Feared food in dieting and non-dieting young women: a preliminary validation of the Food Phobia Survey

Appetite 43 (2004) 155–173 www.elsevier.com/locate/appet

Feared food in dieting and non-dieting young women: a preliminary validation of the Food Phobia Survey Vivian M.M. Gonzalez*, Kelly M. Vitousek1 Department of Psychology, University of Hawaii, 2430 Campus Road, Honolulu, HI 96822, USA Received 31 March 2003; revised 2 July 2003; accepted 31 March 2004

Abstract The Food Phobia Survey (FPS) is a recently developed clinical instrument designed to identify foods that are avoided out of fear or guilt by eating disordered individuals. The measure has potential utility in clinical settings for several purposes: the assessment of current food selection and food-related concerns; the construction of individual hierarchies for graded exposure; and the evaluation of treatment outcomes with reference to fear and avoidance of food items. It is comprised of 180 commonly eaten foods rated on three dimensions: fear/guilt, appeal in the absence of weight concern, and frequency of consumption. Dieting and non-dieting college women were compared to provide preliminary data on the FPS from a non-clinical population. The FPS yielded findings convergent with other data on forbidden foods and discriminated between dieters and non-dieters. For both groups, the perception that foods were fattening was correlated with increased fear/guilt, with dieters showing significantly greater increases in ratings of fear/guilt and number of feared foods with increments in the perceived ‘fatteningness’ of food items. q 2004 Elsevier Ltd. All rights reserved. Keywords: Forbidden food; Dietary restraint; Exposure; Measurement; Validity; Reliability; Anorexia; Bulimia

Introduction Dieting can involve restricting the quantity, frequency, or variety of foods eaten. Weight-concerned individuals often limit their dietary selections to foods that are considered ‘safe’, on the basis of beliefs about the likelihood that specific items will promote weight gain or disinhibit dietary restriction (Knight & Boland, 1989; Rosen, Leitenberg, Fondacaro, Gross, & Willmuth, 1985). The terms ‘feared foods’ and ‘forbidden foods’ have been used to characterize those foods that restrained and eating disordered individuals attempt to avoid because of self-imposed rules (e.g. Gattellari & Huon, 1997; Kales, 1990; Rosen et al., 1985). Although there is some correlation between the objective properties of foods and the likelihood that they will be feared or forbidden, individual beliefs about the weightpromoting effects of a food item are better predictors of associated distress or avoidance than its actual caloric or macronutrient properties. In the 1960s and 1970s it was * Corresponding author. E-mail address: [email protected] (V.M.M. Gonzalez). 1 The former name of the second author is Kelly M. Bemis. 0195-6663/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.appet.2004.03.006

reported that anorexic patients were ‘carbohydrate phobic’ (Crisp & Kalucy, 1974; Russell, 1979); in subsequent decades, research has ascertained that eating disordered individuals selectively avoid fats rather than carbohydrates (e.g. Beumont, Chambers, Rouse, & Abraham, 1981) and designate foods as forbidden largely on the basis of fat content (Kales, 1990). These observations have been considered contradictory by some authors (e.g. Van Binsbergen, Hulshof, Wedel, Odink, & Bennink, 1988), but the discrepant data probably reflect changes in popular beliefs about which property is most ‘fattening’ (Drewnowski, Pierce, & Halmi, 1988). Since the identification of forbidden foods appears to covary with popular beliefs, the dietary profiles of restrained and eating disordered individuals may also be subject to change over time, tracking shifts in available nutritional information and dieting fads. Because feared foods are often excluded by weight-concerned individuals on the basis of idiosyncratic beliefs and learning histories, there is considerable variability in the specific items considered ‘bad’ or ‘dangerous’ and in the degree of distress each item evokes. Eating disordered patients often justify their refusal to eat feared or forbidden foods by claiming to dislike them or by

156

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

expressing overwhelming anxiety at the prospect of eating them. In cognitive-behavioral therapy (CBT), anxiety, expressed dislike, and behavioral avoidance associated with feared foods are important treatment targets. CBT has been studied extensively in the treatment of bulimia nervosa and is considered the ‘gold standard’ treatment for this disorder (Wilson, 1998). This modality has demonstrated efficacy for the treatment of binge eating disorder as well (Wilfley et al., 2002), and a related approach has yielded some encouraging results in the treatment of anorexia nervosa (Pike, Walsh, Vitousek, Wilson, & Bauer, 2003). CBT is focused on the modification of abnormal attitudes about weight and shape, as well as the replacement of dysfunctional dieting with normal eating (Wilson & Fairburn, 1993). During the course of treatment, feared foods are reintegrated into the patient’s diet, and erroneous beliefs and fears attached to the ingestion of these foods are addressed. Previously excluded food items are typically added to the patient’s diet in a hierarchical fashion from least to most feared and/or avoided foods (i.e. graded exposure). Self-monitoring of dietary intake is a central component of CBT for the eating disorders (Wilson & Vitousek, 1999). Although daily records provide important data about what individuals are currently eating, they obviously cannot yield information about items that patients preferred prior to the initiation of dieting, but now exclude for fear of weight gain. No satisfactory measures exist for the purpose of facilitating the identification of foods that are desired but avoided out of fear or guilt associated with their consumption. For clinical purposes three dimensions are relevant in identifying these feared or forbidden foods: the degree of avoidance associated with a given item; the amount of fear and/or guilt it elicits; and the appeal it would hold in the absence of concerns about its putative danger. Restrained individuals tend to categorize foods according to whether they are ‘guilt-inducing’ versus ‘guilt-free’ (King, Herman, & Polivy, 1987) or ‘forbidden’ versus ‘nonforbidden’ (Knight & Boland, 1989). Dichotomous thinking regarding the presence/absence of guilt in association with specific foods is a distinguishing characteristic of dieters (King et al., 1987). These individuals also associate more guilt, pain, and anxiety with food in general, and associate significantly more guilt with forbidden than allowed foods (Gattellari & Huon, 1997). For restrained individuals, the mere anticipation of eating prohibited foods is associated with feelings of failure and dietary abandonment (Knight & Boland, 1989). Although non-restrained individuals also express guilt over eating some foods, they report thinking of foods in more neutral terms than restrained individuals and may be more likely to associate guilt with foods that they consider nutritionally poor, rather than fattening or diet breaking (King et al., 1987). As feared or forbidden foods are associated with fear and/or guilt reactions in restrained and eating disordered individuals, ratings of these emotional reactions to food

items can aid in the task of identifying foods that are considered ‘off-limits’. Also, subjective ratings of the fear or guilt elicited by the actual or imagined consumption of avoided foods help to clarify the basis for dietary exclusions, and in the case of eating disorders, can be used to guide the construction of hierarchies of feared foods for graded exposure. Taste preferences must also be assessed in order to determine whether a food is being avoided because of weight-related concerns, or because the individual simply does not find it palatable. However, ratings of ‘preference’ for food items in weight-concerned individuals are not necessarily equivalent to ratings of hedonic appeal. When asked to imagine that foods are calorie free, eating disordered individuals assign significantly higher ratings for appeal, suggesting that weight-related concerns affect reported preference (Sunday, Einhorn, & Halmi, 1992). This study examines some of the properties and correlates of a recently developed instrument, the Food Phobia Survey (FPS) (Vitousek, 1998). The FPS is intended for clinical work with eating disordered patients and is designed to facilitate the identification of patients’ feared and forbidden foods for the purpose of reintegrating these foods into their diets during the course of treatment. The present study is part of the preliminary investigation of the reliability and construct validity of the scale. Data were collected from dieting and non-dieting female college students to provide comparison data to responses from eating disordered patients. Dieters were utilized in this investigation as they represent a group intermediate between non-dieting and eating disordered individuals on many behavioral and attitudinal indices (Polivy & Herman, 1987), and it was anticipated that the instrument would discriminate between dieters and non-dieters. A second purpose of this study was to validate its use as a research instrument for studying feared and forbidden foods in restrained eaters.

Method Participants For the principal study, 101 female participants were recruited from undergraduate psychology courses at the University of Hawaii. After 22 participants were excluded on the basis of exclusionary criteria outlined below, the sample used for analyses included 34 dieters and 45 nondieters. A separate sample of 37 female participants was recruited to examine the test-retest reliability of the FPS at a 1-week interval. Classification of participants as dieters or non-dieters The current study identified dieting participants on the basis of a positive response to a direct question about

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

whether participants were currently restricting their food intake for the purpose of weight control. The dieting question was adapted from the ‘Restraint over eating’ item of the Eating Disorder Examination (Fairburn & Cooper, 1993). Negative responses to the dieting question were used in conjunction with the Cognitive Restraint subscale of the Eating Inventory (EI-R; Stunkard & Messick, 1985) to designate participants as non-dieters. The EI-R assesses habitual behaviors related to the restriction of dietary intake (French, Jeffery, & Wing, 1994). Although habitual dieters are more likely to be dieting at any given time than nonhabitual dieters, they may not be dieting at the time of data collection (Lowe, 1993). Therefore, in the present study, participants were categorized as non-dieters if they indicated that they were not currently restricting their eating and scored below the 50th percentile on the EI-R. Participants who were not currently dieting but scored above the 50th percentile of the EI-R were eliminated, as were participants who indicated that they has been diagnosed with anorexia or bulimia nervosa at any point in the past. Procedure Self-report measures The Food Phobia Survey (FPS; Vitousek, 1998) is a 180item questionnaire composed of two separately administered sections (see Appendix A). The first section asks participants to indicate the frequency with which they have eaten each of 180 foods items over the last year and the extent to which they have feared and/or felt guilty about eating each food over the last year. Frequency is indicated on a five-point scale (from 1 ‘never’ to 5 ‘very often’); fear and guilt are rated on a five-point scale (from 1 ‘none’ to 5 ‘very strong’). The second section asks participants to indicate how desirable or appealing they find each of the foods on a five-point scale (from 1 ‘not at all’ to 5 ‘extremely’). Participants are instructed to make this judgment independently from other considerations, such as perceived healthfulness, dangerousness, fatteningness, expense, ease, or ethical considerations. The frequency and fear/guilt sections are administered separately from the ratings of appeal in order to prevent contamination of ratings through simultaneous consideration of all variables. The food items included on the FPS were chosen to represent the food preferences of contemporary young women with and without an eating disorder. The range of foods is extensive because of the clinical desirability of including (a) ‘diet’ foods that are highly likely to be represented in the intake of eating disordered and dieting individuals (e.g. rice cakes), (b) foods that are likely to be avoided because of their perceived high fat and/or calorie content (e.g. ice cream bar), and (c) foods that represent ‘transitional’ choices for many individuals (e.g. frozen yogurt). Because one of the main purposes of the scale is to identify specific target foods that are liked but avoided out

157

of fear or guilt by a diverse population of eating disordered individuals, it was considered crucial to incorporate numerous specific items rather than selecting a few representative items per category. The items were drawn from 12 food categories: dairy, breads and grains, snack foods, meats, entrees with meat, entrees without meat, salads and soups, vegetables, fruits, beverages, desserts, and condiments and sauces. Since perceived ‘fatteningness’ was the construct of interest, rather than absolute amounts of fat or calories per serving, the selection of foods to represent low, medium, and high caloric and/or fat content was not based on the actual food values of these items. Instead, items were chosen and classified by expert judges familiar with the food-related beliefs and eating practices of dieting and eating disordered individuals. Fatteningness was generally based on a combination of the caloric density and the amount of fat each food contains relative to other foods in the same category. As this survey is primarily intended for clinical use during the treatment of eating disorders, some fairly fine distinctions were made during categorization of level of fatteningness. For example, apples were assigned to the low level and bananas to the medium/high level within the category of fruit. These fine distinctions are meaningful in view of the extreme degrees of avoidance practiced by many eating disorder patients. During treatment, the goal is to restore normal eating by helping patients become comfortable with eating a variety of foods, representing a range of perceived levels of fatteningness within and across categories. Background Questionnaire. Participants completed a background questionnaire developed for this study that included demographic questions, as well as items assessing history of dieting, possible eating disorder diagnoses, and current level of hunger rated on a five-point scale from 1 (not hungry) to 5 (very hungry). Eating Inventory. The Eating Inventory (EI; Stunkard & Messick, 1985) measures three dimensions of eating behavior: cognitive restraint of eating, disinhibition, and hunger. It is a two-part self-report inventory with 36 truefalse items and 15 four-point Likert rating-scale items. As noted previously, the EI-R subscale was used in this study to identify non-dieting participants, in combination with selfreport of current absence of dieting behavior. Eating Disorder Inventory. The eating disorder inventory (EDI; Garner, Olmstead, & Polivy, 1983) is a 64-item multidimensional instrument designed to assess symptoms of anorexia and bulimia. The EDI includes eight subscales: Drive for Thinness, Bulimia, Body Dissatisfaction, Ineffectiveness, Perfectionism, Interpersonal Distrust, Interoceptive Awareness, and Maturity Fears. The eating disordered-specific subscales (Drive for Thinness, Bulimia, Body Dissatisfaction) were used in this study for the purpose of sample description. Food Choice Questionnaire. The Food Choice Questionnaire (FCQ; Steptoe, Pollard, & Wardle, 1995) is a 36-item

158

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

multidimensional measure of motives related to food choice. Participants are asked to endorse the importance of various considerations in determining the selection of foods. Considerations in food choice are rated on a 4-point scale as ‘not at all important’, ‘a little important’, ‘moderately important’, or ‘very important’. The FCQ includes nine factors: Health, Mood, Convenience, Sensory Appeal, Natural Content, Price, Weight Control, Familiarity, and Ethical Concern (Steptoe et al., 1995). For the purpose of this study Health, Mood, Convenience, Sensory Appeal, and Weight Control subscales were used to investigate factors associated fear/guilt regarding the Food Phobia Scale foods and number of feared foods. Designation of foods as feared Food item ratings on the Food Phobia Survey were used to designate foods as feared or not feared. Foods were operationally defined as ‘feared’ if a participant: (a) rated a food item 1 or 2 for frequency of consumption, indicating that it was never or rarely eaten; (b) rated the same food item 4 or 5 for associated fear/guilt, indicating that the item aroused strong or very strong feelings of fear/guilt; and (c) rated the same food item as 4 or 5 for appeal, indicating that the item was considerably or extremely appealing. The designation of an item as a ‘feared food’ thus suggests that substantial avoidance, fear/guilt, and appeal are simultaneously associated with that food item. While one can fear a food that is disliked, the purpose of this operational definition is to identify foods that are feared and avoided despite the fact that they are appealing to the participant. For the purpose of this study, all foods that failed to meet the operational definition above were not classified as feared foods. Card sorting task In addition to filling-out self-report measures, participants completed a card-sorting task to examine perceived fatteningness of the food items included on the FPS. The card sort task was comprised of FPS foods printed on cards that were presented to participants by food category. Participants were instructed to sort cards into low, medium, and high levels of fatteningness based on their own judgment of what constitutes a fattening food. This allowed an examination of dieting and non-dieting participants’ agreement with the FPS’s categorizations and of whether the scale samples a variety of foods that are perceived by dieting and non-dieting woman to be at various levels of fatteningness. Each participant’s categorizations of foods during the card sort task into low, medium, and high fatteningness were used as objective measures of the individual participant’s perception of how fattening each food item was. For each participant, the food items she categorized at each level of fatteningness were examined for frequency of consumption, fear/guilt, appeal, and number of feared foods. Means were calculated for each variable for each

participant-designated level of fatteningness. For example, if a participant categorized carrots and apples as low in fatteningness, then her ratings for frequency of consumption for those items were summed and divided by the number of foods categorized as low in fatteningness. Continuing the example, if carrots had been rated as eaten at a frequency of 5 and apples at a frequency of 3, then the mean for frequency of consumption of low perceived fatteningness for this individual would be 4. Analyses A 2 £ 3 mixed between-within-subjects multivariate analysis of variance (MANOVA) was conducted on four dependent variables: frequency of consumption, fear/guilt, appeal, and number of feared foods. The within-subjects independent variable was perceived level of fatteningness (low, medium, and high) and the between-subjects independent variable was dieting status (dieter, non-dieter). Mixed between-within-subjects analyses of variance (ANOVAs) were conducted to compare dieters to nondieters for the three levels of fatteningness for the various dependent variables described above. Subsequent pairwise contrasts following ANOVAs were conducted with alpha levels (0.05 used for all statistical tests) corrected with the Bonferroni procedure or Scheffe’s correction for multiple comparisons. Standard multiple regressions were performed using number of feared foods and fear ratings as dependent variables and the FCQ factors (Weight Control, Health, Convenience, Sensory Appeal, and Health) as independent variables.

Results Demographics The sample consisted of approximately 76% AsianAmericans (39% Japanese, 14% Chinese, 10% Filipino, 5% Vietnamese, 4% Korean, 3.5% other Asian-American), 15% Caucasians, 5% Native-Hawaiians, 3% Latinos, and 1% African-Americans. Thirty-five percent of participants were multi-ethnic, and were categorized according to the self-reported group of closest identification. No significant differences were found between the dieting and non-dieting groups for the EDI subscale Bulimia, age, body mass index, or level of hunger (see Table 1). Dieters had significantly higher means than non-dieters for the EDI subscales of Drive for Thinness and Body Dissatisfaction. Food Phobia Survey Test –retest Test – retest correlations were examined for frequency of consumption, fear, and appeal for each food using a separate sample of 37 participants. For frequency of consumption,

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173 Table 1 Group differences for EDI subscales, age, body mass index (BMI), and current level of hunger Variables

Dietersa

Non-dietersb

tc

Table 2 Comparison of percent of foods rated as low, medium, and high fatteningness by participants compared to Food Phobia Survey categorizations Groups

Age BMI Current hunger Drive for Thinnessd Bulimiad Body Dissatisfactiond a b c d

Mean

SD

Mean

SD

22.00 22.61 2.56 8.35 1.62 13.18

6.46 4.01 1.35 5.36 0.93 9.78

20.91 22.46 2.84 2.69 3.28 7.51

4.74 4.12 1.35 3.35 1.42 7.11

1.09 1.25 20.93 5.42* 1.16 2.05*

*p , 0:05. n ¼ 34: n ¼ 45: df ¼ 77: Eating Disorder Inventory subscale.

test – retest reliability ranged from 0.31 to 0.96 across individual food items. Three percent of the frequency of consumption items had a reliability coefficient between 0.30 and 0.49, 28% between 0.50 and 0.69, and 69% had a coefficient of 0.70 or higher. In calculating the test-retest correlations for fear/guilt, 40 foods were identified for which the majority of the entire sample indicated they experienced no or ‘slight’ guilt. In these circumstances, insufficient variance was present to calculate a Pearson’s correlation coefficient that accurately portrayed the temporal stability of the given item. Test – retest correlations for fear/guilt ranged from 0.14 to 0.95. Fourteen percent of test – retest correlations for fear/guilt items (for which correlations could be calculated, i.e. 140 out of the 180 items) ranged between 0.14 and 0.49, 44% between 0.50 and 0.69, and 42% were 0.70 or higher. For appeal ratings, 1-week test-retest varied from 0.39 to 0.93 across individual food items. One appeal item had a reliability coefficient below 0.50 (0.01%), 22% between 0.50 and 0.69, and 78% had a coefficient of 0.70 or higher. The mean number of foods meeting the operational definition of feared foods for the test – retest sample at time 1 was 2.68 (SD ¼ 4.03); 1 week later, at time 2, the mean number was 2.22 (SD ¼ 3.45), with a correlation of 0.75 between sum of feared foods for participants at time 1 and 2. Categorization of food items by levels of fatteningness On the sorting task, the mean percentage of agreement with the categorizations used in the FPS for low fatteningness was 60.6% for dieters, and 59.2% for non-dieters. Mean percentage of agreement for the medium level was 44.6% for dieters and 46.1% for non-dieters, and for the high level 81.1 and 73.1%, respectively. Dieters and nondieters did not differ in agreement with FPS categorization of foods ðFð1; 77Þ ¼ 3:01; p ¼ 0:087Þ; however, significant differences were found for agreement by level of fatteningness ðFð2; 154Þ ¼ 135:39; p ¼ 0:000Þ: Both groups agreed significantly more with expert raters’ categorizations of

159

Dieters Non-dieters

x2 (3)

Rated level of fatteningness Low (%)

Medium (%)

High (%)

28.30 30.42

27.08 29.64

44.62 39.94

2.92a 0.73a

Note: percentages of items at each level of fatteningness on the Food Phobia Survey are as follows: low ¼ 33.33%, medium ¼ 27.78%, and high ¼ 38.89%. Foods from categories with only low and medium/high were categorized as either low or high for scoring purposes. a Not significant.

high fatteningness than with low or medium, and agreed more with low than with medium fatteningness. A chi-square analysis was conducted on categorizations to compare the numbers of foods participants assigned to the three levels of fatteningness with those fixed in the FPS, to gauge participants’ perceptions of foods when they were allowed to distribute items among the levels of fatteningness in any way they wished. No significant differences were found between either dieters or non-dieters in how many foods were perceived as low, medium, and high compared with the categorizations of the FPS (see Table 2). Effects of perceived level of fatteningness on consumption, fear/guilt, appeal, and number of feared foods The combined dependent variables were significantly affected by perceived level of fatteningness (Wilks’ L ¼ 0:46; Fð2; 76Þ ¼ 44:96; p ¼ 0:000Þ; and an interaction was found between perceived level of fatteningness and dieting status (Wilks’ L ¼ 0:92;Fð2; 76Þ ¼ 3:12; p ¼ 0:050Þ: A significant interaction was also yielded for dependent variables by dieting status by perceived level of fatteningness (Wilks’ L ¼ 0:66; Fð6;72Þ ¼ 6:17; p ¼ 0:000Þ: The between subjects test yielded significant differences between dieters and non-dieters on the combined dependent variables ðFð1; 77Þ ¼ 13:95;p ¼ 0:000Þ: Comparisons of key variables between dieters and nondieters by level of fatteningness are summarized in Table 3. Dieters and non-dieters differed in their frequency of consumption depending on how fattening they perceived foods to be. Frequency of consumption for dieters was significantly affected by the level of perceived fatteningness, with lower fat foods tending to be consumed more frequently than higher fat foods. For non-dieters, frequency of consumption was not affected by perceived fatteningness. Dieters and non-dieters also differed in reported fear/guilt depending on how fattening they perceived foods to be. For both dieters and non-dieters, each increase from low to medium to high perceived fatteningness was accompanied by a significant increase in fear/guilt

160

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

Table 3 Mean dependent variable scores according to dieting status and perceived level of fatteningness Variables

Consumption Fear Appeal Feared foods a b c

Dieters ðN ¼ 34Þ perceived fatteningness

Non-dieters ðN ¼ 45Þ perceived fatteningness

F Value Dieta F (2,154) Fatb F (1,77) Diet £ fat F (2,154)

Low

Medium

High

Low

Medium

High

2.59 ^ 0.53c 1.54 ^ 0.55 2.86 ^ 0.56 0.01 ^ 0.02

2.42 ^ 0.45 2.20 ^ 0.81 2.98 ^ 0.66 0.03 ^ .05

2.32 ^ 0.42 3.08 ^ 0.99 3.19 ^ 0.70 0.12 ^ 0.15

2.31 ^ 0.54 1.24 ^ 0.28 2.69 ^ 0.73 0.00 ^ 0.00

2.37 ^ 0.50 1.42 ^ 0.43 2.90 ^ 0.70 0.00 ^ 0.01

2.44 ^ 0.55 0.57 1.80 ^ 0.83 34.26*** 3.06 ^ 0.72 0.85 0.01 ^ 0.02 20.80***

0.90 130.08*** 14.54*** 28.14***

5.90** 27.77*** 0.25 17.87***

**p , 0:01; ***p , 0:001: Dieting status. Perceived fatteningness. Mean ^ SD.

(see Table 4). Dieters showed significantly greater fear/guilt at each perceived level of fatteningness compared to non-dieters, and a greater increase in fear/guilt at each increment in perceived fatteningness, accounting for the interaction. or both dieters and non-dieters, appeal was affected by perceived level of fatteningness, and groups did not differ in ratings of appeal. Subsequent pairwise comparisons with dieters’ and non-dieters’ scores combined showed that appeal increased significantly with each increase in perceived fatteningness for both dieters and non-dieters (low versus medium fatteningness ðtð78Þ ¼ 22:75; p ¼ 0:007Þ; low versus high fatteningness ðtð78Þ ¼ 24:36; p ¼ 0:000Þ; and medium versus high fatteningness ðtð78Þ ¼ 24:05; p ¼ 0:000Þ: Overall, across perceived levels of fatteningness, dieters had a mean of 10.85 (SD ¼ 13.60) items meeting the operational definition of feared foods, compared with a mean for non-dieters of 1.29 (SD ¼ 2.58). Dieters and nondieters differed in number of feared foods depending on how fattening they perceived foods to be. For both dieters and non-dieters, each increase from low to medium to high perceived fatteningness was accompanied by a significant increase in mean number of feared foods (see Table 4). Dieters had significantly more feared foods at each perceived level of fatteningness, and showed a greater

increase in the mean number of feared foods at each increase in perceived fatteningness. Factors associated with feared foods and fear/guilt The standard multiple regression conducted with mean proportion of feared foods as the dependent variable revealed that the FCQ variables accounted for a significant portion of variance in number of feared foods ðFð5; 73Þ ¼ 6:96; p ¼ 0:000Þ: Intercorrelations, means, and standard deviations of dependent and independent variables are summarized in Table 5. Only Weight Control contributed significantly to the prediction of number of feared foods ðtð73Þ ¼ 4:13; p ¼ 0:000Þ; accounting for 19% of the variance (see Table 6). A significant interaction was found between dieting status and Weight Control ðtð75Þ ¼ 3:70; p ¼ 0:000Þ: Weight Control accounted for 30% of the variance in feared foods for dieters, while for non-dieters it accounted for only 1% of the variance. The independent variables accounted for a significant portion of the variance in fear/guilt ðFð5; 73Þ ¼ 4:61; p ¼ 0:001Þ; but only Weight Control contributed significantly to the prediction of fear ðtð73Þ ¼ 3:45; p ¼ 0:001Þ; accounting for 14% of the variance (see Table 7). Dieters and nondieters were not found to differ in the amount of variance in fear accounted for by Weight Control.

Table 4 Computed t values of pairwise comparisons of fear/guilt and feared foods Variables

Within subjects contrasts for perceived levels of fatteningness Dietersb

Fear/guilt Feared foods a b c

Between subjects contrast for perceived levels of fatteningnessa

Non-dietersc

Low vs. Med.

Low vs. High

Low vs. Med.

Low vs. Med.

Low vs. High

Low vs. Med.

Low

Medium

High

27.79*** 22.48*

211.03*** 24.25***

210.72*** 24.28***

23.78*** 22.73*

25.07*** 23.78***

25.18*** 23.55***

22.91** 23.66***

25.08*** 22.86*

26.08*** 23.94***

*p , 0:05; **p , 0:01; ***p , 0:001: df ¼ 77: df ¼ 33: df ¼ 44:

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

161

Table 5 Intercorrelations, means, and standard deviations for fear, feared foods, and Food Choice Questionnaire variables ðN ¼ 79Þ Variables

1

2

Feara Feared Foodsa Weight Controlb Moodb Convenienceb Sensory Appealb Healthb

– 0.56 0.47 0.28 0.10 0.09 0.25

– 0.56 0.14 0.06 20.06 0.39

a b

3

4

– 0.33 0.20 0.07 0.63

– 0.33 0.59 0.22

5

– 0.28 0.04

6

– 20.03

7

Mean

SD

–

1.90 5.35 7.46 16.04 14.70 11.84 16.10

0.79 10.21 2.60 4.27 3.66 2.41 4.04

Dependent variables. Independent variables.

A bivariate regression was performed with number of feared foods as the dependent variable and percent of time participants had spent in the past 6 months dieting as the independent variable, as it was hypothesized that time spent dieting would positively affect the participants’ number of feared foods. Amount of time spent dieting in the 6 months preceding the study accounted for 38% of the variance in feared foods ðFð1; 77Þ ¼ 47:02; p ¼ 0:000Þ:

Discussion Dieters were found to associate significantly more fear/guilt with food items than non-dieting participants, and rated more food items in ways that met study criteria for designation as feared foods. For both dieters and nondieters, increases in the perceived fatteningness of foods were accompanied by significant increases in fear/guilt, and by increases in the number of foods meeting criteria as feared foods; however, perceived fatteningness produced a steeper increase in fear/guilt and number of feared foods for dieting participants. There was no support for the hypothesis that fear/guilt and the number of feared foods would be linked to different food-related concerns for dieting and non-dieting participants on the FCQ. For both groups, weight-related concerns accounted for significantly more variance in fear/guilt than did health concerns, appeal, convenience, or mood. For dieters, weight-related concerns accounted for a significant portion of the variance in feared foods while none of the variables accounted for a significant portion of the variance Table 6 Standard multiple regression of Food Choice Questionnaire variables on number of feared foods

for non-dieters. Non-dieting participants did, however, report significantly less fear and avoidance than dieters. Interestingly, although the fear/guilt reported by non-dieters increased as their perception of fatteningness increased, non-dieters reported an increase in their consumption with increasing fatteningness, in contrast to dieters who showed the opposite tendency. As hypothesized, no differences were found between the ratings of appeal produced by dieters and non-dieters. While it could be conjectured that weight-concerned individuals might not be forthcoming in ratings of appeal for foods that they consider off-limits, self-reported appeal was found to be strongly associated with level of fatteningness, with significant increases in appeal for each successive increase in perceived fatteningness for both dieters and non-dieters. These data suggest that for non-dieters appeal was a stronger factor than fear in determining frequency of consumption. Preliminary data on the use of the FPS with eating disorder patients suggest that these individuals are also candid in acknowledging the appeal of foods that they consider off-limits. The findings of this study are consistent with previous results on forbidden foods in restrained and non-restrained populations. For example, the current study found that the percentage of time spent dieting in the last 6 months was significantly associated with the number of foods meeting the operational definition of feared foods. Similarly, Knight and Boland (1989) reported that restraint scores were positively correlated with forbidden foods. Consistent with results obtained by Sunday et al. (1992) for non-restrained individuals, this investigation found that non-dieters’ fear/guilt was positively correlated with the perceived Table 7 Standard multiple regression of Food Choice Questionnaire variables on fear

Variables

B

SEB

b

sr 2

Variables

B

SEB

b

sr 2

Weight control Mood Convenience Sensory appeal Health

2.11*** 0.08 20.07 20.46 20.09

0.51 0.31 0.29 0.52 0.32

0.54 0.03 20.03 20.11 20.04

0.19 0.00 0.00 0.01 0.00

Weight control Mood Convenience Sensory appeal Health

0.14*** 0.03 20.01 20.01 20.02

0.04 0.03 0.02 0.04 0.03

0.48 0.18 20.04 20.04 20.09

0.14 0.02 0.00 0.00 0.00

***p , 0:001; R ¼ 0:57; R2 ¼ 0:32:

***p , 0:001; R ¼ 0:49; R2 ¼ 0:24:

162

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

fatteningness of foods. The findings that dieters had significantly more guilt and a higher number of feared foods compared to non-dieters, across all levels of perceived fatteningness, are consonant with other studies that used different procedures to identify foods as feared or forbidden (e.g. Gattellari & Huon, 1997). It should be noted, however, that in this study only participants who were currently dieting were classified as ‘dieters’, while the FPS calls for ratings of consumption to be made over the past year. Although it cannot be assumed that these individuals were actively dieting throughout the period of time specified on the scale, their pattern of responses appears similar to that of chronically restrained eaters. For dieters, the positive relationship found between weight concern and fear/guilt was expected, as restrained individuals have been shown to categorize foods according to whether they are guilt-inducing or guilt-free (King et al., 1987). The positive relationship found between weight concern and fear/guilt for non-dieters was not predicted, as investigators such as King et al. (1987) and Sunday et al. (1992) have suggested that non-restrained individuals associate guilt with nutritionally poor rather than fattening foods; however, neither included analyses of this postulated relationship. The findings of Gattellari and Huon (1997) provide some evidence that both restrained and nonrestrained individuals consider foods perceived as forbidden to be unhealthy. These investigators reported that restrained and non-restrained individuals rated forbidden foods significantly lower on health and higher on illness-related variables compared to non-forbidden foods. The findings of the current study suggest that for both dieters and nondieters, fear and guilt associated with foods are primarily due to weight concern, rather than health concern. Overall, the findings of this study suggest that weight concern is significantly linked to fear and/or guilt associated with food items, even among contemporary young women who deny efforts to restrict their intake for purposes of weight-control. The salience of this variable for dieters and non-dieters alike speaks to the pervasive nature of weight concern among at least this demographic subgroup. As a research tool for use with dieting and restrained individuals, the FPS appears to have adequate reliability, as evidenced by high test-retest reliability for the mean number of foods designated as feared. While the overall number of foods identified as feared was consistent from one week to the next, the specific food items identified as feared tended to shift. This problem was likely exacerbated by strict criteria for what constituted a feared food and by the use of a non-clinical sample. In individuals with eating and weight concerns of clinical severity, it is anticipated that ratings of frequency of consumption and fear/guilt would be more stable. As discussed above, the validity of the FPS for dieting and restrained populations is also supported by the convergence of these results with those obtained by researchers who tested similar phenomena through different

means. Dieting and non-dieting participants showed moderate levels of agreement with experts’ specific categorizations of FPS food items and with the number of foods experts placed at high, medium, and low fatteningness levels on the FPS. However, as the current study utilized a non-clinical sample for the purpose of collecting normative data, it can only provide initial information about the properties of this scale with non-clinical weight-concerned individuals. Data collected with other populations (e.g. ethnic or national groups, eating disordered diagnoses) would further add to the validity of this scale for various populations. Data are currently being collected from eating disordered patients to further investigate the reliability and validity of the scale for this population. For the purpose of selecting target foods for clinical interventions, the FPS has several advantages over other measures developed to identify forbidden foods. One existing instrument that could be used for the identification of target foods during the treatment of eating disorders is the Forbidden Food Survey (Ruggiero, Williamson, Davis, Schlundt, & Carey, 1988). The scale is a 45-item self-report inventory designed to measure the ‘anticipated emotional reaction’ to consuming various foods by individuals with bulimia, as these reactions are assumed to influence decisions to avoid the food or to purge it if eaten. The scale does not, however, measure self-reported avoidance of these foods nor provide information about respondents’ hedonic preference for the food item. Both of these dimensions provide crucial information for the clinical purpose of selecting target foods for exposure tasks. In addition, the Forbidden Food Survey does not include enough food items to cover a wide variety of individual hedonic preferences in different food categories. While utilizing a representative sampling of food items may provide adequate information for the purpose of investigating research questions regarding eating disorder patients as a group, a more extensive array of specific choices is desirable in order to identify target foods for individual patients. Moreover, the Forbidden Food Survey omits many popular food choices made by contemporary young women (e.g. taco or hamburger), while including items that are rarely consumed by this demographic group (e.g. porterhouse steak or veal cutlet). In addition, the scale includes only a few of the diet foods that are often prominently represented in the food records kept by individuals with eating disorders. The FPS, in contrast, does incorporate an extensive array of contemporary food items, including items that are likely to be eaten by contemporary young women with eating disorders (e.g. foods low in calories, fat, and/or carbohydrates). It also provides patients with the opportunity to write in and rate items not listed on the measure. The importance of including an extensive list of specific items is underscored by the continuing shifts in what is considered a diet food in the popular media (e.g. foods low in carbohydrates versus fat), as the choices considered

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

off-limits by an individual patient vary widely depending on the dietary advice she has elected to follow. By including an array of common diet foods the FPS allows a rapid general assessment of an eating disordered individual’s current consumption pattern. The results of the sorting task suggest that the scale does include a wide sample of foods that are perceived to be high, medium, and low in perceived fatteningness; however, participants’ perception of which foods belonged in a given category differed from those assigned during scale construction. The FPS provides a standardized means of assessing idiosyncratic avoidance, fear/guilt, and preference for a breadth of food items in clinical populations. By including items that are more typically eaten by eating disordered individuals, as well as items that are typically avoided, the survey allows measurement of current eating patterns and comfort levels with a wide array of foods, and provides data on avoidance. The scale makes it possible to differentiate between items avoided from fear/guilt or from absence of hedonic appeal. The measure has potential utility in clinical settings for several purposes: the assessment of current food selection and food-related concerns; the construction of individual hierarchies for graded exposure; and the evaluation of treatment outcomes with reference to fear and avoidance of food items. In the construction of fear hierarchies for graded exposure tasks, the scale may be used to generate lists of food items that vary in appeal, avoidance, and associated fear/guilt for an individual patient. For example, for initial exposure tasks, the scale can facilitate the identification of target items that are at least moderately appealing and never or rarely eaten, but only moderately fear or guilt inducing. Later exposure tasks could include items that induce higher levels of fear or guilt and/or are more strictly avoided. Generation of such lists are facilitated through the use of computer programs that select items based on clinician set parameters (e.g. highly appealing, moderately feared and strictly avoided foods).

Appendix A Food survey: I Please rate each of the foods on the following pages according to: –

–

how frequently, on average, you have eaten that food over the past year. If you have been hospitalized at any point during the past year, rate each item according to how frequently you have eaten it when you were not in the hospital. how much you have feared eating or felt guilty about eating that food over the past year. Please rate this item independently from how often you have eaten the food and from how much you like

163

the food: † for example, if you never eat spinach because you don’t like spinach, but you would not feel at all fearful or guilty about eating spinach, you would assign a ‘1’ for frequency and a ‘1’ for fear/guilt in rating that item. † for example, if you never eat fudge because you feel very fearful about eating it or would feel very guilty if you ate it, you would assign a ‘1’ for frequency and a ‘5’ for fear/guilt in rating that item. Interpreting food items Note that some of the items specify the type of food very precisely—for example, the questionnaire asks separately about whole milk, 2% milk, and skim milk. Other items are not specific about particular characteristics of the food. In these cases, you should assume that you are being asked about the most usual or typical version of the food. – for example, if the item simply says ‘hot dog’, you should assume that the meat consists of beef and pork, since that is the most ‘typical’ kind of hot dog. – for example, if the item says ‘tuna sandwich’, you should assume that the sandwich is made of two pieces of regular bread and is filled with tuna mixed with mayonnaise, since that is the most ‘usual’ kind of tuna sandwich. You should also assume that the items refer to usual or typical portion sizes. Please check one of the following categories to describe yourself: 1. non-vegetarian (I eat a range of foods that includes a variety of meats). Please check this category if you are someone who eats meat infrequently, but who does eat meat occasionally; the categories below should be endorsed only by people who never eat meat or almost never (no more than once or twice a year) eat meat. 2. poultry/fish only (I eat poultry and fish, but do not eat other meats) 3. fish only (I eat fish, but do not eat poultry or other meats) 4. lacto-ovo vegetarian (I eat dairy products and eggs, but do not eat any meat or fish) 5. vegan (I do not eat animal products of any kind, including dairy products or eggs) 6. other (please explain) Note: If you are a vegetarian or partial vegetarian, please do not include any ethical concerns in making the ‘fear/guilt’ ratings on the following pages—rate the items according to the fear/guilt (if any) that you feel or would feel about eating specific foods for other reasons.

164

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

Please list below any foods to which you have an established allergic response:

(you do not need to list foods that upset your stomach— only those foods to which you have a known allergy)

Please use these scales in rating each item for frequency and fear/guilt, circling one number in both columns:

Frequency 1. Regular cheese (e.g., cheddar, Swiss, Monterey jack) 2. Oatmeal (plain) 3. Fried rice 4. Potato chips (low fat) 5. Baked chicken breast 6. Bacon 7. Grilled chicken sandwich 8. Spaghetti with meat sauce 9. Garden burger 10. Tomato or cucumber salad 11. Saimin 12. Snow peas (Chinese pea pods) 13. Candied yams 14. Avocado 15. Tomato juice or V-8 juice 16. Butterscotch or vanilla pudding (regular) 17. Ketchup 18. Carrot cake with cream cheese frosting 19. Corn 20. Turkey sandwich 21. Rice cake 22. 2% milk 23. Pretzels 24. Caramel-coated popcorn (regular) 25. Taco with shredded beef 26. Tomato soup 27. Cappucino or cocoa (low fat) 28. Chocolate chip cookies 29. Frozen yogurt 30. Apple 31. Taco salad 32. Sushi 33. Power bar or cereal bar 34. Skim milk 35. Soft or semi-soft cheese (e.g. brie, camembert) 36. Hamburger 37. Green salad with regular salad dressing 38. Tea or iced tea 39. Fudgesicle (low fat) 40. Cantaloupe or honeydew melon 41. Milkshake 42. Cheese lasagna 43. Muffin (blueberry or cranberry)

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

Fear/Guilt 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

(continued on next page)

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

165

Table (continued)

Frequency 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89.

White rice Omelet made with egg whites or low-fat egg substitute Dried vegetable chips Fried calamari Steak Tuna sandwich Cheese/veggie wrap or pita Miso soup Mixed vegetables with butter sauce Baked potato Grapes Cappucino or cocoa (regular milk) Fruit juice popsicle Mustard Berries with cream, whipped cream, or sour cream Macaroni salad Meatless chili and rice Fruit yogurt Cereal Doughnut Potato chips (regular) Plate lunch with teriyaki beef Asparagus French fries Fruit juice (apple, orange, guava, passion fruit) Hot fudge sundae Snack cakes (e.g. Twinkies, HoHos, Ding Dongs) Margarine (regular) Green salad with vinegar or lemon juice Grilled mahi burger Bean burrito Soy milk or rice dream Cheese omelet (whole eggs, regular cheese) Whole wheat bread Scone Popcorn (plain, air-popped) Pastrami or corned beef sandwich Fettucine alfredo Green beans Mango or papaya Protein drink Beer Brownies Tabasco or chili sauce Strawberries Peas

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

Fear/Guilt 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 (continued on next

4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 page)

166

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

Table (continued)

Frequency 90. Manapua (steamed, with pork filling) 91. Cottage cheese (low fat) 92. Brown rice 93. Cinnamon roll 94. Crackers (e.g., saltines, soda crackers) 95. Turkey (white meat) 96. Fried fish 97. Spam 98. Macaroni and cheese 99. Chicken noodle soup 100. Jello (diet) 101. Fruit smoothie 102. Hot dog and bun 103. Sun chips or wheat chips 104. Trail mix (nuts, sunflower seeds, dried fruit) 105. Pork or beef chow mein with noodles 106. Potato salad 107. Coffee or iced coffee 108. Oatmeal cookies 109. Margarine (low fat) 110. Tofu with vegetables 111. Hard-boiled egg (whole) 112. English muffin 113. Popcorn (regular, buttered) 114. Shoyu chicken 115. Kalua pork 116. Pasta with vegetables and olive oil 117. Cauliflower 118. Banana 119. Ice cream bar (e.g. Dove Bar) 120. Butter 121. Orange 122. Bean salad 123. Carrots 124. Wine 125. Jello (regular) 126. Plain yogurt 127. Cream cheese 128. Bagel (plain) 129. Waffle or pancakes with syrup 130. Nuts (peanuts, macadamias, cashews, pecans) 131. Fried chicken 132. Portuguese sausage 133. Stuffed peppers with rice and tomatoes 134. Cheese and veggie pizza 135. Chicken broth

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

Fear/Guilt 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5

(continued on next page)

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

167

Table (continued)

Frequency 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180.

Grilled chicken caesar salad Portuguese bean soup Creamed spinach Raisins or dried fruit Diet Coke or Pepsi Mocha (regular milk, whipped cream) Shave ice Malasadas Premium ice cream (e.g. Ben and Jerry’s) Soy sauce Sour cream (regular) Chicken or shrimp stir fry Crackers (e.g. Ritz, Triscuits) Cottage cheese (regular) Whole milk White bread Granola bar Canned tuna (water-packed) Lean ground beef Barbecued ribs Baked potato stuffed with cheese Peanut butter sandwich Lima beans Kahlua and cream Mayonnaise Low-fat cheese (cheddar, Swiss, Monterey jack) Muffin (bran) Broiled fish Lean roast beef Ham and cheese sandwich Vegetable quiche Nachos with cheese New England clam chowder Onion rings Low-fat ice cream Candy bar (e.g. Milky Way, Snickers) Hollandaise or bearnaise sauce Ham Broccoli with cheese sauce Vanilla wafers Cheesecake Regular Coke or Pepsi Corned beef Turkey frankfurter Cocktail shrimp

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

Fear/Guilt 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4

5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 5 1 2 3 (continued on next

4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 4 5 page)

168

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

Table (continued)

Frequency

Food survey: II Please rate each of the foods on the following pages according to how desirable or appealing you find this food. In this case, we are asking about how much you like the food purely in terms of how much it appeals to you, independently from any other considerations about

Fear/Guilt

whether you consider the food healthy/ unhealthy, safe/dangerous, fattening/slimming, cheap/expensive, easy/difficult, or ethical/unethical. If there were no relationship between this food and health, weight, or any other kinds of considerations, how appealing or desirable would this food be for you?

Appeal 1. Regular cheese (e.g. cheddar, Swiss, Monterey jack) 2. Oatmeal (plain) 3. Fried rice 4. Potato chips (low fat) 5. Baked chicken breast 6. Bacon 7. Grilled chicken sandwich 8. Spaghetti with meat sauce 9. Garden burger 10. Tomato or cucumber salad 11. Saimin 12. Snow peas (Chinese pea pods) 13. Candied yams 14. Avocado 15. Tomato juice or V-8 juice 16. Butterscotch or vanilla pudding (regular) 17. Ketchup 18. Carrot cake with cream cheese frosting 19. Corn

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 (continued on next page)

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

169

Table (continued)

Appeal 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68.

Turkey sandwich Rice cake 2% milk Pretzels Caramel-coated popcorn (regular) Taco with shredded beef Tomato soup Cappucino or cocoa (low fat) Chocolate chip cookies Frozen yogurt Apple Taco salad Sushi Power bar or cereal bar Skim milk Soft or semi-soft cheese (e.g. brie, camembert) Hamburger Green salad with regular salad dressing Tea or iced tea Fudgesicle (low fat) Cantaloupe or honeydew melon Milkshake Cheese lasagna Muffin (blueberry or cranberry) White rice Omelet made with egg whites or low-fat egg substitute Dried vegetable chips Fried calamari Steak Tuna sandwich Cheese/veggie wrap or pita Miso soup Mixed vegetables with butter sauce Baked potato Grapes Cappucino or cocoa (regular milk) Fruit juice popsicle Mustard Berries with cream, whipped cream, or sour cream Macaroni salad Meatless chili and rice Fruit yogurt Cereal Doughnut Potato chips (regular) Plate lunch with teriyaki beef Asparagus French fries Fruit juice (apple, orange, guava, passion fruit)

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 (continued on

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 next page)

170

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

Table (continued)

Appeal 69. Hot fudge sundae 70. Snack cakes (e.g. Twinkies, HoHos, Ding Dongs) 71. Margarine (regular) 72. Green salad with vinegar or lemon juice 73. Grilled mahi burger 74. Bean burrito 75. Soy milk or rice dream 76. Cheese omelet (whole eggs, regular cheese) 77. Whole wheat bread 78. Scone 79. Popcorn (plain, air-popped) 80. Pastrami or corned beef sandwich 81. Fettucine alfredo 82. Green beans 83. Mango or papaya 84. Protein drink 85. Beer 86. Brownies 87. Tabasco or chili sauce 88. Strawberries 89. Peas 90. Manapua (steamed, with pork filling) 91. Cottage cheese (low fat) 92. Brown rice 93. Cinnamon roll 94. Crackers (e.g. saltines, soda crackers) 95. Turkey (white meat) 96. Fried fish 97. Spam 98. Macaroni and cheese 99. Chicken noodle soup 100. Jello (diet) 101. Fruit smoothie 102. Hot dog and bun 103. Sun chips or wheat chips 104. Trail mix (nuts, sunflower seeds, dried fruit) 105. Pork or beef chow mein with noodles 106. Potato salad 107. Coffee or iced coffee 108. Oatmeal cookies 109. Margarine (low fat) 110. Tofu with vegetables 111. Hard-boiled egg (whole) 112. English muffin 113. Popcorn (regular, buttered) 114. Shoyu chicken 115. Kalua pork 116. Pasta with vegetables and olive oil 117. Cauliflower

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 (continued on next page)

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

171

Table (continued)

Appeal 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166.

Banana Ice cream bar (e.g. Dove Bar) Butter Orange Bean salad Carrots Wine Jello (regular) Plain yogurt Cream cheese Bagel (plain) Waffle or pancakes with syrup Nuts (peanuts, macadamias, cashews, pecans) Fried chicken Portuguese sausage Stuffed peppers with rice and tomatoes Cheese and veggie pizza Chicken broth Grilled chicken caesar salad Portuguese bean soup Creamed spinach Raisins or dried fruit Diet Coke or Pepsi Mocha (regular milk, whipped cream) Shave ice Malasadas Premium ice cream (e.g. Ben and Jerry’s) Soy sauce Sour cream (regular) Chicken or shrimp stir fry Crackers (e.g. Ritz, Triscuits) Cottage cheese (regular) Whole milk White bread Granola bar Canned tuna (water-packed) Lean ground beef Barbecued ribs Baked potato stuffed with cheese Peanut butter sandwich Lima beans Kahlua and cream Mayonnaise Low-fat cheese (cheddar, Swiss, Monterey jack) Muffin (bran) Broiled fish Lean roast beef Ham and cheese sandwich Vegetable quiche

1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 (continued on

5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 next page)

172

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173

Table (continued)

Appeal 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180.

Nachos with cheese New England clam chowder Onion rings Low-fat ice cream Candy bar (e.g. Milky Way, Snickers) Hollandaise or bearnaise sauce Ham Broccoli with cheese sauce Vanilla wafers Cheesecake Regular coke or pepsi Corned beef Turkey frankfurter Cocktail shrimp

1 1 1 1 1 1 1 1 1 1 1 1 1 1

2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3

4 4 4 4 4 4 4 4 4 4 4 4 4 4

5 5 5 5 5 5 5 5 5 5 5 5 5 5

Please list and rate below up to three foods not noted on this questionnaire that you find appealing/desirable. Rate each one to indicate how frequently you eat it (1 ¼ never; 5 ¼ very often), how much fear/guilt you feel about eating it (1 ¼ none; 5 ¼ very strong), and how appealing you find it (using the scale above).

References Beumont, P. J., Chambers, T. L., Rouse, L., & Abraham, S. F. (1981). The diet composition and nutritional knowledge of patients with anorexia nervosa. Journal of Human Nutrition, 35, 265 –273. Crisp, A. H., & Kalucy, R. S. (1974). Aspects of the perceptual disorder in anorexia nervosa. British Journal of Medical Psychology, 47, 349–361. Drewnowski, A., Pierce, B., & Halmi, K. A. (1988). Fat aversion in eating disorders. Appetite, 10, 119– 131. Fairburn, C. G., & Cooper, Z. (1993). The Eating Disorder Examination. In G. T. Wilson, & C. G. Fairburn (Eds.), Binge eating: Nature, assessment, and treatment (pp. 317–360). New York, NY: Guilford Press. French, S. A., Jeffery, R. W., & Wing, R. R. (1994). Food intake and physical activity: a comparison of three measures of dieting. Addictive Behaviors, 19, 401 –409. Garner, D. M., Olmstead, M. P., & Polivy, J. (1983). Development and validation of a multidimensional eating disorder inventory for anorexia nervosa and bulimia. International Journal of Eating Disorders, 2, 15–34. Gattellari, M., & Huon, G. F. (1997). Restrained and unrestrained females’ positive and negative associations with specific foods and body parts. International Journal of Eating Disorders, 21, 377–383. Kales, E. F. (1990). Macronutrient analysis of binge eating in bulimia. Physiology and Behavior, 48, 837–840. King, G. A., Herman, C. P., & Polivy, J. (1987). Food perception in dieters and non-dieters. Appetite, 8, 147 –158.

Knight, L. J., & Boland, F. J. (1989). Restrained eating: an experimental disentanglement of the disinhibiting variables of perceived calories and food type. Journal of Abnormal Psychology, 98, 412 –420. Lowe, M. R. (1993). The effects of dieting on eating behavior: a threefactor model. Psychological Bulletin, 114, 100 –121. Pike, K.M., Walsh, B.T., Vitousek, K.B., Wilson, G.T., & Bauer, J. (2003). Cognitive behavioral therapy in the post-hospital treatment of anorexia nervosa. American Journal of Psychiatry, 160, 2046–2049. Polivy, J., & Herman, C. P. (1987). Diagnosis and treatment of normal eating. Journal of Consulting and Clinical Psychology, 55, 635–644. Rosen, J. C., Leitenberg, H., Fondacaro, K. M., Gross, J., & Willmuth, M. (1985). Standardized test meals in assessment of eating behavior in bulimia nervosa: consumption of feared foods when vomiting is prevented. International Journal of Eating Disorders, 4, 59– 70. Ruggiero, L., Williamson, D., Davis, C. J., Schlundt, D. G., & Carey, M. P. (1988). Forbidden Food Survey: measure of bulimic’s anticipated emotional reactions to specific foods. Addictive Behaviors, 13, 267–274. Russell, G. (1979). Bulimia nervosa: an ominous variant of anorexia nervosa. Psychological Medicine, 9, 429 –448. Steptoe, A., Pollard, T. M., & Wardle, J. (1995). Development of a measure of the motives underlying the selection of food: the Food Choice Questionnaire. Appetite, 25, 267–284. Stunkard, A. J., & Messick, S. (1985). The three-factor eating questionnaire to measure dietary restraint, disinhibition and hunger. Journal of Psychosomatic Research, 29, 71 –83. Sunday, S., Einhorn, A., & Halmi, K. A. (1992). Relationship of perceived macronutrient and caloric content to affective cognitions about food in

V.M.M. Gonzalez, K.M. Vitousek / Appetite 43 (2004) 155–173 eating disordered, restrained, and unrestrained subjects. American Journal of Clinical Nutrition, 55, 362–371. Van Binsbergen, C. J., Hulshof, K. F., Wedel, M., Odink, J., & Coelingh Bennink, H. J. (1988). Food preferences and aversions and dietary pattern in anorexia nervosa patients. European Journal of Clinical Nutrition, 42, 671–678. Vitousek, K.M (1998). The Food Phobia Survey. Unpublished scale. Wilfley, D. E., Welch, R. R., Stein, R. I., Spurrell, E. B., Cohen, L. R., Saelens, B. E., Dounchis, J. Z., Frank, M. A., Wiseman, C. V., & Matt, G. E. (2002). A randomized comparison of group cognitive-behavioral

173

therapy and group interpersonal psychotherapy for the treatment of overweight individuals with binge-eating disorder. Archives of General Psychiatry, 59, 713 –721. Wilson, G. T. (1998). The clinical utility of randomized controlled trials. International Journal of Eating Disorders, 24, 13–29. Wilson, G. T., & Fairburn, C. G. (1993). Cognitive treatments for eating disorders. Journal of Consulting and Clinical Psychology, 61, 261– 269. Wilson, G. T., & Vitousek, K. M. (1999). Self-monitoring in the assessment of eating disorders. Psychological Assessment, 11, 480 –489.