Validity of test meals for determining binge eating

Validity of test meals for determining binge eating

Eating Behaviors 2 (2001) 105 ± 112 Validity of test meals for determining binge eating Drew A. Andersona,*, Donald A. Williamsonb, William G. Johnso...

72KB Sizes 0 Downloads 4 Views

Eating Behaviors 2 (2001) 105 ± 112

Validity of test meals for determining binge eating Drew A. Andersona,*, Donald A. Williamsonb, William G. Johnsonc, Cheryl O. Grievec a

Department of Psychology, University at Albany, SUNY, Social Sciences 112, Albany, NY 12222, USA b Pennington Biomedical Research Center, 6400 Perkins Rd., Baton Rouge, LA 70808, USA c University of Mississippi Medical Center, Department of Psychiatry & Human Behavior, 2500 N. State St., Jackson, MS 39216-4505, USA

Abstract Assessment of binge eating has been criticized because of serious doubts concerning the accuracy of self-report. This experiment tested the validity of a laboratory test meal as an indicator of binge eating. Eight individuals diagnosed with binge-eating disorder (BED), eight obese non-binge-eaters, and eight normal-weight non-binge-eaters ate a test meal under conditions designed to increase the likelihood of inducing a binge episode. Non-binge-eaters, regardless of weight, felt in control of their eating and ate a relatively small amount of the test meal, while participants with BED ate significantly more food and felt significantly more out of control. Eating behavior during test meals can be a useful indicator of BED diagnostic status and may be a useful method for objectively defining binge eating. D 2001 Elsevier Science Ltd. All rights reserved. Keywords: Binge eating; Obesity; Assessment; Food intake

1. Introduction The assessment of binge eating presents several challenges. Assessment of binge eating via self-report questionnaires has been criticized because of serious doubts concerning the accuracy of self-reported binge episodes. Self-report questionnaires typically ask individuals directly how often they engage in binge eating. However, there is evidence that nonprofessionals do not use strict DSM-IV criteria (American Psychiatric Association, 1994) to define * Corresponding author. Tel.: +1-518-442-4835; fax: +1-518-442-4867. E-mail address: [email protected] (D.A. Anderson). 1471-0153/01/$ ± see front matter D 2001 Elsevier Science Ltd. All rights reserved. PII: S 1 4 7 1 - 0 1 5 3 ( 0 1 ) 0 0 0 2 2 - 8

106

D.A. Anderson et al. / Eating Behaviors 2 (2001) 105±112

the term ``binge'' (Beglin & Fairburn, 1992; Johnson, Boutelle, Torgrud, Davig, & Turner, 2000; Telch, Pratt, & Niego, 1998). While DSM-IV criteria (American Psychiatric Association, 1994) require that the amount of food eaten be objectively large and be accompanied by a loss of control, nonprofessionals rely more on feelings of loss of control and violation of dietary standards than the amount of food eaten to define an eating episode as a binge. Thus, ``binge episodes'' determined via self-report questionnaires might not be true binge episodes. As a result of the limitations of self-report measures of binge eating, researchers developed interview-based formats for assessing binge episodes, most notably the Eating Disorders Examination (EDE; Fairburn & Cooper, 1993). The EDE is a semistructured interview designed to assess a broad range of the specific psychopathology of the eating disorders. During the EDE, the interviewee describes his or her self-defined binge episodes in detail, including the amount of food eaten. The interviewer evaluates these episodes and decides whether they are truly large enough to be considered a binge episode according to DSM-IV criteria (American Psychiatric Association, 1994). This methodology considerably reduces the potential bias of allowing nonprofessionals to define the term ``binge.'' Despite this advantage, interview-based assessments of binge eating are still dependent on interviewees providing accurate reports of food intake in order for the interviewer to rate eating episodes. Unfortunately, self-report of food intake has been shown to be extremely poor. A growing body of literature suggests that individuals significantly overestimate food intake when asked to estimate intake using volumetric portions (e.g., cups). Lansky and Brownell (1982) found that in a laboratory setting, obese individuals overestimated the quantity of 10 foods, from 6% for cola to 260% for potato chips, with a mean error of 63.9%. Zegman (1984) found that obese females overestimated quantities of 10 food models 37% of the time when asked to respond in a multiple choice format. More recently, Anderson, Williamson, Johnson, and Grieve (1999) found that obese and nonobese women overestimated the volumetric intake of an ice cream test meal by almost 70%, even though the ice cream was presented in a pint container. If individuals overestimate volumetric portions when reporting food intake, then interview methods such as the EDE may significantly overestimate the number of binge episodes an individual reports. One alternative to the use of self-reported food intake for documenting the presence of binge-eating episodes is the use of laboratory test meals. Several studies have found that individuals with bulimia nervosa (Hadigan, Kissileff, & Walsh, 1989; LaChaussee, Kissileff, Walsh, & Hadigan, 1992; Walsh, Kissileff, Cassidy, & Dantzic, 1989) and binge-eating disorder (BED) (Cooke, Guss, Kissileff, Devlin, & Walsh, 1997; Goldfein, Walsh, LaChaussee, Kissileff, & Devlin, 1993; Telch & Agras, 1996; Yanovski et al., 1992) will eat significantly more than control participants when asked to binge eat in a laboratory situation. The present experiment was designed as an extension of previous laboratory studies of binge eating by utilizing a normal-weight control sample and arranging test conditions to increase the likelihood of a binge-eating episode in a laboratory setting.

D.A. Anderson et al. / Eating Behaviors 2 (2001) 105±112

107

2. Method 2.1. Participants A total of 24 women were recruited from a larger study of eating behavior and BED. From this larger study, eight obese participants (mean body mass index, BMI = 43.2, range 33.0± 60.0) were identified who met DSM-IV provisional criteria for BED (American Psychiatric Association, 1994). Also from the larger study, eight obese non-binge-eaters were identified who matched the BED participants on BMI within 2.8 units (mean BMI = 41.7, range 33.0± 57.2); eight normal-weight non-binge-eaters (BMI = 21.6, range 18.7±24.6) were randomly selected from the larger study for inclusion in the present study. Exclusion criteria for the study included male gender, diabetes, pregnancy, amenorrhea, diagnosis of anorexia or bulimia nervosa, or current use of any hormone medication (including oral contraceptives). Demographic characteristics of the sample are shown in Table 1. 2.2. Procedure In order to increase the likelihood of inducing a binge episode, participants were scheduled so that the date of the experiment fell within the luteal phase of their menstrual cycle. Both female infrahumans (Czaja, 1978; Wade, 1972, 1976) and humans (Dalvit, 1981; Johnson, Corrigan, Lemmon, Bergeron, & Crusco, 1994) have been found to increase overall caloric intake during the luteal phase of the menstrual cycle, when progesterone levels are at their peak. Female infrahumans (Geiselman, Martin, VanderWeele, & Novin, 1981) and humans (Johnson et al., 1994) also increase preference for fat during the luteal phase of their menstrual cycle. Participants came into the lab for the test meal at approximately 5:30 p.m. This time period was selected because food intake tends to increase in women over the day, and females have been found to consume a relatively higher proportion of calories from fat during this time of day (De Castro, 1987). In order to ensure that participants were hungry at the time of the test meal, they were instructed to eat a normal-sized lunch that day and then to fast for 6 h before the experimental procedure. Upon arrival at the laboratory, participants were administered the Table 1 Participant demographics Variable Height (cm) Weight (kg) BMI (kg/m2) Age (years) Liking for chocolate* Race Caucasian African American

BED

Obese control a

161.0 ‹ 5.0 110.4 ‹ 30.4a 43.2 ‹ 9.2a 35.3 ‹ 11.1a 90.3 ‹ 11.9a 6 2

a

163.1 ‹ 7.4 109.8 ‹ 32.3a 41.7 ‹ 9.4a 36.6 ‹ 10.3a 79.3 ‹ 18.7a 6 2

Normal-weight control 169.3 ‹ 7.2a 60.4 ‹ 5.8b 21.6 ‹ 1.9b 24.3 ‹ 2.7b 90.1 ‹ 13.3a 8 0

Numbers in a row with different superscripts are different at P < .05, using Scheffe post hoc procedure. * Rating on a 100-point scale (0 = totally disgusting; 100 = extremely enjoyable).

108

D.A. Anderson et al. / Eating Behaviors 2 (2001) 105±112

Interview for the Diagnosis of Eating Disorders Ð Fourth Revision (Kutlesic, Williamson, Gleaves, Barbin, & Murphy-Eberenz, 1998) to determine an eating disorder diagnosis. To determine whether they would find the test meal palatable, participants were asked to indicate how enjoyable they found chocolate on a 100-point scale (ranging from 0 = totally disgusting to 100 = extremely enjoyable). All participants found chocolate at least moderately enjoyable (mean score = 86.5 ‹ 15.2). For the test meal, participants were provided with a pint container of HaÈagen-Dazs chocolate ice cream with nutrition and brand information covered, as well as a bowl, spoon, and napkin. This test food was chosen specifically to enhance the possibility of inducing a binge episode. Ice cream is a high-fat food (for HaÈagen-Dazs chocolate, 15.7% fat by volume, with 59% of total calories coming from fat; HaÈagen-Dazs, 1994) and chocolate has been found to be the food most craved among females (Rozin, Levine, & Stoess, 1991; Weingarten & Elston, 1991). To encourage binge eating, participants were told that they were allowed unlimited quantities of ice cream and were instructed to ``let go and eat as much as you can.'' To minimize reactivity and self-consciousness on the part of the participants, the experimenter was not present in the room during the test meal. Any participant desiring more than 1 pint of ice cream was provided with a second pint immediately upon request. When participants indicated that they had finished eating they were asked to indicate the degree to which they found the test meal enjoyable on a 100-point scale (0 = totally disgusting; 100 = extremely enjoyable), their degree of hunger on a 100-point scale (0 = ravenous; 100 = completely stuffed), the degree to which they felt, as judged on a 100-point scale, that the amount of food they had consumed constituted a binge (0 = not at all a binge; 100 = an extremely large binge), and the degree to which they felt out of control while eating on a 100-point scale (0 = totally in control; 100 = totally out of control). The amount of food consumed was determined by subtracting the weight of the ice cream remaining postmeal from the weight of the ice cream premeal. The volume of ice cream eaten was calculated from the weight of the ice cream eaten and the caloric value of ice cream eaten was calculated from nutritional information obtained from the food label. 2.3. Statistics Statistical calculations were performed with SPSS 7.5.1 for Windows (SPSS, Chicago, IL). Data are presented as mean ‹ S.D. Analyses of continuous variables were carried out with parametric analysis of variance procedures with a post hoc Scheffe correction for multiple comparisons. Analyses were also carried out using nonparametric procedures (Kruskal± Wallis). There were no differences in the pattern of results between parametric and nonparametric procedures, so only the parametric results are discussed. In all analyses, an alpha level of P =.05 was established as statistically significant; all tests were two-tailed. 2.4. Results As can be seen in Table 1, obese participants had a higher BMI than normal-weight participants ( P < .05), but there were no differences in BMI between BED and obese non-

D.A. Anderson et al. / Eating Behaviors 2 (2001) 105±112

109

Table 2 Results of the test meal Variable

BED

Obese control

Normal-weight control

Amount eaten Calories Ounces Meal duration (min) Enjoyment rating for the test meal* Hunger rating postmeal** Binge rating postmealy Out of control rating postmealz

1308.0 ‹ 532.2a 17.1 ‹ 7.0a 15.0 ‹ 7.1a 87.5 ‹ 16.5a 69.4 ‹ 21.8a 61.9 ‹ 27.8a 50.1 ‹ 34.0a

661.3 ‹ 367.3b 8.7 ‹ 4.8b 8.9 ‹ 4.9a 81.0 ‹ 18.6a 75.4 ‹ 23.7a 34.8 ‹ 30.9a 9.5 ‹ 20.6b

515.2 ‹ 258.2b 6.7 ‹ 3.4b 8.6 ‹ 3.7a 93.1 ‹ 11.6a 56.5 ‹ 23.3a 13.4 ‹ 26.3b 8.6 ‹ 11.9b

Numbers in a row with different superscripts are different at P < .05, using Scheffe post hoc procedure. * Rating on a 100-point scale (0 = totally disgusting; 100 = extremely enjoyable). ** Rating on a 100-point scale (0 = ravenous; 100 = completely stuffed). y Rating on a 100-point scale (0 = not at all a binge; 100 = an extremely large binge). z Rating on a 100-point scale (0 = totally in control; 100 = totally out of control).

binge-eating participants. Obese participants were older than normal-weight control participants ( P < .05). There were no differences between groups in premeal estimate of liking for chocolate ( P > .1; for all participants, mean rating = 86.5 ‹ 15.2). Table 2 shows the results of the test meal. Participants with BED ate significantly more during the test meal than non-binge-eating participants. While participants with BED ate slightly more than 1 pint of ice cream, non-binge-eating participants, regardless of weight, ate only approximately one-half of that amount. Participants with BED also reported feeling significantly more out of control during the test meal than non-binge-eating participants, regardless of weight, ate only approximately one-half of that amount. Participants with BED also reported feeling significantly more out of control during the test meal than nonbinge-eating participants, with non-binge-eaters reporting almost no loss of control. There were no differences in enjoyment, postmeal hunger, or time to eat the test meal across groups. Both groups of obese participants reported significantly higher binge ratings than normal-weight participants. 3. Discussion This study found that participants with BED reported significantly greater loss of control and ate significantly more food than non-binge-eating participants in a laboratory test meal situation designed to increase the likelihood of a binge episode. In contrast, non-binge-eating participants, regardless of weight, felt in control of their eating and ate a relatively small amount of ice cream despite being asked to eat as much as possible, after a 6-h fast during a time of day likely to produce greater intake of a high-fat food, and during the phase of their menstrual cycle (luteal) most likely to lead to uncontrolled eating of a high-fat food. Interestingly, the eating behavior of obese non-binge-eaters was remarkably similar to normal-weight non-binge-eaters. The one exception was that obese non-binge-eating partic-

110

D.A. Anderson et al. / Eating Behaviors 2 (2001) 105±112

ipants rated the meal as a binge to a significantly higher degree than normal-weight bingeeating participants, despite eating comparable amounts of food and feeling in control to a similar degree. One explanation for this finding is that obese non-binge-eaters may have viewed the high-fat chocolate ice cream as a ``forbidden food'' and considered consumption of any amount of it to be a binge episode. A recent study found that the violation of dietary standards plays an important role in the definition of an eating episode as a binge (Johnson et al., 2000). Although obese non-binge-eating participants were as likely to define the test meal as a binge as BED participants, the absolute difference between these groups was large. BED group ratings were almost twice as high as obese non-binge-eating group ratings (see Table 2). This nonsignificant result is most likely due to low statistical power. This study utilized a single-item test meal paradigm, which limits the generalizability of the findings somewhat. Although several studies have shown strong correlations between patterns of eating in single- and multiple-item test meals in persons with bulimia nervosa (Kissileff, Walsh, Kral & Cassidy, 1986; LaChaussee et al., 1992; Walsh et al., 1989), this pattern is less clear for persons with BED (Goldfein et al., 1993). Further studies utilizing different single-item test meals as well as multiple-item meals may provide more information on the robustness of this phenomenon in persons with BED. Despite this limitation, the study extends the findings of previous laboratory test meal studies of BED. Furthermore, this is the first study to utilize a normal-weight control group in addition to an obese control group. The test meal methodology for this study was unique as well. The test meal conditions were designed so that all participants completed the test meal during the luteal phase of their menstrual cycle, which is associated with increased food intake and increased likelihood of a binge episode. To our knowledge, no previous laboratory study of binge eating has controlled for phase of menstrual cycle. Although it is possible that more extreme circumstances would induce binge eating in normally non-binge-eating individuals, the circumstances in this experiment represent the most extreme conditions likely to be found in typical test meal conditions. Thus, if an individual engages in binge eating in a laboratory setting, it is unlikely that this eating is due to the circumstances of the test meal, but rather to a true problem with binge eating. Non-binge-eaters appear to exercise restraint in laboratory test meal situations, even under conditions designed to disrupt restraint. Given the significant problems in the accuracy of estimating portion sizes, laboratory test meals should be considered as an alternative to selfreport and interview-based methods of assessing binge eating. While individuals may not be able to accurately report the amount of food they have eaten, their behavior in a test meal situation provides information about their likelihood of engaging in binge eating outside the laboratory. Acknowledgments Portions of this paper are based on the first author's doctoral dissertation. The authors would like to thank Paula J. Geiselman, PhD, for her assistance in conducting the study.

D.A. Anderson et al. / Eating Behaviors 2 (2001) 105±112

111

References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. Anderson, D. A., Williamson, D. A., Johnson, W. G., & Grieve, C. O. (1999). Estimation of food intake: effects of the unit of estimation. Eating and Weight Disorders, 4, 6 ± 9. Beglin, S. J., & Fairburn, C. G. (1992). What is meant by the term ``binge''? American Journal of Psychiatry, 149, 123 ± 124. Cooke, E. A., Guss, J. L., Kissileff, H. R., Devlin, M. J., & Walsh, B. T. (1997). Patterns of food selection during binges in women with binge eating disorder. International Journal of Eating Disorders, 22, 187 ± 193. Czaja, J. A. (1978). Ovarian influences on primate food intake: assessment of progesterone actions. Physiology and Behavior, 21, 923 ± 928. Dalvit, S. P. (1981). The effect of the menstrual cycle on patterns of food intake. American Journal of Clinical Nutrition, 34, 1811 ± 1815. De Castro, J. M. (1987). Circadian rhythms of the spontaneous meal pattern, macronutrient intake and mood of humans. Physiology and Behavior, 40, 437 ± 446. Fairburn, C. G., & Cooper, Z. (1993). The eating disorder examination. In: C. G. Fairburn, & G. T. Wilson (Eds.), Binge eating: nature, assessment and treatment (12th ed.) (pp. 317 ± 360). New York: Guilford. Geiselman, P. J., Martin, J. R., VanderWeele, D. A., & Novin, D. (1981). Dietary self-selection in cycling and neonatally ovariectomized rats. Appetite: Journal for International Research, 2, 87 ± 101. Goldfein, J. A., Walsh, B. T., LaChaussee, J. L., Kissileff, H. R., & Devlin, M. J. (1993). Eating behavior in binge eating disorder. International Journal of Eating Disorders, 14, 427 ± 431. HaÈagen-Dazs. (1994). HaÈagen-Dazs chocolate ice cream product information. Minneapolis, MN: Pillsbury. Hadigan, C. M., Kissileff, H. R., & Walsh, B. T. (1989). Patterns of food selection during meals in women with bulimia. American Journal of Clinical Nutrition, 50, 759 ± 776. Johnson, W. G., Boutelle, K. N., Torgrud, L., Davig, J. P., & Turner, S. (2000). What is a binge? The influence of amount, duration and loss of control criteria on judgments of binge eating. International Journal of Eating Disorders, 27, 471 ± 479. Johnson, W. G., Corrigan, S. A., Lemmon, C. R., Bergeron, K. B., & Crusco, A. H. (1994). Energy regulation over the menstrual cycle. Physiology and Behavior, 56, 523 ± 527. Kissileff, H. R., Walsh, B. T., Kral, J. G., & Cassidy, S. M. (1986). Laboratory studies of eating behavior in women with bulimia. Physiology and Behavior, 38, 563 ± 570. Kutlesic, V., Williamson, D. A., Gleaves, D. H., Barbin, J. M., & Murphy-Eberenz, K. P. (1998). The Interview for Diagnosis of Eating Disorders Ð IV: application to DSM-IV diagnostic criteria. Psychological Assessment, 10, 41 ± 48. LaChaussee, J. L., Kissileff, H. R., Walsh, B. T., & Hadigan, C. M. (1992). The single-item meal as a measure of binge-eating behavior in patients with bulimia nervosa. Physiology and Behavior, 51, 593 ± 600. Lansky, D., & Brownell, K. D. (1982). Estimates of food quantity and calories: errors in self-report among obese patients. American Journal of Clinical Nutrition, 35, 727 ± 732. Rozin, P., Levine, E. L., & Stoess, C. (1991). Chocolate craving and liking. Appetite, 17, 199 ± 212. Telch, C. F., & Agras, W. S. (1996). Do emotional states influence binge eating in the obese? International Journal of Eating Disorders, 20, 271 ± 279. Telch, C. F., Pratt, E. M., & Niego, S. H. (1998). Obese women with binge eating disorder define the term binge. International Journal of Eating Disorders, 24, 313 ± 317. Wade, G. N. (1972). Gonadal hormones and behavioral regulation of body weight. Physiology and Behavior, 8, 523 ± 534. Wade, G. N. (1976). Sex hormones and behavioral regulation of body weight. Advances in the Study of Behavior, 6, 201 ± 279. Walsh, B. T., Kissileff, H. R., Cassidy, S. M., & Dantzic, D. (1989). Eating behavior of women with bulimia. Archives of General Psychiatry, 46, 54 ± 58.

112

D.A. Anderson et al. / Eating Behaviors 2 (2001) 105±112

Weingarten, H. P., & Elston, D. (1991). Food cravings in a college population. Appetite, 17, 167 ± 175. Yanovski, S. Z., Leet, M., Yanovski, J. A., Flood, M., Gold, P. W., Kissileff, H. R., & Walsh, B. T. (1992). Food selection and intake of obese women with binge eating disorder. American Journal of Clinical Nutrition, 56, 975 ± 980. Zegman, M. A. (1984). Errors in food recording and calorie estimation: clinical and theoretical implications for obesity. Addictive Behaviors, 9, 347 ± 350.