cognitive changes to preloading in restrianed and unrestrained eaters as measured by the Stroop task more

Cognitive Changes to Preloading in Restrained and Unrestrained Eaters as Measured by the Stroop Task Jane Ogden Lisa Greville (Accepted 26 January 1993) Restraint theory has identified overeating in response to a high calorie preload as characteristic of restrained eaters. The present study evaluated cognitive and motivational changes to preloading using both self-report rating scales and Stroop tasks. The results suggest that the restrained eaters responded to a high calorie preload with increased feelings of rebelliousness, defiance, and a desire to challenge the limitations set by the diet, described as an active state of mind, as measured by both the Stroop task and the rating scales. The restrained eaters also showed retardation in the color naming of the body size words and food words after the high calorie preload, suggesting that the consumption of a forbidden food may increase the dieter's concern about food and her shape and weight. The results are discussed in terms of the effects of preloading and possible mediating variables involved in the transition from successful restraint to overeating. 0 1993 by john Wiley & Son, Inc. Restraint theory (e.g., Herman & Mack, 1975; Herman & Polivy, 1984) was developed to evaluate factors associated with attempts to restrict food intake for the purpose of weight maintenance. Although at times restrained eaters are successful in their attempts to restrict their food intake (Kirkley, Burge, & Ammerman, 1988; Laessle, Tuschl, Kotthaus, & Pirke, 1989) restraint theory suggests that restrained eating is characterized by episodes of both successful undereating and compensatory overeating (Herman & Mack, 1975; Herman, Polivy, & Esses, 1987). In particular, restraint theory has identified the disinhibition of restraint that has been illustrated both in the laboratory and naturalistic setting. Disinhibition or the what the hell effect constitutes relative overeating in response to high calorie preloads, preloads believed to be high in calories, and lowered mood (Herman & Mack, 1975; Herman & Polivy, 1975; Spencer & Fremouw, 1979). The association between restraint and overeating resulted in a reconceptualization of restrained eating with research suggesting that restraint is best conceptualized lane Ogden, Ph.D., is Lecturer in Health Psychology, London University, UMDS, London, U.K. and Lisa Greville, B.Sc., is affiliated with the Springfield Hospital, London. Address reprint requests to Dr. lane Ogden, UMDS, Department of General Practice, 80 Kennington Road, London SEI. United Kingdom. Jnternational journal of Eating Disorders, Vol. 14, No. 2, 185- 195 (1993) (c) 1993 by john Wiley & Sons, Inc. CCC 0276-3478/93/020185-11 186 Ogden and Creville as an intention to eat less, which is only sporadically realized (Heatherton, Polivy, & Herman, 1991; Ogden, 1993). Although restraint is clearly associated with overeating, the mechanisms mediating the transition from successful to unsuccessful restraint are still unclear. In particular, the role of preloading in precipitating disinhibition requires clarification. Preloading may trigger disinhibition via changes in mood and cognitive state. Lowered mood has been implicated as a precursor to overeating both in restrained eaters and bulimics (Herman & Polivy, 1975; Cooper & Bowskill, 1986). It is possible that lowering mood may independently trigger overeating or act indirectly as a response to preloading. In addition, cognitive changes have also been implicated as triggers to disinhibition (Jansen, Merckelbach, Oosterlaan, Tuiten, & Van der Hout, 1988; Herman & Polivy, 1984; Ogden & Wardle, 1991). Such changes in the restrained eater’s cognitive state may also occur as a result of preloading. Preloading may also trigger overeating via changes in the restrained eater’s motivational state. Laboratory studies of alcoholics suggest that alcohol craving can be induced by priming with a small dose of alcohol (Hodgson, Rankin, & Stockwell, 1979). Preloading could prime hunger and consequently enhance the probability of overeating. In an attempt to identify possible mediating variables in the restraint/overeating link, Ogden and Wardle (1991) evaluated cognitive and emotional responses to high calorie preloads in restrained and unrestrained eaters. Checklists, visual analogue scales, and an interview were used to measure emotional and cognitive states before and after the consumption of either a high or low calorie food. The results suggested that consumption of a high calorie food increased feelings of rebelliousness and defiance in the restrained eaters, similar to the state of what the hell described by Herman and Polivy (1984). In addition, the interview data suggest that the restrained eaters also responded to the high calorie food with increased feelings of giving in, passivity, and resignation suggesting motivational collapse. Ogden and Wardle (1991)labelled these the active and passive states and suggest that such changes in cognitive state following preloading may mediate the overeating characteristic of restrained eaters. The results also suggested that preloading may result in decreased anxiety supporting psychosomatic theories of overeating and of eating for comfort (Bruch, 1974). The preoccupation with food is also characteristic of restrained eaters. Research suggests that increasing restraint results in an increase in the preoccupation with food (Warren & Cooper, 1988; Ogden, 1992) and that this increase may be a contributing factor to associated increases in episodes of overeating. Preloading may increase the restrained eater’s preoccupation with food triggering disinhibitory behavior. The aim of the present study was to examine possible changes in cognitive, emotional, and motivational states in restrained eaters following preloading with a high calorie forbidden food. The study aimed to examine factors that may mediate the transition from restraint to the disinhibition of restraint. Previous research has predominantly used self-report measures to assess changes in the subjective state of the individual. Although sensitive, such measures are open to factors such as denial, and rely upon a degree of intellectualization by the subjects to enable them to consciously label their subjective experience. The present study used self-report rating scales and an adaptation of the Stroop task (Stroop, 1935) to assess possible effects of preloading. The Stroop paradigm has been used extensively in research with anxious subjects, phobics, and sufferers of post-traumatic stress disorder (Mathews & MacLeod, 1985; Mcnally, Reimann, & Eunsil, 1990; Foa, Feske, Murdock, Kozak, & McCarthy, 1991; Cognitive Changes to Preload 187 see MacLeod, 1991 for review of the Stroop literature). The research suggests that processing resources are directed towards material that is relevant to the patients’ concerns, resulting in retarded color naming on the Stroop task. This approach has also been applied to the evaluation of patients with eating disorders. The research suggests that anorexics consistently show retardation on tasks involving food-related words (Channon, Hemsley, & de Silva, 1988; Ben-Tovim, Walker, Fok, & Yap, 1989; Ben-Tovim & Walker, 1991; Fairburn, Cooper, Cooper, McKenna, et al., 1991) and suggest a retardation effect if the words are associated with body size (Ben-Tovim & Walker, 1991; Fairburn et al., 1991). In addition, research suggests that food deprivation may result in retardation of color naming of food-related words in normal subjects (Channon & Hayward, 1990). The present study used rating scales to assess changes in depression, anxiety, hunger, and passive and active cognitive states, and an adaptation of the Stroop task to assess changes in passive and active cognitive states following preloading in restrained and unrestrained eaters. In addition the Stroop task was used to assess changes in the preoccupation with food and body size as it was predicted that such changes may also contribute to the disinhibition of restraint. It was predicted that restrained eaters would show a different profile of responses to a high calorie forbidden preload than to a low calorie neutral preload, and a different profile of response to unrestrained eaters. METHODOLOGY Subjects Fifty-six female students from a London University took part in the study. All subjects were categorized as restrained or unrestrained eaters according to a median split in their scores on the restrained eating section of the Dutch Eating Behaviour Scale (DEBQ; Van Strien, Frijters, Bergers, & Defares, 1986). This scale was used as reports suggest that it is easier to complete by British subjects than other scales (Wardle, 1986) and that it selects a population of restrained eaters with a tendency to show both successful and failed restraint (Ogden, 1993). The restrained eaters (n = 28) had a mean restraint score of 35.75 k 6.2, and the unrestrained eaters ( n = 28) had a mean restraint score of 17.6 2 4.5. Subjects were also categorized according to their scores on the restraint scale (Herman & Mack, 1975; Heatherton et al., 1991). Two subjects did not complete the questionnaire and the rest remained in the same subject group as when categorized by the DEBQ. Subjects ranged in age from 19 to 25 years. Design The study consisted of both a within and between subjects design with two experimental conditions. All subjects took part in the study once and completed a set of Stroop tasks and rating scales before and after eating either a high or low calorie preload. Procedure On arrival subjects were given instructions concerning the Stroop task and asked to perform the task as quickly and accurately as possible. They then completed eight Stroop tasks and a set of rating scales. Each subject was then given either a high or low calorie 188 Ogden and Creville preload. The high calorie preload was a single Twix bar that was consideredforbidden but not too substantial. The low calorie preload was a cream cracker that was considered a neutral food. The subject was then left alone to eat the preload for 5 minutes. The subject then completed the same set of Stroop tasks and rating scales as before, and the additional restrained eating questionnaire (Van Strien et al., 1986). Measures The following measures were administered before (Time 1)and after (Time 2) the preload. Stroop Tasks The Stroop tasks were run using a BBC computer program. The program uses a matching procedure whereby the subject is presented with a stimulus word on the lefthand side of the monitor printed in color and a color word on the righthand side printed in white. The subject is asked to decide whether the color of the stimulus word is either the same or different to the color word. This presents the subject with two response options: same or different, which can be selected by pressing either of the response keys. The computer recorded total reaction times for each block of words. Each Stroop variable (active, passive, food words, body words) was represented by six stimulus words and six neutral control words. The stimulus and control words were matched according to word frequency and syllable length. The stimulus words were presented in a counterbalanced block followed by a counterbalanced block of their matched neutral words. The presentation of one variable consisted of the six words in randomized order presented six times. Therefore one set of Stroop words consisted of 36 stimulus words that had to be matched as either same or different to 36 colors also presented in randomized blocks of six. The order of presentation of the variables was also counterbalanced from Time 1 (before preload) and Time 2 (after preload) and between subjects. The following words were selected: Cognitive State The following cognitive state words were selected from those used by Ogden and Wardle (1991). The matched neutral words were selected from Kucera and Francis (1967). The active state and matched words were rebellious, hilarious; defiant, longstanding; challenge, object; daring, tower; reckless, compact; wild, pure. The passive state and matched words were submit, unlock; quit, rock; abandon, encounter; passive, vacant; defeat, double; resign, patron. Food and Body Size Words The following food and body size words and their matched neutral words were adapted from those used by Channon et al. (1988) the food words and matched words were cream, clock; cheese, bags; dinner, record; butter, powder; cake, boot; sugar, pencil. The body size words and matched words were large, form; chest, crowd; fat, grass; thigh, rust; shape, attic; weight, names. Cognitive Changes to Preload 189 Rating Scales Motivational State Hunger and fullness were measured using 100-mm visual analogue scales. The extremes were labelled ”not at all hungry/full” (0 mm) and “as hungry/full as I’ve ever been” (100 mm). Hunger symptoms were measured using a symptom checklist of six hunger symptoms (Monello & Mayer, 1969). Emotional State Anxiety and depression were rated using the appropriate factors from the Profile of Mood States (POMS; McNair, Lorr, & Droppleman, 1981). Cognitive State The active and passive states were measured using active and passive words (Ogden & Wardle, 1991). These words were intermingled with the POMS questionnaire. All subjects also completed the Restrained eating section of the DEBQ (Van Strien et al., 1986) at the end of the study. RESULTS The Stroop task results were analyzed by computing the difference between the reaction time for the stimulus words and their neutral matched words (Stimulus-control) at Time 1 and at Time 2 to create a pure reaction time. The pure reaction times at Time 1 and Time 2 were compared by condition (high preload vs. low preload) and restraint level (restrained eaters vs. unrestrained eaters) using analysis of variance (ANOVA) (SPSSPC). A change score for these pure reaction times was then calculated for both subject groups in both conditions (Time 2 - Time 1).The change score was also analyzed by condition and restraint level using ANOVA (SPSSPC). As specific predictions were made it was deemed justified to use one-tailed tests. The rating scales were also analyzed both at Time 1 and Time 2 and by computing change scores (Time 2 - Time 1). ANOVA (SPSSPC) was used to compare scores by condition and by restraint level. The results for both the Stroop tasks and the rating scales were also analyzed using analysis of covariance (ANCOVA) however this had no effect on the outcome. Stroop Tasks The change scores for the Stroop tasks are shown in Table 1. Active State At Time 1 the results showed no main effects of either condition or restraint level and no interactions for active state. However at Time 2, the results suggest a significant main effect of both condition [F(1,52) = 18.53, p < .001] and restraint level [F(1,52) = 12.15, p < .001] and a significant condition by restraint level interaction [F(1,52) = 2.74, p < .05]. Examination of the means suggests that at Time 2 the restrained eaters took 190 Ogden and Creville longer to color name the active state words after the high calorie preload (mean reaction time; 17.52 t 8.32) than after the low calorie preload (mean reaction time; 6.21 t 6.24), (t = 4.07, p < .001), and longer than the unrestrained eaters in the same condition (mean reaction time; 7.88 t 6.62), ( t = 3.44, p < .005). Analysis of the change scores (see Table 1) from pre to post preload as measured by the Stroop task for the active state words showed a significant main effect of both restraint level [F(1,52) = 7.46, p < .01] and condition [F(1,52) = 6.77, p < .01] but no significant restraint level by condition interaction. However, because the means were in the predicted direction, post hoc comparisons were carried out. The mean reaction times for color naming the active state words suggest that the restrained eaters showed a greater increase in their reaction time after the high calorie preload than after the low calorie preload (t = 2.82, p < .01) and a greater increase in their reaction time than the unrestrained eaters in the same condition ( t = 2.48, p < .05). Passive State At Time 1 the results showed no main effects and no interactions. At Time 2 the results suggest a significant main effect of condition [F(1,52) = 2.83, p < .05] and a trend for a condition by restraint level interaction [F(1,52) = 2.4, p = .063] but no significant main effect of restraint level. Examination of the means suggests that the restrained eaters showed greater retardation of their reaction time after the high calorie preload (mean reaction time; 9.29 t 7.4) than after the low calorie preload (mean reaction time; 3.45 +- 6.6), ( t = 2.21, p < .05), but a similar degree of retardation as the unrestrained eaters in the same condition. Analysis of the change scores (see Table 1) for the passive cognitive state as measured by the Stroop task showed no significant main effects of either condition or restraint level. The results suggest a trend for a condition by restraint level interaction [F(1,52) = 1.77, p = .095] but this failed to reach significance. Food Words At Time 1 the results showed no main effects or interactions. However, at Time 2 the results for reaction times for the food words show a significant main effect of both condition [F(1,52) = 7.91, p < .001] and restraint level [F(1,52) = 3.96, p < .05] and a significant condition by restraint level interaction [F(1,52) = 2.69, p < .05]. The means suggest that at Time 2 the restrained eaters took longer to color name the food words after the high calorie preload (mean reaction time; 12.88 ? 12.69) than after the low calorie preload (mean reaction time; 3.55 + 5.35), (t = 2.53, p < .05), and longer than the unrestrained eaters in the same condition (mean reaction time; 5.38 t 5.39), (t = 2.04, p < .05). Analysis of the change scores (see Table 1)for the food words as measured by the Stroop task showed no significant main effects or interactions. Body Size Words At Time 1 the results for the reaction times for color naming the body size words showed no main effects of either condition or restraint level, but showed a significant restraint level by condition interaction [F(1,52) = 4.48, p < .05]. At Time 2 the results suggest a significant effect of condition [F(1,52) = 14.47, p < .001] with all subjects in the high calorie preload condition, regardless of restraint level, showing the greatest retardation of their reaction times. Analysis of the change scores (see Table 1) for the body size words showed a significant main effect of condition [F(1,52) = 5.44, p < .05], a trend for a main effect of restraint level [F(1,52) = 1.9, p = .08], and a significant restraint level by condition interaction [F(1,52) = 7.33, p < .01]. The means suggest that Cognitive Changes to Preload 191 - Table 1. Change scores (time 2 time 1) for the Stroop tasks Restrained Low Preload (n = 14) -2.63 ? 6.09 -4.53 t 5.15 -1.19 ? 10.58 3.6 k 13.2 Restrained High Preload (n = 14) Change in food words Change in body words Change in passive words Change in active words 5.23 ? 23.17 9.41 2 12.08 4.67 2 12.39 14.75 rC_ 6.5 Unrestrained High Preload (n = 15) -3.70 t 8.32 -1.67 ? 12.52 0.17 5 10.17 3.29 ? 16.5 Unrestrained Low Preload (n = 13) -3.22 2 -0.61 ? 1.82 ? -2.03 2 6.56 9.7 8.5 7.25 the restrained eaters showed a greater increase in their time to color name the body size words after the high calorie preload than after the low calorie preload ( t = 3.97, p < .001) and a greater increase than the unrestrained eaters (t = 2.42, p < .05) in the same condition. Rating Scales The change scores for the rating scales are shown in Table 2. Motivational State Hunger. At Time 1, Time 2, and the change scores the results showed no main effects and no condition by restraint level interactions. Fullness. At Time 1, Time 2, and the change scores the results showed no significant main effects or interactions. Hunger symptoms. At Time 1 the results for the hunger symptoms suggest a significant main effect of restraint level [F(1,52) = 5.17, p < .05] with the restrained eaters reporting more hunger symptoms than the unrestrained eaters regardless of condition. The results at Time 1 show no main effect of condition and no restraint level by condition interaction. At Time 2 the results suggest a significant main effect of restraint level [F(1,52) = 3.02, p < .05] and a significant restraint level by condition interaction [F(1,52) = 3.65, p < .05]. Examination of the means suggests that the restrained eaters reported more hunger symptoms after the high calorie preload (1.14 ? 0.95) than the unrestrained eaters in the same condition (0.40 4 0.63), (t = 2.46, p < .05), but showed no difference in their reports of hunger symptoms between the two conditions. Analysis of the change scores for the hunger symptoms showed no significant main effects or interactions. Emotional State Anxiety. At Time 1 the results for anxiety showed no main effects or interactions. At Time 2 the results showed no significant main effects, but suggest a significant restraint level by condition interaction [F(1,52) = 6.52, p < .01]. The results suggest that the restrained eaters reported higher levels of anxiety after the high calorie preload (10.93 6.3) than after the low calorie preload (6.71 4.4), (t = 2.05, p < .05), but similar levels of anxiety to the unrestrained eaters in the same condition. Analysis of the change scores showed no significant main effects or interactions. Depression. At both Time 1 and Time 2 the results showed no main effects or interactions. Analysis of the change scores suggests a significant main effect of restraint level [F(1,52) = 4.23, p < .05], and a trend for a restraint level by condition interaction [F(1,52) = 1.8, p = .09]. The results show no main effect of condition. Examination of the means * * 192 Ogden and Creville suggests that the restrained eaters became significantly less depressed after the high calorie preload compared to the unrestrained eaters (t = 1.98, p < .05) but showed a similar profile of response to both high and low preloads. Cognitive State Active state. At Time 1 the results for active state showed no main effects or interactions. At Time 2 the results suggest a significant main effect of condition [F(1,52) = 4.12, p < .05], a significant main effect of restraint level [F(1,52) = 5.2, p < .05], and a significant condition by restraint level interaction [F(1,52) = 5.4, p < ,051. Examination of the means suggests that at Time 2 the restrained eaters reported higher levels of active state and feelings of rebelliousness after the high calorie preload (23.93 ? 8.9) than after the low calorie preload (13.21 7.12), (t = 3.52, p < .005) and higher levels of rebelliousness than the unrestrained eaters in the same condition (12.8 2 10.9), (t = 3.01, p < .01). Analysis of the change scores for active state suggests a significant main effect of condition [F(1,52) = 10.67, p < .005], a significant main effect of restraint level [F(1,52) = 9.42, p < .005], and a significant condition by restraint level interaction [F(1,52) = 10.7, p < .005]. The means suggest that the restrained eaters reported a significant increase in feelings of rebelliousness after the high calorie preload than the low calorie preload (t = 4.51, p < .0001), and a significant increase compared to the unrestrained eaters in the same condition ( t = 3.52, p < .005). Passive state. At Time 1 and Time 2 the results showed no significant main effects of either condition or restraint level. However, at both Time 1 and Time 2 the results suggest a significant restraint level by condition interaction [F(1,52) = 5.01, p < .05] and [F(1,52) = 5.6, p < .05], respectively. The results for the change scores of the passive state ratings showed a significant main effect of restraint level [F(1,52) = 2.99, p < .05], but no main effect of condition ur condition by restraint level interaction. * DISCUSSlON The results of the present study suggest that restrained eaters reported an increase in their active state after the high calorie forbidden food as measured by both the Stroop task and the rating scales. This provides support for previous research (Ogden & Wardle, 1991) that indicated that restrained eaters respond to high calorie preloads with increased feelings of rebelliousness and defiance and a desire to challenge the limitations set by the diet. This active state of mind is similar to the what the hell effect described by Herman and Polivy (1984) and to the description of bingeing as a way to Table 2. Change scores (time 2 Restrained High Preload (n = 14) Hunger Fullness Symptoms Anxiety Depression Active Passive 0.86 f 10.48 -1.14 2 6.84 0.07 5 0.92 1.57 f 6.22 -2.86 2 6.05 11.00 f 7.63 1.07 f 5.43 - time 1) for the rating scales Unrestrained High Preload (n = 15) -1.73 f 12.11 -0.4 2 10.15 0.00 f 0.65 -0.8 2 4.97 0.53 f 2.21 1.13 ? 7.46 -1.2 f 3.81 Unrestrained Low Preload ( n = 13) 0.00 f 9.52 1.23 f 4.58 0.08 .+ 0.49 -1.38 .+ 3.27 -0.46 f 3.45 1.15 f 2.72 -1.38 5 4.48 Restrained Low Preload (n = 14) 1.64 f 19.63 0.71 2 9.57 -0.14 f 0.66 -2.14 f 4.84 -1.14 f 2.14 0.64 f 3.99 0.29 2 3.22 Cognitive Changes to Preload 193 unleash resentment, as described by obese bingers (Loro & Orleans, 1981). Although the design of the present study did not include a measure of food intake, the results suggest that the overeating behavior following preloading shown by restrained eaters (Herman & Mack, 1975; Herman et al., 1987) may be a response to an active decision to overeat as a means of rebelling against external and internal pressures to restrict food intake. The results also suggest that restrained eaters may respond to preloading with increased feelings of giving in and a change in their passive state of mind. Such feelings are similar to the state of motivational collapse described by Herman and Polivy (1984, 1989) and suggest that following the consumption of a forbidden diet-breaking food the drives to eat may out weigh the drives to restrain. However, changes in passive state were not as profound as predicted and were only apparent when comparisons were made between the high and low preload conditions in the restrained eaters, and not when comparisons were made between restrained and unrestrained eaters. This profile of results is similar to that found by Ogden and Wardle (1991) with changes in passive state only being apparent in the interview reports. The restrained eaters also showed a retardation of their reaction time in color naming food words on the Stroop task after the high calorie food compared to after the low calorie food. This suggests that preloading may increase the preoccupation with food. Previous research has shown similar retardation in patients with anorexia nervosa and bulimia nervosa (Channon et al., 1988; Ben-Tovim et al., 1989; Fairburn et al., 1991; Ben-Tovim & Walker, 1991) suggesting a similarity between subjects with eating disorders and preloaded restrained eaters. Perhaps eating disordered patients are constantly in a state of selective processing that is only matched in restrained eaters when confronted with a high calorie diet-breaking food. Retardation of food words has also been reported in individuals who have been food deprived for 24 hours (Channon & Hayward, 1990). The results from the present study suggest that retardation did not occur in the restrained eater until preloaded with a high calorie food. This provides some support for the conceptualization of restrained eating as an intention that is only occasionally realized and suggests that the restrained eater is not usually food deprived (Heatherton et al., 1991; Ogden, 1993). It also suggests that the preoccupation with food found in restrained eaters may be triggered by the consumption of a forbidden food that may then trigger an episode of overeating. Women who attempt to lose weight are noted for their preoccupation and dissatisfaction with their body size (Cooper, Taylor, Cooper, & Fairburn, 1987). The results from the present study suggest that the consumption of a forbidden diet-breaking food may increase this preoccupation as measured by the selective processing of body size words. Ben-Tovim and Walker (1991) and Fairburn et al. (1991) reported selective processing of body size words in anorexics and bulimics suggesting some similarity in the selective processing patients with eating disorders and restrained eaters. However, these subject groups were not tested following a high calorie food and showed a preoccupation with body size in the absence of food cues. The present study also examined the possible role of hunger priming as a response to preloading. The results suggest that preloading had no effect on hunger and fullness ratings but resulted in a different profile of results in restrained and unrestrained eaters as measured by hunger symptoms. The results suggest that restrained eaters reported greater hunger symptoms after the high calorie preload compared to the unrestrained eaters suggesting that their desire to eat may have increased in response to a forbidden food. This provides some support for the predicted similarity between 194 Ogden and Greville hunger priming in the restrained eater and alcohol priming in the alcoholic (Hodgson et al., 1979). However changes in the subjects’ motivation to eat were only reported in one of the relevant measures and were only apparent when comparing restrained and unrestrained eaters in the same condition and not when comparing restrained eaters responses to the high and low calorie preload. The final variable that was predicted to change fo!lowing preloading was mood. The results suggest that preloading had no effect on the ratings of depression and only had some effect on their ratings of anxiety. The results indicate that restrained eaters may respond to preloading with an increase in anxiety, which supports previous research suggesting that anxiety may trigger overeating (Herman & Polivy, 1975). However, the results contradict theories of comfort eating and psychosomatic theories of excess food intake (Bruch, 1974). In addition the results contradict a previous study that reported that preloading resulted in a decrease in anxiety (Ogden & Wardle, 1991). The present study used both subjective rating scales and Stroop tasks to assess changes foIIowing preloading. Retardation of reaction times in the Stroop tasks was found for color naming of the active state, body size words, and food words. Previous research using the Stroop suggests that retardation is a result of the subject’s processing resources being diverted to process subjectively relevant information. The results from the present study suggest that preloading with a high calorie forbidden food may result in words such as rebelliousness and defiance, and food and body words becoming specifically relevant to the restrained eater causing their attention to be diverted to these words when attempting to color name them. Such diversion may play a role in causing overeating behavior. Although the design of the present study did not measure changes in eating behavior following preloading, the results suggest that disinhibitory behavior in response to preloading may be mediated by changes in the restrained eaters’ state of mind, specifically, the decision to rebel against the limitations of the diet and a preoccupation with food and body size may be particularly important. An understanding of the overeating behavior of restrained eaters is essential not only as a means of understanding the negative effects of restrained eating, but also of understanding both the diet-breaking behavior of the obese on weight reduction programs and the excessive overeating of patients with bulimia nervosa. REFERENCES Ben-Tovim, D. I., & Walker, M. K. (1991). Further evidence for the stroop test as a quantitative measure of psychopathology in eating disorders. International Journal of Eating Disorders, 10, 609-613. Ben-Tovim, D. I., Walker, M. K.,Fok, D., & Yap, E. (1989). An adaptation of the stroop test for measuring shape and food concerns in eating disorders: A quantitative measure of psychopathology? International Journal of Eating Disorders, 8, 681-687. Bruch, H. (1974). Eating disorders: Anorexia, obesity and the person within. London and Boston: Routledge and Kegan Paul. Channon, S., & Hayward, A. (1990). The effect of short term fasting on processing of food cues in normal subjects. lnternutionul journal of Eating Disorders, 9, 447-452. Channon, S., Hemsley, D., & de Silva, P. (1988). Selective processing of food words in anorexia nervosa. British \ournu/ of Clinical Psychology, 27, 259-260. Cooper, P. J., & Bowskill, R. (1986). Dysphoric mood and overeating. British Journal of Clinical Psychology, 25, 155-156. Cooper, P. J., Taylor, M. J., Cooper, Z., & Fairbum, C. G. (1987). The development and validation of the body shape questionnaire. International lournu1 of Eating Disorders, 6, 485-494. Fairburn, C. G . , Cooper, P. J., Cooper, M. J . , McKenna, F. P., et al. (1991). Selective information processing in bulimia nervosa. International Journal of Eating Disorders, 10, 415-422. Cognitive Changes to Preload 195 Foa, E. D., Feske, U., Murdock, T. B., Kozak, M. J., & McCarthy (1991). Processing of threat related information in rape victims. Journal of Abnormal Psychology, 100, 156-162. Heatherton, T. F., Polivy, J., & Herman, C. P. (1991). Restraint, weight loss and variability of body weight. journal of Abnormal Psychology, 100, 78-83. Herman, P., & Mack, D. (1975). Restrained and unrestrained eating. journal of Personality, 43, 646-660. Herman, C. P., & Polivy, J. (1975). Anxiety, restraint and eating behaviour. Journal of Abnormal Psychology, 84, 666-672. Herman, C. P., & Polivy, J. A. (1984). A boundary model for the regulation of eating. In A. J. Stunkard & E. Stellar (Eds.), Eating and its disorders (pp. 141-156). New York: Raven Press. Herman, C. P., & Polivy, J. A. (1989). Restraint and excess in dieters and bulimics. In K. M. Pirke., W. Vanderycken, & D. Ploog The psychobiology of bulimia nervosa (pp. 33-41). Heidelberg: Springer. Herman, C. P., Polivy J. A., & Esses V. M. (1987). The illusion of countenegulation. Appetite, 9, 161-169. Hodgson, R. J., Rankin, H. J., & Stockwell, T. R. (1979). Alcohol dependence and the priming effect. Behaviour Research and Therapy, 17, 379-387. Jansen, A., Merckelbach, H., Oosterlaan, J., Tuiten, A., & Van der Hout, M. (1988). Cognitions and self talk during food intake of restrained and unrestrained eaters. Behaviour Research and Therapy, 26, 393-398. Kirkley, B. G., Burge, J . C . , & Ammerman, M. P. H. (1988). Dietary restraint, binge eating and dietary behaviour patterns. International Journal of Eating Disorders, 7, 771-778. Kucera, A., & Francis, J. (1967). Conversational analysis of present day American English. Boston: University Press. Laessle, R. G., Tuschl, R. J., Kotthaus, B. C., & Pirke, K. M. (1989). Behavioural and biologcal correlates of dietary restraint in normal life. Appetite, 12, 83-94. Loro, A. D., & Orleans, C. S. (1981). Binge eating in obesity: Preliminary findings and guidelines for behavioural analysis and treatment. Addictive Behaviours, 7, 155- 166. Macleod, C. M. (1991). Half a century of research on the stroop effect: An integrative review. Psychological Bulletin, 109, 163-203. Mathews, A., & MacLeod, C. (1985). Selective processing of threat cues in anxiety states. Behaviour Research and Therapy, 23, 563-569. McNair, D. M., Lorr, M., & Droppleman, L. F. (1981). Manual: Profile of mood states. San Diego, CA: Educational and Industrial testing Service. Mcnally, R. J., Reimann, 8 . C., & Eunsil, K. (1990). Selective processing of threat cues in panic disorder. Behaviour Research and Therapy, 28, 407-412. Monello, L. F., & Mayer, J. (1969) Hunger and satiety sensations in men, women, boys and girls. American Journal of Clinical Nutrition, 20, 253-261. Ogden, J. (1993). The measurement of restraint-confounding success and failure? International Journal of Eating Disorders, 13, 69-76. Ogden, J. (1992) Fat chance! The myth of dieting explained. U.WUSA/Canada: Routledge. Ogden, J., & Wardle, J. (1991). Cognitive and emotional responses to food. Znternational journal of Eating Disorders, 10, 297-311. Spencer, J. A,, & Fremouw, M. J. (1979). Binge eating as a function of restraint and weight classification. journal of Abnormal Psychology, 88, 262-267. Stroop, J. R. (1935). Studies of interference in serial verbal reactions. journal of Experimental Psychology, 18, 643-662. Van Strien, T., Frijters, J. E., Bergers, G. P., & Defares, P. B. (1986). Dutch eating behaviour questionnaire for the assessment of restrained, emotional, and external eating behaviour. Znternational journal of Eating Disorders, 5, 295-315. Wardle, J. (1986). The assessment of restrained eating. Behaviour Research and Theraw, 24, 213-215. Warren, C., & Cooper, P. J. (1988). Psychological effects of dieting. British journal of Clinical Psychology, 27, 269-270.