Further statistical and clinical validity for the Weight Efficacy Lifestyle Questionnaire-Short Form

Eating Behaviors 18 (2015) 115–119

Contents lists available at ScienceDirect

Eating Behaviors

Further statistical and clinical validity for the Weight Efficacy Lifestyle Questionnaire-Short Form Gretchen E. Ames a,⁎, Michael G. Heckman b, Nancy N. Diehl b, Karen B. Grothe c, Matthew M. Clark c a b c

Department of Psychiatry and Psychology, Mayo Clinic, 4500 San Pablo Road, Jacksonville, FL 32224, USA Division of Biomedical Statistics and Informatics, Mayo Clinic, 4500 San Pablo Road, Jacksonville, FL 32224, USA Department of Psychiatry and Psychology, Mayo Clinic, 200 First Street, SW, Rochester, MN 55905, USA

a r t i c l e

i n f o

Article history: Received 3 February 2015 Received in revised form 5 April 2015 Accepted 13 May 2015 Available online 21 May 2015 Keywords: Obesity Eating Self-Efficacy Weight Loss

a b s t r a c t Identifying barriers to long-term adherence to reduced energy intake and increased physical activity level is critically important for obese patients seeking weight loss treatment. Previous research has identified that one such barrier is low eating self-efficacy or poor confidence in one's ability to control eating behavior in the presence of challenging situations. Accordingly, a valid, brief measure of eating self-efficacy for longitudinal assessment of weight loss and regain is needed. The purpose of this study was to test the internal consistency and clinical validity of the Weight Efficacy Lifestyle Questionnaire-Short Form (WEL-SF). Participants were 1740 consecutive obese patients who presented for a psychological evaluation in consideration for bariatric surgery. Median BMI was 44.9 (range: 35.0–111.9), age 48.7 years (range: 18.9–77.3 years), and patients were predominantly female (71.1%) and Caucasian (90.8%). The median WEL-SF total score was 56 (range: 0–80) and Cronbach's alpha measuring internal consistency was 0.92 with a one-factor structure. In terms of clinical validation, lower WEL-SF total scores were significantly associated with higher rates of binge eating episodes (P b 0.0001), food addiction severity and dependence (P b 0.0001), night eating syndrome (P b 0.0001), depression (P b 0.0001), and anxiety (P b 0.0001). In contrast, higher WEL-SF total scores were associated with higher weight management selfefficacy (P b 0.0001) and motivation to make positive lifestyle changes (P b 0.0001). Taken together, these findings suggest that the WEL-SF is a psychometrically valid clinically meaningful measure of eating self-efficacy. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction In spite of many advances in medical and surgical treatments for obesity, maintaining lost weight remains a significant challenge for obese patients, and many patients relapse (MacLean et al., 2014). A complex interaction of biological adaptation, environmental factors, and behavioral recidivism contribute to weight regain following weight loss treatment (MacLean et al., 2014; Ochner, Tsai, Kushner, & Wadden, 2015). One possible explanation for behavioral recidivism is that when weight loss plateaus after 6–8 months of treatment, the perceived effort required to adhere to a healthy diet and exercise prescription is greater than the perceived benefit (Perri, 1998). Thus, identifying barriers to long-term adherence to reduced energy intake and increased physical activity is critically important. One such barrier is low eating selfefficacy (Clark, Abrams, Niaura, Eaton, & Rossi, 1991). Self-efficacy is defined as confidence in one's ability to perform specific behaviors in the presence of challenging situations (Bandura, 1977).

⁎ Corresponding author. Tel.: +1 904 953 2523; fax: +1 904 953 6586. E-mail addresses: [email protected] (G.E. Ames), [email protected] (M.G. Heckman), [email protected] (N.N. Diehl), [email protected] (K.B. Grothe), [email protected] (M.M. Clark).

http://dx.doi.org/10.1016/j.eatbeh.2015.05.003 1471-0153/© 2015 Elsevier Ltd. All rights reserved.

Previous research has consistently demonstrated that eating selfefficacy improves with weight loss and remains high when lost weight is maintained (Batsis et al., 2009; Clark, Cargill, Medeiros, & Pera, 1996; Clark et al., 1991; DePue, Clark, Ruggiero, Medeiros, & Pera, 1995). Given the pragmatics of a busy clinical practice, a brief valid measure of eating self-efficacy has the potential to improve patient screening and care. The Weight Efficacy Lifestyle Questionnaire-Short Form (WEL-SF), measures patients' confidence in their ability to control eating behavior in challenging situations and was developed as a brief measure for use in research and clinical practice (Ames, Heckman, Grothe, & Clark, 2012). The two aims of this study were to provide further statistical and clinical validation of the WEL-SF in a large multisite sample of obese patients seeking bariatric surgery. First, normative values and internal consistency of the WEL-SF were examined. We hypothesized that the WEL-SF items would demonstrate strong internal consistency and a one-factor structure as found in previous studies (Ames et al., 2012; Flolo, Andersen, Nielsen, & Natvig, 2014). Second, a recent study revealed high rates of disordered eating behaviors prior to bariatric surgery in patients who participated in the Longitudinal Assessment of Bariatric Surgery-2 (LABS-2) (Belle et al., 2013). These behaviors included binge eating, night eating, grazing, skipping breakfast and evening hyperphagia and were associated with greater levels of psychopathology (Mitchell, King, Courcoulas, et al., 2014b). In terms of clinical

116

G.E. Ames et al. / Eating Behaviors 18 (2015) 115–119

validation, the associations between WEL-SF total scores and problematic eating behaviors and mood were examined. We hypothesized that lower total scores on the WEL-SF would be significantly associated with higher levels of disordered eating behaviors and lower mood. In contrast, we hypothesized that higher WEL-SF total scores would be significantly associated with higher ratings of self-efficacy for exercise, weight management, and motivation to make lifestyle changes. 2. Methods 2.1. Study participants This study included a total of 1740 consecutive patients who presented for a psychological evaluation at a multisite medical center in the south-east (N = 381) and mid-west (N = 1359) for consideration of bariatric surgery between January, 2012 and October, 2014. The evaluation consisted of a structured clinical interview and completion of validated screening questionnaires recommended by LABS-2 (Belle et al., 2013). Patients were excluded if they were younger than 18, if they had a body mass index (BMI) of less than 35, or if they did not have complete data on all 8 items that comprise the WEL-SF. At the time of the evaluation, demographic information including age, gender, race, and BMI were collected. Information was unavailable regarding race for 22 patients. This study was approved by our Institutional Research Review Board. 2.2. Measures 2.2.1. Eating self-efficacy The Weight Efficacy Lifestyle Questionnaire-Short Form (WEL-SF) is an 8-item self-report measure of confidence for controlling eating behavior in a variety of challenging situations (Ames et al., 2012). Items are rated on a 0 to 10 scale, with a score of 0 indicating “not confident at all that I can resist overeating” and a score of 10 indicating “very confident that I can resist overeating.” The measure yields a total score of 80 with lower scores revealing less confidence in one's ability to control eating behavior. 2.2.2. Binge eating disorder The Questionnaire of Eating and Weight Patterns-Revised (QEWP-R) is a self-report measure (Spitzer, Yanovski, S., & Marcus, M.D., 1993) that provides acceptable sensitivity in identifying individuals with Binge Eating Disorder based on Diagnostic and Statistical Manual of Mental Disorders-IV (DSM-IV) criteria (American Psychiatric Association, 1994). The items used to assess symptoms of Binge Eating Disorder included consumption of a large amount of food and sense of loss of control on at least 2–3 days per week during the previous 6 months. Patients who endorsed “yes” on these items were considered to have a positive screen for Binge Eating Disorder. Information regarding the QEWP-R was not available for 51 patients. 2.2.3. Food addiction The Yale Food Addiction Scale (YFAS) is a 25-item self-report measure of food addiction (Gearhardt, Corbin, & Brownell, 2009). Patients answered items on a scale ranging from “never” to “4 or more times per day” and items with a dichotomous “yes” or “no” answer. Answers on these items are assigned a value of 0 or 1. The measure yields a severity index ranging from 0 to 7 that combines items designed to assess tolerance, withdrawal, continued use, activities given up, loss of control, and unsuccessful attempts at cutting down. The measure also yields a dependence score of 1 (diagnosis met) or 0 (diagnosis not met). The dependence score consists of items that assess level of distress and impaired functioning combined with a severity score of N3. The YFAS has previously been validated with patients seeking bariatric surgery (Meule, Heckel, & Kubler, 2012). The YFAS severity index was not

available for 3 patients, while the dependence score was not assessed for 31 patients. 2.2.4. Night eating syndrome The questions used in LABS-2 to measure Night Eating Syndrome were used in this study (Belle et al., 2013). The question, “During the past 3-months how much of your daily intake did you consume after supper time?” (None, Up to 25%, About half, More than half, Almost all), was used to measure evening hyperphagia. The question, “During the past 3-months if you got up in the middle of the night, how often did you snack?” (Never, Sometimes, About half the time, Usually, Always), was used to measure nocturnal eating. Patients who endorsed about half or greater on either item were considered to have a positive screen for Night Eating Syndrome (Mitchell, King, Courcoulas, et al., 2014b). These two questions were not answered by 58 and 30 patients, respectively. 2.2.5. Depressive symptoms The Patient Health Questionnaire-9 (PHQ-9) is a 9-item self-report measure of depression symptoms with excellent internal consistency and test–retest reliability (Kroenke, Spitzer, & Williams, 2001). Patients answer items on a scale 0 (not at all) to 3 (nearly every day) and responses are categorized according to level of depression including minimal (0–4), mild (5–9), moderate (10–14), moderately severe (15–19), and severe (20–27). Information was not available for the PHQ-9 for 242 patients. 2.2.6. Anxiety level The Generalized Anxiety Disorder-7 (GAD-7) is a 7-item self-report measure designed to identify probable cases of GAD with a score of 10 or greater (Spitzer, Kroenke, Williams, & Lowe, 2006). Patients answer items on a scale 0 (not at all) to 3 (nearly every day) and responses are categorized according to level of anxiety including minimal (0–4), mild (5–9), moderate (10–14), and severe (15–21). The GAD-7 was not assessed for 125 of the study patients. 2.2.7. Alcohol use The Alcohol Use Disorders Identification Test (AUDIT) is a 10-item self-report measure of three domains including hazardous alcohol use, dependence symptoms, and harmful alcohol use (Saunders, Aasland, Babor, de la Fuente, & Grant, 1993). Each item contains a set of responses to choose from where scores range from 0 to 4. A score of 8 or greater is an indicator of harmful or hazardous alcohol use and possible dependence. A total of 118 patients did not have information available regarding the AUDIT. 2.2.8. Victim of childhood sexual abuse The Adverse Childhood Experiences (ACE) scale is a 10-item selfreport measure including domains of abuse, neglect, and household dysfunction (Dube et al., 2003). Sexual abuse was defined by question number three from the ACE, “Did an adult or a person at least 5 years older than you touch or fondle you or have you them touch in sexual way OR try to have oral, anal, or vaginal sex with you?” (Yes or No). During the study period, the ACE replaced the previous measure of childhood sexual abuse to reduce cost. Thus, it was collected for only half of the study period; the ACE was not assessed for 768 of the study patients. 2.2.9. Smoking status Patients self-reported their smoking status with the following question. “Please categorize your history of tobacco use: never smoked, former smoker, current smoker.” Smoking history was not available for 380 patients.

G.E. Ames et al. / Eating Behaviors 18 (2015) 115–119

2.2.10. General weight management The pre-surgical assessment battery contained an additional four questions pertaining to exercise and weight management self-efficacy: 1. “How confident are you that you can have a physically active lifestyle (30 min of physical activity on most days of the week)?” 2. “How confident are you that you will be able to successfully manage your weight?” 3. “How motivated are you to begin making lifestyle changes for weight management?” Patients answered these items using a scale ranging from 0 (least) to 10 (most) confident. The final question was “How often do you weigh yourself?” (never, once per year, every couple of months, once a week, every couple of days, every day, greater than one time per day). Single-item measures of self-efficacy have been validated previously (Clark et al., 2007). Information was unavailable regarding these four questions for a maximum of 39 patients. 2.3. Statistical analysis Internal consistency of the WEL-SF was evaluated using Pearson's correlation coefficient r between individual WEL-SF items as well as Cronbach's alpha. Associations of demographics (age, race, gender, BMI, clinic site) and clinical questionnaires with WEL-SF total score were evaluated using linear regression models. Single variable (i.e. unadjusted) models were utilized, as well as multivariable models that were adjusted for any demographic characteristic that was associated with WEL-SF total score with a P-value of 0.05 or lower in single variable analysis. The continuous variables of age and BMI were considered as categorical variables based on approximate sample quartiles for age, and based on pre-defined categories that were clinically meaningful for BMI (class 2 obesity [35–39.9], class 3 obesity [40–49.9], and super morbid obesity [≥50]). To account for the number of statistical tests performed in linear regression association analysis, the Holm step-down method was utilized (Holm, 1979), after which P-values of 0.0063 or lower were considered as statistically significant. All statistical analyses were performed using SAS (version 9.2; SAS Institute, Inc., Cary, North Carolina) and R Statistical Software (version 2.14.0; R Foundation for Statistical Computing, Vienna, Austria). 3. Results 3.1. Patient characteristics Median age at evaluation was 48.7 years (range: 18.9–77.3 years) for the 1740 study patients, who were predominantly female (71.1%). Most patients were Caucasian (90.8%) and median BMI was 44.9 (range: 35.0–111.9). 3.2. Normative values and internal consistency of the WEL-SF For the WEL-SF total score (range 0 to 80) the median was 56, the first quartile was 42, the third quartile was 69, the mean was 54.33, and the standard deviation was 18.46. The distribution of the WEL-SF total score is shown in Supplemental Fig. 1. Responses of each of the 8 individual items were spread across the 0–10 scale though it tended to be most concentrated at the higher scores. A summary of the individual WEL-SF items and WEL-SF total score for the overall cohort of 1740 patients is provided in Supplemental Table 1. Cronbach's alpha measuring internal consistency of the WEL-SF was 0.92 and this remained consistent when omitting each individual WEL-SF item one at a time, with Cronbach's alpha estimates ranging from 0.90 to 0.92 for these 8 different scenarios. Furthermore, each individual WEL-SF item was strongly correlated with the WEL-SF total score, with Pearson's r estimates ranging from 0.71 to 0.85. A one-factor structure for the WEL-SF was supported as the first eigenvalue of the correlation matrix of the 8 WEL-SF items was 5.08 (explaining 63.5% of the variance) while all remaining eigenvalues were less than 0.70 in magnitude.

117

3.3. Associations between demographic variables and WEL-SF total score Associations were observed with WEL-SF total score for age (P = 0.007) and BMI (P = 0.004), where the WEL-SF total score was higher for younger patients and heavier patients, although only the association with BMI was significant after multiple testing adjustment (P ≤ 0.0063 considered significant; see Table 1). There was no evidence of a difference in WEL-SF total score for gender (P = 0.48), race (P = 0.97), or clinic site (P = 0.30) (Table 1). These results remained consistent in multivariable analysis adjusting for age and BMI (data not shown).

3.4. Clinical measures and construct validity Lower WEL-SF total scores were significantly associated with higher rates of binge eating episodes (P b 0.0001), food addiction severity and dependence (P b 0.0001), night eating syndrome (P b 0.0001) depressive symptoms (P b 0.0001), and anxiety level (P b 0.0001). In contrast, higher WEL-SF total scores were associated with the single-item higher

Table 1 Associations of demographics and psychological variables with WEL-SF total score. Variable Age ≤40.00 40.01–50.00 50.01–60.00 N60.00 Gender Female Male Race Caucasian Non-Caucasian BMI 35.00–39.99 40.00–49.99 ≥50 Clinic site Mid-West South East History of smoking Never Former Current PHQ-9 None/Minimal (0–4) Mild (5–9) Moderate (10–14) Moderately severe or severe (15–27) GAD-7 Minimal (0–4) Mild (5–9) Moderate (10–14) Severe (≥15) AUDIT 0 1 2 3 ≥4 ACE history of sexual abuse No Yes

N

482 443 465 350

Mean change in WEL-SF total score (95% CI)

P-value

Overall test of difference: P = 0.007 0.00 (reference) N/A −1.84 (−4.21, 0.54) 0.13 −3.64 (−5.99, −1.29) 0.002 −3.68 (−6.21, −1.14) 0.005

1238 502

0.00 (reference) 0.70 (−1.22, 2.61)

1560 158

0.00 (reference) N/A 0.05 (−2.97, 3.08) 0.97 Overall test of difference: P = 0.004 0.00 (reference) N/A 0.64 (−1.59, 2.88) 0.57 3.66 (1.17, 6.16) 0.004

371 886 483 1359 381 743 488 129 686 481 217 114

N/A 0.48

0.00 (reference) N/A −1.11 (−3.21, 0.99) 0.30 Overall test of difference: P = 0.030 0.00 (reference) N/A −0.32 (−2.40, 1.77) 0.77 4.33 (0.92, 7.74) 0.013 Overall test of difference: P b 0.0001 0.00 (reference) N/A −6.95 (−9.03, −4.86) b0.0001 −10.75 (−13.47, −8.02) b0.0001 −14.67 (−18.22, −11.13) b0.0001

581 502 216 132 191

Overall test of difference: P b 0.0001 0.00 (reference) N/A −7.34 (−9.61, −5.06) b0.0001 −10.28 (−13.62, −6.94) b0.0001 −9.98 (−14.36, −5.59) b0.0001 Overall test of difference: P = 0.46 0.00 (reference) N/A −0.89 (−3.11, 1.32) 0.43 −1.61 (−4.51, 1.28) 0.28 −3.15 (−6.65, 0.36) 0.079 −0.73 (−3.76, 2.31) 0.64

799 173

0.00 (reference) −1.29 (−4.36, 1.78)

1105 314 126 70

N/A 0.41

Mean changes in WEL-SF total score, 95% CIs, and P-values result from linear regression models. P-values of 0.0063 or lower were considered statistically significant after adjusting for multiple testing (Holm, 1979). Missing information included race (N = 22), smoking history (N = 380), PHQ-9 (N = 242), GAD-7 (N = 125), AUDIT (N = 118), ACE history of sexual abuse (N = 768), CI = confidence interval; N = sample size; and P = P-value.

118

G.E. Ames et al. / Eating Behaviors 18 (2015) 115–119

exercise and weight management self-efficacy (P b 0.0001) and motivation to make lifestyle changes (P b 0.0001) shown in Table 2. There was no evidence of a difference in WEL-SF total score for alcohol use (P = 0.46) or history of being the victim of childhood sexual abuse (P = 0.41). Some degree of association with WEL-SF total score was evident for history of cigarette smoking (P = 0.030) and self-weighing (P = 0.025); however, these findings did not remain significant after accounting for multiple testing shown in Table 1. Results were consistent in multivariable analysis adjusting for age and BMI (data not shown).

Table 2 Associations between eating behavior and WEL-SF total score. Variable

QEWP-R binge eating disorder No Yes YFAS severity 0–1 2 3 4 5 6–7 YFAS dependence Diagnosis not met Diagnosis met Night eating intake after suppertime None Up to a quarter About half More than half or almost all Night eating in middle of night Never Sometimes About half the time, usually, or always Physical activity self-efficacy 0–2 3–5 6–8 9–10 Weight management self-efficacy 0–2 3–5 6–8 9–10 Motivation to make lifestyle changes 0–5 6–8 9–10 How often do you weigh yourself Never or once a year Every couple months Once a month Once a week Every couple days Every day or more than once per day

Single variable analysis N

Mean change in WEL-SF total score (95% CI)

1501 188

0.00 (reference) N/A −19.06 (−21.70, −16.40) b0.0001 Overall test of difference: P b 0.0001 0.00 (reference) N/A −9.65 (−11.67, −7.64) b0.0001 −13.02 (−15.32, −10.73) b0.0001 −16.93 (−19.75, −14.12) b0.0001 −21.53 (−24.48, −18.59) b0.0001 −29.83 (−32.90, −26.76) b0.0001

606 420 279 160 143 129

P-value

1417 292

0.00 (reference) N/A −16.82 (−19.01, −14.62) b0.0001 Overall test of difference: P b 0.0001

416 999 160 107

0.00 (reference) N/A −11.02 (−13.04, −9.00) b0.0001 −13.17 (−16.39, −9.95) b0.0001 −18.41 (−22.16, −14.66) b0.0001 Overall test of difference: P b 0.0001 0.00 (reference) N/A −6.49 (−8.66, −4.32) b0.0001 −10.78 (−13.86, −7.69) b0.0001

1215 346 149

277 392 579 463

Overall test of difference: P b 0.0001 0.00 (reference) N/A 2.19 (−0.57, 4.94) 0.12 6.60 (4.04, 9.16) b0.0001 12.61 (9.95, 15.28) b0.0001 Overall test of difference: P b 0.0001

180 423 640 458

0.00 (reference) N/A 7.80 (4.94, 10.66) b0.0001 17.14 (14.42, 19.85) b0.0001 26.31 (23.49, 29.14) b0.0001 Overall test of difference: P b 0.0001

131 425 1157

0.00 (reference) N/A 7.19 (3.77, 10.60) b0.0001 18.35 (15.20, 21.50) b0.0001 Overall test of difference: P = 0.025

283 248 294 417 265 194

0.00 (reference) 0.67 (−2.46, 3.80) 2.02 (−0.98, 5.01) 4.14 (1.37, 6.91) 0.09 (−2.99, 3.17) 1.12 (−2.23, 4.48)

N/A 0.68 0.19 0.004 0.95 0.51

Mean changes in WEL-SF total score, 95% CIs, and P-values result from linear regression models. Means represent the mean change in WEL-SF total score in comparison to the reference category. P-values b 0.0063 were considered as statistically significant after adjusting for multiple testing (Holm, 1979). Missing data included YFAS (N = 34), food intake after suppertime (N = 58), night snacking (N = 30), physical activity self-efficacy (N = 29), weight management self-efficacy (N = 39), motivation to make lifestyle changes (N = 27), and how often do you weigh yourself (N = 39). CI = confidence interval; N = sample size; P = P-value.

4. Discussion Consistent with the hypothesis of the primary study aim, the WEL-SF demonstrated strong internal consistency including very high Cronbach's alpha values with single item deletion. Recently the WELSF was translated into a Norwegian version (Flolo et al., 2014) using a small sample of obese patients who were selected to undergo bariatric surgery. The large sample in this study revealed nearly identical reliability coefficients to those found in the Norwegian translation and adaptation of the WEL-SF as well as a nearly identical eigenvalues consistent with a one factor structure. Furthermore, means and standard deviations for WEL-SF total score were comparable between patients seeking bariatric surgery in this sample (54.3 ± 18.4) and the Norwegian sample (53.5 ± 16.2). Taken together, these findings suggest that the WEL-SF is a psychometrically valid measure of eating self-efficacy. The second study aim was to examine clinical correlates and construct validity of the WEL-SF including eating behavior, psychopathology, and other single-item measures of self-efficacy for both eating and physical activity level. Problematic eating behaviors observed in this sample were binge eating episodes (13%), food addiction (21%), evening hyperphagia (16%), and eating in the middle of the night (9%). The relatively high rates of disordered eating behaviors prior to bariatric surgery were consistent with a previous investigation of eating behavior from LABS-2 (Mitchell, King, Courcoulas, et al., 2014b). As predicted, the WEL-SF total scores decreased significantly among patients who reported higher rates of binge eating episodes, food addiction severity and dependence, and night eating syndrome. Likewise, symptoms suggestive of moderate to severe depression level (22%) and anxiety (12%) were common in this sample and WEL-SF total scores significantly decreased with higher rates of mood problems. Moreover, greater levels of self-efficacy for physical activity, weight management, and motivation were all associated with greater total scores on the WEL-SF. The total score associations with disordered eating behavior, psychopathology, and self-efficacy were all in the predicted direction, which provides strong evidence of construct validity of the WEL-SF. The strengths of this study include a large sample size, a multisite design, and standardization of assessment measures with LABS-2 (Belle et al., 2013), which may enhance generalizability of the findings. Limitations of the study include lack of diversity in the sample and the use of self-report measures as opposed to structured diagnostic clinical interviews. Self-report measures typically result in inflated rates of disordered eating and mood problems in obese patients (Mitchell et al., 2012). Nevertheless, this study provides further evidence that the WEL-SF has strong statistical and clinical validity and is a useful brief measure of eating self-efficacy for pre-surgical assessment of obese patients. As demonstrated in this study, patients who report low eating self-efficacy prior to bariatric surgery are likely to exhibit higher levels of problem eating behaviors and lower mood. They may then be at greater risk for experiencing the return of these problems and weight regain after initial successful weight loss (Mitchell, King, Chen, et al., 2014a) (Colles, Dixon, & O'Brien, 2008). Thus, in terms of future directions, longitudinal investigations of changes in eating self-efficacy before bariatric surgery and after successful weight loss may improve understanding of what factors underlie behavioral recidivism and weight regain. Role of funding sources Funding for this study was provided by the Mayo Clinic (90085006) Comprehensive Research Team. The Research Team had no role in the study design, collection, analysis, interpretation of the data, writing the manuscript, or the decision to submit the paper for publication

Contributors Gretchen E. Ames and Michael G. Heckman designed the study and wrote the protocol. Nancy N. Diehl initiated data collection and management. Michael G. Heckman conducted statistical analysis and wrote the data analysis and results sections of the manuscript. Gretchen E. Ames wrote the first complete draft of the manuscript. Karen B.

G.E. Ames et al. / Eating Behaviors 18 (2015) 115–119 Grothe and Matthew M. Clark edited subsequent drafts of the manuscript. All authors have contributed to and have approved the final manuscript. Conflict of interest All authors declare that they have no conflicts of interest.

Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.eatbeh.2015.05.003. References American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington D.C.: Author. Ames, G. E., Heckman, M. G., Grothe, K. B., & Clark, M. M. (2012). Eating self-efficacy: Development of a short-form WEL. [Research support, non-U.S. gov't validation studies]. Eating Behaviors, 13(4), 375–378 http://dx.doi.org/10.1016/j.eatbeh.2012.03.013. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavior change. Psychological Review, 84(2), 191–215. Batsis, J. A., Clark, M. M., Grothe, K., Lopez-Jimenez, F., Collazo-Clavell, M. L., Somers, V. K., et al. (2009). Self-efficacy after bariatric surgery for obesity. A population-based cohort study. Appetite, 52(3), 637–645 http://dx.doi.org/10.1016/j.appet.2009.02.017. Belle, S. H., Berk, P. D., Chapman, W. H., Christian, N. J., Courcoulas, A. P., Dakin, G. F., et al. (2013). Baseline characteristics of participants in the Longitudinal Assessment of Bariatric Surgery-2 (LABS-2) study. [Evaluation Studies Research Support, N.I.H., extramural]. Surgery for Obesity and Related Diseases: Official Journal of the American Society for Bariatric Surgery, 9(6), 926–935 http://dx.doi.org/10.1016/j.soard.2013. 01.023. Clark, M. M., Abrams, D. B., Niaura, R. S., Eaton, C. A., & Rossi, J. S. (1991). Self-efficacy in weight management. Journal of Consulting and Clinical Psychology, 59(5), 739–744. Clark, M. M., Cargill, B. R., Medeiros, M. L., & Pera, V. (1996). Changes in self-efficacy following obesity treatment. Obesity Research, 4(2), 179–181. Clark, M. M., Vickers, K. S., Hathaway, J. C., Smith, M., Looker, S. A., Petersen, L. R., et al. (2007). Physical activity in patients with advanced-stage cancer actively receiving chemotherapy. The Journal of Supportive Oncology, 5(10), 487–493. Colles, S. L., Dixon, J. B., & O'Brien, P. E. (2008). Grazing and loss of control related to eating: Two high-risk factors following bariatric surgery. [Clinical trial research support, non-U.S. gov't]. Obesity, 16(3), 615–622 http://dx.doi.org/10.1038/oby.2007.101. DePue, J. D., Clark, M. M., Ruggiero, L., Medeiros, M. L., & Pera, V., Jr. (1995). Maintenance of weight loss: A needs assessment. Obesity Research, 3(3), 241–248. Dube, S. R., Felitti, V. J., Dong, M., Chapman, D. P., Giles, W. H., & Anda, R. F. (2003). Childhood abuse, neglect, and household dysfunction and the risk of illicit drug use: The adverse childhood experiences study. [Comparative study]. Pediatrics, 111(3), 564–572.

119

Flolo, T. N., Andersen, J. R., Nielsen, H. J., & Natvig, G. K. (2014). Translation, adaptation, validation and performance of the American Weight Efficacy Lifestyle Questionnaire Short Form (WEL-SF) to a Norwegian version: A cross-sectional study. PeerJ, 2, e565 http://dx.doi.org/10.7717/peerj.565. Gearhardt, A. N., Corbin, W. R., & Brownell, K. D. (2009). Preliminary validation of the Yale Food Addiction Scale. Appetite, 52(2), 430–436 http://dx.doi.org/10.1016/j.appet. 2008.12.003. Holm, S. (1979). A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics, 6, 65–70. Kroenke, K., Spitzer, R. L., & Williams, J. B. (2001). The PHQ-9: Validity of a brief depression severity measure. [Research support, non-U.S. gov't validation studies]. Journal of General Internal Medicine, 16(9), 606–613. MacLean, P. S., Wing, R. R., Davidson, T., Epstein, L., Goodpaster, B., Hall, K. D., et al. (2014). NIH working group report: Innovative research to improve maintenance of weight loss. Obesity http://dx.doi.org/10.1002/oby.20967. Meule, A., Heckel, D., & Kubler, A. (2012). Factor structure and item analysis of the Yale Food Addiction Scale in obese candidates for bariatric surgery. [Research support, non-U.S. gov't]. European Eating Disorders Review: The Journal of the Eating Disorders Association, 20(5), 419–422 http://dx.doi.org/10.1002/erv.2189. Mitchell, J. E., King, W. C., Chen, J. Y., Devlin, M. J., Flum, D., Garcia, L., et al. (2014a). Course of depressive symptoms and treatment in the longitudinal assessment of bariatric surgery (LABS-2) study. [Research support, N.I.H., extramural]. Obesity, 22(8), 1799–1806 http://dx.doi.org/10.1002/oby.20738. Mitchell, J. E., King, W. C., Courcoulas, A., Dakin, G., Elder, K., Engel, S., et al. (2014b). Eating behavior and eating disorders in adults before bariatric surgery. The International Journal of Eating Disorders http://dx.doi.org/10.1002/eat.22275. Mitchell, J. E., Selzer, F., Kalarchian, M. A., Devlin, M. J., Strain, G. W., Elder, K. A., et al. (2012). Psychopathology before surgery in the longitudinal assessment of bariatric surgery-3 (LABS-3) psychosocial study. [Research support, N.I.H., extramural]. Surgery for Obesity and Related Diseases: Official Journal of the American Society for Bariatric Surgery, 8(5), 533–541 http://dx.doi.org/10.1016/j.soard.2012.07.001. Ochner, C. N., Tsai, A. G., Kushner, R. F., & Wadden, T. A. (2015). Treating obesity seriously: When recommendations for lifestyle change confront biological adaptations. The Lancet. Diabetes & Endocrinology. http://dx.doi.org/10.1016/S2213-8587(15)00009-1. Perri, M. G. (1998). The maintenance of treatment effects in the long-term management of obesity. Clinical Psychology: Science and Practice, 5, 526–543. Saunders, J. B., Aasland, O. G., Babor, T. F., de la Fuente, J. R., & Grant, M. (1993). Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption–II. [Comparative study]. Addiction, 88(6), 791–804. Spitzer, R. L., Kroenke, K., Williams, J. B., & Lowe, B. (2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. [Comparative study multicenter study research support, non-U.S. gov't]. Archives of Internal Medicine, 166(10), 1092–1097 http://dx.doi.org/10.1001/archinte.166.10.1092. Spitzer, R. L., Yanovski, S., & Marcus, M. D. (1993). The Questionnaire of Eating and Weight Patterns-Revised (QEWP-R).