Research reportMultiple types of dieting prospectively predict weight gain during the freshman year of college
Introduction
The rising prevalence of obesity is taking an increasingly severe toll on the nation's health. Approximately two-thirds of American adults are overweight or obese (Hedley et al., 2004). In addition, obesity is recognized by the World Health Organization as one of the top 10 global health problems (Kellner & Helmuth, 2003).
Research has shown that most people who successfully lose weight in lifestyle modification programs regain it within several years (Sarwer & Wadden, 1999). Poor long-term treatment outcome, combined with the fact that huge numbers of people are affected, indicates that prevention may be a more productive approach to fighting the obesity epidemic. Because obesity often produces detrimental physical and psychological consequences across the life span, it is important to intervene to prevent obesity before adult behavioral patterns (e.g., high fat intake, sedentary behavior) are solidified.
A recent meta-analysis of obesity prevention studies involving children, adolescents, or young adults (Stice, Shaw, & Marti, in press) found that only one in five studies showed a significant weight gain prevention effect. Since interventions that target those at known risk for weight gain are likely to be most cost-effective, it is important to empirically establish predictors of future weight gain. One strategy for identifying predictors of weight gain is to examine periods of life that have been associated with a high risk of weight gain.
The freshman year of college has been identified as one such high-risk period. Prospective studies have found that college freshmen show significant increases in BMI, fat mass, weight, and rates of obesity (Anderson, Shapiro, & Lundgren, 2003; Butler, Black, Blue, & Gretebeck, 2004; Hodge, Jackson, & Sullivan, 1993) and that these increases are significantly greater than those observed in similarly aged adolescents that do not attend college (Hovell, Mewborn, Randle, & Fowler-Johnson, 1985). For example, Hovell et al. found that 11% of female first-year college students showed onset of overweight or obesity during the first year of college, relative to 2% among similarly aged adolescents not attending college. The average female student gains between 1.7 and 3.1 kg during the freshman year (Anderson et al., 2003; Garay, 2004; Levitsky, Nussbaum, Halbmaier, & Mrdjenovic, 2003). The transition from high school to college is thought to be associated with an increased risk for weight gain because it is often accompanied by a reduction in physical activity (Butler et al., 2004) and increased consumption of high-fat foods and alcohol (Anderson et al., 2003). Conceptually, this is a critical developmental transition because, for many adolescents, this is the first time they are solely responsible for self-regulation regarding caloric intake and physical activity. Because weight gain occurs in a relatively short period of time, the freshman year of college is a desirable time to study risk factors for weight gain.
Ironically, some of the most robust prospective predictors of weight gain are measures of restrained eating (Drapeau et al., 2003; Klesges, Isbell, & Klesges, 1992; Stice, Cameron, Killen, Hayward, & Taylor, 1999; Stice, Presnell, Shaw & Rohde, 2005). The Restraint Scale (RS, Herman & Mack, 1975) is a widely used measure of dietary restraint that has been associated with a variety of anomalies in appetite (Herman & Polivy, 1984; Lowe (1993), Lowe (2003)). Some authors have argued that the chronic on-again, off-again dieting pattern shown by restrained eaters contributes to appetite disregulation, disordered eating, increased metabolic efficiency, and eventual weight gain (Garner & Wooley, 1991; Polivy & Herman, 1985; Tuschl, Platte, Laessle, Stichler, & Pirke, 1990).
Lowe (1993) has noted that restrained eating is a multidimensional construct that can be decomposed into three factors. The first factor is called Frequency of Dieting and Overeating, which reflects a history of going on and off diets. Heatherton, Herman, Polivy, King, and McGree (1988) referred to this pattern as “unsuccessful dieting.” The second factor is called Current Dieting. The consequences of current dieting have been investigated by asking if an individual is currently on a diet to lose weight (e.g., Lowe, Whitlow & Bellwoar, 1991) and by putting people on diets that produce measurable weight loss (e.g., Presnell & Stice, 2003). The third factor, Weight Suppression, reflects the long-term maintenance of a weight loss. Since measures of restrained eating predict weight gain, examining these specific types of dieting could identify which one(s) might account for the prediction of weight gain by measures of restraint. If some dieting types predict weight gain better than others, it may help clarify the reasons behind the counter-intuitive finding that measures of restrained eating predict weight gain rather than weight loss.
We also administered measures of disinhibited eating and emotional eating. Both eating patterns could produce energy intake that is beyond energy needs and result in weight gain over time. The Disinhibition scale from the Eating Inventory (Stunkard and Messick, 1985) measures a tendency to overeat due to a wide variety of stimuli (e.g. social events, seeing others eating, negative moods) and has been shown to prospectively predict weight gain (McGuire, Wing, Klem, Lang, & Hill, 1999). The Emotional Eating Scale from the Dutch Eating Behaviors Questionnaire (van Strien, Frijters, Bergers, & Defares, 1986) measures a tendency to eat when under emotional stress.
The purpose of the current study is threefold. First, we will reexamine two measures of restrained eating as predictors of weight gain. Second, we will examine the three types of dieting identified in Lowe's (1993) three-factor model as predictors of weight gain. Third, we will determine if measures of disinhibition or emotional eating predict weight gain (possibly as a result of restrained eaters’ greater risk for one or both of these types of eating). This study was carried out with college freshmen because of their previously reported susceptibility to weight gain. It involved only women because past findings in the restraint literature have focused primarily on women.
Section snippets
Participants
Seventy-two incoming freshman females between the ages of 18 and 19 (M=18.06, SD=0.23) years were recruited at Drexel University in Philadelphia. Participants’ mean body mass index (BMI) was 21.9 (SD=2.4; range=17.4–26.6). Participants were not taking any medications known to affect appetite or weight and had not suffered from an eating disorder in the year prior to enrollment in the study. Recruitment was performed through (1) announcements made during an orientation held for freshmen the
Restraint scale
The Revised Restraint Scale (Herman & Polivy, 1980) is a 10-item measure of dietary concerns and weight fluctuations. The reliability and validity of the scale have been supported in a number of studies (Heatherton et al., 1988). The Restraint Scale predicts different eating patterns of restrained and unrestrained eaters in response to negative affect and high-calorie preloads (Herman & Polivy, 1980).
Anthropometry
Body weight, measured while the subject was barefoot and wearing light clothing, was determined
Results
Sixty-nine of the original 72 participants completed all four-assessment points. All analyses are based on the 69 participants with complete data. From the baseline assessment to the 1-month assessment in October, 83% of the participants gained weight and 17% lost weight. The average weight gain across the full sample from September to October was 0.91 kg, from September to January was 1.90 kg, and from September to May was 2.08 kg.
To address the aim of testing whether restraint, overeating or
Discussion
None of the self-report measures of restraint or overeating predicted weight gain during the freshman year of college. Previous studies have found that measures of restrained eating prospectively predict weight gain (Klesges et al., 1992; Stice et al., 1999; Stice, Presnell, et al., 2005). These studies, however, had many more subjects (average N=290) than the current study and therefore had more power to detect a relationship.
In addition, while two of these studies used the RS as a predictor (
Acknowledgment
Supported by a Synergy grant from MCP Hahnemann University.
References (49)
- et al.
Assessment of emotional, externally induced and restrained eating behaviour in nine to twelve-year-old obese and non-obese children
Behaviour Research and Therapy
(1997) - et al.
The effects of repeated cycles of weight loss and regain in rats
Physiology and Behavior
(1986) - et al.
Is restrained eating a risk factor for weight gain in a general population
American Journal of Clinical Nutrition
(2006) - et al.
Current dieting, weight loss history, and weight suppression: Behavioral correlates of three dimensions of dieting
Addictive Behaviors
(1997) - et al.
Confronting the failure of behavioral and dietary treatments for obesity
Clinical Psychology Review
(1991) - et al.
Risk of excess weight gain in university women: A three-year community controlled analysis
Addictive Behaviors
(1985) - et al.
Weight loss and sweetness preferences: the effects of recent versus past weight loss
Physiology and Behavior
(1991) Restrained eating and dieting: Replication of their divergent effects on eating regulation
Appetite
(1995)- et al.
Weight suppression predicts weight gain during inpatient treatment of bulimia nervosa
Physiology and Behavior
(2006) - et al.
Cognitive restraint, weight suppression, and the regulation of eating
Appetite
(1988)