating fatalism (or increasing perceived control). We suggest further research on how discussions in the public might influence people’s perceptions of epigenetics, particularly relating to epigenetics changes outside of the preconception period. Results suggest it may be important to further explore the influence of science trust on receptivity to genetics messages more generally. Further investigation of methods of tailoring risk and educational information to audiences with different levels of trust is warranted. Ultimately, thoughtful communication about epigenetics is the most obvious approach toward mitigating negative reactions at an individual level, and also working against societal-level stigmatization. This could be realized by, for example, explicitly pointing out reasons why parents should not be blamed for epigenetic transmission and by presenting epigenetics information accurately as one piece among many factors that contribute to obesity risk. The current study suggests that genomic risk communication yet another venue wherein systemic efforts to bolster or repair public trust in science could benefit individuals and families.
Statement of Ethics
This study protocol was reviewed and approved by NIH Centralized IRB, Approval No. 10-HG-0076. Written informed consent was obtained from participants.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
This study was funded by intramural research funds from the National Human Genome Research Institute. The funder had no role in the design, data collection, data analysis, and reporting of this study. The effort of S.E.-T. has been funded by the Division of Intramural Research, National Institute on Minority Health and Health Disparities.
Author Contributions
E.M.S. conducted the analyses and contributed to writing of the manuscript. R.A.F. and S.E.-T. contributed to development of survey, interpretation of results, and manuscript writing. S.P. conceptualized the research, led data collection, and contributed to writing of the manuscript.
Data Availability Statement
The data that support the findings of this study are not publicly available due to containing information that could compromise the privacy of research participants but are available from the corresponding author upon reasonable request.
References
