- Email: [email protected]
Effects of disordered eating and obesity on weight, craving, and food intake during ad libitum smoking and abstinence
Eating Behaviors 5 (2004) 353 – 363
Effects of disordered eating and obesity on weight, craving, and food intake during ad libitum smoking and abstinence Karen K. Saules a,*, Cynthia S. Pomerleau b, Sandy M. Snedecor b, Rebecca Namenek Brouwer c, Erin E.M. Rosenberg b a
Psychology Department, EMU Psychology Clinic, Eastern Michigan University, 611 W. Cross Street, Ypsilanti, MI 48197, USA b Nicotine Research Laboratory, Department of Psychiatry, University of Michigan, Ann Arbor, MI, USA c Duke Comprehensive Cancer Center, Cancer Prevention, Detection and Control Research Program, Duke University Medical Center, Durham, NC, USA Received 18 November 2003; received in revised form 14 April 2004; accepted 22 April 2004
Abstract Although there is empirical support for the association between smoking, disordered eating, and subsequent weight gain upon smoking cessation, there have been no prospective studies to track changes in eating patterns during smoking abstinence and explore underlying biobehavioral processes. To help fill these gaps, we recruited four groups of women (N=48, 12/group) based on presence vs. absence of obesity and on low vs. high risk of severe dieting and/or binge-eating to participate in a laboratory study of eating in the context of ad libitum smoking and smoking abstinence. Participants [mean age 31.3 years; Fagerstrom Test of Nicotine Dependence (FTND) 4.3; smoking rate 18.7 cigarettes/day] completed two sessions: one after ad libitum smoking, the other after 2 days’ smoking abstinence, in counterbalanced order. After a half-day’s restricted eating, participants watched a video, with measured amounts of preselected preferred food available throughout. Cigarettes were available during the ad libitum smoking session. High-risk women weighed more after 2 days’ abstinence than during the ad libitum smoking condition, whereas low-risk women did not differ across conditions. Nicotine craving changed significantly more in anticipation of nicotine deprivation for high-BMI women than their low-BMI counterparts. Caloric intake was marginally attenuated during abstinence for low-BMI compared with high-BMI participants ( P < .10), an effect primarily accounted for by differences in protein intake ( P < .10). These findings suggest that low-BMI women may be less prone to weight gain during early abstinence, possibly because they compensate for metabolic changes induced by nicotine washout by eating less. Craving increases experienced by high-BMI
* Corresponding author. Tel.: +1-734-487-4988; fax: +1-734-487-4989. E-mail address: [email protected] (K.K. Saules). 1471-0153/$ - see front matter D 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.eatbeh.2004.04.011
354
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
women during abstinence under conditions of food deprivation may contribute to difficulty quitting in these women. D 2004 Elsevier Ltd. All rights reserved. Keywords: Nicotine; Smoking; Obesity; Eating disorder; Women; Weight
1. Introduction Because nicotine is an addictive drug that also suppresses body weight, in part via anorectic effects, the relationships among smoking, eating, and body weight in women are highly complex on both the biological and behavioral level. Moreover, the culture clash between those who are concerned about the health risks of body image distortion and those who are concerned about the health risks of obesity has further complicated our thinking about these issues. On the one hand, many women appear to use smoking as a ‘‘compensatory behavior.’’ Although not necessarily diagnosable as bulimic, they may evince considerable body image dissatisfaction focused on thinness and engage in a complex of unhealthful dieting and nutritional practices and compensatory behaviors, including severe dieting, fasting, diet pills, purging, and excessive exercise—behaviors intended to control weight but in fact often having the opposite effect because of a characteristic pattern of fluctuation between bingeing or overeating and dietary restraint. For such women, smoking may also constitute part of their arsenal of weight management tools, as evidenced by their overrepresentation among smokers (Pomerleau & Krahn, 1993). On the other hand, women smokers who are also obese are at increased risk of morbidity and mortality. Obesity predisposes to or exacerbates hypertension and diabetes, serves as a deterrent to fitness, and undermines quality of life in a variety of ways. Although there is evidence that smoking has an attenuated effect on metabolic rate in obese women (Audrain, Klesges, & Klesges, 1995), and although obese women may on average experience less postcessation weight gain (Williamson et al., 1991), even a modest weight gain is an undesirable outcome for women for whom reduction and not mere control of body mass is a health imperative. Moreover, for obese women who engage in unhealthful dieting and bingeing practices (23 –46% among those seeking treatment for weight reduction; de Zwaan & Mitchell, 1992), smoking cessation may increase the difficulties of following sound nutritional practices. These considerations present a confusing picture not only for the woman smoker but also for the service provider. Many are rightly concerned about the damage to both psychological health and nutritional status inflicted by the internalization of societal pressures to achieve thinness and the pattern of behaviors it engenders; to the extent that smoking is integrated into this pattern, the dangers are compounded. Yet, efforts to promote smoking cessation at the expense of weight control, while laudable (if often unsuccessful) in normal-weight women, may in fact be misguided in women who are already obese and whose interests are not well served by minimizing the health risks of either. Unfortunately, little evidence-based guidance is available from the literature. Although there is support for the existence of an association between smoking and eating disorders (Pomerleau & Krahn, 1993; Welch & Fairburn, 1998), as well as of a relationship between self-reported disinhibited eating styles and subsequent weight gain upon smoking cessation (Hudmon, Gritz, Clayton, & Nisenbaum, 1999), to date there have been no prospective studies specifically designed to track the emergence of
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
355
such eating patterns during smoking abstinence, to examine systematically the link between binge eating and early weight gain, or to explore the underlying biobehavioral processes in individuals susceptible to binge eating. A better understanding of this issue may have important implications for the design of cessation strategies in this population. To help fill these gaps, we recruited four groups of women based on presence vs. absence of obesity and on low vs. high risk of severe dieting and/or binge eating to participate in a laboratory study of eating in the context of ad libitum smoking and smoking abstinence. Presession weight, craving, and withdrawal symptomatology were also measured. We hypothesized that (1) overall, participants would eat significantly more, including a larger proportion of high-fat/high-sweet/high-caloric food, during the abstinence session than during the ad libitum smoking session; (2) high-risk women would eat significantly more, including a larger proportion of high-fat/high-sweet/high-caloric food, than low-risk women at the ad libitum smoking session; (3) there would be an interaction effect such that the increased eating, including increased intake of high-fat/high-sweet/high-caloric food, would be more pronounced in the high-risk group, particularly those who were also obese.
2. Method 2.1. Participants Forty-eight female smokers were recruited through local newspaper advertisements and posters. This study was limited to female participants in light of a large body of literature suggesting important gender differences in body image (Sorbara & Geliebter, 2002), weight concerns, and bingeing/dieting behaviors (George & Johnson, 2001)—particularly with respect to the use of smoking as a weight control strategy (Cawley, Markowitz, & Tauras, 2004; Klesges, Elliott, & Robinson, 1997; Perkins, Epstein, Fonte, & Grobe, 1994; Pomerleau et al., 1993). To be included in the study, candidates had to qualify for one of four cells of 12 participants each, based on measures of Body Mass Index (BMI) and dieting and bingeing behavior as described below. They also had to be (a) between 18 and 55 years old; (b) in good health; (c) not pregnant or breastfeeding within the past 3 months; (d) not taking medication that could affect appetite, weight, or mood; (e) currently smoking at least 10 cigarettes per day; and (f) free of current anorexia and psychosis. Although women with bulimia were eligible, those who reported engaging in dangerous purging behaviors that could be carried out in the laboratory were excluded. To obtain the final sample of 48 completers, a total of 62 women were enrolled, of whom 14 were unable to abstain from smoking and were thus dropped from the study. 2.1.1. Classification of participants Those with BMI V25 were considered to be of normal weight, while those with BMI z27 considered obese (WHO, 2002). Dieting and binge-eating behavior was assessed using the Dieting and Bingeing Severity Scale (DBSS), adapted from an instrument developed and validated by Krahn et al. (Drewnowski, Yee, Kurth, & Krahn, 1994; Krahn, Kurth, Demitrack, & Drenowski, 1992; Kurth, Krahn, Nairn, & Drewnowski, 1995). This scale classifies people into one of the following categories: 1 = not dieting at all, 2 = casual dieters, 3 = intense dieters, 4 = severe dieters, 5 = at risk dieters, and 6 = consistent with a
356
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
clinical diagnosis of bulimia nervosa. To qualify for inclusion, candidates had to fall within the following ranges: (1) BMI 25 or lower, DBSS category = 1–2; (2) BMI 25 or lower, DBSS category = 4–6; (3) BMI 27 or higher, DBSS category = 1–2; and (4) BMI 27 or higher, DBSS category = 4–6. 2.2. Procedure The study protocol was approved by the University of Michigan Medical School Institutional Review Board. During the screening appointment, the study was explained, and candidates were informed that the purpose of this research project was to gain a better understanding of the biological and behavioral aspects of smoking cessation in women. They were not explicitly told that food intake would be monitored but were aware that it would be freely available. Following collection of informed consent, height and weight were measured and a general health assessment was conducted. The Computerized Diagnostic Interview Schedule (C-DIS; Blouin, 1991) was used to rule out past or current psychosis, and manic or bipolar disorder, and the eating disorders portion of the Structured Clinical Interview for DSMIV (SCID; First, Spitzer, Gibbon, & Williams, 1995) was administered by a clinician to rule out past or current anorexia. A smoking history questionnaire, which included demographic data, smoking history, family smoking history, and the Fagerstrom Test of Nicotine Dependence (FTND; Fagerstro¨m, Heatherton, & Kozlowski, 1991; Heatherton, Kozlowski, Frecker, & Fagerstro¨m, 1991; Pomerleau, Carton, Lutzke, Flessland, & Pomerleau, 1994), was administered at baseline. In addition, in consideration of strong empirical support for a high degree of comorbidity of both depression and anxiety with binge eating (e.g., Brewerton et al., 1995; Johnsen, Gorin, Stone, & le Grange, 2003; Linde et al., 2004; Schwalberg, Barlow, Alger, & Howard, 1992; Speranza, Corcos, Atger, Pareneti, & Jeammet, 2003) and smoking (e.g., Black, Zimmerman, & Coryell, 1999; Breslau, 1995; Breslau, Novak, & Kessler, 2004), we administered the Center for Epidemiological Studies-Depression (CES-D; Radloff, 1977) and State Trait Anxiety Inventory (STAI-trait; Spielberger, 1988). Two sessions, one following ad libitum smoking and one following 2.5 days of abstinence from smoking, were scheduled approximately a week apart. To avoid changes in appetite and food intake associated with the premenstruum (Laessle, Tuschl, Schweiger, & Pirke, 1990; Pliner & Fleming, 1983; Tomelleri & Grunewald, 1987), regularly menstruating women were scheduled for their first session in the early-to-middle follicular phase. To control for possible order effects, sessions were counterbalanced so that participants were randomly assigned to one of two possible orders: abstinence first or ad libitum smoking first. The schedule for the 2 days preceding the session was as follows. On Day 0 at midnight (abstinence week only), the participant stopped smoking. On Days 1 and 2, the participant reported to the laboratory to provide an expired breath sample (to verify abstinence during nonsmoking week and to verify smoking during the ad libitum smoking week). The participant completed daily diaries both evenings (assessing smoking, alcohol and caffeine intake, caloric restriction, nicotine withdrawal symptomatology, and binge-eating episodes). On Day 2, the participant selected (1) the two foods from each of four categories to be sampled during the session (see Table 1) and (2) one of seven 2-h videos with limited emphasis on weight, eating, and smoking. In addition, she received a prepackaged breakfast consisting of Hostess mini muffins and Minute Maid juice, to be consumed no later than 9:00 a.m. the following morning. On Day 3, the day of the session, the participant reported to the laboratory at 2:30 p.m. after eating nothing since the prepackaged breakfast, a procedure intended to simulate a typical binge-eating
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
357
Table 1 Participants were asked to choose two foods from each of the following four groups: Group A (high sugar, high fat) 5 M&Ms (5 plain or 5 peanut) 5 Cookies (5 chocolate chip or 5 oatmeal) 5 Pie (5 chocolate cream or 5 banana cream) 5 Cake (5 carrot cake or 5 chocolate fudge) Group B (low sugar, high fat) 5 Ruffles potato chips with sour cream and onion dip 5 Doritos with nacho cheese dip 5 Salted peanuts 5 Club crackers with cheese spread Group C (high sugar, low fat) 5 Jelly beans 5 Jell-o 5 Licorice (5 black or 5 red) 5 Rice Krispie treats Group D (low sugar, low fat) 5 Pretzels 5 White cheddar rice cakes 5 Saltine crackers 5 Chex party mix
‘‘scenario.’’ Upon arrival, she was weighed and asked to complete a bipolar version of the Minnesota Withdrawal Scale (Hughes & Hatsukami, 1998), with scores ranging from 5 to + 5, measuring craving and seven DSM-IV withdrawal symptoms (depressed mood, insomnia, irritability, anxiety, difficulty concentrating, restlessness, and increased appetite). Following completion of preliminary measures, participants were exposed to a laboratory procedure designed to elicit differences in eating behavior that might be indicative of disinhibited eating, binge eating, or exaggerated weight gain after smoking cessation. Similar laboratory procedures have been used to elicit eating in the context of smoking vs. abstinence (Duffy & Hall, 1988), including in bulimic women smokers (Bulik, Dahl, Epstein, & Kaye, 1991). To prime eating, participants were first instructed to sample 2- to 3-g portions of each of the eight foods she had previously selected and rate them for taste, texture, and how often she typically ate this food. After completing this taste test, the participant was seated on a comfortable couch in a softly lighted room in front of a television and VCR. Beside her was a table on which measured amounts of the chosen foods were placed for the duration of the movie. A pitcher full of water and empty glass were available as well. A blood pressure cuff and a stopwatch were available for measuring pulse (every 15 min) and blood pressure (every 30 min). (The procedure of ostensibly measuring blood pressure enabled us to determine whether refills of food or water were required without overemphasizing the fact that we were interested in how much was consumed and possibly thereby confounding results). An ashtray and cigarettes were available during the ad libitum smoking session, and number of cigarettes smoked was monitored. If the participant was scheduled for an ad libitum smoking session the following week, she was asked to resume smoking as usual. Remaining portions of food were weighed
358
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
following the session. Caloric and nutritional content of all foods eaten was calculated using the Minnesota Nutritional Data System (MNDS; Nutrition Coordinating Center, 1994; Schakel, Sievert, & Buzzard, 1988).
3. Results Participant characteristics are shown in Table 2. Although the dieting groups differed in scores on the DBSS and the weight groups differed in BMI because of participant selection, no significant interaction effect was detected for either of these variables. The overweight women were significantly older than the normal-weight women; high-risk women had significantly more years of education than low-risk women. Although differences between groups in mean CES-D scores did not reach significance, highDBSS participants were significantly more likely to score 16 or above, the standard cutoff for clinical depression. An interaction effect emerged for nicotine dependence, with normal-weight low-risk women and overweight high-risk women scoring higher than women in the other two groups. 3.1. Smoking Twelve participants (two to five per group) did not smoke during the laboratory session that took place on their ‘‘smoking day,’’ although they were permitted to do so. (Several participants indicated that they had become used to not smoking indoors and indeed felt uncomfortable in doing so.) There were no Table 2 Participant characteristics (means F S.D.) Low risk
Normal weight DBSS (range 1 – 6)
High risk
Overweight
Normal weight
Overweight
Significance (disordered eating)
Significance (weight)
Significance (disordered eating weight)
( P=)
( P=)
( P=)
NS
NS NS
1.8 F 0.4
1.9 F 0.3
4.3 F 0.5
4.5 F 0.5
BMI (kg/m2)
21.1 F 1.5
31.3 F 3.2
22.6 F 1.7
33.3 F 6.3
F = 398.9, P=.000 NS
Age (years)
29.8 F 10.7
34.0 F 9.8
25.3 F 7.2
36.2 F 10.5
NS
Race (% White) Smoking rate (cigarettes/day) FTND (range 0 – 10) CES-D (range 0 – 60) CES-D (% scored z 16) STAI-trait (range 20 – 80)
83.3 20.2 F 8.4
91.7 17.2 F 5.2
100 17.0 F 7.6
83.3 20.2 F 8.3
NS NS
F = 94.9, P=.000 F = 7.3, P=.01 NS NS
5.2 F 2.4
3.0 F 2.1
3.2 F 2.4
5.6 F 2.2
NS
NS
6.2 F 3.4
8.6 F 8.4
12.2 F 10.3
9.6 F 6.9
NS
NS
F = 12.1, P=.001 NS
v2 = 4.9, P=.03 NS
NS
–
NS
NS
0
16.7
36.4
33.3
36.4 F 7.2
35.8 F 8.0
37.6 F 15.3
37.8 F 9.6
NS – NS
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
359
significant group differences in the proportion of participants who elected not to smoke. The remaining participants smoked from 1 to 7 cigarettes during the session in which they were permitted to smoke. There were no significant between-groups differences in number of cigarettes smoked. 3.2. Weight after 2 days’ abstinence and after ad libitum smoking Presession weight was tested via MANOVA, using session (ad libitum vs. abstinence) as a repeated measure and BMI (low vs. high) and DBSS (low vs. high) as grouping variables. A significant Session DBSS Group interaction emerged [ F(1,42) = 4.2, P < .05], such that high-DBSS women weighed more after 2 days’ abstinence than during ad libitum smoking (means of 165.6 vs. 164.3 lb, respectively), whereas low-DBSS women showed little difference (156.1 vs. 156.3 lb, respectively). 3.3. Presession withdrawal and craving Initially, presession withdrawal symptomatology and craving were tested via MANOVA, using session (ad libitum vs. abstinence) as a repeated measure and BMI (low vs. high) and DBSS (low vs. high) as grouping variables. Significant main effects for session were observed for depression [ F(1,43) = 17.1, P < .001], irritability [ F(1,43) = 45.5, P < .001], anxiety [ F(1,43) = 35.6, P < .001], difficulty concentrating [ F(1.43) = 18.5, P < .001], and restlessness [ F(1,43) = 38.0, P < .001], with levels prior to ad libitum smoking lower than those prior to the abstinence session. No other significant main or interaction effects emerged for withdrawal symptoms. Results for cigarette craving are shown in Fig. 1. A significant main effect emerged for session [ F(1,42) = 50.6, P < .001], with significant interaction effects for Session BMI Group [ F(1,42) = 4.9, P < .05] and Session DBSS Group BMI Group [ F(1,42) = 4.3, P < .05]. Specifically, relative to their low-BMI counterparts, those in the high-BMI group reported less cigarette craving during ad libitum smoking and more cigarette craving during the abstinence condition. The effects of our stratifying variables on craving appeared to be swamped by group differences in nicotine dependence, resulting in interaction effects that were difficult to interpret. Therefore, because we were interested in women’s reactivity to anticipated nicotine deprivation, craving was analyzed as a difference score (abstinence minus ad libitum ratings), using BMI (low vs. high) and DBSS (low vs. high) as grouping variables. Again, a significant effect for BMI group [ F(1) = 4.9, P < .05] emerged, as did a BMI Group DBSS Group interaction [ F(1) = 4.3, P < .05].
Fig. 1. Changes in craving from smoking to abstinence conditions by BMI and dieting risk (DBSS) categories. BMI group (F [1] = 4.9, P < .05); BMI group by DBSS group interaction (F [1] = 4.3, P < .05). Post-hocs: P < .05 for 2 < 1,3,4 and for Low < High BMI, overall.
360
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
3.4. Food and nutrient intake Food intake during the session was tested via MANOVA, using session (ad libitum vs. abstinence) as a repeated measure and BMI (low vs. high) and DBSS (low vs. high) as grouping variables. For caloric intake, low-BMI participants consumed marginally less [ F(1,45) = 3.14; P=.08] during abstinence than during ad libitum smoking (means of 1250 vs. 1484 cal, respectively), while the caloric intake of highBMI participants did not change across smoking conditions. Analysis of protein intake revealed a marginally significant Session BMI Group interaction [ F(1,45) = 3.4, P=.07], with low-BMI women consuming more protein calories during the smoking phase than during abstinence. For carbohydrate intake, a significant main effect emerged for smoking condition, with more carbohydrates consumed, overall, in the smoking condition [ F(1,45) = 4.05, P=.05]. No significant effects were observed in carbohydrate intake for smoking condition by BMI group, DBSS group, or their interaction. No significant differences or trends emerged for fat or sugar intake. The DBSS groups did not differ significantly in caloric intake.
4. Discussion The overweight women were significantly older than the normal-weight women, and the high-risk women were more likely to score in the ‘‘clinically depressed range’’ than the low-risk women. Because weight tends to increase until middle age (Sheehan, DuBrava, DeChello, & Fang, 2003), and because depression is often associated with disordered eating (Killen et al., 1987; Telch & Agras, 1994) and with weight concerns in smokers (Levine, Marcus, & Perkins, 2003), these differences are probably associated with group membership. The significant interaction on the measure of nicotine dependence was probably adventitious, as cigarettes per day did not differ significantly between groups. Interestingly, women at high risk for disordered eating weighed significantly more in anticipation of the abstinence condition, relative to when they were anticipating ad libitum availability of cigarettes. This finding suggests that, for such women, smoking may serve as a ‘‘recurrent inappropriate compensatory behavior. . .to prevent weight gain’’ (APA, 2000) that has not yet been recognized as a symptom of disordered eating. Overweight women showed significantly greater craving increases before the abstinence condition vs. ad libitum smoking, compared to their low-BMI counterparts. This pattern of craving reactivity in anticipation of abstinence was not observed in high-risk women. The caloric intake of overweight women was remarkably stable across conditions, while the normalweight women showed a reduction during the abstinence condition. Our findings for normal-weight women are consistent with a study by Perkins, Epstein, Sexton, Stiller, and Jacob (1992), who reported an unexpected decrease in caloric intake following placebo administration relative to nicotine administration (nasal spray). In contrast, because we did not observe this type of compensation amongst our high-BMI women, our results suggest that obese women may not adjust their eating behavior in response to metabolic shifts that may be induced by varying smoking patterns that typically occur over the course of a day, potentially contributing to increased problems with weight maintenance when embarking on cessation attempts. In combination, overweight women smokers’ high level of craving reactivity and their failure to modulate caloric intake as a function in smoking status may complicate smoking cessation efforts for
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
361
this population. In contrast, women at high risk for disordered eating may not experience comparable difficulties with craving when attempting cessation, but their tendency to gain weight during short-term abstinence suggests that they may use maladaptive weight control strategies during smoking cessation. Some caveats are in order. (1) Because this was a laboratory study, further research will be needed to determine if results will generalize to the natural environment. (2) Because we did not constrain smoking during the ad libitum session, it is possible that differences in smoking could have affected food intake. (3) Differences in age, nicotine dependence, and depression may have had differential effects upon dependent variables. (4) Entering the room and providing refills, rather than presenting large amount of foods and then leaving the participant alone, may have undercut our ability to elicit true binge eating. (5) Our sample may have been too small to reveal the full range of interactions between obesity and disordered eating. Nonetheless, our findings suggest that women smokers at risk for eating disorders and overweight women smokers may be more prone to maladaptive behavioral and/or biological responses during early smoking abstinence than normal-weight women without disordered eating patterns. In particular, the observed patterns of craving reactivity, enhanced weight gain, and/or disinhibited eating during early abstinence may compromise smoking cessation efforts during the first few fragile days after quitting.
Acknowledgements Preparation of this manuscript was supported by Grant HL52981 to Cynthia S. Pomerleau from the National Heart, Lung, and Blood Institute. A preliminary version of this manuscript was presented at the Annual Meeting of the Society for Research on Nicotine and Tobacco (February, 2004).
References APA (2000). Diagnostic & Statistical Manual of Mental Disorders (4th ed.) Text Revision. Washington, DC: American Psychiatric Association. Audrain, J. E., Klesges, R. C., & Klesges, L. M. (1995). Relationship between obesity and the metabolic effects of smoking in women. Health Psychology, 14, 116 – 1123. Black, D. W., Zimmerman, M., & Coryell, W. H. (1999). Cigarette smoking and psychiatric disorder in a community sample. Annals of Clinical Psychiatry, 11, 129 – 136. Blouin, A. (1991). C-DIS operations manual. Ottawa, ON: C-DIS Management Group. Breslau, N. (1995). Psychiatric comorbidity of smoking and nicotine dependence. Behavior Genetics, 25, 95 – 101. Breslau, N., Novak, S. P., & Kessler, R. C. (2004). Daily smoking and the subsequent onset of psychiatric disorders. Psychologie in der Medizin, 34, 323 – 333. Brewerton, T. D., Lydiard, R. B., Herzog, D. B., Brotman, A. W., O’Neil, P. M., & Ballenger, J. C. (1995). Comorbidity of axis I psychiatric disorders in bulimia nervosa. Journal of Clinical Psychiatry, 56, 77 – 80. Bulik, C. M., Dahl, R. E., Epstein, L. H., & Kaye, W. H. (1991). The effects of smoking deprivation on caloric intake in women with bulimia nervosa. International Journal of Eating Disorders, 10, 451 – 459. Cawley, J., Markowitz, S., & Tauras, J. (2004). Lighting up and slimming down: The effects of body weight and cigarette prices on adolescent smoking initiation. Health Economics, 23, 293 – 311. de Zwaan, M., & Mitchell, J. E. (1992). Binge eating in the obese. Annals of Medicine, 24, 303 – 308. Drewnowski, A., Yee, D. K., Kurth, C. L., & Krahn, D. D. (1994). Eating pathology and DSM-III-R bulimia nervosa: A continuum of behaviors. American Journal of Psychiatry, 151, 1217 – 1219.
362
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
Duffy, J., & Hall, S. M. (1988). Smoking abstinence, eating style, and food intake. Journal of Consulting and Clinical Psychology, 56, 417 – 421. Fagerstro¨m, K. O., Heatherton, T. F., & Kozlowski, L. T. (1991). Nicotine addiction and its assessment. Ear, Nose, and Throat Journal, 69, 763 – 768. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1995). The structured clinical interview for DSM-IV. New York: Biometrics Research Department New York State Psychiatric Institute. George, V. A., & Johnson, P. (2001). Weight loss behaviors and smoking in college students of diverse ethnicity. American Journal of Health Behavior, 25, 115 – 124. Heatherton, T. F., Kozlowski, L. T., Frecker, R. C., & Fagerstro¨m, K. O. (1991). The Fagerstro¨m Test for Nicotine Dependence: A revision of the Fagerstro¨m Tolerance Questionnaire. British Journal of Addiction, 86, 1119 – 1127. Hudmon, K. S., Gritz, E. R., Clayton, S., & Nisenbaum, R. (1999). Eating orientation, postcessation weight gain, and continued abstinence among female smokers receiving an unsolicited smoking cessation intervention. Health Psychology, 18, 29 – 36. Hughes, J. R., & Hatsukami, D. K. (1998). Errors in using tobacco withdrawal scales. Tobacco Control, 7, 92 – 93 (Letter to the Editor). Johnsen, L. A., Gorin, A., Stone, A. A., & le Grange, D. (2003). Characteristics of binge eating among women in the community seeking treatment for binge eating or weight loss. Eating Behaviors, 3, 295 – 305. Killen, J. D., Taylor, C. B., Telch, M. J., Robinson, T. N., Maron, D. J., & Saylor, K. E. (1987). Depressive symptoms and substance use among adolescent binge eaters and purgers: A defined population study. American Journal of Public Health, 77, 1539 – 1541. Klesges, R. C., Elliott, V. E., & Robinson, L. A. (1997). Chronic dieting and the belief that smoking controls body weight in a biracial, population-based adolescent sample. Tobacco Control, 6, 89 – 94. Krahn, D., Kurth, C. L., Demitrack, M., & Drewnowski, A. (1992). The relationship of dieting severity and bulimic behaviors to alcohol and other drug use in young women. Journal of Substance Abuse, 4, 341 – 353. Kurth, C. L., Krahn, D. D., Nairn, K., & Drewnowski, A. (1995). The severity of dieting and bingeing behaviors in college women: Interview validation of survey data. Journal of Psychiatric Research, 29, 211 – 225. Laessle, R. G., Tuschl, R. J., Schweiger, U., & Pirke, K. M. (1990). Mood changes and physical complaints during the normal menstrual cycle in healthy young women. Psychoneuroendocrinology, 15, 131 – 138. Levine, M. D., Marcus, M. D., & Perkins, K. A. (2003). A history of depression and smoking cessation outcomes among women concerned about post-cessation weight gain. Nicotine & Tobacco Research, 5, 69 – 76. Linde, J. A., Jeffery, R. W., Levy, R. L., Sherwood, N. E., Utter, J., Pronk, N. P., & Boyle, R. G. (2004). Binge eating disorder, weight control self-efficacy, and depression in overweight men and women. International Journal of Obesity and Related Metabolic Disorders, 28, 418 – 425. Nutrition Coordinating Center (1994). Minnesota nutrition data system software. Minneapolis, MN: University of Minnesota. Perkins, K. A., Epstein, L. H., Fonte, C., & Grobe, J. E. (1994). The reinforcing value of food during brief smoking abstinence in male and female smokers. Presented at the 15th Annual Meeting of SBM, Boston, MA, April, 1994. Perkins, K. S., Epstein, L. H., Sexton, J. E., Stiller, R. L., & Jacob, R. G. (1992). Effects of dose, gender, and level of physical activity on acute metabolic response to nicotine. Pharmacology, Biochemistry & Behavior, 40, 203 – 208. Pliner, P., & Fleming, A. S. (1983). Food intake, body weight, and sweetness preferences over the menstrual cycle in humans. Physiology & Behavior, 30, 663 – 666. Pomerleau, C. S., Carton, S. M., Lutzke, M. L., Flessland, K. A., & Pomerleau, O. F. (1994). Reliability of the Fagerstro¨m Tolerance Questionnaire and the Fagerstro¨m Test for Nicotine Dependence. Addictive Behaviors, 19, 33 – 39. Pomerleau, C. S., Ehrlich, E., Tate, J. C., Marks, J. L., Flessland, K. A., & Pomerleau, O. F. (1993). The female weight-control smoker: A profile. Journal of Substance Abuse, 5, 391 – 400. Pomerleau, C. S., & Krahn, D. (1993). Smoking and eating disorders: A connection? Journal of Addictive Diseases, 12, 169. Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385 – 401. Schakel, S. F., Sievert, Y. A., & Buzzard, I. M. (1988). Sources of data for developing and maintaining a nutrient database. Journal of the American Dietetic Association, 88, 1268 – 1271. Schwalberg, M. D., Barlow, D. H., Alger, S. A., & Howard, L. J. (1992). Comparison of bulimics, obese binge eaters, social phobics, and individuals with panic disorder on comorbidity across DSM-III-R anxiety disorders. Journal of Abnormal Psychology, 101, 675 – 681.
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
363
Sheehan, T. J., DuBrava, S., DeChello, L. M., & Fang, Z. (2003). Rates of weight change for black and white Americans over a twenty year period. International Journal of Obesity and Related Metabolic Disorders, 27, 498 – 504. Sorbara, M., & Geliebter, A. (2002). Body image disturbance in obese outpatients before and after weight loss in relation to race, gender, binge eating, and age of onset of obesity. International Journal of Eating Disorders, 31, 416 – 423. Speranza, M., Corcos, M., Atger, F., Paterniti, S., & Jeammet, P. (2003). Binge eating behaviours, depression and weight control strategies. Eating and Weight Disorders, 8, 201 – 206. Spielberger, C. D. (1988). State-trait anxiety inventory (form Y). In M. Hersen, & A. S. Ballack (Eds.), Dictionary of behavioral assessment ( pp. 448 – 450). New York: Pergamon. Telch, C. F., & Agras, W. S. (1994). Obesity, binge eating and psychopathology: Are they related? International Journal of Eating Disorders, 15, 53 – 61. Tomelleri, R., & Grunewald, K. K. (1987). Menstrual cycle and food cravings in young college. Journal of the American Dietetic Association, 87, 311 – 315. Welch, S. L., & Fairburn, C. G. (1998). Smoking and bulimia nervosa. International Journal of Eating Disorders, 23, 433 – 437. Williamson, D. F., Madans, J., Anda, R. F., Kleinman, J. C., Giovino, G. A., & Byers, T. (1991). Smoking cessation and severity of weight gain in a national cohort. New England Journal of Medicine, 324, 739 – 745. World Health Organization (WHO) (2002, January – February). Diet, nutrition, and the prevention of chronic diseases. Geneva, Switzerland: Author (Technical Report Series 916).
Effects of disordered eating and obesity on weight, craving, and food intake during ad libitum smoking and abstinence Karen K. Saules a,*, Cynthia S. Pomerleau b, Sandy M. Snedecor b, Rebecca Namenek Brouwer c, Erin E.M. Rosenberg b a
Psychology Department, EMU Psychology Clinic, Eastern Michigan University, 611 W. Cross Street, Ypsilanti, MI 48197, USA b Nicotine Research Laboratory, Department of Psychiatry, University of Michigan, Ann Arbor, MI, USA c Duke Comprehensive Cancer Center, Cancer Prevention, Detection and Control Research Program, Duke University Medical Center, Durham, NC, USA Received 18 November 2003; received in revised form 14 April 2004; accepted 22 April 2004
Abstract Although there is empirical support for the association between smoking, disordered eating, and subsequent weight gain upon smoking cessation, there have been no prospective studies to track changes in eating patterns during smoking abstinence and explore underlying biobehavioral processes. To help fill these gaps, we recruited four groups of women (N=48, 12/group) based on presence vs. absence of obesity and on low vs. high risk of severe dieting and/or binge-eating to participate in a laboratory study of eating in the context of ad libitum smoking and smoking abstinence. Participants [mean age 31.3 years; Fagerstrom Test of Nicotine Dependence (FTND) 4.3; smoking rate 18.7 cigarettes/day] completed two sessions: one after ad libitum smoking, the other after 2 days’ smoking abstinence, in counterbalanced order. After a half-day’s restricted eating, participants watched a video, with measured amounts of preselected preferred food available throughout. Cigarettes were available during the ad libitum smoking session. High-risk women weighed more after 2 days’ abstinence than during the ad libitum smoking condition, whereas low-risk women did not differ across conditions. Nicotine craving changed significantly more in anticipation of nicotine deprivation for high-BMI women than their low-BMI counterparts. Caloric intake was marginally attenuated during abstinence for low-BMI compared with high-BMI participants ( P < .10), an effect primarily accounted for by differences in protein intake ( P < .10). These findings suggest that low-BMI women may be less prone to weight gain during early abstinence, possibly because they compensate for metabolic changes induced by nicotine washout by eating less. Craving increases experienced by high-BMI
* Corresponding author. Tel.: +1-734-487-4988; fax: +1-734-487-4989. E-mail address: [email protected] (K.K. Saules). 1471-0153/$ - see front matter D 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.eatbeh.2004.04.011
354
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
women during abstinence under conditions of food deprivation may contribute to difficulty quitting in these women. D 2004 Elsevier Ltd. All rights reserved. Keywords: Nicotine; Smoking; Obesity; Eating disorder; Women; Weight
1. Introduction Because nicotine is an addictive drug that also suppresses body weight, in part via anorectic effects, the relationships among smoking, eating, and body weight in women are highly complex on both the biological and behavioral level. Moreover, the culture clash between those who are concerned about the health risks of body image distortion and those who are concerned about the health risks of obesity has further complicated our thinking about these issues. On the one hand, many women appear to use smoking as a ‘‘compensatory behavior.’’ Although not necessarily diagnosable as bulimic, they may evince considerable body image dissatisfaction focused on thinness and engage in a complex of unhealthful dieting and nutritional practices and compensatory behaviors, including severe dieting, fasting, diet pills, purging, and excessive exercise—behaviors intended to control weight but in fact often having the opposite effect because of a characteristic pattern of fluctuation between bingeing or overeating and dietary restraint. For such women, smoking may also constitute part of their arsenal of weight management tools, as evidenced by their overrepresentation among smokers (Pomerleau & Krahn, 1993). On the other hand, women smokers who are also obese are at increased risk of morbidity and mortality. Obesity predisposes to or exacerbates hypertension and diabetes, serves as a deterrent to fitness, and undermines quality of life in a variety of ways. Although there is evidence that smoking has an attenuated effect on metabolic rate in obese women (Audrain, Klesges, & Klesges, 1995), and although obese women may on average experience less postcessation weight gain (Williamson et al., 1991), even a modest weight gain is an undesirable outcome for women for whom reduction and not mere control of body mass is a health imperative. Moreover, for obese women who engage in unhealthful dieting and bingeing practices (23 –46% among those seeking treatment for weight reduction; de Zwaan & Mitchell, 1992), smoking cessation may increase the difficulties of following sound nutritional practices. These considerations present a confusing picture not only for the woman smoker but also for the service provider. Many are rightly concerned about the damage to both psychological health and nutritional status inflicted by the internalization of societal pressures to achieve thinness and the pattern of behaviors it engenders; to the extent that smoking is integrated into this pattern, the dangers are compounded. Yet, efforts to promote smoking cessation at the expense of weight control, while laudable (if often unsuccessful) in normal-weight women, may in fact be misguided in women who are already obese and whose interests are not well served by minimizing the health risks of either. Unfortunately, little evidence-based guidance is available from the literature. Although there is support for the existence of an association between smoking and eating disorders (Pomerleau & Krahn, 1993; Welch & Fairburn, 1998), as well as of a relationship between self-reported disinhibited eating styles and subsequent weight gain upon smoking cessation (Hudmon, Gritz, Clayton, & Nisenbaum, 1999), to date there have been no prospective studies specifically designed to track the emergence of
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
355
such eating patterns during smoking abstinence, to examine systematically the link between binge eating and early weight gain, or to explore the underlying biobehavioral processes in individuals susceptible to binge eating. A better understanding of this issue may have important implications for the design of cessation strategies in this population. To help fill these gaps, we recruited four groups of women based on presence vs. absence of obesity and on low vs. high risk of severe dieting and/or binge eating to participate in a laboratory study of eating in the context of ad libitum smoking and smoking abstinence. Presession weight, craving, and withdrawal symptomatology were also measured. We hypothesized that (1) overall, participants would eat significantly more, including a larger proportion of high-fat/high-sweet/high-caloric food, during the abstinence session than during the ad libitum smoking session; (2) high-risk women would eat significantly more, including a larger proportion of high-fat/high-sweet/high-caloric food, than low-risk women at the ad libitum smoking session; (3) there would be an interaction effect such that the increased eating, including increased intake of high-fat/high-sweet/high-caloric food, would be more pronounced in the high-risk group, particularly those who were also obese.
2. Method 2.1. Participants Forty-eight female smokers were recruited through local newspaper advertisements and posters. This study was limited to female participants in light of a large body of literature suggesting important gender differences in body image (Sorbara & Geliebter, 2002), weight concerns, and bingeing/dieting behaviors (George & Johnson, 2001)—particularly with respect to the use of smoking as a weight control strategy (Cawley, Markowitz, & Tauras, 2004; Klesges, Elliott, & Robinson, 1997; Perkins, Epstein, Fonte, & Grobe, 1994; Pomerleau et al., 1993). To be included in the study, candidates had to qualify for one of four cells of 12 participants each, based on measures of Body Mass Index (BMI) and dieting and bingeing behavior as described below. They also had to be (a) between 18 and 55 years old; (b) in good health; (c) not pregnant or breastfeeding within the past 3 months; (d) not taking medication that could affect appetite, weight, or mood; (e) currently smoking at least 10 cigarettes per day; and (f) free of current anorexia and psychosis. Although women with bulimia were eligible, those who reported engaging in dangerous purging behaviors that could be carried out in the laboratory were excluded. To obtain the final sample of 48 completers, a total of 62 women were enrolled, of whom 14 were unable to abstain from smoking and were thus dropped from the study. 2.1.1. Classification of participants Those with BMI V25 were considered to be of normal weight, while those with BMI z27 considered obese (WHO, 2002). Dieting and binge-eating behavior was assessed using the Dieting and Bingeing Severity Scale (DBSS), adapted from an instrument developed and validated by Krahn et al. (Drewnowski, Yee, Kurth, & Krahn, 1994; Krahn, Kurth, Demitrack, & Drenowski, 1992; Kurth, Krahn, Nairn, & Drewnowski, 1995). This scale classifies people into one of the following categories: 1 = not dieting at all, 2 = casual dieters, 3 = intense dieters, 4 = severe dieters, 5 = at risk dieters, and 6 = consistent with a
356
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
clinical diagnosis of bulimia nervosa. To qualify for inclusion, candidates had to fall within the following ranges: (1) BMI 25 or lower, DBSS category = 1–2; (2) BMI 25 or lower, DBSS category = 4–6; (3) BMI 27 or higher, DBSS category = 1–2; and (4) BMI 27 or higher, DBSS category = 4–6. 2.2. Procedure The study protocol was approved by the University of Michigan Medical School Institutional Review Board. During the screening appointment, the study was explained, and candidates were informed that the purpose of this research project was to gain a better understanding of the biological and behavioral aspects of smoking cessation in women. They were not explicitly told that food intake would be monitored but were aware that it would be freely available. Following collection of informed consent, height and weight were measured and a general health assessment was conducted. The Computerized Diagnostic Interview Schedule (C-DIS; Blouin, 1991) was used to rule out past or current psychosis, and manic or bipolar disorder, and the eating disorders portion of the Structured Clinical Interview for DSMIV (SCID; First, Spitzer, Gibbon, & Williams, 1995) was administered by a clinician to rule out past or current anorexia. A smoking history questionnaire, which included demographic data, smoking history, family smoking history, and the Fagerstrom Test of Nicotine Dependence (FTND; Fagerstro¨m, Heatherton, & Kozlowski, 1991; Heatherton, Kozlowski, Frecker, & Fagerstro¨m, 1991; Pomerleau, Carton, Lutzke, Flessland, & Pomerleau, 1994), was administered at baseline. In addition, in consideration of strong empirical support for a high degree of comorbidity of both depression and anxiety with binge eating (e.g., Brewerton et al., 1995; Johnsen, Gorin, Stone, & le Grange, 2003; Linde et al., 2004; Schwalberg, Barlow, Alger, & Howard, 1992; Speranza, Corcos, Atger, Pareneti, & Jeammet, 2003) and smoking (e.g., Black, Zimmerman, & Coryell, 1999; Breslau, 1995; Breslau, Novak, & Kessler, 2004), we administered the Center for Epidemiological Studies-Depression (CES-D; Radloff, 1977) and State Trait Anxiety Inventory (STAI-trait; Spielberger, 1988). Two sessions, one following ad libitum smoking and one following 2.5 days of abstinence from smoking, were scheduled approximately a week apart. To avoid changes in appetite and food intake associated with the premenstruum (Laessle, Tuschl, Schweiger, & Pirke, 1990; Pliner & Fleming, 1983; Tomelleri & Grunewald, 1987), regularly menstruating women were scheduled for their first session in the early-to-middle follicular phase. To control for possible order effects, sessions were counterbalanced so that participants were randomly assigned to one of two possible orders: abstinence first or ad libitum smoking first. The schedule for the 2 days preceding the session was as follows. On Day 0 at midnight (abstinence week only), the participant stopped smoking. On Days 1 and 2, the participant reported to the laboratory to provide an expired breath sample (to verify abstinence during nonsmoking week and to verify smoking during the ad libitum smoking week). The participant completed daily diaries both evenings (assessing smoking, alcohol and caffeine intake, caloric restriction, nicotine withdrawal symptomatology, and binge-eating episodes). On Day 2, the participant selected (1) the two foods from each of four categories to be sampled during the session (see Table 1) and (2) one of seven 2-h videos with limited emphasis on weight, eating, and smoking. In addition, she received a prepackaged breakfast consisting of Hostess mini muffins and Minute Maid juice, to be consumed no later than 9:00 a.m. the following morning. On Day 3, the day of the session, the participant reported to the laboratory at 2:30 p.m. after eating nothing since the prepackaged breakfast, a procedure intended to simulate a typical binge-eating
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
357
Table 1 Participants were asked to choose two foods from each of the following four groups: Group A (high sugar, high fat) 5 M&Ms (5 plain or 5 peanut) 5 Cookies (5 chocolate chip or 5 oatmeal) 5 Pie (5 chocolate cream or 5 banana cream) 5 Cake (5 carrot cake or 5 chocolate fudge) Group B (low sugar, high fat) 5 Ruffles potato chips with sour cream and onion dip 5 Doritos with nacho cheese dip 5 Salted peanuts 5 Club crackers with cheese spread Group C (high sugar, low fat) 5 Jelly beans 5 Jell-o 5 Licorice (5 black or 5 red) 5 Rice Krispie treats Group D (low sugar, low fat) 5 Pretzels 5 White cheddar rice cakes 5 Saltine crackers 5 Chex party mix
‘‘scenario.’’ Upon arrival, she was weighed and asked to complete a bipolar version of the Minnesota Withdrawal Scale (Hughes & Hatsukami, 1998), with scores ranging from 5 to + 5, measuring craving and seven DSM-IV withdrawal symptoms (depressed mood, insomnia, irritability, anxiety, difficulty concentrating, restlessness, and increased appetite). Following completion of preliminary measures, participants were exposed to a laboratory procedure designed to elicit differences in eating behavior that might be indicative of disinhibited eating, binge eating, or exaggerated weight gain after smoking cessation. Similar laboratory procedures have been used to elicit eating in the context of smoking vs. abstinence (Duffy & Hall, 1988), including in bulimic women smokers (Bulik, Dahl, Epstein, & Kaye, 1991). To prime eating, participants were first instructed to sample 2- to 3-g portions of each of the eight foods she had previously selected and rate them for taste, texture, and how often she typically ate this food. After completing this taste test, the participant was seated on a comfortable couch in a softly lighted room in front of a television and VCR. Beside her was a table on which measured amounts of the chosen foods were placed for the duration of the movie. A pitcher full of water and empty glass were available as well. A blood pressure cuff and a stopwatch were available for measuring pulse (every 15 min) and blood pressure (every 30 min). (The procedure of ostensibly measuring blood pressure enabled us to determine whether refills of food or water were required without overemphasizing the fact that we were interested in how much was consumed and possibly thereby confounding results). An ashtray and cigarettes were available during the ad libitum smoking session, and number of cigarettes smoked was monitored. If the participant was scheduled for an ad libitum smoking session the following week, she was asked to resume smoking as usual. Remaining portions of food were weighed
358
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
following the session. Caloric and nutritional content of all foods eaten was calculated using the Minnesota Nutritional Data System (MNDS; Nutrition Coordinating Center, 1994; Schakel, Sievert, & Buzzard, 1988).
3. Results Participant characteristics are shown in Table 2. Although the dieting groups differed in scores on the DBSS and the weight groups differed in BMI because of participant selection, no significant interaction effect was detected for either of these variables. The overweight women were significantly older than the normal-weight women; high-risk women had significantly more years of education than low-risk women. Although differences between groups in mean CES-D scores did not reach significance, highDBSS participants were significantly more likely to score 16 or above, the standard cutoff for clinical depression. An interaction effect emerged for nicotine dependence, with normal-weight low-risk women and overweight high-risk women scoring higher than women in the other two groups. 3.1. Smoking Twelve participants (two to five per group) did not smoke during the laboratory session that took place on their ‘‘smoking day,’’ although they were permitted to do so. (Several participants indicated that they had become used to not smoking indoors and indeed felt uncomfortable in doing so.) There were no Table 2 Participant characteristics (means F S.D.) Low risk
Normal weight DBSS (range 1 – 6)
High risk
Overweight
Normal weight
Overweight
Significance (disordered eating)
Significance (weight)
Significance (disordered eating weight)
( P=)
( P=)
( P=)
NS
NS NS
1.8 F 0.4
1.9 F 0.3
4.3 F 0.5
4.5 F 0.5
BMI (kg/m2)
21.1 F 1.5
31.3 F 3.2
22.6 F 1.7
33.3 F 6.3
F = 398.9, P=.000 NS
Age (years)
29.8 F 10.7
34.0 F 9.8
25.3 F 7.2
36.2 F 10.5
NS
Race (% White) Smoking rate (cigarettes/day) FTND (range 0 – 10) CES-D (range 0 – 60) CES-D (% scored z 16) STAI-trait (range 20 – 80)
83.3 20.2 F 8.4
91.7 17.2 F 5.2
100 17.0 F 7.6
83.3 20.2 F 8.3
NS NS
F = 94.9, P=.000 F = 7.3, P=.01 NS NS
5.2 F 2.4
3.0 F 2.1
3.2 F 2.4
5.6 F 2.2
NS
NS
6.2 F 3.4
8.6 F 8.4
12.2 F 10.3
9.6 F 6.9
NS
NS
F = 12.1, P=.001 NS
v2 = 4.9, P=.03 NS
NS
–
NS
NS
0
16.7
36.4
33.3
36.4 F 7.2
35.8 F 8.0
37.6 F 15.3
37.8 F 9.6
NS – NS
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
359
significant group differences in the proportion of participants who elected not to smoke. The remaining participants smoked from 1 to 7 cigarettes during the session in which they were permitted to smoke. There were no significant between-groups differences in number of cigarettes smoked. 3.2. Weight after 2 days’ abstinence and after ad libitum smoking Presession weight was tested via MANOVA, using session (ad libitum vs. abstinence) as a repeated measure and BMI (low vs. high) and DBSS (low vs. high) as grouping variables. A significant Session DBSS Group interaction emerged [ F(1,42) = 4.2, P < .05], such that high-DBSS women weighed more after 2 days’ abstinence than during ad libitum smoking (means of 165.6 vs. 164.3 lb, respectively), whereas low-DBSS women showed little difference (156.1 vs. 156.3 lb, respectively). 3.3. Presession withdrawal and craving Initially, presession withdrawal symptomatology and craving were tested via MANOVA, using session (ad libitum vs. abstinence) as a repeated measure and BMI (low vs. high) and DBSS (low vs. high) as grouping variables. Significant main effects for session were observed for depression [ F(1,43) = 17.1, P < .001], irritability [ F(1,43) = 45.5, P < .001], anxiety [ F(1,43) = 35.6, P < .001], difficulty concentrating [ F(1.43) = 18.5, P < .001], and restlessness [ F(1,43) = 38.0, P < .001], with levels prior to ad libitum smoking lower than those prior to the abstinence session. No other significant main or interaction effects emerged for withdrawal symptoms. Results for cigarette craving are shown in Fig. 1. A significant main effect emerged for session [ F(1,42) = 50.6, P < .001], with significant interaction effects for Session BMI Group [ F(1,42) = 4.9, P < .05] and Session DBSS Group BMI Group [ F(1,42) = 4.3, P < .05]. Specifically, relative to their low-BMI counterparts, those in the high-BMI group reported less cigarette craving during ad libitum smoking and more cigarette craving during the abstinence condition. The effects of our stratifying variables on craving appeared to be swamped by group differences in nicotine dependence, resulting in interaction effects that were difficult to interpret. Therefore, because we were interested in women’s reactivity to anticipated nicotine deprivation, craving was analyzed as a difference score (abstinence minus ad libitum ratings), using BMI (low vs. high) and DBSS (low vs. high) as grouping variables. Again, a significant effect for BMI group [ F(1) = 4.9, P < .05] emerged, as did a BMI Group DBSS Group interaction [ F(1) = 4.3, P < .05].
Fig. 1. Changes in craving from smoking to abstinence conditions by BMI and dieting risk (DBSS) categories. BMI group (F [1] = 4.9, P < .05); BMI group by DBSS group interaction (F [1] = 4.3, P < .05). Post-hocs: P < .05 for 2 < 1,3,4 and for Low < High BMI, overall.
360
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
3.4. Food and nutrient intake Food intake during the session was tested via MANOVA, using session (ad libitum vs. abstinence) as a repeated measure and BMI (low vs. high) and DBSS (low vs. high) as grouping variables. For caloric intake, low-BMI participants consumed marginally less [ F(1,45) = 3.14; P=.08] during abstinence than during ad libitum smoking (means of 1250 vs. 1484 cal, respectively), while the caloric intake of highBMI participants did not change across smoking conditions. Analysis of protein intake revealed a marginally significant Session BMI Group interaction [ F(1,45) = 3.4, P=.07], with low-BMI women consuming more protein calories during the smoking phase than during abstinence. For carbohydrate intake, a significant main effect emerged for smoking condition, with more carbohydrates consumed, overall, in the smoking condition [ F(1,45) = 4.05, P=.05]. No significant effects were observed in carbohydrate intake for smoking condition by BMI group, DBSS group, or their interaction. No significant differences or trends emerged for fat or sugar intake. The DBSS groups did not differ significantly in caloric intake.
4. Discussion The overweight women were significantly older than the normal-weight women, and the high-risk women were more likely to score in the ‘‘clinically depressed range’’ than the low-risk women. Because weight tends to increase until middle age (Sheehan, DuBrava, DeChello, & Fang, 2003), and because depression is often associated with disordered eating (Killen et al., 1987; Telch & Agras, 1994) and with weight concerns in smokers (Levine, Marcus, & Perkins, 2003), these differences are probably associated with group membership. The significant interaction on the measure of nicotine dependence was probably adventitious, as cigarettes per day did not differ significantly between groups. Interestingly, women at high risk for disordered eating weighed significantly more in anticipation of the abstinence condition, relative to when they were anticipating ad libitum availability of cigarettes. This finding suggests that, for such women, smoking may serve as a ‘‘recurrent inappropriate compensatory behavior. . .to prevent weight gain’’ (APA, 2000) that has not yet been recognized as a symptom of disordered eating. Overweight women showed significantly greater craving increases before the abstinence condition vs. ad libitum smoking, compared to their low-BMI counterparts. This pattern of craving reactivity in anticipation of abstinence was not observed in high-risk women. The caloric intake of overweight women was remarkably stable across conditions, while the normalweight women showed a reduction during the abstinence condition. Our findings for normal-weight women are consistent with a study by Perkins, Epstein, Sexton, Stiller, and Jacob (1992), who reported an unexpected decrease in caloric intake following placebo administration relative to nicotine administration (nasal spray). In contrast, because we did not observe this type of compensation amongst our high-BMI women, our results suggest that obese women may not adjust their eating behavior in response to metabolic shifts that may be induced by varying smoking patterns that typically occur over the course of a day, potentially contributing to increased problems with weight maintenance when embarking on cessation attempts. In combination, overweight women smokers’ high level of craving reactivity and their failure to modulate caloric intake as a function in smoking status may complicate smoking cessation efforts for
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
361
this population. In contrast, women at high risk for disordered eating may not experience comparable difficulties with craving when attempting cessation, but their tendency to gain weight during short-term abstinence suggests that they may use maladaptive weight control strategies during smoking cessation. Some caveats are in order. (1) Because this was a laboratory study, further research will be needed to determine if results will generalize to the natural environment. (2) Because we did not constrain smoking during the ad libitum session, it is possible that differences in smoking could have affected food intake. (3) Differences in age, nicotine dependence, and depression may have had differential effects upon dependent variables. (4) Entering the room and providing refills, rather than presenting large amount of foods and then leaving the participant alone, may have undercut our ability to elicit true binge eating. (5) Our sample may have been too small to reveal the full range of interactions between obesity and disordered eating. Nonetheless, our findings suggest that women smokers at risk for eating disorders and overweight women smokers may be more prone to maladaptive behavioral and/or biological responses during early smoking abstinence than normal-weight women without disordered eating patterns. In particular, the observed patterns of craving reactivity, enhanced weight gain, and/or disinhibited eating during early abstinence may compromise smoking cessation efforts during the first few fragile days after quitting.
Acknowledgements Preparation of this manuscript was supported by Grant HL52981 to Cynthia S. Pomerleau from the National Heart, Lung, and Blood Institute. A preliminary version of this manuscript was presented at the Annual Meeting of the Society for Research on Nicotine and Tobacco (February, 2004).
References APA (2000). Diagnostic & Statistical Manual of Mental Disorders (4th ed.) Text Revision. Washington, DC: American Psychiatric Association. Audrain, J. E., Klesges, R. C., & Klesges, L. M. (1995). Relationship between obesity and the metabolic effects of smoking in women. Health Psychology, 14, 116 – 1123. Black, D. W., Zimmerman, M., & Coryell, W. H. (1999). Cigarette smoking and psychiatric disorder in a community sample. Annals of Clinical Psychiatry, 11, 129 – 136. Blouin, A. (1991). C-DIS operations manual. Ottawa, ON: C-DIS Management Group. Breslau, N. (1995). Psychiatric comorbidity of smoking and nicotine dependence. Behavior Genetics, 25, 95 – 101. Breslau, N., Novak, S. P., & Kessler, R. C. (2004). Daily smoking and the subsequent onset of psychiatric disorders. Psychologie in der Medizin, 34, 323 – 333. Brewerton, T. D., Lydiard, R. B., Herzog, D. B., Brotman, A. W., O’Neil, P. M., & Ballenger, J. C. (1995). Comorbidity of axis I psychiatric disorders in bulimia nervosa. Journal of Clinical Psychiatry, 56, 77 – 80. Bulik, C. M., Dahl, R. E., Epstein, L. H., & Kaye, W. H. (1991). The effects of smoking deprivation on caloric intake in women with bulimia nervosa. International Journal of Eating Disorders, 10, 451 – 459. Cawley, J., Markowitz, S., & Tauras, J. (2004). Lighting up and slimming down: The effects of body weight and cigarette prices on adolescent smoking initiation. Health Economics, 23, 293 – 311. de Zwaan, M., & Mitchell, J. E. (1992). Binge eating in the obese. Annals of Medicine, 24, 303 – 308. Drewnowski, A., Yee, D. K., Kurth, C. L., & Krahn, D. D. (1994). Eating pathology and DSM-III-R bulimia nervosa: A continuum of behaviors. American Journal of Psychiatry, 151, 1217 – 1219.
362
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
Duffy, J., & Hall, S. M. (1988). Smoking abstinence, eating style, and food intake. Journal of Consulting and Clinical Psychology, 56, 417 – 421. Fagerstro¨m, K. O., Heatherton, T. F., & Kozlowski, L. T. (1991). Nicotine addiction and its assessment. Ear, Nose, and Throat Journal, 69, 763 – 768. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1995). The structured clinical interview for DSM-IV. New York: Biometrics Research Department New York State Psychiatric Institute. George, V. A., & Johnson, P. (2001). Weight loss behaviors and smoking in college students of diverse ethnicity. American Journal of Health Behavior, 25, 115 – 124. Heatherton, T. F., Kozlowski, L. T., Frecker, R. C., & Fagerstro¨m, K. O. (1991). The Fagerstro¨m Test for Nicotine Dependence: A revision of the Fagerstro¨m Tolerance Questionnaire. British Journal of Addiction, 86, 1119 – 1127. Hudmon, K. S., Gritz, E. R., Clayton, S., & Nisenbaum, R. (1999). Eating orientation, postcessation weight gain, and continued abstinence among female smokers receiving an unsolicited smoking cessation intervention. Health Psychology, 18, 29 – 36. Hughes, J. R., & Hatsukami, D. K. (1998). Errors in using tobacco withdrawal scales. Tobacco Control, 7, 92 – 93 (Letter to the Editor). Johnsen, L. A., Gorin, A., Stone, A. A., & le Grange, D. (2003). Characteristics of binge eating among women in the community seeking treatment for binge eating or weight loss. Eating Behaviors, 3, 295 – 305. Killen, J. D., Taylor, C. B., Telch, M. J., Robinson, T. N., Maron, D. J., & Saylor, K. E. (1987). Depressive symptoms and substance use among adolescent binge eaters and purgers: A defined population study. American Journal of Public Health, 77, 1539 – 1541. Klesges, R. C., Elliott, V. E., & Robinson, L. A. (1997). Chronic dieting and the belief that smoking controls body weight in a biracial, population-based adolescent sample. Tobacco Control, 6, 89 – 94. Krahn, D., Kurth, C. L., Demitrack, M., & Drewnowski, A. (1992). The relationship of dieting severity and bulimic behaviors to alcohol and other drug use in young women. Journal of Substance Abuse, 4, 341 – 353. Kurth, C. L., Krahn, D. D., Nairn, K., & Drewnowski, A. (1995). The severity of dieting and bingeing behaviors in college women: Interview validation of survey data. Journal of Psychiatric Research, 29, 211 – 225. Laessle, R. G., Tuschl, R. J., Schweiger, U., & Pirke, K. M. (1990). Mood changes and physical complaints during the normal menstrual cycle in healthy young women. Psychoneuroendocrinology, 15, 131 – 138. Levine, M. D., Marcus, M. D., & Perkins, K. A. (2003). A history of depression and smoking cessation outcomes among women concerned about post-cessation weight gain. Nicotine & Tobacco Research, 5, 69 – 76. Linde, J. A., Jeffery, R. W., Levy, R. L., Sherwood, N. E., Utter, J., Pronk, N. P., & Boyle, R. G. (2004). Binge eating disorder, weight control self-efficacy, and depression in overweight men and women. International Journal of Obesity and Related Metabolic Disorders, 28, 418 – 425. Nutrition Coordinating Center (1994). Minnesota nutrition data system software. Minneapolis, MN: University of Minnesota. Perkins, K. A., Epstein, L. H., Fonte, C., & Grobe, J. E. (1994). The reinforcing value of food during brief smoking abstinence in male and female smokers. Presented at the 15th Annual Meeting of SBM, Boston, MA, April, 1994. Perkins, K. S., Epstein, L. H., Sexton, J. E., Stiller, R. L., & Jacob, R. G. (1992). Effects of dose, gender, and level of physical activity on acute metabolic response to nicotine. Pharmacology, Biochemistry & Behavior, 40, 203 – 208. Pliner, P., & Fleming, A. S. (1983). Food intake, body weight, and sweetness preferences over the menstrual cycle in humans. Physiology & Behavior, 30, 663 – 666. Pomerleau, C. S., Carton, S. M., Lutzke, M. L., Flessland, K. A., & Pomerleau, O. F. (1994). Reliability of the Fagerstro¨m Tolerance Questionnaire and the Fagerstro¨m Test for Nicotine Dependence. Addictive Behaviors, 19, 33 – 39. Pomerleau, C. S., Ehrlich, E., Tate, J. C., Marks, J. L., Flessland, K. A., & Pomerleau, O. F. (1993). The female weight-control smoker: A profile. Journal of Substance Abuse, 5, 391 – 400. Pomerleau, C. S., & Krahn, D. (1993). Smoking and eating disorders: A connection? Journal of Addictive Diseases, 12, 169. Radloff, L. S. (1977). The CES-D Scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385 – 401. Schakel, S. F., Sievert, Y. A., & Buzzard, I. M. (1988). Sources of data for developing and maintaining a nutrient database. Journal of the American Dietetic Association, 88, 1268 – 1271. Schwalberg, M. D., Barlow, D. H., Alger, S. A., & Howard, L. J. (1992). Comparison of bulimics, obese binge eaters, social phobics, and individuals with panic disorder on comorbidity across DSM-III-R anxiety disorders. Journal of Abnormal Psychology, 101, 675 – 681.
K.K. Saules et al. / Eating Behaviors 5 (2004) 353–363
363
Sheehan, T. J., DuBrava, S., DeChello, L. M., & Fang, Z. (2003). Rates of weight change for black and white Americans over a twenty year period. International Journal of Obesity and Related Metabolic Disorders, 27, 498 – 504. Sorbara, M., & Geliebter, A. (2002). Body image disturbance in obese outpatients before and after weight loss in relation to race, gender, binge eating, and age of onset of obesity. International Journal of Eating Disorders, 31, 416 – 423. Speranza, M., Corcos, M., Atger, F., Paterniti, S., & Jeammet, P. (2003). Binge eating behaviours, depression and weight control strategies. Eating and Weight Disorders, 8, 201 – 206. Spielberger, C. D. (1988). State-trait anxiety inventory (form Y). In M. Hersen, & A. S. Ballack (Eds.), Dictionary of behavioral assessment ( pp. 448 – 450). New York: Pergamon. Telch, C. F., & Agras, W. S. (1994). Obesity, binge eating and psychopathology: Are they related? International Journal of Eating Disorders, 15, 53 – 61. Tomelleri, R., & Grunewald, K. K. (1987). Menstrual cycle and food cravings in young college. Journal of the American Dietetic Association, 87, 311 – 315. Welch, S. L., & Fairburn, C. G. (1998). Smoking and bulimia nervosa. International Journal of Eating Disorders, 23, 433 – 437. Williamson, D. F., Madans, J., Anda, R. F., Kleinman, J. C., Giovino, G. A., & Byers, T. (1991). Smoking cessation and severity of weight gain in a national cohort. New England Journal of Medicine, 324, 739 – 745. World Health Organization (WHO) (2002, January – February). Diet, nutrition, and the prevention of chronic diseases. Geneva, Switzerland: Author (Technical Report Series 916).