Abstract
Purpose
This study was designed to examine (1) the types of technologies or other applications (apps) emerging adults use to track their eating, physical activity, or weight; (2) who uses these apps and (3) whether eating and weight-related concerns during adolescence predict app use in emerging adulthood.
Methods
Longitudinal survey data were obtained from EAT 2010–2018 (Eating and Activity over Time study, N = 1428), a population-based sample of ethnically/racially and socioeconomically diverse adolescents (mean age: 14.5 ± 2.0 years), who were followed into emerging adulthood (mean age: 22.0 ± 2.0 years). Data were used to examine sociodemographic correlates of physical activity- and dietary-focused app use. Adjusted, gender-stratified logistic regressions were used to investigate longitudinal relationships between eating and weight-related concerns in adolescence and app use in emerging adulthood.
Results
Compared to men, women were more likely to use physical activity- (23.2 versus 12.5%, p < 0.001) and dietary-focused apps (16.1 versus 5.5%, p < 0.001). Among women, eating and weight-related concerns in adolescence, particularly unhealthy muscle-building behaviors (OR = 1.73, 95% CI 1.03–2.92), were associated with later dietary-focused app use. Among men, use of other muscle-building behaviors and body dissatisfaction in adolescence predicted use of physical activity- (ORother muscle-building = 1.60, 95% CI 1.03–2.49 and ORbody dissatisfaction = 1.67, 95% CI 1.06–2.65) and dietary-focused (ORother muscle-building = 2.18, 95% CI 1.07–4.47 and ORbody dissatisfaction = 2.35, 95% CI 1.12–4.92) apps 8 years later.
Conclusions
Eating and weight-related concerns may predict later use of physical activity- and dietary-focused apps; future research is needed to understand whether use of such apps further increases eating and weight-related concerns.
Level of evidence
III, well-designed longitudinal cohort study.
Keywords: Self-monitoring, Eating behaviors, Physical activity, Eating disorders, Epidemiology, Prevention
Introduction
Weight-related self-monitoring (WRSM) involves tracking one’s weight or behaviors that may affect weight, such as physical activity or dietary intake. Technologies and smartphone applications (apps) designed to help individuals engage in WRSM have gained popularity in recent years and amassed millions of users. Wearable and all-in-one fitness trackers such as Fitbit have over 27 million active users a month and popular dietary-focused self-monitoring apps have nearly 20 million active users a month in the USA alone [1–4]. The popularity of WRSM apps is likely due to increased accessibility and ease of tracking that WRSM apps offer, high levels of weight and health concerns among the general population, and recommendations or incentives for WRSM by health-care professionals [5–7]. Prior research has suggested that emerging adults (ages: 18–29 years) are the most frequent users of such WRSM apps, but little is known about what types of apps emerging adults use or what demographic subgroups are most likely to use WRSM apps, particularly outside of college samples [8, 9].
Understanding who uses WRSM apps among the general population of emerging adults is important, given that the use of WRSM apps has been linked to both positive and negative health outcomes in this age range [10, 11]. WRSM is a key component in behavioral weight loss interventions and has been shown to lead to weight maintenance over time in structured weight management programs [12]. Notably, evidence suggests that use of technology-based WRSM yields greater weight loss and dietary quality compared to traditional forms of WRSM such as using paper food logs [13, 14]. However, it is unknown whether using WRSM apps outside of clinical programs, where there are beneficial components including social support and assistance from clinicians, yields the same effects on health behaviors [15]. Cross-sectional studies among non-clinical samples of college students have shown that dietary self-monitoring is associated with intake of a more nutrient-dense diet and that physical activity self-monitoring is associated with higher motivation for physical activity [10, 16, 17]. However, WRSM has also been linked to negative health outcomes when used outside of structured weight management programs. For example, among non-clinical samples of college students, WRSM app use has been cross-sectionally associated with elevated eating disorder risk, including higher rates of disordered eating and problematic alcohol use [11, 18, 19]. Given observed associations between WRSM app use and eating disorder risk among non-clinical samples, high incidence and prevalence of eating disorders and disordered eating in this age range [20–23], and the severe consequences of eating disorders [24], it is crucial to identify populations that may be more likely to use WRSM apps during emerging adulthood to identify vulnerable populations at risk of developing eating disorders.
Because of the cross-sectional nature of existing literature between WRSM and eating disorder risk, it is unclear which comes first; it is possible that WRSM contributes to eating disorder risk or that those who have elevated eating disorder risk use WRSM because of their elevated eating and weight-related concerns, or potentially both. If eating and weight-related concerns precede the use of WRSM apps, it may explain the cross-sectional findings between WRSM app use and eating disorder risk, as well as other eating and weight-related outcomes. If WRSM app use precedes eating and weight-related concerns, then WRSM may be a by-product or maintenance factor for eating disorder risk. Specifically, those who are interested in improving their health may opt using WRSM apps to improve nutritional intake or physical activity, while those already using unhealthy weight control behaviors (e.g., fasting) may begin using WRSM apps in an effort to gain further control, which can maintain or exacerbate their eating disorder risk. However, if elevated eating and weight-related concerns do not precede WRSM app use, it may provide further support to the theory that WRSM contributes to elevated eating disorder risk. Therefore, understanding whether eating and weight-related concerns precede the use of WRSM apps is an important first step in understanding who uses WRSM apps and why, as well as understanding the temporality of relationships between eating and weight concerns and WRSM app use. Further, because eating and weight-related concerns are more common among individuals with higher BMIs, and WRSM is often recommended clinically for weight loss among those with higher BMIs, we wanted to explore the extent to which BMI may explain associations between eating and weight-related concerns and later WRSM app use [25–27].
Given the widespread use of WRSM apps and known associations between WRSM app use and health outcomes such as dietary intake and eating disorder risk in emerging adulthood [10, 11, 18], it is important to understand who uses WRSM apps and what predicts WRSM app use. Therefore, the present study aims to understand (1) what types of WRSM apps emerging adults used to manage their eating, physical activity, and weight, (2) who uses WRSM apps, and (3) whether eating and weight-related concerns in adolescence predicted the use of WRSM apps in emerging adulthood. We hypothesized that measures of eating and weight-related concerns in adolescence would predict use of WRSM apps in emerging adulthood; we also aimed to explore the role of BMI in observed associations. Study findings will help inform future research examining the potential consequences and implications of WRSM app use among non-clinical samples, as well as inform public health recommendations regarding engagement with WRSM apps among the general population.
Methods
Study design and sample
Longitudinal data were collected as part of the EAT 2010–2018 (Eating and Activity over Time) study, a population-based investigation of eating and weight-related health behaviors and associated factors in a diverse sample of young people. Participants enrolled in the EAT 2010 study were adolescents (mean age = 14.4 ± 2.0 years) during the 2009–2010 academic year and completed 8-year follow-up EAT 2018 surveys as emerging adults (mean age = 22.0 ± 2.0 years). For EAT 2010 data collection, middle and senior high school students at 20 urban public schools in Minneapolis-St. Paul, Minnesota completed classroom surveys and anthropometric measures in a private area of their school. Invitations to participate in the online EAT 2018 survey were mailed to baseline participants. To further encourage participation, non-responders were mailed up to eight reminders and additional contact attempts were made using email, phone calls, text messages, messaging through social media, and home visits. The response rate at follow-up was 65.8% among participants with available contact information (n = 1568). The sample for the present study included those with data at both 2010 and 2018 who were identified as the same gender at both time points, were not high school students at 2018, and were not missing data for covariates (n = 1428). All study protocols were approved by the University of Minnesota Institutional Review Board Human Subjects Committee.
To account for missing data, inverse probability weighting was used for all analyses [28, 29]. Inverse probability weighting is the recommended method for handling missing data in longitudinal studies, where individuals who do not respond to surveys at various assessment time points or have missing values on many variables. Inverse probability weighting minimizes potential response bias due to missing data and allows for extrapolation of EAT 2018 data back to the original EAT 2010 school-based sample. Participants who were identified as male, non-white, and having parents with lower educational attainment at 2010 were less likely to respond to the EAT 2018 survey. Weights for inverse probability weighting were therefore derived as the inverse of the estimated probability that an individual responded to the EAT 2018 survey based on several characteristics reported in 2010, including demographics, past year frequency of dieting, and weight status. After weighting, there were no significant differences between the EAT 2018 sample and the full EAT 2010 sample on demographic characteristics, dieting, or weight status (p > 0.9). In the final weighted sample for the present study, 54% were female and 46% were male. Of the final weighted sample, 28.4% were identified as African American or Black, 20.3% White, 20.0% Asian American, 17.1% Hispanic/Latinx, and 14.2% mixed or other ethnicity/race”. Socioeconomic status (SES) was based primarily on baseline parental educational attainment, with 38.9% of participants with low SES, 40.2% were of low-middle/middle SES, and 21.0% were of middle–high/high SES. Because participants were under 18 years of age at baseline but over 18 at follow-up, BMI was assessed via BMI percentiles in adolescence and BMI in emerging adulthood. The average BMI percentile in adolescence was 69.1 (standard deviation [SD] = 27.7%) and the average BMI at emerging adulthood was 27.2 (SD = 7.0).
Measures
The EAT 2018 survey retained key items from the EAT 2010 survey and included several additional items to address areas of new interest [30–33]. In particular, questions pertaining to WRSM apps were added in the 2018 survey and were defined as smartphone applications or other technologies participants used to manage their weight or behaviors that may affect weight, such as eating and physical activity, by tracking their behaviors (i.e., self-monitoring). Focus groups with 29 emerging adults were used to pretest the EAT 2018 survey. The test–retest reliability of measures was assessed at each time point; at baseline (EAT 2010), test–retest reliability of survey measures was assessed in a diverse sample of 129 adolescents. At follow-up (EAT 2018), test–retest reliability was assessed in a subgroup of 112 emerging adult participants. All measures used in the current analysis can be found in Table 1 with psychometrics, such as test–retest reliability, reported for the relevant year that the measure was assessed.
Table 1.
Measures used in the current study examining eating and weight concern and WRSM app use
| Measure | Description | Times collected |
|---|---|---|
| Weight-related self-monitoring (WRSM) | ||
| Application/technology (App) use | Weight-related self-monitoring (WRSM) app use was assessed using an open-ended question, “Thinking about the mobile apps, tracker devices, and web-based programs you’ve used in the past year to help you manage your eating, activity or weight, please list up to 3 and tell us how often you currently use them.” Participants could write in up to three technologies/applications (apps), and for each indicated how often they used them with response options of, “Never”, “Rarely”, “Sometimes”, and “Often”. WRSM apps written by participants were then coded by a researcher based on the specific WRSM app functionality and the categorizations were reviewed by a second coder. Nine distinct types of WRSM apps were reported by more than one participant: smartphone/smartwatch all-in-one tracker (automatically tracks physical activity and allows for tracking of other weight-related factors, e.g., FitBit, Samsung Health), physical activity/fitness specific apps (e.g., Nike Run, 7 min workout), apps for diet and physical activity combined (e.g., MyFitness-Pal, 30-day challenge), diet or weight loss-specific apps (e.g., MyPlate, LOSE IT!), social media (e.g., Facebook, YouTube), body building apps (e.g., BodyBuilder), gamification or incentive using apps (e.g., Sweatcoin), water intake tracking apps, and all others were classified as miscellaneous. Each type of app use was dichotomized based on any use Two derived variables were then created: (1) physical activity-focused apps and (2) dietary-focused apps. Physical activity-focused apps included any use of all-in-one trackers which automatically record physical activity, or physical activity/fitness specific apps. Dietary-focused apps included apps that allowed for tracking of both diet and physical activity, but have an emphasis on diet and do not automatically track physical activity (e.g., MyFitnessPal), and those for diet or weight loss only. The two derived variables were not mutually exclusive, and each compared users to non-users |
EAT 2018 |
| Eating and weight-related concerns | ||
| Unhealthy weight control behaviors | Unhealthy weight control behaviors were assessed with the question, “Have you done any of the following things in order to lose weight or keep from gaining weight during the past year?” [41]. Assessed behaviors included fasting, eating very little food, using a food substitute, skipping meals, smoking more cigarettes, taking diet pills, vomiting, abusing laxatives, and using diuretics. Participants who endorsed one or more of the behaviors were considered to have engaged in unhealthy weight control behaviors (test-retest agreement = 85%) | EAT 2010 |
| Unhealthy muscle-building behaviors | Unhealthy muscle-building behaviors were assessed with a question adapted from previous studies [42–44], “How often have you done each of the following things in order to increase your muscle size or tone during the past year?” Assessed behaviors included using protein powder or shakes, using steroids, or using other muscle-building substances (such as creatine or growth hormone), and for each behavior response options included “never”, “rarely”, “sometimes”, and “often”. Participants who responded that they used any of the assessed behaviors rarely or more were considered to have engaged in unhealthy muscle-building behaviors. Protein powders or shakes were considered unhealthy muscle-building behaviors based on evidence that they are frequently contaminated by other dangerous substances and may lead to more extreme muscle-building behaviors, such as taking steroids [45, 46] | EAT 2010 |
| Binge eating with loss of control | Binge eating with loss of control was assessed with the following two questions from the Questionnaire on Eating and Weight Patterns-Revised [47]: “In the past year, have you ever eaten so much food in a short period of time that you would be embarrassed if others saw you (binge-eating)?” (test-retest agreement = 90%) and, “During the times when you ate this way, did you feel you couldn’t stop eating or control what or how much you were eating?” (test-retest agreement = 75%). Both were yes/no questions, and participants who endorsed both constructs were considered to have binge eaten with loss of control | EAT 2010 |
| Compulsive exercise | Compulsive exercise was assessed using three questions from the Obligatory Exercise Questionnaire [48]. The three questions were: 1) “When I miss a scheduled exercise session, I may feel tense, irritable, or depressed” 2) “If I feel I have overeaten I will try to make up for it by increasing the amount I exercise” and 3) “When I don’t exercise, I feel guilty”. Response options ranged from 0 = never to 3 = always, scores from the three questions were summed with higher scores indicating higher levels of compulsive exercise (test-retest r = 0.81). Scores were dichotomized with scores above the median being considered positive for compulsive exercise | EAT 2010 |
| Recommended weight control behaviors | Behaviors recommended by the Centers for Disease Control for weight management were included in this construct [49]. These behaviors were assessed with the following question, “How often have you done each of the following things in order to lose weight or keep from gaining weight in the past year?” The four assessed behaviors were exercising, eating more fruit and vegetables, eating less high-fat foods, and eating less sweets. Response options for each of the behaviors were, “Never”, “Rarely”, “Sometimes”, and “On a regular basis”. Individuals who reported using any of the four behaviors on a regular basis were considered to have regularly engaged in recommended weight control behaviors [50] | EAT 2010 |
| Other muscle-building behaviors | Other muscle-building behaviors were assessed with the same parent question as unhealthy muscle-building behaviors: “Have you done any of the following things in order to increase your muscle size or tone during the past year?”. Individuals who endorsed changing their eating (test-retest r = 0.60, p < 0.001) or exercising (test-retest r = 0.64, p < 0.001) often were considered to have used other muscle-building behaviors | EAT 2010 |
| Dieting | Dieting in the last year was assessed with the following question, “How often have you gone on a diet during the last year? By “diet” we mean changing the way you eat so you can lose weight”. Possible response options included, “Never”, “1–4 times”, “5–10 times”, “More than 10 times”, and “I am always dieting”. The variable was dichotomized to group those who had dieted one or more times in the last year and those who had not dieted in the last year (test-retest agreement = 82%) | EAT 2010 |
| Body dissatisfaction | Body dissatisfaction was assessed with the question, “How satisfied are you with your:” with sub-questions for various body parts. The present variable assessed satisfaction of “weight” (test-retest r = 0.62, p < 0.001) and “body shape” (test-retest r = 0.60, p < 0.001). There were five response options for both weight and body shape on a scale that was anchored with “very dissatisfied” (1) and “very satisfied” (5). Those who responded, “very dissatisfied” (1) or (2) for weight and/or shape were considered body dissatisfied | EAT 2010 |
| Perceived overweight | Perceived overweight was assessed using the question, “At this time, do you feel you are:” with response options of “very underweight”, “somewhat underweight”, “about the right weight”, “somewhat overweight” and “very overweight”. Those who endorsed feeling either somewhat or very overweight were considered to perceive themselves as overweight (test-retest agreement = 90%) | EAT 2010 |
| BMI ≥ 85th percentile | BMI ≥ 85th percentile was determined using baseline age and sex-specific body mass index (BMI) percentiles based on Centers for Disease Control and Prevention growth charts [51]. Percentiles were calculated using anthropometric data assessed by trained staff following standardized procedures [52] | EAT 2010 |
| Covariates | ||
| Sociodemographic characteristics | Structurally racialized categories labeled as ethnicity/race (test-retest agreement range 98–100%) and socioeconomic status (SES; test-retest r = 0.90, p < 0.001) were reported at baseline and included in all models [53–55]. Gender (test-retest agreement: 99%), and age (test-retest r = 0.99, p < 0.001) were assessed at both time points. In the 2018 follow-up survey, participants self-reported their current student status (test-retest agreement = 92%) and educational attainment (r = 0.89,p < 0.001) | EAT 2010: ethnicity/race, SES, age, gender EAT 2018: age, gender, educational/student status |
| BMI | BMI percentile was determined using height and weight measured by trained staff and calculated according to CDC growth charts in 2010. At 2018, BMI was assessed using self-reported height and weight | EAT 2010 and 2018 |
Test–retest agreement was calculated in a subsample of participants at the relevant time point. Agreement was calculated as the percentage of participants who reported the same response on the EAT survey and the retest survey approximately 2 weeks later
Statistical analysis
Analyses were stratified by gender due to previously observed gender differences in associations between WRSM and eating disorder symptoms [9]. Univariate and bivariate statistics were conducted for all WRSM app use outcome variables. To examine associations between sociodemographic variables and derived WRSM app use variables, we utilized Chi-square analyses. Chi-square analysis tested whether prevalence of use for the respective type of WRSM app differed by demographics. Because types of WRSM app use were not mutually exclusive, independent Chi-square tests were run for both physical activity- and dietary-focused app use. The longitudinal relationships were examined using logistic regressions adjusting for ethnicity/race, baseline parental SES, baseline age, and educational/student status at follow-up. Baseline/adolescent SES was included in all models, as SES in adolescence is known to be a predictor of health outcomes in adulthood, irrespective of adult experiences [34]. Models were run with and without adjustment for adolescent BMI to better understand the role of BMI in the longitudinal associations between eating and weight-related concerns and WRSM [11, 35, 36]. Analyses were conducted using SAS version 9.4 and results were considered statistically significant at p < 0.05.
Results
Types of WRSM apps used by emerging adults
Use of any WRSM apps was more common among women (31.7%) than men (20.1%, p < 0.001) (Table 2). All-in-one trackers, or devices that automatically track physical activity and allow for tracking of other weight-related factors (e.g., Samsung Health or wearable devices with accompanying apps like Fitbit), were the most commonly used app (14.7%), followed by physical activity/fitness (e.g., exercise video apps) (7.8%), diet/physical activity combined (e.g., MyFitnessPal) (6.0%), and diet/weight loss (e.g., MyPlate) (4.3%) apps. All-in-one trackers, diet/physical activity combined apps, and diet/weight loss apps were more commonly used by women than men, but there was no significant difference by gender for physical activity/fitness apps. All other WRSM apps were used by less than 5% of men and women.
Table 2.
Prevalence of each type of weight-related self-monitoring (WRSM) during emerging adulthood overall and by gender
| Overall (n = 1428) | Women (n = 836) | Men (n = 592) | p value | |
|---|---|---|---|---|
| % (n) | % (n) | % (n) | ||
| Any app use | 26.4 (393) | 31.7 (268) | 20.1 (125) | < 0.001 |
| Individual app usea | ||||
| Smartphone/smartwatch all-in-one tracker | 14.7 (222) | 16.8 (143) | 12.3 (79) | 0.020 |
| Physical activity/fitness | 7.8 (115) | 8.8 (76) | 6.7 (39) | 0.136 |
| Diet/physical activity | 6.0 (91) | 7.9 (67) | 3.6 (24) | < 0.001 |
| Diet/weight loss | 4.3 (61) | 5.7 (46) | 2.6 (15) | 0.005 |
| Social media | 1.7 (24) | 2.1 (17) | 1.2 (7) | 0.228 |
| Body building | 0.7 (8) | 0.3 (3) | 1.0 (5) | 0.122 |
| Gamification/incentives | 0.5 (8) | 0.9 (7) | 0.1 (1) | 0.052 |
| Water intake tracking | 0.5 (6) | 0.5 (4) | 0.4 (2) | 0.776 |
| Miscellaneous | 1.2 (16) | 1.3 (10) | 1.1 (6) | 0.780 |
| App use by typeb | ||||
| Physical activity-focused apps | 20.0 (299) | 23.2 (199) | 16.1 (100) | < 0.001 |
| Dietary-focused apps | 9.3 (140) | 12.5 (135) | 5.5 (35) | < 0.001 |
Percentage is weighted to account for attrition over time and allow for extrapolation to the original population-based sample, while n represents observed count. Accordingly, ns may not directly match weighted percentages. Bold font indicates p < .05
Types of app use are not mutually exclusive. Description of app types can be found in Table 1
Categorized app use based on type. Collapsed types of app use were not mutually exclusive
When examining types of WRSM app use, 20.0% of participants were found to use a physical activity-focused app (all-in-one tracker and/or physical activity/fitness), with more women (23.2%) than men (16.1%, p < 0.001) reporting use. Dietary-focused app use (diet/physical activity combined and/or diet/weight loss) was less common than physical activity-focused app use in the overall sample (9.3%), but was similarly more common among women (12.5%) than men (5.5%, p < 0.001).
Who uses WRSM apps
Women
WRSM app use was consistently associated with educational/student status (Table 3), such that individuals with a high school degree or less had the lowest prevalence of both physical activity-focused and dietary-focused app use (ps < 0.01). There was also a strong bivariate relationship between BMI category and dietary-focused app use, with 18.7% of participants with a BMI ≥ 30 kg/m2 using dietary-focused apps compared to 9.2% of those with a BMI of 18.5–24.9 kg/m2, and only 3.5% of participants with a BMI < 18.5 kg/m2. WRSM app use did not differ by SES, ethnicity/race, or age.
Table 3.
Frequency (%) of weight-related self-monitoring (WRSM) during emerging adulthood by sociodemographics*
| Women | Men | |||
|---|---|---|---|---|
| Physical activity-focused apps | Dietary-focused apps | Physical activity-focused apps | Dietary-focused apps | |
| % (n) | % (n) | % (n) | % (n) | |
| Educational/student status | ||||
| High school degree or less | 14.1 (31) | 4.7 (10) | 9.5 (18) | 2.6 (6) |
| Some college, associate, vocational | 25.3 (39) | 14.2 (22) | 12.2 (18) | 3.3 (5) |
| Current Community/Technical College Student | 22.4 (37) | 11.0 (18) | 23.6 (28) | 5.4 (7) |
| Current Bachelor or Graduate Student | 31.1 (72) | 20.4 (46) | 18.3 (20) | 11.3 (12) |
| Bachelor, graduate degree | 24.8 (20) | 12.3 (9) | 29.7 (16) | 9.5 (5) |
| p value from X2 for column* | 0.003 | < 0.001 | < 0.001 | 0.008 |
| Ethnicity/race | ||||
| White | 30.5 (53) | 14.5 (25) | 20.7 (38) | 9.1 (17) |
| Black/African American | 18.9 (35) | 11.9 (23) | 10.1 (11) | 3.8 (4) |
| Hispanic/Latinx | 28.2 (43) | 14.2 (21) | 22.4 (23) | 4.8 (5) |
| Asian American | 21.6 (42) | 10.9 (21) | 19.0 (24) | 6.4 (8) |
| Mixed/other | 20.0 (26) | 11.8 (15) | 6.9 (4) | 1.3 (1) |
| p value from X2 for column* | 0.063 | 0.863 | 0.004 | 0.105 |
| Socioeconomic status | ||||
| Low | 19.8 (70) | 9.9 (35) | 15.5 (28) | 3.6 (7) |
| Low-middle/middle | 25.7 (83) | 15.1 (47) | 14.7 (37) | 4.1 (11) |
| High-middle/high | 26.2 (46) | 13.4 (23) | 19.1 (35) | 10.1 (17) |
| p value from X2 for column* | 0.148 | 0.133 | 0.464 | 0.001 |
| BMI | ||||
| < 18.5 | 13.0 (4) | 3.5 (1) | 15.5 (3) | 0 (0) |
| 18.5–24.9 | 23.2 (85) | 9.2 (33) | 13.1 (31) | 5.5 (14) |
| 25–29.9 | 25.1 (55) | 13.0 (28) | 15.9 (29) | 6.8 (12) |
| ≥ 30 | 24.3 (55) | 18.7 (43) | 22.6 (37) | 5.0 (9) |
| p value from X2 for column* | 0.516 | 0.005 | 0.084 | 0.678 |
| Age (years) | ||||
| 18–20 | 24.2 (64) | 12.3 (32) | 12.1 (24) | 8.0 (15) |
| 21–22 | 23.0 (113) | 12.7 (60) | 16.9 (55) | 2.9 (10) |
| 23–30 | 22.3 (22) | 12.3 (13) | 19.7 (21) | 9.2 (10) |
| p value from X2 for column* | 0.921 | 0.985 | 0.182 | 0.007 |
Percentage is weighted to account for attrition over time and allow for extrapolation to the original population-based sample, while n represents observed count. Accordingly, ns may not directly match weighted percentages. Ethnicity/race and socioeconomic status were measured at 2010 and all other variables were measured in 2018. Percentages represent the percent of individuals in that demographic category who used the respective type of app, for example 14.1% of women with a high school degree or less reported using physical activity-focused apps. Physical activity- and dietary-focused app use was not mutually exclusive. Bold font indicates p < .05
BMI body mass index
X2 tests tested for difference in use of the respective type of app across the demographic category
Men
There was also a relationship between WRSM app use and educational/student status among men. Current college students and those who had a bachelor’s or graduate degree had higher prevalence of physical activity-focused app use compared to those who were not currently students or did not have a college degree (p < 0.001). Similarly, those with a bachelor or graduate degree or who were current bachelor or graduate students were more likely to use dietary-focused apps (p = 0.008) than those who with lower educational attainment. There was a significant association between ethnicity/race and physical activity-focused app use (p = 0.004) with the highest prevalences among Hispanic/Latino men (22.4%), and lowest prevalence among mixed/other ethnicity/race (6.9%). There was also an association between SES and dietary focused-app use (p = 0.001), with the highest prevalence of use among high-middle/high SES (10.1%) compared to low SES (3.6%) and low–middle/middle SES (4.1%). Prevalence of dietary self-monitoring use was also lowest among individuals 21–22 years of age (2.9%) compared to those who were 18–20 (8.0%) or 23–30 years of age (9.2%, p = 0.007). There were no associations between WRSM app use and BMI among men.
Relationships between eating and weight-related concerns in adolescence and use of WRSM apps in emerging adulthood
Women
Dietary-focused app use
Use of unhealthy weight control behaviors (odds ratio [OR] = 1.81, 95% confidence interval [CI]: 1.14–2.87), engagement in unhealthy muscle-building behaviors (OR = 1.84, 95% CI 1.10–3.08), perceiving oneself to be overweight (OR = 2.22, 95% CI 1.41–3.50), and having a BMI at or above the 85th percentile (OR = 2.11, 95% CI 1.33–3.33) during adolescence predicted greater use of dietary-focused apps in emerging adulthood. Associations were further examined with adjustment for BMI during adolescence and the associations only remained statistically significant for prior unhealthy muscle-building behaviors (OR = 1.73, 95% CI 1.03–2.92).
Physical activity-focused app use
None of the eating and weight-related concern measures assessed in adolescence predicted use of physical activity-focused apps 8 years later, irrespective of adjustment for BMI (Table 4).
Table 4.
Prospective associations of eating and weight-related measures during adolescence and weight-related self-monitoring (WRSM) 8 years later, by gender (adjusted for age, race/ethnicity, educational/student status, socioeconomic status, with and without adjustment for BMI)
| WRSM | ||||||||
|---|---|---|---|---|---|---|---|---|
| Physical activity-focused Apps | Dietary-focused Apps | |||||||
| Adjusted for sociodemographics | Adjusted for sociodemographics and BMI | Adjusted for sociodemographics | Adjusted for sociodemographics and BMI | |||||
| OR (95% CI) | p value | OR (95% CI) | p value | OR (95% CI) | p value | OR (95% CI) | p value | |
| Women | ||||||||
| Unhealthy weight control behaviors | 1.04 (0.73, 1.49) | 0.820 | 1.00 (0.69, 1.46) | 0.985 | 1.81 (1.14, 2.87) | 0.012 | 1.48 (0.91, 2.40) | 0.112 |
| Unhealthy muscle-building behaviors | 0.91 (0.59, 1.42) | 0.679 | 0.90 (0.58, 1.40) | 0.641 | 1.84 (1.10, 3.08) | 0.021 | 1.73 (1.03, 2.92) | 0.040 |
| Binge eating with loss of control | 1.70 (0.81, 3.53) | 0.159 | 1.68 (0.81, 3.50) | 0.167 | 1.93 (0.79, 4.69) | 0.147 | 1.85 (0.75, 4.53) | 0.180 |
| Compulsive exercise | 0.96 (0.67, 1.36) | 0.800 | 0.91 (0.63, 1.32) | 0.630 | 1.43 (0.90, 2.25) | 0.127 | 1.16 (0.72, 1.87) | 0.536 |
| Recommended weight control behaviors | 1.29 (0.91, 1.84) | 0.151 | 1.28 (0.89, 1.83) | 0.179 | 1.28 (0.82, 2.00) | 0.285 | 1.12 (0.70, 1.76) | 0.643 |
| Other muscle-building behaviors | 1.22 (0.84, 1.76) | 0.297 | 1.20 (0.83, 1.74) | 0.325 | 1.50 (0.95, 2.38) | 0.082 | 1.42 (0.89, 2.25) | 0.141 |
| Dieting | 1.17 (0.82, 1.66) | 0.394 | 1.13 (0.77, 1.65) | 0.542 | 1.56 (1.00, 2.45) | 0.053 | 1.18 (0.73, 1.92) | 0.505 |
| Body dissatisfaction | 1.20 (0.84, 1.70) | 0.320 | 1.16 (0.80, 1.68) | 0.425 | 1.43 (0.91, 2.23) | 0.121 | 1.15 (0.72, 1.85) | 0.551 |
| Perceive overweight | 1.16 (0.81, 1.66) | 0.418 | 1.15 (0.75, 1.77) | 0.518 | 2.22 (1.41, 3.50) | 0.001 | 1.66 (0.97, 2.83) | 0.064 |
| BMI ≥ 85th percentile | 1.27 (0.89, 1.83) | 0.194 | -- | - | 2.11 (1.33, 3.33) | 0.001 | -- | - |
| Men | ||||||||
| Unhealthy weight control behaviors | 1.18 (0.73, 1.91) | 0.497 | 1.06 (0.65, 1.75) | 0.806 | 1.85 (0.85, 4.05) | 0.123 | 1.61 (0.72, 3.61) | 0.244 |
| Unhealthy muscle-building behaviors | 1.12 (0.70, 1.78) | 0.633 | 1.11 (0.70, 1.77) | 0.665 | 1.94 (0.94, 4.00) | 0.072 | 1.96 (0.95, 4.05) | 0.069 |
| Binge eating with loss of control | 1.40 (0.48, 4.09) | 0.534 | 1.35 (0.46, 3.97) | 0.583 | 3.01 (0.78, 11.57) | 0.110 | 3.08 (0.81, 11.71) | 0.100 |
| Compulsive exercise | 1.07 (0.69, 1.66) | 0.777 | 1.01 (0.64, 1.58) | 0.984 | 1.50 (0.73, 3.08) | 0.267 | 1.41 (0.68, 2.93) | 0.360 |
| Recommended weight control behaviors | 1.09 (0.70, 1.69) | 0.704 | 1.03 (0.66, 1.61) | 0.904 | 1.60 (0.77, 3.29) | 0.207 | 1.54 (0.74, 3.22) | 0.254 |
| Other muscle-building behaviors | 1.63 (1.05, 2.53) | 0.029 | 1.60 (1.03, 2.49) | 0.038 | 2.22 (1.09, 4.54) | 0.029 | 2.18 (1.07, 4.47) | 0.033 |
| Dieting | 1.65 (1.03, 2.66) | 0.038 | 1.49 (0.89, 2.48) | 0.130 | 1.76 (0.82, 3.81) | 0.150 | 1.42 (0.62, 3.28) | 0.409 |
| Body dissatisfaction | 1.74 (1.11, 2.73) | 0.016 | 1.67 (1.06, 2.65) | 0.028 | 2.45 (1.18, 5.07) | 0.017 | 2.35 (1.12, 4.92) | 0.024 |
| Perceive overweight | 1.45 (0.90, 2.32) | 0.125 | 1.33 (0.76, 2.34) | 0.322 | 1.85 (0.89, 3.85) | 0.100 | 1.90 (0.76, 4.76) | 0.169 |
| BMI ≥ 85th percentile | 1.59 (1.01, 2.50) | 0.046 | - | - | 1.34 (0.65, 2.76) | 0.435 | - | - |
All statistics are weighted to account for attrition over time and allow for extrapolation to the original population-based sample. All models are adjusted for adolescent reported race/ethnicity, age and socioeconomic status, and emerging adulthood educational/student status; models adjusted for BMI included adolescent BMI. Bold font indicates p < .05
WRSM weight-related self-monitoring, OR odds ratio, CI confidence interval, BMI body mass index
Men
Dietary-focused app use
Other muscle-building behaviors (OR = 2.22, 95% CI 1.09–4.54) and body dissatisfaction (OR = 2.45, 95% CI 1.18–5.07) in adolescence were the only significant predictors of dietary-focused app use in emerging adulthood. In models with additional adjustment for BMI, both other muscle-building behaviors (OR = 2.18, 95% CI 1.07–4.47) and body dissatisfaction (OR = 2.35, 95% CI 1.12–4.92) remained significant predictors of dietary-focused app use.
Physical activity-focused app use
Dieting in the last year (OR = 1.65, 95% CI 1.02–2.65), other muscle-building behaviors (OR = 1.63, 95% CI 1.05–2.53) and body dissatisfaction (OR = 1.71, 95% CI 1.09–2.68) in adolescence predicted increased likelihood of using physical activity-focused apps in emerging adulthood, as did having a BMI at or above the 85th percentile (OR = 1.59, 95% CI 1.01–2.50). After adjusting for BMI during adolescence, only the association between prior use of other muscle-building behaviors (OR = 1.60, 95% CI 1.03–2.49) and body dissatisfaction (OR = 1.65, 95% CI: 1.04–2.61) and emerging adulthood use of physical activity apps remained significant.
Discussion
In a community-based sample of emerging adults from diverse ethnic/racial and socioeconomic backgrounds, the present study examined what types of WRSM apps participants used, who used WRSM apps, and whether eating and weight-related concerns in adolescence predicted use of WRSM apps in emerging adulthood. Results indicated that use of physical activity-focused apps was more common than use of other WRSM apps. Furthermore, women were more likely than men to be using WRSM apps. Students and participants who had graduated from college were more likely to use both physical activity- and dietary-focused apps than those with a high school education or who were not students. Among women, many forms of eating and weight-related concerns in adolescence predicted use of dietary-focused app use eight years later. Unhealthy muscle-building behaviors remained a salient predictor of later dietary-focused app use among women after controlling for BMI. Among men, other muscle-building behaviors and body dissatisfaction predicted later dietary-focused app use, irrespective of BMI adjustment. There were no observed associations between eating and weight-related concerns in adolescence and physical activity-focused app use in emerging adulthood among women. However, among men there was evidence that use of other muscle-building behaviors, dieting, body dissatisfaction, and having a higher BMI in adolescence were associated with increased use of physical activity-focused apps in emerging adulthood. Thus, results indicated WRSM apps were common among emerging adults, particularly women, and that eating and weight-related concerns in adolescence may predict dietary self-monitoring app use but be a less salient predictor of physical activity-focused self-monitoring app use.
The present study expands upon prior research examining WRSM app use, which has been conducted primarily among college students [9, 18], by showing that WRSM app use is common among a highly diverse community sample of emerging adults. Notably, education/student status had the most consistent relationship with WRSM app use of any sociodemographic characteristic with the highest level of use among college students and those with a college education. A prior study of college students found no associations between WRSM and parental education, which was similar to our findings across SES in women [9]. However, we found dietary-focused app use to be less common among men of low SES, which may be due, in part, to the cost of certain dietary-focused apps. The present study also found that dietary self-monitoring was more common among women with higher BMI, similar to previous findings among college students [11]. Higher prevalence of dietary self-monitoring among women with higher BMI is not surprising given that WRSM is often recommended to persons with weight-related concerns [6]. It is possible that a portion of women with higher BMI were attempting to lose weight and independently sought out or were introduced to dietary-focused WRSM apps to assist in their attempts at weight loss. Our results indicating that WRSM apps are common among individuals of all SES, ethnicity/race and weight status has important clinical and public health implications. Because WRSM apps are widely used, the potential ramifications of WRSM app use are therefore also widespread and may be impacting vulnerable populations.
One such potential consequence of using WRSM apps is increased eating disorder risk. Among non-clinical samples, there has been a growing body of literature focused on the hypothesis that WRSM app use increases eating disorder risk. However, prior work examining whether WRSM app use increases eating disorder risk in non-clinical populations has, to our knowledge, been exclusively cross-sectional, limiting the ability to understand the directionality of the relationships between WRSM app use and eating disorder risk. It has been hypothesized that eating and weight concerns may actually precede the use of WRSM apps rather than being the results of WRSM [18]. The present study adds to this literature by showing that eating and weight concerns in adolescence may indeed predict the use of WRSM apps in emerging adulthood, particularly dietary self-monitoring apps. However, the relationships between eating and weight-related concerns in adolescence and later WRSM app use are not consistent across genders or by types of eating and weight-related concerns suggesting that these relationships are nuanced and complex. Interestingly, WRSM app use was more common among women as compared to men. That said, there were more consistent associations between eating and weight-concerns in adolescence predicting later WRSM app use among men as compared to women; this finding stands in contrast to the current literature, which shows stronger cross-sectional associations in emerging adulthood among women, as compared to men [37]. There may be many reasons for these seemingly discrepant results. For example, it may be that higher prevalence of WRSM app use in women is a by-product of ubiquitously high eating and weight-related concerns in adolescence or increases in concerns over time, whereas WRSM app use in men may be more indicative of long-standing eating and weight-related concerns, which are less common among men [38]. Future research is needed to disentangle how the relationships between WRSM app use and eating and weight-related concerns vary over time and by gender. Additionally, elevated BMI during adolescence predicted future WRSM app use, and many of the relationships between eating and weight-related concerns and WRSM app use were no longer significant after adjusting for baseline BMI. Because those of higher BMI are more likely to have eating and weight-related concerns and also gain more weight over time, it is also possible that WRSM apps are used in emerging adulthood as a form of weight management. Understanding that individuals who have eating and weight-related concerns may be more likely to use WRSM apps is important in that it contributes to our understanding of who uses WRSM and how the use of WRSM apps might impact those who choose to use them. While the present study takes an important first step toward the complex relationships between eating and weight-related concerns and WRSM app use, future longitudinal research should examine more immediate predictors of use and whether eating and weight-related concerns increase following use of WRSM apps. Clinicians should also be cognizant of the complex relationships between WRSM app use and eating and weight-related concerns and should screen for eating and weight-related concern among those who use WRSM apps. Further, clinicians recommending the use of WRSM apps to their patients should do so with caution until research can determine whether WRSM apps are effective in eliciting beneficial behavior change in the general population, and whether WRSM apps can cause increased eating disorder risk.
The only measure of eating and weight-related concern in adolescence that predicted future dietary-focused app use among both men and women was other muscle-building behaviors. A relationship between muscle-building behaviors and dietary restraint has been demonstrated previously [39]; therefore, it could be that those engaging in muscle-building behaviors choose to engage in dietary restraint, using dietary-focused WRSM apps as a tool to do so. Surprisingly, muscle-building behaviors did not predict physical activity-focused app use among men or women. The lack of association may be a result of the types of physical activity tracked using these apps (e.g., steps, calories burned), in that individuals using muscle-building behaviors are engaging in forms of physical activity that are not generally monitored via apps, such as weightlifting or resistance exercises [40]. Future work is needed to further understand how those engaging in muscle-building behaviors are using dietary-focused apps, and whether there are other forms of physical activity self-monitoring that those using muscle-building behaviors are engaging in.
The current study has many strengths, most notably the longitudinal study design and large and socioeconomically and ethnically/racially diverse population-based sample of emerging adults. Additionally, participants self-generated responses to apps used for managing their eating, activity, and weight, allowing us to build understanding of the breadth of WRSM apps used among the population. However, we had to make some assumptions about types of use as participants reported the WRSM apps that they used and not what features of the WRSM apps that they used. Additionally, because of the lower prevalence of WRSM apps, particularly among men, we may not have been able to statistically detect all associations despite potentially meaningful effect estimates. Further, due to low numbers of gender minorities and the use of gender-stratified analyses, gender minorities were excluded from the present analyses and warrant future research. Moreover, it is possible that there may have been other unmeasured factors that influenced WRSM app use over time such as nutritional knowledge or being introduced to WRSM for health reasons. Future research should examine smaller increments of time and explore unmeasured potential reasons for WRSM app use.
The findings of this study indicate that WRSM apps are popular among emerging adults in general, but more common among women and those with higher educational attainment or who are college students. The present study suggests that eating and weight-related concerns in adolescence may predict use of dietary-focused apps in emerging adulthood, particularly for women, but that BMI may explain many of these associations. Further, eating and weight concerns may not be the most salient predictors of future WRSM app use, but that youth with elevated BMI may be particularly likely to engage in future WRSM. Future research is needed to identify potential mental and physical health consequences of using WRSM apps, particularly whether eating and weight-related concerns increase because of using WRSM apps. Clinicians, parents, and public health professionals should exhibit caution when recommending or encouraging the use of WRSM apps until we are able to further understand the temporality of WRSM app use and negative health consequences.
What is already known on this subject?
Use of weight-related self-monitoring (WRSM) apps is cross-sectionally associated with eating and weight-related concerns. However, little is known whether eating and weight-related concerns in adolescence predict use of WRSM apps in emerging adulthood.
What does the study add?
This study adds valuable information on the longitudinal relationships between eating and weight-related concerns in adolescence predicting increased likelihood of using weight-related self-monitoring apps in emerging adulthood.
Funding
This study was supported by the National Heart, Lung, and Blood Institute (Grant numbers: R35HL139853 and R01HL127077, PI: D. Neumark-Sztainer). SLH and VMH were funded by the National Institute of Mental Health (Grant number: T32MH082761, PI: S. Crow). KAL was funded by the National Institute of Child Health and Human Development (Grant number: K23HD090324–02, PI: Katie Loth). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute, the National Institute of Mental Health, the National Institute of Child Health and Human Development, or the National Institutes of Health.
Footnotes
Conflict of interest The authors declare that they do not have any conflicts of interest.
Code availability Code for this study is not publicly available.
Ethics approval EAT 2010–2018 was approved by the University of Minnesota Institutional Review Board Human Subjects Committee.
Consent for participation Informed consent was obtained from all participants included in the study.
Consent for publication No identifying details were included, and thus consent for publication is not applicable.
Availability of data and material
Data from EAT 2010–2018 are not publicly available.
References
- 1.MyFitnessPal (2014) Announcing 75 million MyFitnessPal users! hellohealthy, 2014 [Google Scholar]
- 2.MyFitnessPal (2016) How an App got 165 million users. In: Jones D (ed) Mobile engagement podcast. iTunes, 2016 [Google Scholar]
- 3.Fitbit I (2016) Fitbit reports $712M Q415 and $1.86B FY15 revenue; guides to $2.4 to $2.5B revenue in FY16. Guides FY16 non-GAAP gross margin of 48.5% to 49.0% [Google Scholar]
- 4.Verto Analytics (2018) Most popular health and fitness apps in the United States as of May 2018, by monthly active users (in millions) [Google Scholar]
- 5.Larson N, Haynos AF, Roberto CA, Loth KA, Neumark-Sztainer D (2018) Calorie labels on the restaurant menu: is the use of weight-control behaviors related to ordering decisions? J Acad Nutr Diet 118:399–408 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Jensen MD, Ryan DH, Apovian CM, Ard JD, Comuzzie AG, Donato KA, Hu FB, Hubbard VS, Jakicic JM, Kushner RF, Loria CM, Millen BE, Nonas CA, Pi-Sunyer FX, Stevens J, Stevens VJ, Wadden TA, Wolfe BM, Yanovski SZ (2014) 2013 AHA/ACC/TOS guideline for the management of overweight and obesity in adults: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and The Obesity Society. J Am Coll Cardiol 63:2985–3023 [DOI] [PubMed] [Google Scholar]
- 7.Blue Cross Blue Shield Association (2018) Blue Cross Blue Shield Association Partners with Fitbit to Deliver Special Offer on Fitbit Devices to over 60 Million Members. Partnership enables and encourages members to lead healthy and active lifestyles. Chicago, BlueCross BlueShield [Google Scholar]
- 8.Fox S, Duggan M (2012) Mobile Health 2012. Mobile Health, Pew Research Center, 2012 [Google Scholar]
- 9.Hahn SL, Bauer KW, Kaciroti N, Eisenberg D, Lipson SK, Sonneville KR (2021) Relationships between patterns of weight-related self-monitoring and eating disorder symptomology among undergraduate and graduate students. Int J Eat Disord 54:595–605 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Christoph MJ, Ellison B (2017) A cross-sectional study of the relationship between nutrition label use and food selection, servings, and consumption in a university dining setting. J Acad Nutr Diet 117:1528–1537 [DOI] [PubMed] [Google Scholar]
- 11.Plateau CR, Bone S, Lanning E, Meyer C (2018) Monitoring eating and activity: Links with disordered eating, compulsive exercise, and general wellbeing among young adults. Int J Eat Disord 51:1270–1276 [DOI] [PubMed] [Google Scholar]
- 12.Burke LE, Wang J, Sevick MA (2011) Self-monitoring in weight loss: a systematic review of the literature. J Am Diet Assoc 111:92–102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Burke LE, Conroy MB, Sereika SM, Elci OU, Styn MA, Acharya SD, Sevick MA, Ewing LJ, Glanz K (2011) The effect of electronic self-monitoring on weight loss and dietary intake: a randomized behavioral weight loss trial. Obesity (Silver Spring) 19:338–344 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Acharya SD, Elci OU, Sereika SM, Styn MA, Burke LE (2011) Using a personal digital assistant for self-monitoring influences diet quality in comparison to a standard paper record among overweight/obese adults. J Am Diet Assoc 111:583–588 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Benn Y, Webb TL, Chang BP, Harkin B (2016) What is the psychological impact of self-weighing? A meta-analysis. Health Psychol Rev 10:187–203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Graham DJ, Laska MN (2012) Nutrition label use partially mediates the relationship between attitude toward healthy eating and overall dietary quality among college students. J Acad Nutr Diet 112:414–418 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sarcona A, Kovacs L, Wright J, Williams C (2017) Differences in eating behavior, physical activity, and health-related lifestyle choices between users and nonusers of mobile health apps. Am J Health Educ 48:298–305 [Google Scholar]
- 18.Simpson CC, Mazzeo SE (2017) Calorie counting and fitness tracking technology: associations with eating disorder symptomatology. Eat Behav 26:89–92 [DOI] [PubMed] [Google Scholar]
- 19.Hahn SL, Lipson SK, Sonneville KR (2020) Dietary self-monitoring is associated with increased likelihood of problematic alcohol use among college students. J Am Coll Health. 10.1080/07448481.2020.1741592 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hudson JI, Hiripi E, Pope HG Jr, Kessler RC (2007) The prevalence and correlates of eating disorders in the National Comorbidity Survey replication. Biol Psychiatry 61:348–358 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kessler RC, Berglund PA, Chiu WT, Deitz AC, Hudson JI, Shahly V, Aguilar-Gaxiola S, Alonso J, Angermeyer MC, Benjet C, Bruffaerts R, de Girolamo G, de Graaf R, Maria HJ, Kovess-Masfety V, O’Neill S, Posada-Villa J, Sasu C, Scott K, Viana MC, Xavier M (2013) The prevalence and correlates of binge eating disorder in the World Health Organization World Mental Health Surveys. Biol Psychiatry 73:904–914 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lipson SK, Sonneville KR (2017) Eating disorder symptoms among undergraduate and graduate students at 12 US colleges and universities. Eat Behav 24:81–88 [DOI] [PubMed] [Google Scholar]
- 23.Hazzard VM, Loth KA, Berge JM, Larson NI, Fulkerson JA, Neumark-Sztainer D (2020) Does exposure to controlling parental feeding practices during adolescence predict disordered eating behaviors 8 years later in emerging adulthood? Pediatr Obes 15:e12709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Economics DA (2020) The social and economic cost of eating disorders in the United States of America: a report for the strategic training initiative for the prevention of eating disorders and the academy for eating disorders. Australia: Deloitte Access Economics. Retrieved from https://www.hsph [Google Scholar]
- 25.Neumark-Sztainer D, Wall M, Story M, Standish AR (2012) Dieting and unhealthy weight control behaviors during adolescence: associations with 10-year changes in body mass index. J Adolesc Health 50:80–86 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Schvey NA, Marwitz SE, Mi SJ, Galescu OA, Broadney MM, Young-Hyman D, Brady SM, Reynolds JC, Tanofsky-Kraff M, Yanovski SZ, Yanovski JA (2019) Weight-based teasing is associated with gain in BMI and fat mass among children and adolescents at-risk for obesity: a longitudinal study. Pediatr Obes 14:e12538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Goldfield GS, Moore C, Henderson K, Buchholz A, Obeid N, Flament MF (2010) Body dissatisfaction, dietary restraint, depression, and weight status in adolescents. J Sch Health 80:186–192 [DOI] [PubMed] [Google Scholar]
- 28.Little RJA (1986) Survey nonresponse adjustments for estimates of means. Int Stat Rev 54:139–157 [Google Scholar]
- 29.Seaman S, White I (2011) Review of inverse probability weighting for dealing with missing data. Stat Methods Med Res 22:278–295 [DOI] [PubMed] [Google Scholar]
- 30.Bucchianeri MM, Eisenberg ME, Neumark-Sztainer D (2013) Weightism, racism, classism, and sexism: shared forms of harassment in adolescents. J Adolesc Health 53:47–53 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Larson NI, Wall MM, Story MT, Neumark-Sztainer DR (2013) Home/family, peer, school, and neighborhood correlates of obesity in adolescents. Obesity (Silver Spring) 21:1858–1869 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Neumark-Sztainer D, Wall M, Fulkerson JA, Larson N (2013) Changes in the frequency of family meals from 1999 to 2010 in the homes of adolescents: trends by sociodemographic characteristics. J Adolesc Health 52:201–206 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Neumark-Sztainer D, Wall MM, Choi J, Barr-Anderson DJ, Telke S, Mason SM (2020) Exposure to adverse events and associations with stress levels and the practice of yoga: survey findings from a population-based study of diverse emerging young adults. J Altern Complement Med 26:482–490 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Montez JK, Hayward MD (2014) Cumulative childhood adversity, educational attainment, and active life expectancy among US adults. Demography 51:413–435 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ro O, Reas DL, Rosenvinge J (2012) The impact of age and BMI on Eating Disorder Examination Questionnaire (EDE-Q) scores in a community sample. Eat Behav 13:158–161 [DOI] [PubMed] [Google Scholar]
- 36.Inoue K, Goto A, Sugiyama T, Ramlau-Hansen CH, Liew Z (2020) The confounder-mediator dilemma: should we control for obesity to estimate the effect of perfluoroalkyl substances on health outcomes? Toxics 8(4):125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Hahn SL, Sonneville KR, Kaciroti N, Eisenberg D, Bauer KW (2021) Relationships between patterns of technology-based weight-related self-monitoring and eating disorder behaviors among first year university students. Eat Behav 42:101520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Larson N, Loth KA, Eisenberg ME, Hazzard VM, Neumark-Sztainer D (2021) Body dissatisfaction and disordered eating are prevalent problems among US young people from diverse socioeconomic backgrounds: findings from the EAT 2010–2018 study. Eating Behav 42:101535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Rodgers RF, Slater A, Gordon CS, McLean SA, Jarman HK, Paxton SJ (2020) A biopsychosocial model of social media use and body image concerns, disordered eating, and muscle-building behaviors among adolescent girls and boys. J Youth Adolesc 49:399–409 [DOI] [PubMed] [Google Scholar]
- 40.Yager Z, McLean S (2020) Muscle building supplement use in Australian adolescent boys: relationships with body image, weight lifting, and sports engagement. BMC Pediatr 20:89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Neumark-Sztainer D, Croll J, Story M, Hannan PJ, French SA, Perry C (2002) Ethnic/racial differences in weight-related concerns and behaviors among adolescent girls and boys: findings from Project EAT. J Psychosom Res 53:963–974 [DOI] [PubMed] [Google Scholar]
- 42.Field AE, Austin SB, Camargo CA Jr, Taylor CB, Striegel-Moore RH, Loud KJ, Colditz GA (2005) Exposure to the mass media, body shape concerns, and use of supplements to improve weight and shape among male and female adolescents. Pediatrics 116:e214–220 [DOI] [PubMed] [Google Scholar]
- 43.McCabe MP, Ricciardelli LA (2001) Body image and body change techniques among young adolescent boys. Eur Eat Disorders Rev 9:335–347 [Google Scholar]
- 44.Smolak L, Murnen SK, Thompson JK (2005) Sociocultural influences and muscle building in adolescent boys. Psychol Men Masc 6:227 [Google Scholar]
- 45.Maughan RJ (2012) Quality assurance issues in the use of dietary supplements, with special reference to protein supplements. J Nutr 143:1843S–1847S [DOI] [PubMed] [Google Scholar]
- 46.Hildebrandt T, Harty S, Langenbucher JW (2012) Fitness supplements as a gateway substance for anabolic-androgenic steroid use. Psychol Addict Behav 26:955–962 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.ZelitchYanovski S (1993) Binge eating disorder: current knowledge and future directions. Obes Res 1:306–324 [DOI] [PubMed] [Google Scholar]
- 48.Thompson J, Pasman L (1991) The Obligatory Exercise Questionnaire. Behav Ther 14:137 [Google Scholar]
- 49.Centers for Disease Control and Prevention (2021) Healthy eating for a healthy weight. In: Division of Nutrition P.A., Obesity, National Center for Chronic Disease Prevention and Health Promotion (eds) Healthy weight, nutrition, and physical activity https://www.cdc.gov/obesity/strategies/index.html & https://www.cdc.gov/nccdphp/dnpao/divisioninformation/aboutus/index.htm [Google Scholar]
- 50.Lampard AM, Maclehose RF, Eisenberg ME, Larson NI, Davison KK, Neumark-Sztainer D (2016) Adolescents who engage exclusively in healthy weight control behaviors: who are they? Int J Behav Nutr Phys Act 13:5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS, Wei R, Mei Z, Curtin LR, Roche AF, Johnson CL (2000) CDC growth charts: United States. Adv Data 1–27 [PubMed] [Google Scholar]
- 52.Gibson RS (2005) Principles of nutritional assessment. Oxford University Press, New York [Google Scholar]
- 53.Yuan C, Spiegelman D, Rimm E, Rosner B, Stampfer M, Barnett J, Chavarro J, Subar A, Sampson L, Willett W (2017) Validity of a dietary questionnaire assessed by comparison with multiple weighed dietary records or 24-hour recalls. Am J Epidemiol 185:570–584 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Feskanich D, Rimm E, Giovannucci E, Colditz G, Stampfer M, Litin L, Willett W (1993) Reproducibility and validity of food intake measurements from a semiquantitative food frequency questionnaire. J Am Diet Assoc 93:790–796 [DOI] [PubMed] [Google Scholar]
- 55.Larson N, Wall M, Story M, Neumark-Sztainer D (2013) Home/family, peer, school, and neighborhood correlates of obesity in adolescents. Obesity 21:1858–1869 [DOI] [PMC free article] [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 from EAT 2010–2018 are not publicly available.