Abstract
Food motivation, herein referred to as the relative reinforcing value of food (RRVFood), refers to how hard an individual will work for a food. RRVFood has been linked to the development and maintenance of overweight/obesity (OW/OB). However, there is little known on the relationships between RRVFood, dietary intake, and weight status within children. The current study aims to assess these relationships in children ages 8-11 who are at risk for OW/OB. Exploratory sex-based differences will also be assessed. Data presented are from the first two sessions of a 7-session longitudinal study. One session consisted of the participants completing a RRV task, which measures RRVFood, and the second session consisted of eating at a buffet-style test meal. Data suggests that participants with OW/OB may not have higher RRVFood in comparison to participants with healthy weight (HW), which opposes previous findings. Secondarily, RRVFood positively predicted total kcal, carbohydrate, and fat intake in the total sample and in participants with HW. While the relationship between RRVFood and energy intake has been established, research on adults did not show similar relationships with carbohydrates and fat. This highlights a potential relationship unique to this age range that could be intervened on as OW/OB prevention. As for the sex differences, boys had higher RRVFood, however RRVFood positively predicted total kcal, carbohydrate, fat, and protein intake in girls only. This reflects a unique relationship not studied thus far. Future work can aim to confirm these relationships in a fully powered study, explore other measures of RRVFood, assess natural-setting dietary intake, and explore variables that may contribute to the sex differences observed in the current study.
Keywords: Food reinforcement, children, dietary intake, overweight and obesity
1.0. Introduction
Reinforcing value of food (RRVFood), which is the amount of effort an individual will spend to attain access to a specific food (Epstein, 2007a), has been strongly associated with the development and maintenance of overweight and obesity (OW/OB). As of 2020, 20.7% of children ages 6-11 in the United States have OW/OB (Stierman et al., 2021). Childhood OW/OB have been associated with acute and long-term health consequences such as type II diabetes, asthma, sleep apnea, cardiovascular disease, poor mental health, etc. (Russell-Mayhew et al., 2012; Sahoo et al., 2015). Furthermore, childhood obesity can lead to adult obesity, which is also related to the development of those conditions (Reilly et al., 2003). Given these potential health risks, there is a need to study obesogenic factors and to develop effective interventions. Previous studies found that both children and adults with higher weight status have increased RRVFood (Epstein et al., 2007a; Epstein et al., 2014; Temple et al., 2008b). Similarly, research has identified RRVFood of high energy dense (HED) snack foods as a positive predictor of zBMI increase in non-overweight adolescents (Epstein et al., 2014).
Children with higher RRVFood also have greater HED food intake (Epstein et al., 2009). Increased HED food intake has been associated with a lower diet quality [i.e., less fruit, vegetable, and nutrient-dense snack food intake] and increased weight status (Ambrosini et al., 2014). Increased intake of HED food is associated with both physiological (Tey et al., 2012) and psychological (Temple & Epstein, 2012) changes. For example, individuals’ sensory-specific satiety can be decreased by consistent consumption of HED snack foods, resulting in potential overconsumption of the HED snack food (Tey et al., 2012). In addition, participants with high weight status and high RRVFood show decreased habituation and increased sensitization to a HED snack food (Temple & Epstein, 2012). In sum, these studies found that participants with overweight and high RRVFood experienced an increased drive to consume that same HED snack food and are less likely to habituate to repeated consumption.
Research on adults indicated that greater RRVFood is associated with greater energy and sugar intake (Epstein et al., 2011), however, there are no known similar analyses conducted in children and adolescents. This is a major gap in that middle childhood (ages 5-11) is an essential time for development and establishes trajectories for development and maintenance of overweight and obesity. During this time, the number of adipocytes (Knittle et al., 1979) and obesity prevalence (Stierman et al., 2021) increases, and diet quality tends to decrease (Costa et al., 2024). Therefore, assessing these topics in children, particularly within that middle childhood range, could provide further insight regarding the implications of RRVFood on dietary intake and general health outcomes.
The current study aims to elucidate the relationship(s) between RRVFood and child dietary intake during a test meal. Specifically, whether RRVFood of a preferred HED snack food (i.e., the study food) predicts child intake of: 1) study food calories (kcal); 2) total kcal; 3) total carbohydrates; 4) total protein; and 5) total fat consumption during a buffet-style test meal. Study food intake will only be examined as kcal as macronutrient composition differs between study food options. In addition to examining these within the total sample, given known relationships between RRVFood and weight status (Epstein, 2007a; Epstein, 2007b; Epstein, 2014; Temple, 2008b), these relationships will be examined by weight status. It is hypothesized that children with OW/OB will have higher RRVFood (Epstein et al., 2009; Temple et al., 2008b) and that RRVFood will positively predict the child dietary intake during the test meal throughout both weight statuses, but show stronger relationships in those with OW/OB. Exploratory analyses aim to highlight sex differences between both dietary intake and RRVFood log, and the relationships between those two variables.
2.0. Methods
2.1. Participants.
The data presented are a secondary data analysis of 112 8-11 year old children participating in a 1-year longitudinal study assessing predictors of weight change. Participants were recruited through a variety of methods (e.g., flyers distributed at elementary schools, flyers posted in local community settings, targeted Facebook ads, previous participant lists). Eligibility criteria included: 1) child ages 8-11; 2) child risk for obesity (defined as either child BMI percentile ≥ 85th but ≤ 98th) or parent BMI ≥ 25; 3) parent-reported child liking of at least 14 of the 28 items on the buffet as moderate or greater; 4) parent-reported child consumption of at least two times/week for one of the seven HED snack foods. This definition of “at risk” for obesity was based on review studies assessing childhood weight status and their associations to both short- and long-term health consequences (Daniels, 2006). Children were not eligible to participate if they reported any medical conditions that affect eating or reported taking any medications that impacted appetite. All parents gave informed consent, and all children gave written assent. All procedures were approved by the University at Buffalo Institutional Review Board.
2.2. Study Design.
The study design consisted of seven study visits over 12 months, with four baseline visits, two six-month follow-up visits, and one 12-month follow-up visit. Data presented for this paper are from visits 1 and 2, which were separated by one week. Children completed all the laboratory assessments, including the test meal, in a location separate from their caregiver to limit any parental influence on child behavior during their procedures. Given that children were being served a test meal, visits were scheduled at a typical lunch or dinner time for the child. Participants were instructed not to eat or drink anything, other than water, for two hours prior to each study visit, which was verified using a same-day dietary recall.
Visit 1.
During the first visit, informed consent and assent was taken. Participant basic anthropometric and demographic data were collected after. Participants were then separated from their caregiver. Then, participants were assigned one high energy dense (HED) snack food out of seven options [see all options in Table 1] for the RRV task. This assignment was based on parent-reported participant consumption frequency (i.e., participant ate the item once/month) and participant reported liking of the item. For all HED snack foods with a parent-reported frequency of at least two times/week, participants rated their liking using a 100mm visual analog [sliding] scale, with anchors of “do not like at all,” (0) and “like the most possible,” (100). Their highest rated HED snack food [herein referred to as their “study food”] was then used for the RRV task. The participants tested three different sedentary activities (handheld electronic games, puzzle activity books, and drawing and coloring) in a room separate from their caregiver. Participants then rated these activities using the same 100mm visual analog scale described earlier and selected their preferred activity. The study food and the preferred sedentary activity were used as the options for the RRV task. Participants consumed a 100-kcal soft granola bar as a preload. Throughout the session, the participants completed multiple hunger and thirst 100mm visual analog scales, with anchors of “not [hungry/thirsty] at all,” (0) and “the most [hungry/thirsty] possible,” (100). These surveys measured hunger, thirst, and liking of their study food at three different time points: before the preload, at least 10 minutes after the preload, and then after the RRV task. The participants then completed the RRV task (Figure 1). One computer allowed the participant to earn portions of their study food (Table 1), and the other allowed the participant to earn two- minute increments of their preferred sedentary activity. After the participants indicated that they had finished playing, they were given time alone to eat their portions of snack food and/or use the minutes they earned towards their chosen sedentary activity. A detailed description of the RRV task can be found in ‘Section 2.3. Measures: Reinforcing Value of food (RRVFood).’
Table 1.
RRVFood options, reward portion size, and kcal per portion
| Study food options | Game portion | Kcal |
|---|---|---|
| Lays Potato Chips1 | 10.5 grams | 60 |
| Cool Ranch Doritos2 | 11.1 grams | 59.46 |
| Cheez-its3 | 12 grams | 60 |
| Chips Ahoy4 | 1 cookie | 53.33 |
| Oreos5 | 1 cookie | 53.32 |
| M&M’s6 | 12 grams | 60 |
| Skittles7 | 15 grams | 58.93 |
Lays Potato Chips (Frito-Lay, Plano, TX).
Cool Ranch Doritos (Frito-Lay, Plano, TX).
Cheez-its (WK Kellogg Co & Kellanova, Battle Creek, MI).
Chips Ahoy (Mondelez International, Chicago, IL).
Oreos (Mondelez International, Chicago, IL).
M&M’s (Mars Inc., McLean, VA).
Skittles (Mars Inc., McLean, VA).
Figure 1.
Diagram of RRV task environment.
Visit 2.
The second visit consisted of ad libitum eating at a test meal comprised of 3 beverages and a mix of high- and low- energy dense foods. Child participants engaged in the test meal in a separate room from their parent / guardian. Once in the room, participants reported on their hunger and thirst levels using a 100mm visual analog scale and then engaged in a buffet-style test meal. Participants were then left in that room while research staff went into a connected room and monitored the participant through the previously mentioned baby monitor. This ensured safety and allowed us to monitor behaviors during the test meal.
2.3. Measures
Anthropometry.
Height was measured in triplicate using a wall- mounted SECA stadiometer (Buffalo, NY) and the average height was used. Weight was measured using a SECA digital scale (Buffalo). Participants were measured without shoes and wearing light clothing. BMI percentile was calculated using age- and sex- specific CDC growth charts (Himes, 1994), which was used to classify weight status (underweight: BMI percentile < 5.0; healthy weight: BMI percentile > 5.0 and < 85.0; overweight: BMI percentile ≥ 85.0 and < 95.0; obese: BMI percentile ≥ 95.1. There were no participants with underweight. Participants with OW/OB were grouped together for comparison and split-weight analyses.
Reinforcing Value of food (RRVFood).
RRVFood was measured through the RRV task. This consisted of two side-by-side computers with the same computer-based program loaded on to it (Figure 1). The procedure was adapted from Epstein, 2011 and has been utilized regularly in a variety of age groups (Epstein, 2011; Epstein et al., 2014; Carr et al., 2023; Temple et al., 2008; Temple & Epstein, 2011; Temple et al., 2017; Temple et al., 2018), including children ages 8-12. This program displayed three shapes that would change when the participant clicked the correlating mouse in the rectangle outlining the three different shapes. Participants would get one point when all three shapes and colors matched. Five points on one computer earned the participant a 60 kcal portions of their study food (Table 1), and five points on the other computer allowed the participant to earn two-minute increments of their preferred sedentary activity. Participants then can choose to play on both computers separately to earn portions of their study food and/or minutes to spend on their preferred sedentary activity. This program used a reinforcement schedule set at a progressive ratio schedule with the following fixed ratio response requirements; 20, 40, 80, 160, 320, 640, 1280, 2560, and 5120. Each value refers to the number of clicks on a computer mouse required to complete that round and receive a portion of food or minutes to spend on a sedentary activity. Once participants completed a round, the following round on the same computer requires double the amount of clicks as the last round. This captures how hard an individual is willing to work for the target stimuli [portions of the study food or 2-minute increments to spend on a preferred sedentary activity]. While participants were told they could end the game when they felt like they were done, each computer had a maximum of nine rounds due to time restraints. After the participants indicated that they had finished playing, they were given time alone to eat their portions of snack food and/or use the minutes they earned towards their preferred sedentary activity. RRVFood was measured using the logged value of total number of responses for the study food on computer-based progressive ratio schedules of reinforcement. The total clicks value was positively skewed, so computing it via natural log provides the RRVFood. This process is standard and has been recommended from previous works (Heo & Gabriel, 1998; Carr & Epstein, 2011).
Child dietary intake during the test meal.
The test meal consisted of 26 foods, including chicken nuggets, sandwich materials, condiments, fruits, vegetables, all study foods listed prior, and three drinks. These foods and beverages were weighed to the correct amount (+/− a small margin of error) and recorded. Once the child indicated that they finished eating, research staff recorded length of time eating. Food and beverages were reweighed and recorded. Intake was calculated by subtracting the post-weight (i.e., amount remaining in each container) from the pre-weight. The weight (g) consumed of each food was used to calculate kcal and macronutrient composition. Total calories (kcal), carbohydrates (grams), protein (grams), and fat (grams), and study food kcal were the measures used for all presented analyses. Study food was measured via kcal intake only as the study food options (Lays Potato Chips, Doritos, Cheez-its, Chips Ahoy Cookies, Oreos, M&Ms, Skittles) differ in macronutrients.
2.4. Statistical Analysis.
All data analyses were performed using SPSS version 30. Statistical significance was defined as p ≤ 0.05. Descriptive statistics, frequencies, and t-tests assessed demographic variables and differences based on weight status. A series of t-tests were run to compare RRVFood and dietary intake during the test-meal based on weight status. A series of linear regression analyses were run to assess whether RRVFood predicted total kcal, study food kcal, and macronutrient (grams of carbohydrates, fat, and protein) intake during the test meal. Since there is a well-established relationship between RRVFood and weight status, all models were run three times: with the whole sample, with participants with healthy weight, and with participants with OW/OB. The same series of t-tests and linear regression analyses were used to analyze the exploratory sex-based differences.
3.0. Results
Participants (n=112) had a mean age of 9.18 ±1.02 years. Slightly over half of the participants identified as male (n = 58; 52%), the remainder of participants identified as female. Most participants reported that they were White (n = 92; 82%) and not Hispanic or Latino (n = 98; 88%). Most participants classified as having healthy weight (HW) (n = 81; 72%), and the remainder of participants classified as having overweight or obesity (OW/OB) (n = 31; 28%). Detailed demographics and group differences based on weight status are shown in Table 2.
Table 2.
Participant demographics differences based on weight status.
| HW | OW/OB | ||||
|---|---|---|---|---|---|
| M | SD | M | SD | p | |
| Age | 9.10 | 1.02 | 9.39 | .989 | .180 |
| N | % | N | % | p | |
| Sex | .088 | ||||
| Male | 46 | 56.8 | 12 | 38.7 | |
| Female | 35 | 43.2 | 19 | 61.3 | |
| Race | .688 | ||||
| Asian | 1 | 1.2 | 0 | 0 | |
| Black or African American | 3 | 3.7 | 2 | 6.5 | |
| White or Caucasian | 68 | 84 | 24 | 77.4 | |
| Other or multiracial | 5 | 6.2 | 4 | 12.9 | |
| Unreported | 4 | 4.9 | 1 | 3.2 | |
| Hispanic or Latino: | .551 | ||||
| Hispanic or Latino | 3 | 3.7 | 2 | 6.5 | |
| Non-Hispanic or Latino | 71 | 87.7 | 27 | 87.1 | |
| Unreported | 7 | 8.6 | 2 | 6.5 | |
| Weight status | 81 | 72 | 31 | 28 | |
3.2. Boys had higher RRVFood and ate more total calories and protein in comparison to girls.
RRVFood log was larger in boys (M = 6.28, SD = 1.22) in comparison to girls (M = 5.39, SD = 1.91) [t (110) = 2.902, p = .004]. Boys also ate significantly more total calories (M = 918.57, SD = 428.37) and protein (M = 21.12, SD = 12.99) in comparison to girls (M = 756.25, SD = 434.73; M = 16.31, SD = 10.53) [t (110) = 1.989, ps < .049]. There were no other differences in dietary intake based on sex (Table 3).
Table 3.
T-tests for RRVFood log dietary intake differences based on sex.
| Boys | Girls | |||||
|---|---|---|---|---|---|---|
| Measure: | M | SE | M | SE | t(110) | p |
| RRVFood log | 6.28 | 1.22 | 5.39 | 1.91 | 2.946 | .004 |
| Study food kcal | 79.10 | 100.07 | 54.81 | 73.15 | 1.457 | .148 |
| Total kcal | 918.57 | 428.37 | 756.25 | 434.73 | 1.989 | .049 |
| Carbohydrates (g) | 131.11 | 60.55 | 109.19 | 68.69 | 1.794 | .076 |
| Fat (g) | 35.93 | 20.15 | 29.65 | 17.21 | 1.769 | .080 |
| Protein (g) | 21.12 | 12.99 | 16.31 | 10.53 | 2.141 | .034 |
3.3. There were limited differences in test-meal intake based on weight status.
There was no difference in RRVFood log between participants with HW (M = 5.95, SD = 1.67) and those with OW/OB (M = 5.58, SD = 1.56) [t (110) = 1.074, p = .285]. Participants with OW/OB ate significantly more protein (p = .019) compared to participants with HW (Table 4). There were no other differences in dietary intake between participants with HW and with OW/OB.
Table 4.
T-tests for dietary intake differences based on weight status.
| HW | OW/OB | |||||
|---|---|---|---|---|---|---|
| Measure of dietary intake: | M | SE | M | SE | t(110) | p |
| Study food kcal | 70.50 | 10.49 | 59.25 | 12.93 | .600 | .550 |
| Total kcal | 795.47 | 48.22 | 957.47 | 77.28 | −1.771 | .079 |
| Carbohydrates (g) | 115.28 | 7.12 | 134.28 | 12.10 | −1.385 | .169 |
| Fat (g) | 30.85 | 2.11 | 38.25 | 3.27 | −1.867 | .065 |
| Protein (g) | 17.15 | 1.28 | 23.11 | 12.58 | −2.389 | .019 |
3.4. RRVFood positively predicted total kcal, carbohydrates, and fat in the total sample and in participants with HW, but did not predict any intake in participants with OW/OB.
Total sample.
RRVFood log positively predicted total kcal (β = .232, p = .014), carbohydrates (β = .207, p = .028), and fat (β = .255, p = .021) intake during the test meal. RRVFood log did not predict study food kcal (β = .085, p = .376) and protein (β = .198, p = .077) intake during the test meal (Table 5).
Table 5.
Regression analyses for RRVFood log as a predictor for dietary intake during the test meal in the whole sample.
| Dependent Variables: | F | R | R2 | Adjusted R2 | SE | t(110) | p |
|---|---|---|---|---|---|---|---|
| Study food kcal | 0.791 | .085 | .007 | −.002 | 88.658 | .890 | .376 |
| Total kcal | 6.261 | .232 | .054 | .045 | 427.150 | 2.502 | .014 |
| Carbohydrates (g) | 4.948 | .207 | .043 | .034 | 64.113 | 2.224 | .028 |
| Fat (g) | 6.196 | .231 | .053 | .045 | 18.541 | 2.489 | .014 |
| Protein (g) | 3.400 | .173 | .030 | .021 | 11.929 | 1.844 | .068 |
Participants with HW.
RRVFood log positively predicted total kcal (β = .265, p = .017), carbohydrate (β = .245, p = .028), and fat (β = .231, p = .014) intake for participants with HW during the test meal. RRVFood log did not predict study food kcal (β = .198, p = .734) and protein (β = .173, p = .068) intake for participants with HW during the test meal (Table 6).
Table 6.
Linear regression of RRVFood log as a predictor of dietary intake consumed during the test meal for participants with HW.
| Dependent Variables: | F | R | R2 | Adjusted R2 | SE | t(80) | p |
|---|---|---|---|---|---|---|---|
| Study food kcal | .116 | .038 | .001 | −.011 | 94.897 | .340 | .734 |
| Total kcal | 5.958 | .265 | .070 | .058 | 421.169 | 2.441 | .017 |
| Carbohydrates (g) | 5.032 | .245 | .060 | .048 | 62.504 | 2.243 | .028 |
| Fat (g) | 5.507 | .255 | .065 | .053 | 18.453 | 2.347 | .021 |
| Protein (g) | 3.221 | .198 | .039 | .027 | 11.350 | 1.795 | .077 |
Participants with OW/OB.
RRVFood log did not predict study food kcal (β = .234, p = .206), total kcal (β = .223, p = .229), carbohydrate (β = .169, p = .365), fat (β = .250, p = .175), and protein (β = .214, p = .248) intake during the test meal in participants with OW/OB (Table 7).
Table 7.
Regression analyses for RRVFood log as a predictor for dietary intake during the test meal for participants with OW/OB.
| Dependent Variables: | F | R | R2 | Adjusted R2 | SE | t(30) | p |
|---|---|---|---|---|---|---|---|
| Study food kcal | 1.677 | .234 | .055 | .022 | 71.203 | 1.295 | .206 |
| Total kcal | 1.513 | .223 | .050 | .017 | 426.646 | 1.230 | .229 |
| Carbohydrates (g) | .848 | .169 | .028 | −.005 | 67.525 | .921 | .365 |
| Fat (g) | 1.937 | .250 | .063 | .030 | 17.927 | 1.392 | .175 |
| Protein (g) | 1.390 | .214 | .046 | .013 | 12.500 | 1.179 | .248 |
3.5. RRVFood positively predicted total kcal, carbohydrates, fat, and protein in girls, but did not predict any intake in boys.
RRVFood log positively predicted total kcal (β = .308, p = .023), carbohydrates (β = .268, p = .050), and fat (β = .328, p = .015), and protein (β = .279, p = .041) intake during the test meal in girls only. RRVFood log did not predict study food kcal (β = .163, p = .238) intake during the test meal in girls. RRVFood log did not predict study food kcal (β = −.064, p = .632), total kcal (β = .031, p = .816), carbohydrates (β = .015, p = .912), fat (β = .051, p = .703), and protein (β = −.039, p = .769) intake during the test meal in boys (Table 8).
Table 8.
Regression analyses for RRVFood log as a predictor for dietary intake based on sex.
| F | R | R2 | Adjusted R2 | SE | t(53) | p | |
|---|---|---|---|---|---|---|---|
| Girls | |||||||
| Study food kcal | 1.427 | .163 | .027 | .008 | 72.854 | 1.194 | .238 |
| Total kcal | 5.455 | .308 | .095 | .078 | 417.532 | 2.336 | .023 |
| Carbohydrates (g) | 4.026 | .268 | .072 | .054 | 66.811 | 2.007 | .050 |
| Fat (g) | 6.283 | .328 | .108 | .091 | 16.408 | 2.507 | .015 |
| Protein (g) | 4.392 | .279 | .078 | .060 | 10.207 | 2.096 | .041 |
| F | R | R2 | Adjusted R2 | SE | t(57) | p | |
| Boys | |||||||
| Study food kcal | .231 | .064 | .004 | −.014 | 100.750 | −.481 | .632 |
| Total kcal | .054 | .031 | .001 | −.017 | 431.969 | .233 | .816 |
| Carbohydrates (g) | .012 | .015 | .000 | −.018 | 61.080 | .111 | .912 |
| Fat (g) | .147 | .051 | .003 | −.015 | 20.305 | .703 | .703 |
| Protein (g) | .087 | .039 | .002 | −.016 | 13.094 | −.295 | .769 |
4.0. Discussion
The present study investigated the relationships among RRVFood and dietary intake at a buffet-style test meal in children ages 8-11. Given this population is considered at risk for developing obesity and associated health risks, assessing the relationships between RRVFood and dietary intake can identity relationships to target aiming to decrease these future risks to health. These relationships were assessed in the total sample and split by weight status (HW versus OW/OB). Furthermore, exploratory analyses assessed differences based on sex. The study found no difference in RRVFood based on weight status, which contrasts with previous literature (Temple et al., 2008b). One explanation for this finding is differences in the frequency of at-home study food consumption. As mentioned previously, to be eligible for this study, participants needed to consume their study food at least 2-3 times per week. However, some participants consumed their study food daily, whereas others consumed their study food only 2-3 times per week. Research by Temple et al., (2017) found that RRVFood decreased in adolescents after daily intake of the study food, regardless of the individual’s weight status (Temple et al., 2017). Frequency of consumption was reported by the parent during the screening questionnaire, but this was not reported beyond that or by the participating child. Including a measure of child-reported at-home consumption during the week leading up to the test meal could have provided varying results. Another potential for these different results may be the different populations studied. The current study assessed these relationships in children at risk of developing obesity, which is different from previous works (Epstein et al., 2009; Epstein et al., 2014; Temple et al., 2008b). Given that all participants are at risk for obesity, and high RRVFood has predicted future zBMI gain in non-overweight adolescents (Epstein et al., 2014), no difference in RRVFood based on weight status may not be surprising. Lastly, these differences could have been a result of potentially underpowered groups, confirming in a fully powered sample could provide better elucidation of any potential group differences based on weight status.
While there were no differences in RRVFood based on weight status, there were statistically significant differences based on sex. Boys had significantly higher RRVFood in comparison to girls. There are no known studies that have assessed sex-differences in RRVFood in humans, however these results are inconsistent with an animal study assessing RRVFood. A study by Tapia et al., (2019) found that female rats worked harder for sucrose pellets compared to the male rats. Given the present study had a variety of HED snack foods as the study food option, the results from Tapia et al., (2019), may not be translational as they only utilized sucrose pellets. This variability could have yielded these opposing results. The current study results suggest that boys, specifically within the age range 8-11 years old, are willing to work harder for a preferred HED snack food compared to girls within the same age range. This could be a result of diet culture; previous studies have shown that girls internalize diet culture at a younger age (Loews & Tiggemann, 2003) and girls overall exhibit more dieting behaviors compared to boys (Maloney et al., 1989; Rolls et al., 1991). These results suggest that boys are more food motivated than girls, however other variables, such as dieting behaviors, should be assessed as drivers in this relationship.
RRVFood predicted intake in the full sample and for participants with HW, but not in participants with OW/OB. In the full sample, RRVFood positively predicted total kcal, carbohydrates, and fat intake during the test meal. When assessed by weight status, RRVFood did not predict kcal or macronutrient intake in participants with OW/OB. This could be related to the differences in kcal and macronutrient intake based on weight status. Participants with OW/OB ate more kcal, carbohydrates, fat, and protein. While most of these differences were not statistically significant, when paired with lack of difference in RRVFood based on weight status, this increased intake could have muddled any results by not having standard intake distribution. Regarding participants with HW, RRVFood positively predicted total kcal, carbohydrates, and fat intake during the test meal. This indicates that the relationships seen in the full sample are likely driven by participants with HW. The relationship between RRVFood and kcal intake aligns with the previous results seen in adults (Epstein et al., 2011; Epstein et al., 2007b). However, in the study conducted by Epstein et al., (2011), they found that RRVFood did not predict carbohydrate or fat intake. Fat is not as satiating as protein, is highly palatable, and is related to weight gain when consumed in high amounts in a diet (Blundell et al., 1993; Mela & Sacchetti, 1991). Carbohydrates, particularly when refined, are also proposed to be related to the reward system (Carreiro et al., 2016). Thus, it is unsurprising to see a relationship between fat and carbohydrate intake and RRVFood. We may not see these results in adults because they are more proficient in intertemporal choice in comparison to children (Iknik et al., 2023). This means that they are better at assessing immediate satisfaction versus long-term benefits. Therefore, adults may show inhibition towards these highly palatable foods that children are less likely to exhibit.
Previous studies assessing the relationship between RRVFood and dietary intake did not assess the impact of weight status (Epstein et al., 2011). Based on these results, RRVFood is related to intake only for individuals with HW. This suggests that RRVFood may not be beneficial to intervene on as a means of reducing OW/OB, however these relationships should be assessed in a fully powered sample. However, with the strong relationships between RRVFood and kcal, carbohydrate, and fat intake in participants with HW, it may be beneficial to explore intervening as a form of OW/OB prevention. Since participants in this study had to be considered at risk for OW/OB, paired with the current obesogenic environment we live in, OW/OB prevention efforts should be studied and imposed in this developmental period. Given the previously found relationship between daily intake and RRVFood (Temple et al., 2008a), providing regular access to these HED snack foods to children with HW may decrease RRVFood, which in turn may decrease dietary intake. Thus, consistent access to these foods may be protective against the development of OW/OB in children with HW and high RRVFood. Fully powered data would better support these claims and thus should be studied in the future.
The sex-based exploratory analysis yielded strong results; RRVFood positively predicted total kcal, carbohydrates, fat, and protein during the test meal in girls only. There were no relationships between RRVFood and dietary intake in boys. These findings are novel and highlight a unique relationship between RRVFood, dietary intake, and sex. Despite boys having increased RRVFood, as well as increased overall kcal and protein intake, this did not translate to a direct relationship. Previous works have found that girls’ eating behaviors are more related to external motivators, whereas boys’ eating behaviors are more related to internal motivators (Deslippe et al., 2023). Internal motivators, such as enjoyment, may change based on day, whereas external motivators, such as avoiding excessive weight gain, may be consistent across periods of time. This may contribute to the differences as RRVFood and dietary intake during the test meal were measured one week apart at separate sessions, thus consistent motivators may yield more direct relationships compared to the inconsistent motivators. Furthermore, as previously discussed, girls internalize and exhibit higher levels of dieting behaviors at a younger age and more frequently when compared to boys (Loews & Tiggemann, 2003; Maloney et al., 1989; Rolls et al., 1991). This may yield that direct relationship that we see: girls who exhibit higher dieting behavior may not work as hard for the food and may show self-restriction when engaging in the test-meal. Future work should aim to elucidate further relationships between dieting behaviors, food motivation, and dietary intake.
There were some limitations to this study. As previously discussed, frequency of at-home study food intake was parent-reported, which may not be accurate as children within this age range have increased access to food choices outside of the home. Consumption frequency has a proposed role in RRVFood (Temple, 2017), thus inaccurate rates of consumption may have impacted RRVFood. This could have also impacted kcal intake of RRVFood. One strength of the study is that all children were given the same test meals under the same conditions, so dietary intake is comparable across participants. While laboratory dietary data ensures accurate values, are highly controlled, and can be accurately compared across participants, the eating conditions are generally atypical to real life. Specifically, the test meal takes place with the child alone and in a relatively novel environment. This may mean that their intake does not accurately reflect standard intake. Future studies would benefit from including naturalistic dietary intake and eating behaviors outside of a laboratory setting. Lastly, there was an unequal number of participants with HW and participants with OW/OB, which decreased power and statistical significance. Future studies should be conducted to confirm group differences and relationships in a larger, fully-powered sample. Reassessing with more balanced groups or further splitting between participants with overweight and participants with obesity could provide better insight into the implications of RRVFood on eating behaviors and child health.
5.0. Conclusion
In the current study, RRVFood did not differ based on weight status, which is in contrast with previous literature. This could be based a few reasons: differences in at-home intake of the study food, the specific sample being studied, and unbalanced group sizes. Despite previous works stating that children with OW/OB have higher RRVFood, these data suggest otherwise, but these results should be studied in a fully powered sample. Moving forward, these results could be utilized to adjust future investigative approaches on RRVFood and childhood OW/OB. RRVFood positively predicted kcal, carbohydrates, and fat in participants only with HW. This suggests that these relationships do not exist in participants with OW/OB. This could be related to intake differences during the test meal, thus future work should be powered to assess the differences between those with HW, OW, and OB. Exploratory analyses assessing sex-differences yielded novel results: boys overall had a higher and total kcal and protein intake in comparison to girls. However, RRVFood positively predicted total kcal, carbohydrates, protein, and fat in girls only. These results could be due to a biological difference not yet assessed by previous researchers, or it could be related to the impact of diet culture on girls. These findings suggest that girl’s dietary intake is more related to RRVFood thus they should be targeted for intervention efforts as a means of OW/OB prevention. Future work should aim to assess these relationships further in more natural settings, include other measures, and explore interventions aiming at preventing the development of OW/OB in children with HW.
Funding sources
This work was supported by the National Institute of Kidney and Digestive Diseases (grant number: K01 DK120778 awarded to KNB). The funding source had no role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Footnotes
CRediT authorship contribution statement
Kolpack: Writing – original draft, formal analysis, investigation, project administration; Hatzinger: Writing – review & editing, data processing, investigation, project administration; Temple: Writing – review & editing; Wilding: Writing – review & editing; Balantekin: Writing – review & editing, supervision, resources, project administration, methodology, investigation, funding acquisition, conceptualization.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Research Data
Data have been published on clinicaltrials.gov: https://clinicaltrials.gov/study/NCT05177705 .
Human subjects
All procedures were performed in compliance with relevant laws and institutional guidelines. All procedures were approved by the University at Buffalo Institutional Review Board (STUDY00002806, approved continuously 11/18/18 though 10/2/25). In addition, privacy rights of human subjects have been observed and informed consent was obtained for experimentation with human subjects.
Informed consent and patient details
Written informed consent to take part in the study and to publish the article has been obtained from all participants or their legal representatives. The privacy rights of participants have been observed.
Studies in Human
This study was performed in compliance with relevant laws, regulatory frameworks and guidelines where the research took place.
This study was approved by the University at Buffalo.
(Approval No. STUDY00002806)
Clinical Trials
The results of this clinical trial and any associated work have been posted in a registry.
This clinical trial was registered with number NCT05177705 (Clinicaltrials.gov).
Ethical statement
This study was conducted according to the guidelines established by the institution and the Declaration of Helsinki. The Institutional Review Board at the University at Buffalo approved the study STUDY00002806. Parent / legal guardians provided written informed consent and children provided assent prior to any data collection.
Declaration of competing interest
All authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper.
Declaration of interests
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:
Katherine Balantekin reports financial support was provided by National Institute of Diabetes and Digestive and Kidney Diseases. Given her role as Deputy Editor-in-Chief, Jennifer Temple had no involvement in the peer review of this article and had no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to another journal editor. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data availability
Data will be available upon request to the corresponding author.
References
- Ambrosini GL, Emmett PM, Northstone K, & Jebb SA. (2014). Tracking a Dietary Pattern Associated with Increased Adiposity in Childhood and Adolescence. Obesity, 22(2), 458–465. 10.1002/oby.20542 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Blundell JE, Burley VJ, Cotton JR, & Lawton CL. (1993). Dietary fat and the control of energy intake: evaluating the effects of fat on meal size and postmeal satiety. American Journal of Clinical Nutrition, 57(5 Suppl), 772S–777S; discussion 777S–778S. 10.1093/ajcn/57.5.772S [DOI] [PubMed] [Google Scholar]
- Carr KA, Black W, Guth C, Shapiro L, Leone LA, Temple JL, & Epstein LH. (2023). Reinforcing value of food, enriched home environment, and changes in percent overweight in children. Obesity (Silver Spring), 31(4), 1075–1084. 10.1002/oby.23685 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carr KA & Epstein LH. (2011). Relationship between food habituation and reinforcing efficacy of food. Learning and Motivation, 42(2), 165–172. 10.1016/j.lmot.2011.01.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Carreiro AL, Dhillon J, Gordon S, Higgins KA, Jacobs AG, McArthur BM, Redan BW, Rivera RL, Schmidt LR, & Mattes RD. (2016). The Macronutrients, Appetite, and Energy Intake. Annual Review of Nutrition, 36, 73–103. 10.1146/annurev-nutr-121415-112624 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Costa MPD, Severo M, Araújo J, & Vilela S. (2024). Longitudinal tracking of diet quality from childhood to adolescence: The Interplay of individual and sociodemographic factors. Appetite, 196. 10.1016/j.appet.2024.107279 [DOI] [PubMed] [Google Scholar]
- Daniels DR. (2004). The Consequences of Childhood Overweight and Obesity. The Future of Children 16(1), 47–67. 10.1353/foc.2006.0004 [DOI] [PubMed] [Google Scholar]
- Deslippe AL, Bergeron C, and Cohen TR. (2023) Boys and girls differ in their rationale behind eating: a systematic review of intrinsic and extrinsic motivations in dietary habits across countries. Frontiers in Nutrition. 10:1256189. 10.3389/fnut.2023.1256189 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dietz WH. (1998). Health consequences of obesity in youth: Childhood predictors of adult disease. Pediatrics, 101(3), 518–525. <Go to ISI>://WOS:000072362900004 [PubMed] [Google Scholar]
- Epstein LH, Carr KA, Lin H, & Fletcher KD. (2011). Food reinforcement, energy intake, and macronutrient choice. American Journal of Clinical Nutrition, 94(1), 12–18. 10.3945/ajcn.110.010314 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein LH, Leddy JJ, Temple JL, & Faith MS. (2007a). Food reinforcement and eating: A multilevel analysis. Psychological Bulletin, 133(5), 884–906. 10.1037/0033-2909.133.5.884 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein LH, Robinson JL, Temple JL, Roemmich JN, Marusewski A, & Nadbrzuch R. (2008). Sensitization and habituation of motivated behavior in overweight and nonoverweight children. Learning and Motivation, 39(3), 243–255. 10.1016/j.lmot.2008.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein LH, Robinson JL, Temple JL, Roemmich JN, Marusewski A, & Nadbrzuch R. (2009). Variety influences habituation of motivated behavior for food and energy intake in children. American Journal of Clinical Nutrition, 89(3), 746–754. doi: 10.3945/ajcn.2008.26911 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein LH, Temple JL, Neaderhiser BJ, Salis RJ, Erbe RW, & Leddy JJ. (2007b). Food reinforcement, the dopamine D2 receptor genotype, and energy intake in obese and nonobese humans. Behavioral Neuroscience, 121(5), 877–886. 10.1037/0735-7044.121.5.877 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Epstein LH, Yokum S, Feda DM, & Stice E. (2014). Food reinforcement and parental obesity predict future weight gain in non-obese adolescents. Appetite, 82, 138–142. 10.1016/j.appet.2014.07.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Friedlander SL, Larkin EK, Rosen CL, Palermo TM, & Redline S. (2003). Decreased quality of life associated with obesity in school-aged children. Archives of Pediatrics & Adolescent Medicine, 157(12), 1206–1211. 10.1001/archpedi.157.12.1206 [DOI] [PubMed] [Google Scholar]
- Heo M, Gabriel KR. (1998). A permutation test of association between configurations by means of the RV coefficient. Communication in Statistics Simulations 27 (3), 843–856. [Google Scholar]
- Ikink I, van Duijvenvoorde ACK, Huizenga H, Roelofs K, & Figner B. (2023). Age differences in intertemporal choice among children, adolescents, and adults. Journal of Experimental Child Psychology, 233, 105691. 10.1016/j.jecp.2023.105691 [DOI] [PubMed] [Google Scholar]
- Knittle JL, Timmers K, Ginsbergfellner F, Brown RE, & Katz DP. (1979). Growth of Adipose-Tissue in Children and Adolescents - Cross-Sectional and Longitudinal-Studies of Adipose Cell Number and Size. Journal of Clinical Investigation, 63(2), 239–246. 10.1172/Jci109295 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lappalainen R, & Epstein LH. (1990). A behavioral economics analysis of food choice in humans. Appetite, 14(2), 81–93. 10.1016/0195-6663(90)90002-p [DOI] [PubMed] [Google Scholar]
- Lartey ST, Jayawardene WP, Dickinson SL, Chen X, Gletsu-Miller N, & Lohrmann DK. (2023). Evaluation of Unintended Consequences of COVID-19 Pandemic Restrictions and Obesity Prevalence Among Youths. Jama Network Open, 6(7), e2323596. 10.1001/jamanetworkopen.2023.23596 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lowes J, & Tiggemann M. (2003). Body dissatisfaction, dieting awareness and the impact of parental influence in young children. British journal of health psychology, 8(Pt 2), 135–147. 10.1348/135910703321649123 [DOI] [PubMed] [Google Scholar]
- Maloney MJ, McGuire J, Daniels SR, & Specker B. (1989). Dieting behavior and eating attitudes in children. Pediatrics, 84(3), 482–489. [PubMed] [Google Scholar]
- Mela DJ, & Sacchetti DA. (1991). Sensory preferences for fats: relationships with diet and body composition. American Journal of Clinical Nutrition, 53(4), 908–915. 10.1093/ajcn/53.4.908 [DOI] [PubMed] [Google Scholar]
- Reilly JJ, Methven E, McDowell ZC, Hacking B, Alexander D, Stewart L, Kelnar CJH. (2003). Health consequences of obesity. Achieves of Disease in Childhood, 88, 748–752. 10.1136/adc.88.9.748 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rolls BJ, Fedoroff IC, & Guthrie JF. (1991). Gender differences in eating behavior and body weight regulation. Health psychology : official journal of the Division of Health Psychology, American Psychological Association, 10(2), 133–142. 10.1037//0278-6133.10.2.133 [DOI] [PubMed] [Google Scholar]
- Russell-Mayhew S, McVey G, Bardick A, & Ireland A. (2012). Mental health, wellness, and childhood overweight/obesity. Journal of obesity, 2012(1), 281801. 10.1155/2012/281801 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sahoo B, Bhadoria A, Sahoo K, Kumar R, Choudhury A, & Sofi N. (2015). Childhood obesity: causes and consequences. Journal of Family Medicine and Primary Care, 4(2), 187–192. 10.4103/2249-4863.154628 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stierman B, Afful J, Carroll MD, Chen TC, Davy O, Fink S, Fryar CD, Gu Q, Hales CM, Hughes JP, Ostchega Y, Storandt RJ, & Akinbami LJ. (2021). National Health and Nutrition Examination Survey 2017-March 2020 Prepandemic Data Files-Development of Files and Prevalence Estimates for Selected Health Outcomes. National Health Statistics Report(158). 10.15620/cdc:106273 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tapia MA, Lee JR, Weise VN, Tamasi AM, & Will MJ. (2019). Sex differences in hedonic and homeostatic aspects of palatable food motivation. Behavioural Brain Research (359), 396–400. 10.1016/j.bbr.2018.11.023. [DOI] [PubMed] [Google Scholar]
- Temple JL, Chappel A, Shalik J, Volcy S, & Epstein LH. (2008a). Daily consumption of individual snack foods decreases their reinforcing value. Eating Behaviors, 9(3), 267-276. 10.1016/j.eatbeh.2007.10.001 [DOI] [PubMed] [Google Scholar]
- Temple JL, & Epstein LH. (2012). Sensitization of food reinforcement is related to weight status and baseline food reinforcement. International Journal of Obesity, 36(8), 1102–1107. 10.1038/ijo.2011.210 [DOI] [PubMed] [Google Scholar]
- Temple JL, Giacomelli AM, Roemmich JN, & Epstein LH. (2007). Overweight children habituate slower than non-overweight children to food. Physiology & Behavior, 91(2-3), 250–254. 10.1016/j.physbeh.2007.03.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Temple JL, Legierski CM, Giacomelli AM, Salvy SJ, & Epstein LH. (2008b). Overweight children find food more reinforcing and consume more energy than do nonoverweight children. American Journal of Clinical Nutrition, 87(5), 1121–1127. 10.1093/ajcn/87.5.1121 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Temple JL, Van der Kloet E, Atkins AM, Crandall AK, & Ziegler AM. (2017). Daily exposure to either a high- or low-energy-dense snack food reduces its reinforcing value in adolescents. Obesity (Silver Spring), 25(2), 432–437. 10.1002/oby.21730 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tey SL, Brom RC, Gray AR, Chisholm AW, & Delahunty CM. (2012). Long-term consumption of high energy-dense snack foods on sensory-specific satiety and intake. American Journal of Clinical Nutrition, 95(5), 1038–1047. 10.3945/ajcn.111.030882 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data will be available upon request to the corresponding author.