September 5, 2008

Risk Factors Associated With Overweight and Obesity in College Students

By Desai, Melissa N Miller, William C; Staples, Betty; Bravender, Terrill

Abstract. College obesity is increasing, but to the authors' knowledge, no researchers to date have evaluated risk factors in this population. Objective: The authors assessed whether abnormal eating perceptions and behaviors were associated with overweight in college students. Participants and Methods: A sample of undergraduates (N = 4,201) completed an online survey containing demographic questions and the Eating Attitudes Test-26 (EAT-26) questionnaire. The authors stratified participants into normalweight and overweight/obese groups and evaluated physical activity, EAT-26 score, purging behaviors, and answers to selected overweight screening questions on the basis of body mass index. A modified EAT- 26 score of >/= 11 was associated with overweight (p = .016). Results: Compared with normal-weight peers, overweight participants displayed an increased fear of bingeing, preoccupation with food, desire to be thinner, and engagement in dieting behavior. Mean body mass index also increased with age and physical inactivity (p

Overweight and obesity are increasing nationally. Between 1976 and 2004, the prevalence of overweight among children aged 6 to 11 years tripled from 6% to 18% and the prevalence of overweight and obesity among adults aged 20 to 74 years increased from 47% to 66%.1 Similar trends are seen in college students. The percentage of adolescents aged 12 to 19 years who are overweight or obese more than tripled, from 5% to 17%, between 1976 and 2002.1

The college years are highly influential in shaping adult behaviors,2 particularly with regard to diet, physical activity, and other lifestyle habits. Interventions aimed at the college population may help reduce the rate of overweight during the transition from adolescence to adulthood and thereby prevent some of the long-term health consequences of obesity, which include coronary heart disease, hypertension, type 2 diabetes, and dyslipidemia.3

Obesity in adolescence tends to persist into adulthood. Results from both the National Longitudinal Study of Adolescent Health and the National Health and Nutrition Examination Survey show that during the transition from adolescence to young adulthood, a high proportion of adolescents become obese and remain obese.4 In addition, because obesity is not easily reversible, those who are obese or develop obesity as young adults are at increased risk of obesity through adulthood.5 Although these studies highlight important patterns, they have not identified abnormal eating attitudes or specific behaviors that may contribute to and perpetuate obesity during the transition from adolescence to adulthood. A clearer understanding of these factors may prove useful in the treatment of adolescent obesity by helping in the development of multifaceted interventions that not only address weight loss but also target associated disordered behaviors. Therefore, to characterize the abnormal eating attitudes and behaviors linked to overweight and obesity in early adulthood, we examined data collected via an electronic, cross-sectional survey of undergraduate students at 3 large, geographically similar universities.


Survey Instrument and Administration

In January 2002, undergraduates at 3 North Carolina universities received a recruitment e-mail with nutritional and behavior assessments. The e-mail provided a link to an encrypted Web-based survey. Institutional review board approval was waived for the study, given that participation was voluntary and anonymous. We did not offer incentives for participation.

The survey included questions about age, height, weight, undergraduate institution, extracurricular activities, athletic involvement, and eating disorder history, as well as the Eating Attitudes Test-26 (EAT-26) questionnaire. The EAT-26 questionnaire is based on the original EAT-40 questionnaire that Garner and Garfinkel developed to screen for anorexia symptoms.6 The original questionnaire has strong internal consistency and is both sensitive and specific for identifying respondents at increased risk for eating disorders.6,7 The abbreviated EAT-26 is highly correlated with the EAT-40 (r = .97 for female university students and r = .98 for anorexia nervosa).6,7 Each survey question is scored on a 6- point forced Likert scale ranging from 1 (never) to 6 (always). Scores >/= 20 on the EAT-26 have traditionally been suggestive of anorexia nervosa or bulimia nervosa.8 Recently, scores >/= 11 have been associated with an increased risk for binge eating disorder.9

All undergraduates at one university received the recruitment e- mail, but because of e-mail server difficulties, only a sample of students at the other 2 universities received the e-mail. A total of 5,144 students at all 3 universities completed surveys. We excluded from analysis those who did not report either height or weight and those who did not complete all components of the EAT-26. We also excluded respondents with a body mass index (BMI; in kg/m^sup 2^)


We stratified survey participants into 2 groups on the basis of BMI classifications: normal-weight (18.5 = BMI /= 30). Dependent variables included total EAT-26 score and validated EAT-26 scores of 11 and 20. In addition, we assessed the relationship between BMI and abnormal behaviors, including purging, excessive physical activity, and previous treatment for eating disorders. We also examined self- reported Greek (ie, sorority or fraternity) affiliation to determine whether members were at increased risk of overweight or obesity as compared with the general population of students.

We selected 7 items from the EAT-26 survey that we hypothesized would be useful screening items for disordered eating thoughts and behaviors associated with overweight and obesity:

1. Find myself preoccupied with food

2. Have gone on eating binges where I feel that I may not stop

3. Vomit after I have eaten

4. Am preoccupied with a desire to be thinner

5. Feel that food controls my life

6. Engage in dieting behaviors

7. Have the impulse to vomit after meals

We evaluated participant responses as dichotomous outcomes, with clustering of always, usually, and often responses and sometimes, rarely, and never responses. We then analyzed them by BMI classification to determine whether overweight and obesity were associated with disordered eating perceptions and behaviors.

Statistical Analysis

We analyzed data using Stata 8.0 (StataCorp, College Station, TX). We calculated descriptive statistics for the overall group of respondents on the basis of BMI. For the continuous outcome of BMI, we used the Wilcoxon rank sum test to assess its relationship with total EAT-26 score. We performed multiple linear regression modeling to examine the relationship between BMI and (1) activity level, (2) extreme weight loss measures, and (3) social affiliations. For the dichotomous outcome of overweight, we used Pearson's chi-square tests to assess associations with specific EAT-26 cutoff scores, physical inactivity, purging behaviors, and previous treatment for eating disorders. We used logistic regression to determine the relationship between overweight status and the same independent variables.


A total of 5,144 students completed the survey, and 4,201 met eligibility requirements. Of the eligible students, 1,522 were from Duke University, 1,673 were from the University of North Carolina at Chapel Hill (UNC), and 1,001 were from North Carolina State University (NCSU). Because an unknown number of students at UNC and NCSU actually received the survey, the precise response rates are unknown. However, at Duke University, all undergraduates received the survey, and the calculated response rate for this population was approximately 26%. Of the eligible respondents, 3,574 (67.5%) were female, 4,178 (99%) were aged 18 or older, and 3,457 (82.5%) were Caucasian. These proportions are consistent with the demographics of survey respondents who did not meet inclusion criteria across all 3 universities. Average demographic characteristics for the 3 institutions in 2002 include a population that was 55% female and 73% Caucasian. In regard to BMI, 78.5% of the participants were normal weight, 16.8% were overweight, and 4.7% were obese. Participants were in all years of education, with 18.9% of students reporting a Greek affiliation (see Table 1).

Overweight or obese students had higher average total EAT- 26 scores than did normal-weight students (10 vs 9, p = .0001). We evaluated the EAT-26 as a dichotomous variable to assess whether traditional survey cutoff scores showed an association in our population. We found no such association with the usual cutoff score >/= 20, yet a modified EAT-26 score of >/= 11 was associated with overweight such that 31.9% of those who were overweight had an EAT- 26 score >/= 11, and 27.8% of those with a normal BMI had a score >/ = 11 (p = .016). All disordered eating screening questions selected from the EAT-26 assessed either overweight/obese status or purging behavior (see Table 2). Whereas 11.4% of overweight individuals feared engaging in bingeing episodes without stopping, fewer than 8% of normal-weight participants had the same fear (p = .001). Similarly, overweight respondents had a significantly greater preoccupation with food than did normal-weight respondents (33.3% vs 23.7%) and a greater desire to be thinner (42.4% vs 31.3%). In addition, overweight respondents felt that food controlled their lives and engaged in dieting behaviors more than did those who were normal weight (p

Multivariate linear regression modeling revealed that mean BMI was associated with physical activity, previous treatment for disordered eating, year in college, and Greek affiliation. Overweight participants also were more likely to report complete physical inactivity (46%) than were normal-weight participants (37%). More specifically, those reporting physical inactivity had higher average BMIs (23.3 vs 22.9, p

Multiple logistic regression analysis revealed that students who reported a lack of physical activity were more likely to be overweight (odds ratio [OR] = 1.4, 95% confidence interval [CI] = 1.2-1.6), as were those reporting prior treatment for an eating disorder (OR = 1.7, 95% CI = 1.2-2.5). An EAT-26 score >/= 11 was also associated with an increased risk of overweight or obesity (OR = 1.3, 95% CI = 1.1-1.6).


In the United States, the prevalence of obesity is on the rise, and the college-aged population is at particular risk for developing obesity and maintaining it into adulthood.2,4 We found that college students who reported inactivity were more likely to be overweight or obese than were their peers. These students also reported a higher frequency of being preoccupied with food, feeling that food controlled their lives, and desiring to be thinner. They also engaged in more dieting and bingeing behaviors than did their normal- weight peers and had higher total scores on a validated eating disorders screening test.

These findings suggest that overweight and obesity are associated with distorted eating behaviors and perceptions. Therefore, screening for these attitudes in overweight youths may facilitate the development and success of obesity treatment programs.10 Although the EAT-26 questionnaire has been validated as a screening tool for anorexia nervosa and bulimia nervosa, researchers have not yet used it to assess overweight or obesity. Studies testing the accuracy of the EAT-26 in detecting disordered eating behaviors in obese patients showed that an EAT-26 cutoff value of 11, the same value used in our study, was optimal in screening for disordered eating behaviors.9 However, additional diagnostic analyses with different patient populations are needed to confirm the usefulness of the EAT-26 for routine clinical use in regard to obesity.

Our results suggest that selected questions from the EAT- 26 survey may be useful markers for disordered behaviors in overweight or obese students. Specifically, a preoccupation with food, feeling that food controlled one's life, the desire to be thinner, and bingeing and dieting behaviors were all positively associated with overweight and obesity. The identified affective tendencies have been associated with higher BMI and severity of bingeing behavior, which in turn are correlated with higher levels of body dissatisfaction and dieting frequency.11-13

Overweight and obesity remain a treatment challenge because we have yet to fully understand the barriers to weight loss and how best to approach these challenges. For eating disorders such as anorexia and bulimia nervosa, addressing disordered attitudes, rather than just focusing on weight gain, is an essential component of treatment programs.14-16 Although the eating behaviors leading to obesity are not technically considered disordered eating behaviors, overweight individuals do exhibit disordered eating symptoms, such as feelings of ineffectiveness, perfectionist attitudes, low interoceptive awareness, low selfesteem, emotional eating, and depressive symptoms.14-17 These psychological concerns have implications for the treatment of overweight in that a simple dietary intervention or physical education course may not be sufficient to generate a sustainable reduction in young adult overweight and obesity. As has been shown with eating disorders, an understanding of the psychological component of the illness is crucial in the development of effective treatment interventions. 14- 17 Assessments of weight loss interventions suggest that psychological counseling preceding weight loss efforts may enhance long-term success.16,18,19 Therefore, a multifaceted approach that includes education about healthy eating attitudes as well as emotional strengthening and personality evaluation is likely to be more effective than weight reduction alone in the treatment of overweight populations.

In addition to identifying and treating disordered attitudes in obese youths, interventions aimed at behavior modification are also essential to treatment programs. We suggest that college and university administrators be more proactive about educating students on healthy behaviors. Researchers have found that despite the poor eating habits that manifest in college, the overall mean caloric intake in college freshmen may actually decrease.20-22 Yet at the same time, body weight tends to increase during the college years, suggesting that physical activity must also be decreasing during the transition from high school to college. It is widely believed that the prevalence of a sedentary lifestyle increases with age, and numerous studies have shown that anywhere from 20% to 68% of college students do not meet minimum physical activity recommendations.20- 26 In this study, complete physical inactivity during college was associated with overweight and obesity (p

Researchers27,28 who have found that inactivity tracks from adolescence to adulthood and that decreasing inactivity can positively affect obesity reduction have demonstrated the importance of reducing physical inactivity rates. Even normal-weight college students often resort to dietary weight control mechanisms rather than healthful eating and physical activity, and thus they too may remain inactive throughout their college lives.25 Therefore, to treat overweight youths, programs should focus on increased physical activity along with intensive counseling and behavioral programming.


Because overweight and obesity were secondary outcomes assessed from the original survey data, these results may lack information on weight-related outcomes. In addition, because the study sample was a convenience sample, it may have been biased toward those who were more healthconscious and thus more likely to return the survey. Furthermore, because we excluded those who failed to report their height or weight, obese students may have been underrepresented in the sample because they may have been less willing to report these measures. The large sample size (N = 4,201), multicenter study base, and decent response rates, however, help overcome some of the sampling limitations. Although the study was limited to self- reported measures, previous studies have suggested that college students report weight and height accurately, with correlation coefficients of .987 and .981, respectively.29,30 Of note, the study population showed overweight and obesity rates (16.8% and 4.7%, respectively) similar to the population in the National College Health Risk Behavior Study.31

Future Research

Overweight and obesity in college students is associated with high levels of physical inactivity, as well as a number of disordered eating attitudes and behaviors. BMI increases with time in college, but the risk factors and habits influencing young adult weight gain require further evaluation so that treatment programs can target these behaviors. The data obtained in this study were crosssectional and thus do not track behaviors over time or assess how the college environment directly influences an individual's normal behaviors. Future research should include longitudinal studies that more closely evaluate BMI and the determinants of obesity, including eating behaviors, physical activity trends, and body image perceptions. A greater understanding of how these determinants change from prematriculation through graduation can promote healthy habits that last a lifetime and lead to an effective intervention in the collegiate environment.


The authors thank Dr Margaret Gourlay (University of North Carolina, Chapel Hill, NC) for editing assistance, as well as the students at Duke University, the University of North Carolina, and North Carolina State University who participated in the study.


1. Center for Disease Control and Prevention. Prevalence of Overweight and Obesity Among Adults: United States, 2003-2004. Hyattsville, MD: National Center for Health Statistics; 2006.

2. Racette SB, Deusinger SS, Strube MJ, Highstein GR, Deusinger RH. Weight changes, exercise, and dietary patterns during freshman and sophomore years of college. J Am Coll Health. 2005;53:245-251. 3. Kiess W, Reich A, Muller G, et al. Clinical aspects of obesity in childhood and adolescence-diagnosis, treatment and prevention. Int J Obes Relat Metab Disord. 2001;25(suppl 1):S75-S79.

4. Gordon-Larsen P, Adair LS, Nelson MC, Popkin BM. Fiveyear obesity incidence in the transition period between adolescence and adulthood: the National Longitudinal Study of Adolescent Health. Am J Clin Nutr. 2004;80:569-575.

5. McTigue KM, Garrett JM, Popkin BM. The natural history of the development of obesity in a cohort of young US adults between 1981 and 1998. Ann Intern Med. 2002;136:857-864.

6. Garner DM, Garfinkel PE. The Eating Attitudes Test: an index of the symptoms of anorexia nervosa. Psychol Med. 1979;9:273-279.

7. Garner DM, Olmsted MP, Bohr Y, Garfinkel PE. The Eating Attitudes Test: psychometric features and clinical correlates. Psychol Med. 1982;12:871-878.

8. Pastore DR, Fisher M, Friedman SB. Abnormalities in weight status, eating attitudes, and eating behaviors among urban high school students: correlations with self-esteem and anxiety. J Adolesc Health. 1996;18:312-319.

9. Orbitello B, Ciano R, Corsaro M, et al. The EAT-26 as screening instrument for clinical nutrition unit attenders. Int J Obes (Lond). 2006;30:977-981.

10. Anstine D, Grinenko D. Rapid screening for disordered eating in college-aged females in the primary care setting. J Adolesc Health. 2000;26:338-342.

11. French SA, Perry CL, Leon GR, Fulkerson JA. Dieting behaviors and weight change history in female adolescents. Health Psychol. 1995;14:548-555.

12. Friedman MA, Schwartz MB, Brownell KB. Differential relation of psychological functioning within the history and experience of weight cycling. J Consult Clin Psychol. 1998;66:646-650.

13. Friedman MA, Wilfey dE, Pike KM, Striegel-Moore RH, Rodin J. The relationship between weight and psychological functioning among adolescent girls. Obes Res. 1995;3:57-62.

14. Grissett NI, Fitzgibbon ML. The clinical significance of binge eating in an obese population: support for BED and questions regarding its criteria. Addict Behav. 1996;21:57-66.

15. de Zwaan M, Mitchell JE, Seim HC, et al. Eating-related and general psychopathology in obese females with binge eating disorder. Int J Eat Disord. 1994;15:43-52.

16. Dingemans AE, Bruna MJ, van Furth EF. Binge eating disorder: a review. Int J Obes Relat Metab Disord. 2002;26:299-307.

17. Walsh BT, Devlin MJ. Eating disorders: progress and problems. Science. 1998;280:1387-1390.

18. Yanovski SZ. Binge eating disorder and obesity in 2003: could treating an eating disorder have a positive effect on the obesity epidemic? Int J Eat Disord. 2003;34 (suppl):S117-S120.

19. de Zwaan M, Mitchell JE. Binge eating in the obese. Ann Med. 1992;24:303-308.

20. Dinger MK, Waigandt A. dietary intake and physical activity behaviors of male and female college students. Am J Health Promot. 1997;11:360-362.

21. Haberman S, Luffey D. Weighing in college students' diet and exercise behaviors. J Am Coll Health. 1998;46:189-191.

22. Anding JD, Suminski RR, Boss L. Dietary intake, body mass index, exercise, and alcohol: are college women following the dietary guidelines for Americans? J Am Coll Health. 2001;49:167- 171.

23. Burke SM, Carron AV, Eys MA. Physical activity context and university students' propensity to meet the guidelines Centers for Disease Control and Prevention/American College of Sports Medicine. Med Sci Monit. 2005;11:171-176.

24. July F, Hawthorne D, Elliot J, Robinson W. Weight management behaviors of African American female college students. ABNF J. 2003;14:71-72.

25. Clement JM, Schmidt CA, Bernaix LW, Covington NK, Carr TR. Obesity and physical activity in college women: implications for clinical practice. J Am Acad Nurse Pract. 2004;16:291-299.

26. Suminski RR, Petosa R, Utter AC, Zhang JJ. Physical activity among ethnically diverse college students. J Am Coll Health. 2002;51:75-80.

27. Epstein L, Saelens B, O'Brien J. Effects of reinforcing increases in active versus decreases in sedentary behavior in obese children. Int J Behav Med. 1995;2:41-50.

28. Gordon-Larsen P, McMurray RG, Popkin BM. Adolescent physical activity and inactivity vary by ethnicity: the National Longitudinal Study of Adolescent Health. Pediatrics. 1999;135:301-306.

29. Shapiro J, Anderson D. The effects of restraint, gender, and body mass index on the accuracy of self-reported weight. Int J Eat Disord. 2003;34:177-180.

30. Sciacca JP, Melby CL, Hyner GC, Brown AC, Femea PL. Body mass index and perceived weight status in young adults. J Community Health. 1991;16:159-168.

31. Centers for Disease Control and Prevention. Youth Risk Behavior Surveillance: National College Health Risk Behavior Survey- United States, 1995. MMWR. 1997;46:1-54.

Melissa N. Desai, MD, MPH; William C. Miller, MD, PhD, MPH; Betty Staples, MD; Terrill Bravender, MD, MPH

At the time of the study, Dr Desai was with Duke University Medical Center in Durham, NC. Drs Staples and Bravender are with the Department of Pediatrics at Duke University Medical Center. Dr Miller is with the Department of Medicine at the University of North Carolina, Chapel Hill.

Copyright (c) 2008 Heldref Publications


For comments and further information, address correspondence to Dr Melissa N. Desai, 7 Amherst Way, Princeton Junction, NJ 08550, USA (email: [email protected]).

Copyright Heldref Publications Jul/Aug 2008

(c) 2008 Journal of American College Health. Provided by ProQuest LLC. All rights Reserved.