اثرات متقابل استرس، خویشتن داری غذایی و عدم بازداری بر اشتها

Previous laboratory studies of disinhibited eating in response to stress have had varied outcomes. Since recent research implies that disinhibited eating might be observed when using the Thre-Factor Eating Questionnaire restraint (TFEQ-R) measure when scores on the TFEQ disinhibition (TFEQ-D) scale were also used, the present study investigated the disinhibitory effects of stress on eating in women classified using both TFEQ-R [high R (HR) vs. low R (LR)] and TFEQ-D [high D (HD) vs. low D (LD)] scores. Twenty women in each restraint (R) or disinhibition (D) combination were tested in either a stress or no-stress condition followed by a test lunch. Women classified as LR-HD consumed more than the other groups in the no-stress condition and reduced intake in response to stress, whereas HR-HD and LR-LD both ate more in the stress than no-stress conditions. HD consumed more sweet foods regardless of stress, whereas HR ate less savoury foods than LR. Mood data confirmed the success of the stress manipulation on affective state and also suggested that HD were more responsive to stress. Overall, these data imply that tendency to overeat, as measured by the TFEQ-D scale, is a better predictor than restraint in predicting short-term eating in response to stress.

Dietary restraint describes the cognitively mediated effort individuals make to restrict food intake in order to control body weight (Herman & Mack, 1975). Herman and Polivy (1980) claimed that women who score high on measures of restraint (HR) are ultimately unsuccessful at restricting food intake and develop abnormal eating patterns characterised by dieting followed by periodic overeating. This periodic overeating is believed to be due to certain events, “disinhibitors,” which trigger an overeating episode. The experience of stress has been shown to be one such disinhibitor (for a review, see Greeno & Wing, 1994). However, the outcome of previous research evaluating the effects of stress on eating in HR has varied depending on the measure of restraint used. Studies using the traditional measure of restraint, the Revised Restraint Scale (RRS) (Herman & Polivy, 1980), generally report disinhibited eating by HR women in response to a stressor (e.g., Herman et al., 1987 and Ruderman, 1985); whereas studies using the restraint scale of the Dutch Eating Behaviour Questionnaire (DEBQ) (Van Strein Frijters, Bergers, & Defares, 1986) or the Three-Factor Eating Questionnaire (TFEQ) (Stunkard & Messick, 1985) have not found disinhibited eating in HR (for example, see Oliver et al., 2000, Ridgway & Jeffrey, 1998 and Steere & Cooper, 1993). A possible explanation for the latter results comes from the suggestion that individuals with high TFEQ-R or DEBQ-R scores consist of both successful dieters and individuals who are likely to diet and overeat (i.e., unsuccessful dieters; Lowe & Maycock, 1988 and Tuschl, 1990). Westenhoefer (1991), Westenhoefer, Broeckmann, Munch, and Pudel (1994), Westenhoefer, Pudel, and Maus (1990), and Westenhoefer, Stunkard, and Pudel (1999) suggest that one way to differentiate these two groups of dieters is by assessing scores on the TFEQ disinhibition (D) scale with high D (HD) associated with unsuccessful dieting and low D (LD) associated with successful dieting. Accordingly, Westenhoefer et al. (1994) replicated Herman and Mack's (1975) preload study with subjects categorised into four subgroups: HR-HD, HR-LD, LR-HD, and LR-LD, and reported that only the HR-HD group ate more (disinhibited) in response to a preload, while all other groups ate slightly less after the preload. Moreover, the HR-LD group ate significantly less than the HR-HD group in both preload and no-preload conditions, reinforcing the concept that HR individuals exhibit different behavioural strategies in relation to eating, which are reflected in the level of “disinhibition” as measured by the TFEQ-D. The present study investigated whether the reason for a “failure” to find disinhibited eating in response to stress in HR women categorised according to the TFEQ was due to differences in levels of “disinhibition” within LR and HR individuals. Hence, women were categorised into four groups on the basis of both TFEQ-R and TFEQ-D, and they participated in either a stress condition or control (no stress) condition, after which they consumed a multi-item meal. We predicted that only individuals classified as HR-HD would eat more in response to induced stress. Since the TFEQ-D specifically assesses overeating in response to stress (for example, “When I feel anxious, I find myself eating”), it was also predicted that LR-HD would increase their food intake in the stress compared to control condition. The majority of laboratory studies investigating food intake in response to a stressor have measured intake of a single type of food (for example, different flavours of an ice cream, Heatherton, Herman, & Polivy, 1991). However, prospective studies, for example, Wardle, Steptoe, Oliver, and Lipsey (2000), have found that specific types of food (high energy, high fat, sweet foods) were selected by HR (measured by the DEBQ-R) in response to stress, whereas LR exhibited no difference in food intake or selection of particular foods. Wardle et al. also reported a significant positive relationship between perceived stress and intake of sweet and high-fat foods in HR only. The findings of Wardle et al. indicate that when given the option to consume a variety of foods rather than a single type of food, HR may exhibit stress-induced hyperphagia in the laboratory. Thus, the present study used a multi-item meal, with additional predictions that women classified as HR-HD would preferentially consume snacks/sweet foods in response to stress in line with Westenhoefer et al. (1994). Several different “stress procedures” have been utilised to investigate the effects of an experimental stressor on food intake in individuals differing in restraint (see Greeno & Wing, 1994). The most appropriate procedures appear to be tasks involving ego threats (where there is the possibility of task failure or negative evaluation) as these procedures elicit increases in stress hormones such as cortisol Berger et al., 1987, Hubert & De Jong-Meyer, 1989 and Kirschbaum & Hellhammer, 1989 and induce negative affect (Ruderman, 1985). The stress procedure used in the present study was developed from previous procedures to maximise the likelihood of ego threat and consisted of the concept formation (CF) task (Baucom & Aiken, 1981) and a mental arithmetic (MA) task used in other stress studies (see Kirschbaum, Pirke, & Hellhammer, 1993). Although all subjects were predicted to exhibit increases in dysphoric mood posttest in the stress condition compared to baseline levels, previous reports that HR is associated with poor coping strategies in response to stressors (e.g., Fryer et al., 1997, Mayhew & Edelmann, 1989, Paa & Larson, 1998 and Sanftner & Crowther, 1998) suggest that compared to LR women, HR may exhibit higher levels of posttest negative affect in the stress condition, and this hypothesis was tested.