Abstract
Objective:
Few studies have investigated disordered eating and unhealthy weight control behaviors and cardiovascular health (CVH) outside of adolescence and early adulthood. We examined the cross-sectional and prospective associations of these behaviors and CVH in middle adulthood.
Methods:
2,095 CARDIA participants were assessed at Year 10 (Y10, 1995–1996) and Year 30 (Y30, 2015–2016). The Y10-administered Questionnaire on Eating and Weight Patterns-Revised was used to create the Problematic Relationship to Eating and Food (PREF) score (range 0–8 points). Higher scores indicated greater disordered eating and/or unhealthy weight control behaviors across eight components: anxiety surrounding food, compensatory behaviors, overeating, loss of control, distress surrounding overeating, distress surrounding loss of control, shape concerns, and time spent dieting. PREF was modeled categorically: 0–1 (reference), 2–3, and 4–8. Diet, physical activity, smoking, blood pressure, cholesterol, glucose, and body mass index (BMI) were measured at Y10 and Y30 (except diet at Y7 and Y20) and used to define CVH. CVH was modeled categorically: Poor-to-Intermediate (0–9) and Ideal (10–14; reference). Logistic regression was used to evaluate associations between PREF and CVH categories and components.
Results:
PREF 4–8 was associated with Y10 Poor-to-Intermediate CVH (OR=2.35, 95%CI: 1.78–3.10) but not Y30 (OR=1.34, 95%CI: 0.96–1.87) compared to PREF 0–1. PREF 2–3 was not associated with Y10 or Y30 CVH. Individual PREF components were not uniformly associated with individual CVH components, though all PREF components were associated with Y10 Poor-to-Intermediate BMI.
Conclusions:
Disordered eating and unhealthy weight control behaviors are cross-sectionally but not prospectively associated with poorer CVH during middle age.
Keywords: disordered eating, weight control, cardiovascular health
Public Significance
Middle-aged adults experience a significant loss of cardiovascular health. Our study suggests that higher levels of disordered eating and unhealthy weight control behaviors are associated with lower levels of current, but not future, cardiovascular health behaviors in middle-aged U.S. adults. Disordered eating and unhealthy weight control behaviors are not uncommon in middle-age and may be one contributing factor to the loss of cardiovascular health observed in this age group.
Abstract
Objetivo:
Pocos estudios han investigado los trastornos alimentarios y las conductas poco saludables de control de peso y la salud cardiovascular (CVH, por sus siglas en inglés) fuera de la adolescencia y la edad adulta temprana. Examinamos las asociaciones transversales y prospectivas de estos comportamientos y CVH en la edad adulta media.
Métodos:
2,095 participantes de CARDIA fueron evaluados en el año 10 (Y10, 1995–1996) y en el año 30 (Y30, 2015–2016). Se utilizó el Cuestionario Revisado sobre Patrones de Alimentación y Peso administrado en Y10 para crear la puntuación de la Relación Problemática con la Alimentación y los Alimentos (PREF, por sus siglas en inglés) (rango de 0 a 8 puntos). Las puntuaciones más altas indicaron mayores trastornos alimentarios y/o conductas de control de peso poco saludables en ocho componentes: ansiedad en torno a la comida, conductas compensatorias, comer en exceso, pérdida de control, angustia en torno a comer en exceso, angustia en torno a la pérdida de control, preocupaciones sobre la forma y tiempo dedicado a hacer dieta. PREF se modeló categóricamente: 0–1 (referencia), 2–3, y 4–8. La dieta, la actividad física, el tabaquismo, la presión arterial, el colesterol, la glucosa y el índice de masa corporal (BMI, por sus siglas en inglés) se midieron en Y10 y Y30 (excepto la dieta en Y7 y Y20) y se utilizaron para definir CVH. La CVH se modeló categóricamente: pobre a intermedia (0–9) e ideal (10–14; referencia). Se utilizó regresión logística para evaluar las asociaciones entre las categorías y componentes de PREF y CVH.
Resultados:
PREF 4–8 se asoció con Y10 CVH Pobre a Intermedia (OR=2.35, IC 95%: 1.78–3.10) pero no Y30 (OR=1.34, IC 95%: 0.96–1.87) en comparación con PREF 0– 1. PREF 2–3 no se asoció con Y10 o Y30 CVH. Los componentes individuales de PREF no se asociaron uniformemente con los componentes individuales de CVH, aunque todos los componentes de PREF se asociaron con el BMI de pobre a intermedio en Y10.
Conclusiones:
Los trastornos alimentarios y las conductas poco saludables de control de peso se asocian de manera transversal, pero no prospectiva, con una peor CVH durante la mediana edad.
Introduction
Primary prevention of cardiovascular disease (CVD) is largely dependent on the optimal management of lifestyle and behavioral factors. The American Heart Association created the Life’s Simple 7 (LS7), a guideline for cardiovascular health (CVH) based on optimizing parameters for leading CVD risk factors (e.g., blood pressure, fasting glucose, diet quality) (Lloyd-Jones et al., 2010). Optimal levels of LS7 have been associated with lowered risk of total CVD (Folsom et al., 2011), heart failure (Folsom et al., 2015), atrial fibrillation (Ogunmoroti et al., 2018), and stroke (Kulshreshtha et al., 2013). However, a number of psychosocial factors may negatively impact an individual’s ability to achieve optimal CVH, potentially contributing to a subsequent elevation in CVD risk (Levine et al., 2021).
Disordered eating and unhealthy weight control behaviors have received minimal attention in relation to CVH compared to other negative psychosocial and suboptimal behavioral factors. While eating disorders are uncommon among the general U.S. adult population (Udo & Grilo, 2018), subclinical levels of disordered eating and unhealthy weight control behaviors are not uncommon. An estimated 17% of U.S. young adults age 18–24 endorse subclinical levels of these behaviors (Nagata, Garber, Tabler, Murray, & Bibbins-Domingo, 2018) and 59% of U.S. middle-aged adults age 27–41 endorse one or more components related to a problematic relationship with food and/or weight (Yoon et al., 2018). Much of the research on these behaviors has focused on adolescents and young adults. In these age groups, there appears to be consistent associations between higher levels of disordered eating and/or unhealthy weight control behaviors and lower LS7 components that are less age-dependent, including higher body mass index (BMI) trajectories (Neumark-Sztainer et al., 2012; Yoon et al., 2020), poorer diet quality (de Lauzon et al., 2004), and incidence of hyperlipidemia (Nagata, Garber, Tabler, Murray, Vittinghoff, et al., 2018).
Despite disordered eating and unhealthy weight control behaviors appearing to be a problem of adolescence and young adulthood, there is a distinct need to examine these behaviors and CVH in middle-age and beyond. First, as noted above, a substantial number of middle-aged adults endorse some level of subclinical disordered eating behaviors. Second, evidence suggests that middle-aged adults experience an accelerated loss of CVH at age 37 (Krefman et al., 2021), followed by a steady decline in ideal CVH through late mid-life (Shah et al., 2015).
It is unclear what role subclinical disordered eating and unhealthy weight control behaviors may play in the downward trend in CVH observed during middle-age. Furthermore, individual components of disordered eating and unhealthy weight control behaviors have not been sufficiently examined in relation to CVH or some of its components in the middle-age group. Given the heterogenous nature of disordered eating and unhealthy weight control behaviors that span affective experiences, eating behaviors, and summaries of weight management behaviors, it is unlikely that this group of components uniformly affect CVH. Identifying which of these behaviors are associated with which CVH components is an important step in helping to identify currently underrecognized psychological and behavioral contributors to CVH in middle-age. Thus, the aim of this study was to examine the cross-sectional and prospective associations of disordered eating and unhealthy weight control behaviors (total score and each component) with CVH (total score and each component) among middle-aged adults.
Methods
Study population
Data for the present study came from the Coronary Artery Risk Development in Young Adults (CARDIA) study, a community-based, prospective cohort study of Black and White adults in the U.S. originally designed to study the development and determinants of CVD risk factors and their cardiovascular sequelae. Participants between the ages of 18–30 were recruited between 1985–1986 from four field centers: Birmingham, AL, Chicago, IL, Minneapolis, MN, and Oakland, CA. Through 2016, CARDIA participants have been followed for 30 years, completing a total of nine in-person examinations. Institutional review boards at each participating center approved the study procedures and all participants provided written informed consent. CARDIA methods and study protocols are described in detail elsewhere (Friedman et al., 1988). In the present study, we used the Year 10 (Y10, 1995–1996) exam as the study baseline and followed participants for 20 years until the Year 30 (Y30, 2015–2016) exam.
Of the 5,115 participants enrolled in CARDIA, we excluded participants that (a) did not attend the baseline/Y10 visit (n=1,165), did not attend the last follow-up/Y30 visit (n=950), were pregnant or breastfeeding at baseline (n=83), or were missing data on the exposure (n=38), outcomes (n=778), or covariates (n=6). Our final analytic sample consisted of 2,095 middle-aged adults (Figure 1).
Figure 1.
CARDIA analytic sample.
Exposures
Disordered eating and unhealthy weight control behaviors were derived from the Questionnaire on Eating and Weight Patterns-Revised (QEWP-R) administered at Y10; the QEWP-R is a 23-item measure created to identify Binge Eating Disorder, with a test-rest reliability comparable to structured clinical interviews and a high ability to differentiate between clinical and nonclinical binge eating (Nangle et al., 1994). While the questionnaire’s original purpose was to screen individuals for Binge Eating Disorder, it also provides information on endorsement of disordered eating and weight control behaviors across multiple components of both Binge Eating Disorder and Bulimia Nervosa. The QEWP-R was used to create the Problematic Relationship to Eating and Food (PREF) scale as done in prior CARDIA studies (Yoon et al., 2018, 2019). The PREF scale includes eight components of disordered eating and weight control behaviors, each scored as 0 or 1 (Supplemental Table 1): anxiety surrounding food/eating, use of compensatory behaviors to avoid weight gain or maintain weight loss, episodic overeating, episodic overeating with loss of control, distress about overeating, distress about loss of control, weight/shape concerns, and proportion of time spent dieting to lose weight or prevent weight gain. These components were summed for a PREF total score (range 0–8), with higher scores indicating more disordered eating and/or unhealthy weight control behaviors. We modeled the PREF score categorically using the following groups: 0–1 (reference), 2–3, and 4–8; these cut-points were used to reflect groupings likely to reflect those with little-to-no, subclinical, and clinical levels of disordered eating/unhealthy weight control behaviors, respectively.
Outcomes
CVH was defined by the LS7 (Lloyd-Jones et al., 2010) health factors (blood pressure, fasting glucose, total cholesterol, BMI) and health behaviors (smoking status, diet quality, physical activity) measured at Y10 and Y30 unless otherwise specified. Blood pressure was measured three times in one-minute intervals following a five-minute seated rest period, from which we used the average of the second and third measurements. Fasting glucose and total cholesterol were measured from blood samples taken following a 12-hour fasting period during which participants were asked to refrain from smoking or engaging in heavy physical activity. Medication use for hypertension, diabetes, and hyperlipidemia was assessed by an interviewer-administered Medical History Questionnaire. Height and weight were measured in light clothing and without shoes by trained personnel; BMI was calculated as weight in kilograms/height in meters-squared. Smoking status was assessed using the CARDIA Tobacco Use Questionnaire, a 13-item tool assessing historical and current use of inhaled and smokeless tobacco sources. The LS7 criteria were directly applied to create categories of Poor, Intermediate, or Ideal for blood pressure (Poor: systolic blood pressure [SBP] ≥140 mm Hg or diastolic blood pressure [DBP] ≥90 mm Hg; Intermediate: SBP 120–139, DBP 80–89, or treated to goal; Ideal: <120/<80 mm Hg), fasting glucose (Poor: ≥126 mg/dL; Intermediate: 100–125 mg/dL or treated to goal; Ideal: <100 mg/dL), total cholesterol (Poor: ≥240 mg/dL; Intermediate: 200–239 mg/dL or treated to goal; Ideal: <200 mg/dL), BMI (Poor: ≥30 kg/m2; Intermediate: 25–29.9 kg/m2; Ideal: <25 kg/m2), and smoking status (Poor: current smoker; Intermediate: former smoker ≤12 months; Ideal: never smoker or former smoker >12 months). Diet quality and physical activity had to be approximated due to differences in measurement methods. Diet quality was measured by the Diet History Interview (McDonald et al., 1991), a food-frequency questionnaire created to assess habitual food intake, administered at Year 7 and Year 20 and used to represent diet quality at Y10 and Y30, respectively, as diet was not assessed at Y10 or Y30. Participants were asked open-ended questions about their intake over the past month across 100 different food and drink categories, including serving size, frequency of consumption, and preparation method. The five AHA diet recommendations for fruits/vegetables, fish, whole grains, sodium, and sugar-sweetened beverages were applied to the self-reported intake to create diet categories (Poor: 0–1 components; Intermediate: 2–3 components; Ideal: 4–5 components). Physical activity was measured by the Physical Activity Questionnaire, an interviewer-administered questionnaire created to measure levels of physical activity over the last 12 months to calculate summary estimates of activity called exercise units, which demonstrate reasonable validity and reliability to rank individuals’ physical activity levels and has high accuracy in classifying individuals meeting recommended physical activity guidelines (Gabriel et al., 2014; Jacobs et al., 1993) (Poor: inactive with <100 exercise units; Intermediate: active but below guidelines with 100–299 exercise units; Ideal: active above guidelines with ≥300 exercise units). Measures for the health behaviors are publicly available on the CARDIA website (https://www.cardia.dopm.uab.edu/exam-materials2/data-collection-forms). For participants missing data on CVH components collected at Y30 despite visit attendance, the Y30 values were replaced with Year 25 values (component, n missing and replaced: blood pressure, 0; fasting glucose, 16; total cholesterol, 12; BMI, 2; smoking status, 50; physical activity, 37). Individual CVH components were summed for a CVH total score (range 0–14), with higher scores indicating better CVH. We modeled the CVH score categorically using the following groups: Poor (0–4), Intermediate (5–9), and Ideal (10–14).
Covariates
Covariates, measured via standard structured questionnaires, included: age (years), sex (female, male), education (less than high school, high school graduate/some college, college degree or higher), household income (<$25,000, $25,000-$49,000, ≥$50,000), and CARDIA field center (AL, IL, MN, CA). Though not included as a covariate, race (Black, White), measured via a standard structured questionnaire, was examined as a potential modifier. Hispanic ethnicity was not assessed.
Statistical analyses
Baseline characteristics were summarized as mean (standard deviation) for continuous variables and frequency (percent) for categorical variables by PREF category.
We examined the cross-sectional (Y10) and prospective (Y10-Y30) associations of disordered eating and unhealthy weight control behaviors with CVH using logistic regression analyses estimating odds ratios (OR) and 95% confidence intervals (CI) in two ways. First, we examined the association of PREF category (0–1 [reference], 2–3, and 4–8) at Y10 with CVH category (Poor-to-Intermediate, Ideal [reference]) at Y10 and at Y30. Second, we examined the associations of the eight individual PREF (yes, no [reference]) components at Y10 with the seven individual CVH components (Poor-to-Intermediate, Ideal [reference]) at Y10 and at Y30. The collapsed definition of Poor-to-Intermediate CVH was used given the small number of participants in the Poor CVH category. For cross-sectional analyses, the models were as follows: model 1: unadjusted; model 2: adjusted for age, sex, education, household income, and field center. For prospective analyses, the models were as follows: model 1: adjusted for baseline CVH or baseline CVH component; model 2: model 1 + age, sex, education, household income, and field center. For the PREF component-CVH component analyses, we additionally included model 3: model 2 + baseline BMI (except for BMI CVH component analyses).
To examine the potential for effect measure modification by sex (female, male) and race (Black, White) in the CVH total score analyses, we included interaction terms between the PREF categories and each potential effect measure modifier. We did not examine potential modifiers for the PREF component and CVH component analyses due to small sample size.
All analyses were conducted using SAS software (v.9.4; SAS Institute, Inc., Cary, NC). Due to the high number of analyses, a threshold of p<.01 was used to determine statistical significance.
Results
Participant Characteristics
The baseline characteristics for the CARDIA study population at Y10 are presented in Table 1 by PREF total score category. Participants in the highest PREF category (i.e., higher levels of disordered eating and/or unhealthy weight control behaviors) were more likely to be female, less likely to have a college degree, and less likely to have a household income ≥$50,000 than those in lower PREF categories. The mean age (35.2 years) and distribution of participants across categories of race (43.4% Black) and field center did not differ by PREF category.
Table 1.
Characteristics of the CARDIA analytic sample by PREF category.
| Total Sample (N=2,095) | PREF 0–1 (n=1,361) | PREF 2–3 (n=442) | PREF 4–8 (n=292) | |
|---|---|---|---|---|
|
| ||||
| Age, years | 35.2 (3.6) | 35.3 (3.5) | 35.0 (3.6) | 35.2 (3.6) |
| Sex, female | 1,155 (55.1) | 681 (50.0) | 250 (56.6) | 224 (76.7) |
| Race, Black | 909 (43.4) | 583 (42.8) | 201 (45.5) | 125 (42.8) |
| Education | ||||
| Less than high school | 58 (2.8) | 43 (3.2) | 10 (2.3) | 5 (1.7) |
| High school grad/some college | 1,073 (51.2) | 683 (50.2) | 225 (50.9) | 165 (56.5) |
| College degree or higher | 964 (46.0) | 635 (46.7) | 207 (46.8) | 122 (41.8) |
| Household income | ||||
| <$25,000 | 467 (22.3) | 303 (22.3) | 88 (19.9) | 76 (26.0) |
| $25,000–$49,999 | 727 (34.7) | 476 (35.0) | 145 (32.8) | 106 (36.3) |
| ≥$50,000 | 901 (43.0) | 582 (42.8) | 209 (47.3) | 110 (37.7) |
| Field center | ||||
| Birmingham, AL | 508 (24.3) | 333 (24.5) | 104 (23.5) | 71 (24.3) |
| Chicago, IL | 535 (25.5) | 343 (25.2) | 121 (27.4) | 71 (24.3) |
| Minneapolis, MN | 517 (24.7) | 334 (24.5) | 106 (24.0) | 77 (26.4) |
| Oakland, CA | 535 (25.5) | 353 (25.8) | 111 (25.1) | 73 (24.9) |
| CVH total score (Year 10) | 9.8 (2.0) | 10.0 (2.0) | 9.8 (2.0) | 9.1 (2.2) |
| Ideal | 1,242 (59.3) | 856 (63.0) | 261 (59.1) | 125 (42.8) |
| Intermediate | 833 (39.8) | 494 (36.3) | 179 (40.5) | 160 (54.8) |
| Poor | 20 (0.9) | 11 (0.8) | 2 (0.5) | 7 (2.4) |
| CVH total score (Year 30) | 8.4 (2.3) | 8.6 (2.3) | 8.4 (2.3) | 7.8 (2.3) |
| Ideal | 702 (33.5) | 489 (35.9) | 142 (32.1) | 71 (24.3) |
| Intermediate | 1,303 (62.2) | 816 (60.0) | 282 (63.8) | 205 (70.2) |
| Poor | 90 (4.3) | 56 (4.1) | 18 (4.1) | 16 (5.5) |
Continuous variables presented as mean (SD). Dichotomous/categorical variables presented as frequency (%),
CVH: cardiovascular health; PREF: problematic relationship to eating and food.
The mean PREF total score for the total analytic sample was 1.5, with 65.0%, 21.1%, and 13.9% participants in the 0–1, 2–3, or 4–8 PREF categories, respectively. Although mean PREF total score did not differ by sex, female participants were overrepresented in PREF 4–8 compared to male participants (19.4% versus 7.2%) (Supplemental Table 2). PREF score did not differ by race (Supplemental Table 3).
CARDIA participants had relatively high CVH throughout the study period, with a mean CVH score of 9.8 (out of 14) at baseline and 8.4 after 20 years of follow-up. Only 0.9% and 4.3% of the sample were in the Poor CVH category at baseline and follow-up, respectively.
Cross-sectional: PREF at Y10 and CVH at Y10
PREF 4–8 was associated with a higher odds of Poor-to-Intermediate CVH (OR=2.27, 95%CI: 1.75–2.93) versus Ideal CVH compared to PREF 0–1 in model 1 (unadjusted; Table 2). The association did not alter in magnitude or significance (OR=2.35, 95%CI: 1.78–3.10) in model 2 adjusting for age, sex, education, household income, and field center. PREF 2–3 was not associated with Poor-to-Intermediate versus Ideal CVH compared to PREF 0–1 in model 1 or model 2. The pattern of results did not significantly differ by sex (p-value for interaction=0.256) or race (p-value for interaction=0.349).
Table 2.
Odds ratios and 95% confidence intervals of the association between PREF at Year 10 and Poor-to-Intermediate versus Ideal CVH at Year 10 and Year 30 (N=2,095).
| PREF 0–1 (n=1,361) | PREF 2–3 (n=442) | PREF 4–8 (n=292) | |
|---|---|---|---|
|
| |||
| Cross-sectional/Year 10 | |||
|
| |||
| Poor-to-Intermediate CVH, n | 505 | 181 | 167 |
| Model 1 | 1.00 (Ref) | 1.18 (0.94–1.46) | 2.27 (1.75–2.93) |
| Model 2 | 1.00 (Ref) | 1.26 (0.99–1.59) | 2.35 (l.78–3.10) |
| Prospective/Year 10 to Year 30 | |||
|
| |||
| Poor-to-Intermediate CVH, n | 872 | 300 | 221 |
| Model 1 | 1.00 (Ref) | 1.13 (0.88–1.45) | 1.26 (0.91–1.73) |
| Model 2 | 1.00 (Ref) | 1.18 (0.91–1.52) | 1.34 (0.96–1.87) |
CVH: cardiovascular health; PREF: problematic relationship to eating and food.
Cross-sectional:
Model 1: unadjusted.
Model 2: adjusted for age, sex, education, household income, and field center.
Prospective
Model 1: baseline CVH.
Model 2: model 1 + age, sex, education, household income, and field center.
Bold indicates p < .01
Individual components of PREF were differentially associated with individual components of CVH at Y10 (Table 3). Blood pressure. Anxiety surrounding food/eating, compensatory behaviors, and distress about overeating were associated with higher odds of Poor-to-Intermediate (ORs=1.37–1.53) versus Ideal blood pressure in model 2 adjusting for age, sex, education, household income, and field center. However, these associations became attenuated and non-significant in model 3 after further adjustment for BMI. Fasting glucose. No components of PREF were significantly associated with fasting glucose in any model. Total cholesterol. Episodic overeating was associated with higher odds of Poor-to-Intermediate (OR=1.39) versus Ideal cholesterol in model 1 but not models 2–3. BMI. All eight components of PREF were associated with higher odds of Poor-to-Intermediate (ORs=1.42–3.14) versus Ideal BMI in model 2, with the lowest magnitude for shape/weight concerns and the highest for distress about loss of control. Smoking status. No components of PREF were significantly associated with smoking status in any model. Diet quality. No components of PREF were significantly associated with diet quality in any model. Physical activity. Compensatory behaviors and weight/shape concerns were significantly associated with lower odds of Poor-to-Intermediate (OR=0.49) versus Ideal physical activity in model 3. Episodic overeating with loss of control, distress about overeating, distress about loss of control, and time spent dieting were associated with higher odds of Poor-to-Intermediate (ORs=1.34–1.94) versus Ideal physical activity in model 1, but attenuated and became non-significant in models 2–3.
Table 3.
Odds ratios and 95% confidence intervals of the cross-sectional association between PREF components and Poor-to-Intermediate versus Ideal CVH components at Year 10 (N = 2,095)
| Anxiety surrounding food/eating (n=384) | Compensatory behaviors (n=234) | Episodic overeating (n=549) | Episodic overeating with loss of control (n=116) | Distress about overeating (n=506) | Distress about loss of control (n=348) | Weight/Shape concerns (n=610) | Time spent dieting (n=322) | |
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Poor-to-Intermediate Blood Pressure, n | 112 | 76 | 167 | 36 | 136 | 94 | 152 | 78 |
| Model 1 | 1.24 (0.99–1.63) | 1.49 (1.11–2.00) | 1.43 (1.15–1.77) | 1.35 (0.90–2.03) | 1.12 (0.89–1.40) | 1.11 (0.86–1.44) | 0.97 (0.78–1.21) | 0.93 (0.71–1.23) |
| Model 2 | 1.53 (1.18–1.99) | 1.53 (1.13–2.08) | 1.34 (1.07–1.68) | 1.42 (0.92–2.17) | 1.37 (1.08–1.75) | 1.37 (1.04–1.81) | 1.14 (0.91–1.43) | 1.34 (0.99–1.81) |
| Model 3 | 1.29 (0.98–1.68) | 1.29 (0.94–1.77) | 1.15 (0.91–1.45) | 1.15 (0.75–1.77) | 1.08 (0.84–1.39) | 1.08 (0.81–1.44) | 1.03 (0.82–1.31) | 1.10 (0.81–1.50) |
| Poor-to-Intermediate Fasting Glucose, n | 33 | 18 | 44 | 13 | 38 | 27 | 36 | 27 |
| Model 1 | 1.23 (0.82–1.83) | 1.05 (0.63–1.75) | 1.13 (0.78–1.62) | 1.63 (0.90–2.98) | 1.02 (0.70–1.50) | 1.06 (0.69–1.64) | 0.72 (0.49–1.06) | 1.18 (0.76–1.81) |
| Model 2 | 1.45 (0.96–2.19) | 1.10 (0.65–1.85) | 1.00 (0.69–1.45) | 1.75 (0.94–3.25) | 1.23 (0.83–1.82) | 1.30 (0.83–2.04) | 0.82 (0.55–1.21) | 1.78 (1.11–2.86) |
| Model 3 | 1.20 (0.78–1.81) | 0.90 (0.53–1.52) | 0.84 (0.57–1.22) | 1.42 (0.76–2.65) | 0.95 (0.64–1.42) | 1.01 (0.64–1.60) | 0.73 (0.49–1.09) | 1.43 (0.89–2.31) |
| Poor-to-Intermediate Total Cholesterol, n | 105 | 54 | 155 | 24 | 128 | 79 | 143 | 81 |
| Model 1 | 1.27 (0.99–1.63) | 0.96 (0.70–1.33) | 1.39 (1.11–1.74) | 0.83 (0.53–1.32) | 1.12 (0.89–1.42) | 0.94 (0.71–1.23) | 0.98 (0.79–1.23) | 1.10 (0.84–1.45) |
| Model 2 | 1.39 (1.08–1.80) | 1.01 (0.73–1.40) | 1.33 (1.06–1.66) | 0.84 (0.53–1.35) | 1.24 (0.97–1.57) | 1.03 (0.78–1.37) | 1.06 (0.85–1.33) | 1.34 (1.00–1.80) |
| Model 3 | 1.26 (0.97–1.64) | 0.91 (0.65–1.27) | 1.22 (0.97–1.53) | 0.74 (0.46–1.19) | 1.09 (0.85–1.39) | 0.90 (0.67–1.20) | 1.01 (0.81–1.28) | 1.21 (0.90–1.62) |
| Poor-to-Intermediate Body Mass Index, n | 275 | 174 | 388 | 93 | 375 | 266 | 382 | 224 |
| Model 1 | 2.09 (1.64–2.66) | 2.31 (1.70–3.14) | 2.12 (1.72–2.62) | 3.12 (1.96–4.97) | 2.59 (2.07–3.23) | 2.76 (2.12–3.60) | 1.33 (1.10–1.62) | 1.82 (1.41–2.35) |
| Model 2 | 2.42 (1.88–3.11) | 2.34 (1.71–3.21) | 2.03 (1.64–2.51) | 3.09 (1.92–4.96) | 3.04 (2.41–3.84) | 3.14 (2.38–4.14) | 1.42 (1.16–1.74) | 2.30 (1.76–3.02) |
| Poor-to-Intermediate Smoking Status, n | 83 | 63 | 144 | 35 | 100 | 78 | 146 | 76 |
| Model 1 | 0.91 (0.70–1.19) | 1.27 (0.93–1.73) | 1.28 (1.02–1.60) | 1.49 (0.99–2.24) | 0.78 (0.61–0.99) | 0.96 (0.73–1.27) | 1.08 (0.87–1.35) | 1.04 (0.79–1.38) |
| Model 2 | 0.92 (0.69–1.22) | 1.19 (0.86–1.66) | 1.19 (0.94–1.51) | 1.22 (0.79–1.88) | 0.74 (0.57–0.97) | 0.85 (0.63–1.15) | 1.11 (0.87–1.41) | 1.05 (0.77–1.43) |
| Model 3 | 0.97 (0.72–1.29) | 1.25 (0.90–1.75) | 1.25 (0.98–1.60) | 1.29 (0.83–2.01) | 0.78 (0.60–1.03) | 0.91 (0.67–1.23) | 1.14 (0.89–1.44) | 1.11 (0.81–1.51) |
| Poor-to-Intermediate Diet Quality, n | 374 | 231 | 544 | 113 | 494 | 341 | 601 | 315 |
| Model 1 | 0.51 (0.24–1.08) | 1.26 (0.38–4.17) | 2.01 (0.77–5.22) | 0.58 (0.17–1.93) | 0.55 (0.27–1.13) | 0.74 (0.32–1.71) | 1.10 (0.51–2.37) | 0.67 (0.29–1.56) |
| Model 2 | 0.58 (0.27–1.25) | 1.11 (0.33–3.76) | 1.69 (0.64–4.47) | 0.46 (0.13–1.64) | 0.58 (0.29–1.23) | 0.73 (0.31–1.75) | 1.04 (0.47–2.30) | 0.78 (0.32–1.89) |
| Model 3 | 0.46 (0.21–1.03) | 0.96 (0.28–3.31) | 1.50 (0.56–4.01) | 0.34 (0.09–1.28) | 0.48 (0.22–1.03) | 0.60 (0.25–1.47) | 0.96 (0.43–2.13) | 0.69 (0.28–1.70) |
| Poor-to-Intermediate Physical Activity, n | 222 | 100 | 287 | 79 | 297 | 219 | 318 | 194 |
| Model 1 | 1.25 (1.00–1.57) | 0.62 (0.47–0.82) | 0.95 (0.78–1.15) | 1.94 (1.30–2.89) | 1.34 (1.09–1.64) | 1.61 (1.27–2.04) | 0.94 (0.78–1.13) | 1.40 (1.10–1.78) |
| Model 2 | 1.07 (0.84–1.35) | 0.53 (0.39–0.71) | 1.05 (0.85–1.29) | 1.51 (0.99–2.30) | 1.11 (0.89–1.37) | 1.23 (0.96–1.59) | 0.83 (0.68–1.02) | 0.98 (0.75–1.27) |
| Model 3 | 0.99 (0.78–1.26) | 0.49 (0.36–0.65) | 0.99 (0.80–1.22) | 1.39 (0.91–2.13) | 1.01 (0.81–1.26) | 1.13 (0.88–1.46) | 0.81 (0.66–0.99) | 0.91 (0.70–1.19) |
CVH: cardiovascular health; PREF: problematic relationship to eating and food.
Model 1 : unadjusted; Model 2: adjusted for age, sex, education, household income, and field center; Model 3 : model 2 + baseline body mass index (except body mass index analysis). Bold indicates p < .01
Prospective: PREF at Y10 and CVH at Y30
PREF 4–8 and PREF 2–3 at Y10 were not associated with Poor-to-Intermediate versus Ideal CVH at Y30 compared to PREF 0–1 in model 1 or model 2. The pattern of results did not significantly differ by sex (p-value for interaction=0.185) or race (p-value for interaction=0.617).
Individual components of PREF at Y10 were differentially associated with individual components of CVH at Y30 (Table 4). Blood pressure. No PREF components at Y10 were associated with Poor-to-Intermediate versus Ideal blood pressure at Y30 in models 1–3. Fasting glucose. Episodic overeating and episodic overeating with loss of control at Y10 were associated with higher odds of Poor-to-Intermediate (ORs=1.67–1.75) versus Ideal fasting glucose at Y30 in model 2 adjusting for baseline fasting glucose, age, sex, education, household income, and field center. The association for episodic overeating was attenuated but remained significant in model 3 adjusting for baseline BMI (OR=1.45, 95%CI: 1.17–1.81), while the association for episodic overeating with loss of control was attenuated and became non-significant in model 3 (OR=1.33, 95%CI: 0.88–2.01). Total cholesterol. No PREF components at Y10 were associated with Poor-to-Intermediate versus Ideal cholesterol at Y30 in models 1–3. BMI. Episodic overeating at Y10 was associated with higher odds of Poor-to-Intermediate (OR=1.54, 95%CI: 1.12–2.12) versus Ideal BMI at Y30 in model 1 adjusting for baseline BMI, which became non-significant in model 2 further adjusting for age, sex, education, household income, and field center. Distress about loss of control at Y10 was marginally or significantly associated with higher odds of Poor-to-Intermediate (ORs=1.64–1.82) versus Ideal BMI at Y30 in models 1–2. Smoking status. No PREF components at Y10 were associated with Poor-to-Intermediate versus Ideal smoking status at Y30 in models 1–3. Diet quality. Weight/shape concerns at Y10 were associated with lower odds of Poor-to-Intermediate (ORs=0.34) versus Ideal diet quality at Y30 in models 1–3. In other words, having weight/shape concerns at Y10 was associated with better diet quality at Y30. Physical activity. Episodic overeating and distress about overeating at Y10 were associated with higher odds of Poor-to-Intermediate (ORs=1.37–1.47) versus Ideal physical activity at Y30 in model 1 adjusting for baseline physical activity, but only episodic overeating remained significant (ORs=1.45–1.53) in models 2–3 further adjusting for age, sex, education, household income, field center, and baseline BMI.
Table 4.
Odds ratios and 95% confidence intervals of the prospective association between PREF components at Year 10 and Poor-to-Intermediate versus Ideal CVH components at Year 30 (N = 2,095)
| Anxiety surrounding food/eating (n=384) | Compensatory behaviors (n=234) | Episodic overeating (n=549) | Episodic overeating with loss of control (n=116) | Distress about overeating (n=506) | Distress about loss of control (n=348) | Weight/Shape concerns (n=610) | Time spent dieting (n=322) | |
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Poor-to-Intermediate Blood Pressure, n | 239 | 165 | 354 | 75 | 320 | 229 | 376 | 201 |
| Model 1 | 0.96 (0.75–1.22) | 1.45 (1.06–1.98) | 1.07 (0.86–1.33) | 1.06 (0.70–1.61) | 1.08 (0.87–1.34) | 1.23 (0.96–1.59) | 1.02 (0.83–1.25) | 1.06 (0.82–1.38) |
| Model 2 | 0.98 (0.76–1.26) | 1.38 (1.00–1.90) | 1.05 (0.84–1.31) | 0.93 (0.60–1.43) | 1.09 (0.87–1.37) | 1.19 (0.92–1.56) | 1.03 (0.83–1.27) | 1.09 (0.83–1.44) |
| Model 3 | 0.84 (0.65–1.08) | 1.20 (0.86–1.67) | 0.92 (0.73–1.16) | 0.77 (0.50–1.20) | 0.88 (0.69–1.12) | 0.96 (0.73–1.27) | 0.97 (0.78–1.20) | 0.93 (0.70–1.24) |
| Poor-to-Intermediate Fasting Glucose, n | 153 | 100 | 263 | 58 | 195 | 140 | 219 | 117 |
| Model 1 | 1.09 (0.86–1.38) | 1.29 (0.97–1.71) | 1.83 (1.49–2.25) | 1.64 (1.11–2.43) | 1.05 (0.84–1.30) | 1.14 (0.89–1.45) | 0.94 (0.77–1.15) | 0.90 (0.69–1.17) |
| Model 2 | 1.23 (0.97–1.58) | 1.27 (0.95–1.71) | 1.75 (1.42–2.16) | 1.67 (1.12–2.50) | 1.18 (0.95–1.48) | 1.29 (1.00–1.66) | 1.02 (0.83–1.26) | 1.14 (0.87–1.50) |
| Model 3 | 0.97 (0.75–1.25) | 0.98 (0.72–1.33) | 1.45 (1.17–1.81) | 1.33 (0.88–2.01) | 0.83 (0.65–1.05) | 0.90 (0.68–1.18) | 0.90 (0.72–1.11) | 0.87 (0.65–1.16) |
| Poor-to-Intermediate Total Cholesterol, n | 219 | 123 | 310 | 61 | 298 | 198 | 345 | 193 |
| Model 1 | 1.01 (0.80–1.29) | 0.87 (0.65–1.16) | 0.95 (0.77–1.17) | 0.93 (0.62–1.37) | 1.19 (0.96–1.48) | 1.11 (0.87–1.42) | 1.08 (0.89–1.32) | 1.25 (0.97–1.61) |
| Model 2 | 0.95 (0.75–1.21) | 0.88 (0.66–1.19) | 1.00 (0.81–1.24) | 0.86 (0.57–1.29) | 1.10 (0.88–1.37) | 1.02 (0.79–1.31) | 1.06 (0.86–1.30) | 1.09 (0.83–1.42) |
| Model 3 | 0.96 (0.75–1.22) | 0.89 (0.67–1.20) | 1.01 (0.82–1.26) | 0.87 (0.58–1.30) | 1.13 (0.90–1.41) | 1.04 (0.80–1.35) | 1.06 (0.86–1.30) | 1.10 (0.84–1.44) |
| Poor-to-Intermediate Body Mass Index, n | 326 | 204 | 477 | 107 | 436 | 313 | 476 | 266 |
| Model 1 | 1.08 (0.75–1.55) | 1.22 (0.76–1.97) | 1.54 (1.12–2.12) | 1.92 (0.88–4.21) | 1.09 (0.78–1.52) | 1.64 (1.07–2.51) | 0.78 (0.59–1.03) | 0.89 (0.61–1.31) |
| Model 2 | 1.15 (0.79–1.67) | 1.19 (0.73–1.94) | 1.52 (1.10–2.11) | 1.88 (0.85–4.17) | 1.23 (0.87–1.74) | 1.82 (1.17–2.83) | 0.79 (0.60–1.06) | 1.05 (0.70–1.56) |
| Poor-to-Intermediate Smoking Status, n | 53 | 36 | 90 | 22 | 57 | 47 | 90 | 43 |
| Model 1 | 1.04 (0.69–1.57) | 0.98 (0.60–1.59) | 1.07 (0.76–1.52) | 1.11 (0.59–2.08) | 0.79 (0.54–1.17) | 0.95 (0.62–1.46) | 1.01 (0.71–1.42) | 0.88 (0.57–1.37) |
| Model 2 | 1.08 (0.69–1.68) | 0.91 (0.55–1.50) | 1.00 (0.69–1.44) | 1.01 (0.52–1.97) | 0.86 (0.57–1.30) | 0.96 (0.61–1.52) | 1.13 (0.78–1.62) | 1.09 (0.68–1.75) |
| Model 3 | 1.03 (0.65–1.61) | 0.86 (0.52–1.43) | 0.97 (0.67–1.41) | 0.99 (0.50–1.94) | 0.78 (0.50–1.21) | 0.88 (0.54–1.42) | 1.11 (0.77–1.60) | 1.03 (0.64–1.68) |
| Poor-to-Intermediate Diet Quality, n | 372 | 226 | 539 | 111 | 490 | 335 | 579 | 307 |
| Model 1 | 0.91 (0.47–1.74) | 0.76 (0.35–1.62) | 1.62 (0.81–3.24) | 0.60 (0.23–1.55) | 0.86 (0.48–1.56) | 0.68 (0.36–1.29) | 0.34 (0.20–0.58) | 0.56 (0.31–1.03) |
| Model 2 | 0.97 (0.50–1.88) | 0.70 (0.32–1.52) | 1.57 (0.78–3.16) | 0.55 (0.21–1.44) | 0.93 (0.51–1.70) | 0.71 (0.37–1.37) | 0.34 (0.20–0.59) | 0.65 (0.35–1.23) |
| Model 3 | 0.98 (0.50–1.93) | 0.71 (0.32–1.55) | 1.62 (0.80–3.29) | 0.54 (0.20–1.46) | 0.95 (0.51–1.77) | 0.72 (0.37–1.40) | 0.34 (0.20–0.59) | 0.66 (0.35–1.26) |
| Poor-to-Intermediate Physical Activity, n | 230 | 126 | 335 | 79 | 317 | 224 | 341 | 205 |
| Model 1 | 1.17 (0.92–1.50) | 1.12 (0.83–1.51) | 1.47 (1.19–1.83) | 1.47 (0.96–2.26) | 1.37 (1.10–1.71) | 1.34 (1.03–1.73) | 1.05 (0.85–1.29) | 1.37 (1.05–1.78) |
| Model 2 | 1.14 (0.89–1.46) | 0.98 (0.72–1.34) | 1.53 (1.23–1.92) | 1.21 (0.78–1.88) | 1.32 (1.05–1.66) | 1.19 (0.91–1.56) | 1.03 (0.83–1.28) | 1.30 (0.98–1.73) |
| Model 3 | 1.05 (0.81–1.36) | 0.89 (0.65–1.22) | 1.45 (1.15–1.81) | 1.10 (0.71–1.72) | 1.19 (0.94–1.52) | 1.06 (0.81–1.40) | 1.00 (0.81–1.24) | 1.21 (0.91–1.61) |
CVH: cardiovascular health; PREF: problematic relationship to eating and food.
Model 1: respective baseline CVH component; Model 2: model 1 + age, sex, education, household income, and field center; Model 3: model 2 + baseline body mass index (except body mass index analysis). Bold indicates p < .01
Discussion
In this study of middle-aged adults in CARDIA followed for 20 years from Y10 at 1995–1996 to Y30 at 2015–2016, we found that higher levels of disordered eating and unhealthy weight control behaviors were cross-sectionally but not prospectively associated with Poor-to-Intermediate CVH after adjusting for age, sex, education, household income, field center, and baseline CVH. Our results did not differ by sex or race. To our knowledge, this is the first study to examine disordered eating and unhealthy weight control behaviors and overall CVH in middle-age. These results suggest that these behavioral and psychological factors may be associated with CVH in earlier rather than later stages of adulthood.
In addition, we found that individual components of PREF were not uniformly associated with the health factors and health behaviors of CVH, except for the BMI component at Y10. For the health factors (i.e., blood pressure, fasting glucose, total cholesterol, BMI), we observed that none of the PREF components were associated with blood pressure cross-sectionally at Y10 or prospectively at Y30 in fully-adjusted models. For fasting glucose, we observed that none of the PREF components were associated with fasting glucose at Y10 but that episodic overeating was associated with poorer fasting glucose at Y30 in the fully-adjusted model. For total cholesterol, we observed that none of the PREF components were associated with cholesterol cross-sectionally at Y10 or prospectively at Y30 in our fully-adjusted models. Finally, all components of PREF were associated with higher BMI cross-sectionally at Y10 but only distress about loss of control was associated with higher BMI at Y30 in fully-adjusted models.
For the CVH health behavior factors (i.e., smoking status, diet quality, physical activity), we found that none of the PREF components were associated with smoking status cross-sectionally at Y10 or prospectively from Y10 to Y30 in our fully-adjusted models. For diet quality, we observed that weight/shape concerns were associated with better diet quality prospectively from Y10 to Y30 in fully-adjusted models. For physical activity, we found that compensatory behaviors were associated with more physical activity cross-sectionally at Y10, while episodic overeating was associated with less physical activity prospectively from Y10 to Y30 in fully-adjusted models. Taken together, our findings suggest that the relationship between disordered eating/unhealthy weight control behaviors and CVH is nuanced and complex, with different components of these behavioral and psychological factors associated with different components of CVH. For example, at Y10, compensatory behaviors (extreme efforts to lose weight or maintain weight loss) were associated with higher BMI but also higher physical activity. It is worth noting that excessive exercise can be a form of compensatory behavior, though we do not have the ability to differentiate between high and excessively high physical activity.
There are only three other longitudinal studies available examining subclinical disordered eating/unhealthy weight control behaviors and components of CVH outside of adolescence. Two studies by Yoon et al. also using data from CARDIA, observed that higher PREF total scores were associated with greater BMI increases (Yoon et al., 2018), higher odds of metabolic syndrome, and higher risk of type 2 diabetes throughout middle-age (Yoon et al., 2019). We extended these findings by identifying which components of the PREF measure may be driving some of these relationships disaggregated by components of CVH. Similarly, researchers using data from ELSA-Brazil found that individuals reporting episodic binge eating with loss of control had higher odds of metabolic syndrome, hypertension, and hypertriglyceridemia after mean follow-up time of 3.9 years (Solmi et al., 2021). We replicated parts of this finding with fasting glucose and BMI but not hypertension at Y30, which could be due to a difference in age range (ELSA-Brazil included up to 75 years of age). Available cross-sectional studies are almost exclusively focused on eating behaviors (i.e., binge, uncontrolled, or emotional eating), noting positive associations with metabolic syndrome, blood pressure/hypertension, glucose/diabetes, hypertriglyceridemia, obesity, compensatory smoking, and energy intake (Abraham et al., 2014; de Lauzon et al., 2004; Leone et al., 2016; White, 2012), as well as a negative association with a Mediterranean dietary pattern (Bertoli et al., 2015).
In addition to our study’s strengths (i.e., prospective community-based cohort, racial diversity, comprehensive assessment of LS7 metrics, examination of overall and individual components of disordered eating/unhealthy weight control behaviors and CVH), there are several limitations to note. First, our study was unable to use the updated Life’s Essential 8 CVH definition due to the lack of sleep duration data at our study time points (Lloyd-Jones et al., 2022). Second, the diet assessment did not align with the Y10 baseline and Y30 follow-up exams and was carried forward from prior exams (Year 7 and Year 20). Third and related, a small number of participants in the analytic sample met the definition of Ideal diet quality (n=46 at Y10, n=80 at Y30), leading to a small referent outcome group in these CVH component analyses. Likewise, few participants met the definition of Ideal BMI between those with and without particular PREF components (e.g., episodic overeating with loss of control and compensatory behaviors), leading to a small referent outcome group in some of these component analyses. Thus, diet quality analyses and some of the BMI analyses should be interpreted with caution. Fourth, we addressed missing data on CVH components collected at Y30 despite visit attendance by carrying forward Y25 CVH component data. This affected a small number of participants’ data (≤2.4%). Fifth, the aggregated compensatory behaviors are heterogenous and may be differentially associated with the LS7 metrics, biasing results towards the null for this component. For example, compensatory behaviors of excess laxative and diet pill use can raise blood pressure while compensatory behaviors of excessive exercise and excess diuretic use can lower blood pressure. Sixth, the QEWP-R does not capture comprehensive criteria for anorexia nervosa and thus we may be underestimating disordered eating in this cohort. Seventh, the CARDIA study did not assess Hispanic/Latino ethnicity. This is an important limitation given evidence suggesting that a) acculturation and related constructs may be associated with disordered eating (Song et al., 2023) and b) differences exist in CVH components by ethnicity. As an example of the latter, there were a higher proportion of Hispanic/Latinos categorized as having Poor BMI but also a higher proportion categorized as having Ideal physical activity in the Hispanic Community Health Study/Study of Latinos compared to the general U.S. population (González et al., 2016). Thus, our results may be an underestimation of the PREF-CVH association among Hispanic/Latinos living in the U.S. It is also worth noting that our analyses only examine the initial loss of CVH and thus does not capture whether these behavioral and psychological factors are associated with not just loss of ideal CVH but an exacerbation of poor CVH in ways that lead to the progression of cardiometabolic disease (e.g., first reaching fasting glucose ≥126 mg/dL versus chronic elevation of glucose or experience of hyperglycemic episodes). Because of our use of CARDIA, our results are likely generalizable to middle-aged Black and White adults in the United States. Future work is needed to examine these associations in other racial/ethnic groups and among other social strata not considered in the present study.
In summary, our analysis of middle-aged CARDIA participants provides evidence that disordered eating and unhealthy weight control behaviors are cross-sectionally but not prospectively associated with poorer overall CVH during middle age. In other words, the influence of disordered eating and unhealthy weight control behaviors on CVH may be stronger earlier in the life course with a wanning influence over time. Future studies are needed to further elucidate how the disordered eating and unhealthy weight control behaviors cluster together and affect CVH at different phases throughout the life course. Ultimately, this line of work could extend the utility of intervening on eating/weight control factors beyond the primary prevention of obesity and characterize an additional modifiable risk factor for overall CVH.
Supplementary Material
Acknowledgements:
The authors wish to acknowledge the participants and research staff of the CARDIA study, without whom this research would not be possible.
Financial Disclosure:
The Coronary Artery Risk Development in Young Adults Study (CARDIA) is supported by contracts 75N92023D00002, 75N92023D00003, 75N92023D00004, 75N92023D00005, and 75N92023D00006 from the National Heart, Lung, and Blood Institute (NHLBI). Dr. Polanka acknowledges support from NIH/NHLBI training grant T32HL007779.
Footnotes
Conflicts of Interest: The authors have no conflicts of interest to declare.
Data Availability:
CARDIA data are available upon reasonable request from the CARDIA Coordinating Center. CARDIA investigators are eager to collaborate with investigators interested in using CARDIA data. Please see the CARDIA website (https://www.cardia.dopm.uab.edu) for publications policies and for a list of CARDIA investigators. CARDIA data are also publicly available on the NIH-supported BioLINCC and dbGaP platforms.
References
- Abraham TM, Massaro JM, Hoffmann U, Yanovski JA, & Fox CS (2014). Metabolic characterization of adults with binge eating in the general population: The Framingham Heart Study. Obesity (Silver Spring, Md.), 22(11), 2441–2449. 10.1002/oby.20867 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bertoli S, Spadafranca A, Bes-Rastrollo M, Martinez-Gonzalez MA, Ponissi V, Beggio V, Leone A, & Battezzati A (2015). Adherence to the Mediterranean diet is inversely related to binge eating disorder in patients seeking a weight loss program. Clinical Nutrition (Edinburgh, Scotland), 34(1), 107–114. 10.1016/j.clnu.2014.02.001 [DOI] [PubMed] [Google Scholar]
- de Lauzon B, Romon M, Deschamps V, Lafay L, Borys J-M, Karlsson J, Ducimetière P, & Charles MA (2004). The Three-Factor Eating Questionnaire-R18 is able to distinguish among different eating patterns in a general population. The Journal of Nutrition, 134(9), 2372–2380. 10.1093/jn/134.9.2372 [DOI] [PubMed] [Google Scholar]
- Folsom AR, Shah AM, Lutsey PL, Roetker NS, Alonso A, Avery CL, Miedema MD, Konety S, Chang PP, & Solomon SD (2015). American Heart Association’s Life’s Simple 7: Avoiding Heart Failure and Preserving Cardiac Structure and Function. The American Journal of Medicine, 128(9), 970–976.e2. 10.1016/j.amjmed.2015.03.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Folsom AR, Yatsuya H, Nettleton JA, Lutsey PL, Cushman M, & Rosamond WD (2011). Community prevalence of ideal cardiovascular health, by the American Heart Association definition, and relationship with cardiovascular disease incidence. Journal of the American College of Cardiology, 57(16), 1690–1696. 10.1016/j.jacc.2010.11.041 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Friedman GD, Cutter GR, Donahue RP, Hughes GH, Hulley SB, Jacobs DRJ, Liu K, & Savage PJ (1988). CARDIA: study design, recruitment, and some characteristics of the examined subjects. Journal of Clinical Epidemiology, 41(11), 1105–1116. 10.1016/0895-4356(88)90080-7 [DOI] [PubMed] [Google Scholar]
- Gabriel KP, Sidney S, Jacobs DRJ, Quesenberry CPJ, Reis JP, Jiang S-F, & Sternfeld B (2014). Convergent validity of a brief self-reported physical activity questionnaire. Medicine and Science in Sports and Exercise, 46(8), 1570–1577. 10.1249/MSS.0000000000000278 [DOI] [PMC free article] [PubMed] [Google Scholar]
- González HM, Tarraf W, Rodríguez CJ, Gallo LC, Sacco RL, Talavera GA, Heiss G, Kizer JR, Hernandez R, Davis S, Schneiderman N, Daviglus ML, & Kaplan RC (2016). Cardiovascular health among diverse Hispanics/Latinos: Hispanic Community Health Study/Study of Latinos (HCHS/SOL) results. American Heart Journal, 176, 134–144. 10.1016/j.ahj.2016.02.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jacobs DRJ, Ainsworth BE, Hartman TJ, & Leon AS (1993). A simultaneous evaluation of 10 commonly used physical activity questionnaires. Medicine and Science in Sports and Exercise, 25(1), 81–91. 10.1249/00005768-199301000-00012 [DOI] [PubMed] [Google Scholar]
- Krefman AE, Labarthe D, Greenland P, Pool L, Aguayo L, Juonala M, Kähönen M, Lehtimäki T, Day RS, Bazzano L, Muggeo VMR, Van Horn L, Liu L, Webber LS, Pahkala K, Laitinen TT, Raitakari O, Lloyd-Jones DM, & Allen NB (2021). Influential Periods in Longitudinal Clinical Cardiovascular Health Scores. American Journal of Epidemiology, 190(11), 2384–2394. 10.1093/aje/kwab149 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kulshreshtha A, Vaccarino V, Judd SE, Howard VJ, McClellan WM, Muntner P, Hong Y, Safford MM, Goyal A, & Cushman M (2013). Life’s Simple 7 and risk of incident stroke: The reasons for geographic and racial differences in stroke study. Stroke, 44(7), 1909–1914. 10.1161/STROKEAHA.111.000352 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Leone A, Bedogni G, Ponissi V, Battezzati A, Beggio V, Magni P, Ruscica M, & Bertoli S (2016). Contribution of binge eating behaviour to cardiometabolic risk factors in subjects starting a weight loss or maintenance programme. The British Journal of Nutrition, 116(11), 1984–1992. 10.1017/S0007114516004141 [DOI] [PubMed] [Google Scholar]
- Levine GN, Cohen BE, Commodore-Mensah Y, Fleury J, Huffman JC, Khalid U, Labarthe DR, Lavretsky H, Michos ED, Spatz ES, & Kubzansky LD (2021). Psychological Health, Well-Being, and the Mind-Heart-Body Connection: A Scientific Statement From the American Heart Association. Circulation, 143(10), e763–e783. 10.1161/CIR.0000000000000947 [DOI] [PubMed] [Google Scholar]
- Lloyd-Jones DM, Allen NB, Anderson CAM, Black T, Brewer LC, Foraker RE, Grandner MA, Lavretsky H, Perak AM, Sharma G, & Rosamond W (2022). Life’s Essential 8: Updating and Enhancing the American Heart Association’s Construct of Cardiovascular Health: A Presidential Advisory From the American Heart Association. Circulation, 101161CIR0000000000001078. 10.1161/CIR.0000000000001078 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lloyd-Jones DM, Hong Y, Labarthe D, Mozaffarian D, Appel LJ, Van Horn L, Greenlund K, Daniels S, Nichol G, Tomaselli GF, Arnett DK, Fonarow GC, Ho PM, Lauer MS, Masoudi FA, Robertson RM, Roger V, Schwamm LH, Sorlie P, … Rosamond WD (2010). Defining and setting national goals for cardiovascular health promotion and disease reduction: The American Heart Association’s strategic Impact Goal through 2020 and beyond. Circulation, 121(4), 586–613. 10.1161/CIRCULATIONAHA.109.192703 [DOI] [PubMed] [Google Scholar]
- McDonald A, Van Horn L, Slattery M, Hilner J, Bragg C, Caan B, Jacobs DJ, Liu K, Hubert H, Gernhofer N, Betz E, & Havlik D (1991). The CARDIA dietary history: Development, implementation, and evaluation. Journal of the American Dietetic Association, 91(9), 1104–1112. [PubMed] [Google Scholar]
- Nagata JM, Garber AK, Tabler JL, Murray SB, & Bibbins-Domingo K (2018). Prevalence and Correlates of Disordered Eating Behaviors Among Young Adults with Overweight or Obesity. Journal of General Internal Medicine, 33(8), 1337–1343. 10.1007/s11606-018-4465-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nagata JM, Garber AK, Tabler J, Murray SB, Vittinghoff E, & Bibbins-Domingo K (2018). Disordered eating behaviors and cardiometabolic risk among young adults with overweight or obesity. The International Journal of Eating Disorders, 51(8), 931–941. 10.1002/eat.22927 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nangle DW, Johnson WG, Carr-Nangle RE, & Engler LB (1994). Binge eating disorder and the proposed DSM-IV criteria: Psychometric analysis of the Questionnaire of Eating and Weight Patterns. The International Journal of Eating Disorders, 16(2), 147–157. [DOI] [PubMed] [Google Scholar]
- 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. The Journal of Adolescent Health : Official Publication of the Society for Adolescent Medicine, 50(1), 80–86. 10.1016/j.jadohealth.2011.05.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ogunmoroti O, Michos ED, Aronis KN, Salami JA, Blankstein R, Virani SS, Spatz ES, Allen NB, Rana JS, Blumenthal RS, Veledar E, Szklo M, Blaha MJ, & Nasir K (2018). Life’s Simple 7 and the risk of atrial fibrillation: The Multi-Ethnic Study of Atherosclerosis. Atherosclerosis, 275, 174–181. 10.1016/j.atherosclerosis.2018.05.050 [DOI] [PubMed] [Google Scholar]
- Shah AM, Claggett B, Folsom AR, Lutsey PL, Ballantyne CM, Heiss G, & Solomon SD (2015). Ideal Cardiovascular Health During Adult Life and Cardiovascular Structure and Function Among the Elderly. Circulation, 132(21), 1979–1989. 10.1161/CIRCULATIONAHA.115.017882 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Solmi F, Moreno AB, Lewis G, Angélica Nunes M, de Jesus Mendes da Fonseca M, & Harter Griep R. (2021). Longitudinal association between binge eating and metabolic syndrome in adults: Findings from the ELSA-Brasil cohort. Acta Psychiatrica Scandinavica, 144(5), 464–474. 10.1111/acps.13356 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Song S, Stern CM, Deitsch T, & Sala M (2023). Acculturation and eating disorders: A systematic review. Eating and Weight Disorders : EWD, 28(1), 39. 10.1007/s40519-023-01563-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Udo T, & Grilo CM (2018). Prevalence and Correlates of DSM-5-Defined Eating Disorders in a Nationally Representative Sample of U.S. Adults. Biological Psychiatry, 84(5), 345–354. 10.1016/j.biopsych.2018.03.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- White MA (2012). Smoking for weight control and its associations with eating disorder symptomatology. Comprehensive Psychiatry, 53(4), 403–407. 10.1016/j.comppsych.2011.05.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yoon C, Jacobs DRJ, Duprez DA, Dutton G, Lewis CE, Neumark-Sztainer D, Steffen LM, West DS, & Mason SM (2018). Questionnaire-based problematic relationship to eating and food is associated with 25 year body mass index trajectories during midlife: The Coronary Artery Risk Development In Young Adults (CARDIA) Study. The International Journal of Eating Disorders, 51(1), 10–17. 10.1002/eat.22813 [DOI] [PubMed] [Google Scholar]
- Yoon C, Jacobs DRJ, Duprez DA, Neumark-Sztainer D, Steffen LM, & Mason SM (2019). Problematic eating behaviors and attitudes predict long-term incident metabolic syndrome and diabetes: The Coronary Artery Risk Development in Young Adults Study. The International Journal of Eating Disorders, 52(3), 304–308. 10.1002/eat.23020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yoon C, Mason SM, Hooper L, Eisenberg ME, & Neumark-Sztainer D (2020). Disordered Eating Behaviors and 15-year Trajectories in Body Mass Index: Findings From Project Eating and Activity in Teens and Young Adults (EAT). The Journal of Adolescent Health : Official Publication of the Society for Adolescent Medicine, 66(2), 181–188. 10.1016/j.jadohealth.2019.08.012 [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.
Supplementary Materials
Data Availability Statement
CARDIA data are available upon reasonable request from the CARDIA Coordinating Center. CARDIA investigators are eager to collaborate with investigators interested in using CARDIA data. Please see the CARDIA website (https://www.cardia.dopm.uab.edu) for publications policies and for a list of CARDIA investigators. CARDIA data are also publicly available on the NIH-supported BioLINCC and dbGaP platforms.