Surgery for Obesity and Related Diseases 11 (2015) 637–642
Integrated health original article
Correlation between the Beck Depression Inventory and bariatric surgical procedures Subhashini Ayloo, M.D.a,*, Kara Thompson, M.D.b, Nabajit Choudhury, M.D.a, Raiyah Sheriffdeen, M.D.a a
Division of General Minimally Invasive and Robotic Surgery, Department of Surgery, University of Illinois at Chicago, Chicago, Illinois b Department of Medicine, Research Office, Dalhousie University, Halifax, Nova Scotia Received August 4, 2014; accepted November 10, 2014
Abstract
Background: The Beck Depression Inventory (BDI) is a psychosocial screen for depression in obese patients seeking bariatric surgery. Gastric bypass improves postsurgical BDI scores due to weight loss, which predicts future weight loss. The effect of different bariatric procedures with differences in weight loss on BDI scores is unknown. Objective: To evaluate the relationship between different bariatric procedures and changes in the BDI scores, adjusting for the initial BDI score, and to consider the impact of psychosocial variables. The secondary objective was to assess the relationship between changes in BDI scores and weight loss at 6 to 12 months. Setting: University Hospital, United States. Methods: Bariatric surgical patients were prospectively enrolled and retrospectively reviewed. We assessed changes in BDI after adjusting for the presurgical BDI and analyzed the relationship between patient demographic characteristics/psychological disorders and changes in BDI. Results: We enrolled 137 patients who underwent a gastric band procedure, sleeve gastrectomy, or gastric bypass. We found a significant decrease in BMI and BDI scores across the full sample. Unlike BDI, change in BMI varied with procedure. Normalizing for baseline BDI, change in BDI did not significantly correlate with change in BMI. Patients who were employed and those without psychiatric history experienced even greater improvement in BDI scores. No statistically significant correlation was found between the change in BDI and weight loss at 6–12 months. Conclusions: BDI scores were independent of the type of bariatric procedure and the amount of weight loss. Advantageous psychosocial parameters were associated with greater improvement in BDI scores. (Surg Obes Relat Dis 2015;11:637–642.) r 2015 American Society for Metabolic and Bariatric Surgery. All rights reserved.
Keywords:
Beck Depression Inventory; Psychosocial; Weight loss; Depression; Bariatric surgery
Depression is a highly prevalent co-morbidity occurring in patients with severe obesity [1]. Several studies have reported an association between obesity and mild to moderate depressive symptoms compared with nondepressed controls [2–4]. Depression in bariatric patients is of specific interest, * Correspondence: Subhashini Ayloo, MD FACS, Division of General Minimally Invasive and Robotic Surgery, 840 S. Wood St., Mail Code 958, Suite 435 E, Chicago, IL 60612. E-mail: [email protected]
because the presence of depressive symptoms before surgery has been associated with poorer outcomes after surgery, including regaining of lost weight [5,6]. A prerequisite for bariatric surgery includes a presurgical evaluation of the patient’s psychosocial status, most often by a mental health professional. Most programs use symptom inventories such as the Beck Depression Inventory (BDI) as the most common assessment tool [7]. These assessments also serve as an indicator for further thorough psychological consultation and for comparison of mood
http://dx.doi.org/10.1016/j.soard.2014.11.005 1550-7289/r 2015 American Society for Metabolic and Bariatric Surgery. All rights reserved.
638
S. Ayloo et al. / Surgery for Obesity and Related Diseases 11 (2015) 637–642
changes in the postsurgical period, including either an improvement in or an increased likelihood of feelings of depression. Previous studies have shown a trend toward improvement in symptoms of depression after bariatric surgery that may be correlated with weight loss [8–10]. The objective of this study was to analyze the effect of bariatric surgery, patient demographic characteristics, and clinical characteristics on postoperative BDI scores in urban, severely obese patients who underwent bariatric surgery at a tertiary center. Patients were further stratified by the type of bariatric surgery, including gastric banding, sleeve gastrectomy, and Rouxen-Y gastric bypass, to determine if a correlation was present between the type of procedure and weight loss and the BDI scores. Methods BDI is a multiple-choice questionnaire composed of 21 items. The questions assess symptoms of irritability, feelings of guilt, suicidal ideation, fatigue, and weight loss. Scores of 0–9 are considered to indicate minimal depression, 10–19 mild depression, 19–29 moderate depression, and Z30 severe depression. Sample and procedures A total of 318 patients with available presurgical BDI scores who underwent bariatric surgery between January 2006 and December 2012 were retrospectively reviewed for inclusion in the study. All surgeries were performed by a single surgeon at a university-based tertiary care center that offered Roux-en-Y gastric bypass, sleeve gastrectomy, and gastric banding. Only patients who completed postsurgical follow-up and who completed a postsurgical BDI were included in the study. Of the 318 patients, 137 met these requirements. After Institutional Review Board approval, the charts for the 137 patients were reviewed for age, gender, height, weight, presurgical body mass index (BMI), BDI scores, and postsurgical BMI and BDI scores. Presurgical or baseline data was collected about 4.6 months before surgery. About one half of the sample had their psychiatric assessment performed between 3–6 months before surgery. Postsurgical collection of data are classified into 2 groups: 1 week to 6 months follow-up after surgery as “up to 6 months,” and 46 months until the end of study period (i.e., 12 months from the bariatric procedure) as “6–12 months,” with an assumption that significant weight changes from bariatric procedures usually take about 6 months. The timeline for the collection of data was categorized into 0– 3 months, 3–6 months and 6–12 months. Ideal measure for BMI in up to 6 month group was taken between 3–6 months postsurgery. If that measure was not available, the 0–3 months measure was used. For nearly three quarters of the
sample, the BMI was measured from 3–6 months. Additionally, the type and number of co-morbidities and the psychiatric history, including a history of psychotropic drug use, were noted. Finally, charts were reviewed for social factors including the level of education completed, selfreported perception of family support, marital status, history of physical, mental, or sexual abuse, and employment status. Patients were then stratified by the type of surgery, which included Roux-en-Y gastric bypass, sleeve gastrectomy, and adjustable gastric banding. Statistical analysis Summary and descriptive statistics were generated for each patient’s baseline clinical characteristics. The Wald test and analysis of variance (ANOVA) were used to assess baseline imbalances between patients in the 3 surgical groups. The primary outcome, the change in BDI score from pre- to postsurgery, was analyzed using the paired t test. Analysis of covariance (ANCOVA) was used to assess the absolute change in BDI among the surgical groups after adjusting for the baseline BDI score. The least square mean differences with standard errors (SE) are reported. The relationship between marital status, employment, history of abuse, and psychiatric diagnosis and the change in the BDI score was investigated with exploratory analysis using ANCOVA. Sensitivity analysis was conducted using nonparametric rank order ANCOVA models. The Pearson correlation coefficient was calculated to assess the relationship between the change in BDI score and the change in BMI up to 6 months. Sensitivity analysis was performed carrying forward the BMI measurement up to 6 months for patients with missing data and the BMI measurement at 6– 12 months. A 2-sided P value of less than .05 was considered statistically significant. Statistical analysis was performed with SAS statistical software version 9.3 (SAS Institute, Cary, NC). Results Patient demographic characteristics The patient population in this study included 115 women and 22 men with a mean (⫾ SD) age of 41.7 ⫾ 10.4 years, height of 165.7 ⫾ 10.3 cm, weight of 132.0 ⫾ 24.8 kg, presurgical BMI of 47.7 ⫾ 7.7 kg/m2, and postsurgical BMI of 42.2 ⫾ 7.7 kg/m2. Seventy-eight (57%) patients were African American, 44 (32%) patients were married, 64 (47%) patients were employed, and 55 (40%) patients had completed high school or the equivalent at the time of surgery. Patient demographic characteristics were stratified by the type of bariatric surgery and are presented in Table 1. Of the 137 patients, 58 patients underwent gastric banding, 51 underwent sleeve gastrectomy, and 28 underwent gastric bypass. The Wald test and ANOVA were used to assess the
Depression and Different Bariatric Procedures / Surgery for Obesity and Related Diseases 11 (2015) 637–642
639
Table 1 Patient demographic characteristics Variables
Gender (female) (%) Age (yr) Height (cm) Weight (kg) Race (%) African-American Caucasian Others (Hispanic, Asian) Marital status (%) Married Single, divorced, or widowed Education (%) Less than high school High school or equivalent College or higher Employed (%) Perception of positive family support History of physical, sexual, or mental abuse History of psychiatric diagnosis (%) History of psychiatric medications (%) Medical co-morbidities (%) Hypertension Type 2 diabetes mellitus Obstructive sleep apnea Asthma Dyslipidemia Gastroesophageal reflux disease Hypothyroidism Average number of co-morbidities Body mass index (kg/m2) Presurgery Beck Depression Inventory score (mean) Presurgery
All surgeries
Gastric band
Sleeve gastrectomy
Gastric bypass
n ¼ 137
n ¼ 58
n ¼ 51
n ¼ 28
83.9 (n ¼ 115) 41.65 (⫾10.3) 165.73 (þ10.3) 131.97 (þ24.8)
96 (n ¼ 49) 44.0 (⫾ 11.1) 166.6 (⫾ 9.4) 132.3 (⫾ 28.1)
82 (n ¼ 42) 39.6 (⫾ 9.4) 166.8 (⫾ 9.2) 138.3 (⫾21.3)
86 (n ¼ 24) 40.7 (⫾ 10.0) 161.9 (⫾ 13.3) 119.7 (⫾ 19.0)
57 (n ¼ 78) 25 (n ¼ 34) 18 (n ¼ 25)
66 (n ¼ 38) 21 (n ¼ 12) 13.7 (n ¼ 8)
47 (n ¼ 24) 35 (n ¼ 18) 17.6 (n ¼ 9)
57 (n ¼ 16) 14 (n ¼ 4) 29 (n ¼ 8)
NS
32 (n ¼ 44) 68 (n ¼ 93)
31 (n ¼ 18) 69 (n ¼ 40)
33 (n ¼ 17) 67 (n ¼ 34)
32 (n ¼ 9) 68 (n ¼ 19)
.96
NS
(n (n (n (n (n (n (n (n
¼ ¼ ¼ ¼ ¼ ¼ ¼ ¼
18) 55) 63) 64) 127) 24) 50) 28)
14 45 40 48 93 22 34 21
51 (n 33 (n 34 (n 28 (n 27 (n 20 (n 11 (n 2.7
¼ ¼ ¼ ¼ ¼ ¼ ¼
70) 45) 47) 39) 37) 28) 15)
¼ ¼ ¼ ¼ ¼ ¼ ¼ ¼
P
.91 .09 .004
(n (n (n (n (n (n (n (n
¼ ¼ ¼ ¼ ¼ ¼ ¼ ¼
7) 20) 24) 25) 47) 9) 20) 11)
11 (n ¼ 3) 32 (n ¼ 9) 57 (n ¼ 16) 39 (n ¼ 11) 93 (n ¼ 26) 7 (n ¼ 2) 36 (n ¼ 10) 18 (n ¼ 5)
.67 .98 .24 .87 .92
57 (n ¼ 33) 33 (n ¼ 19) 34 (n ¼ 20) 28 (n ¼ 16) 22 (n ¼ 13) 19 (n ¼ 11) 5 (n ¼ 3) 2.8
41 (n 35 (n 41 (n 33 (n 25 (n 20 (n 20 (n 2.6
¼ ¼ ¼ ¼ ¼ ¼ ¼
21) 18) 21) 17) 13) 10) 10)
57 (n ¼ 16) 29 (n ¼ 8) 21 (n ¼ 6) 21 (n ¼ 6) 39 (n ¼ 11) 25 (n ¼ 7) 7 (n ¼ 2) 2.7
.20 .83 .22 .53 .27 .80 .06 .96
47.7 (⫾7.7)
47.3 (⫾ 7.8)
49.9 (⫾ 7.8)
44.5 (⫾ 6.0)
.017
11.3 (⫾ 8.8)
9.9 (⫾ 7.8)
11.5 (⫾ 8.9)
13.7 (⫾ 1.1)
.25
13 40 46 47 93 18 36 20
association between surgical procedures and demographic variables including age, gender, race, marital status, education, employment, co-morbidities, family support, history of physical/sexual/mental abuse, history of psychiatric disorder, or psychiatric medications. No statistically significant differences were found among the groups. Medical and psychiatric variables Table 1 presents data on co-morbidities present at the time of surgery. Although patients had a wide range of medical histories, the mean number of co-morbidities did not vary significantly among the different surgical groups (2.7 ⫾ 1.9; P ¼ .9636). The most common affective disorder for all patients in this data set was depression (n ¼ 31, 23%), followed by anxiety (n ¼ 11, 8%) which was clinically diagnosed by psychologist dedicated in assessing bariatric surgical patients. The majority of patients, however, did not have clinical psychiatric diagnoses. Patients who underwent gastric banding, sleeve gastrectomy and gastric bypass were
(n (n (n (n (n (n (n (n
8) 26) 23) 28) 54) 13) 20) 12)
14 39 47 49 92 18 39 22
each prescribed a mean of .4, .4, and .3 psychotropic medications, respectively. A mean of .4 psychotropic medications was prescribed for all patients combined. We found a statistically significant difference in the baseline mean preoperative BMI among the different surgical procedures (P ¼ .017) with the highest preoperative BMI in the sleeve gastrectomy group (mean 49.9 ⫾ 7.8, SD). No significant difference was noted in the mean preoperative BDI scores among the different groups (P ¼ .1693). At 6 months postsurgery, we observed significant (P o .0001) improvement in BDI scores from 11.3 ⫾ 8.8 (SD) to 7.2 ⫾ 8.3 (SD) with a mean difference of 4.11 ⫾ 11.69 (SD) and a significant drop in BMI from 47.7 ⫾7.7 to 42.2 ⫾ 7.73 with a mean difference of 5.60 ⫾ 3.80 (SD) (P o .0001). ANCOVA was performed to look at the relationship between different surgical procedures and the change in BDI after adjusting for preoperative BDI (Table 2). Gastric bypass produced the largest decrease in the least squares (LS) adjusted mean BDI change at 5.8 ⫾ 1.58 (SE), followed by sleeve
640
Outcome
Gastric band Baseline
Sleeve gastrectomy Up to 6 mo
Mean (SD) Number of patients Weight (kg) Least square difference (95% versus sleeve gastrectomy Least square difference (95% versus lap band BMI Least square difference (95% versus sleeve gastrectomy Least square difference (95% versus lap band BDI Least square difference (95% versus sleeve gastrectomy Least square difference (95% versus lap band
58 132.3 (28.0) CI) CI)
Change from baseline *Least square mean (SE)
57 57 122.9 (26.9) -8.8 (1.2) 10.22 (6.51, 13.92)
Up to 6 mo Mean (SD)
51 138.3 (21.3)
49 119.5 (21.3) -
Change from baseline *Least square mean (SE) 49 -19.0 (1.3)
Baseline
47.3 (7.8)
28 119.7 (18.9)
Change from baseline *Least square mean (SE)
28 28 97.1 (21.2) -23.5 (1.8) -4.49 (-9.15, .17) -14.71 (-19.14, -10.28)
44.2 (7.8) -3.2 (.4) 3.70 (2.51, 4.88)
CI)
49.8 (7.8)
42.9 (7.2) -
-6.9 (.4)
44.5 (6.0)
9.9 (7.8)
Up to 6 mo
Mean (SD)
-
CI)
CI)
Baseline
Gastric bypass
8.59 (9.9) 2.33(-.83, 5.49)
CI)
BMI ¼ body mass index; BDI ¼ Beck Depression Inventory.
-2.5 (1.0)
-4.98 (-6.38, -3.58) 11.4 (8.8)
-
36.7 (5.6) -8.1 (.5) -1.28 (-2.76, .19)
6.3 (7.3) -
-4.9 (1.1)
13.7 (10.1)
5.6 (5.9) -5.8 (1.5) -0.90(-4.79, 2.97) -3.24(-7.07, .59)
S. Ayloo et al. / Surgery for Obesity and Related Diseases 11 (2015) 637–642
Table 2 Changes in Beck Depression Inventory from baseline among different surgical procedures
Depression and Different Bariatric Procedures / Surgery for Obesity and Related Diseases 11 (2015) 637–642
gastrectomy at 4.91 ⫾ 1.16 (SE) and gastric banding at 2.57 ⫾ 1.09 (SE). We found no statistically significant difference in BDI change among surgical procedures when adjusting for presurgical BDI scores (P ¼ .1739). These results were similar when the model included the change in BMI (P ¼ .4061). There was a statistically significant difference for change in weight between surgical procedures after adjusting for preoperative weight (P o .0001). Gastric bypass had a LS mean weight loss difference of -14.71, 95% CI (-19.14, -10.28) greater compared to gastric band and a LS mean weight loss difference of -4.49 (95% CI -9.15, .17) greater compared to sleeve gastrectomy. Gastric band had 10.22 (95% CI 6.51, 13.92) less weight loss compared to sleeve gastrectomy. As well there was a statistically significant difference for change in BMI between surgical procedures after adjusting for preoperative BMI (P o .0001). Results for BMI are shown in Table 2. Parametric ANCOVA analyses were performed to evaluate potentially influential demographic variables for association with the change in BDI, adjusting for initial BDI. Employment and psychiatric history were statistically significant. Results were consistent with those from the nonparametric analysis. Patients who were employed had a greater adjusted LS mean decrease in BDI of 5.77 ⫾ 1.03 (SE) compared to those not employed, with a mean decrease change in BDI of 2.64 (SE, .96; P ¼ .0284). Patients with no psychiatric diagnosis had a greater decrease in adjusted LS mean BDI change 5.50 (SE, .88) compared to those with a psychiatric diagnosis who had a decrease in BDI change of 1.68 (SE, 1.16; P ¼ .0101) after adjusting for initial BDI score. The correlation between BDI and BMI was examined with the Pearson correlation coefficient considering the difference between baseline and 6 to 12 months after surgery (n ¼ 94). No statistically significant association was identified between the BDI change and the BMI change (P ¼ .1327). A sensitivity analysis was then performed in which the value up to 6 months after surgery was carried forward to include the entire sample. Results were consistent with the original analysis (P ¼ .8790) and were also consistent with those from nonparametric analysis. Discussion According to the American Society for Metabolic and Bariatric Surgery, 179,000 bariatric procedures were performed in 2013 [11]. The most commonly performed bariatric operations are Roux-en-Y gastric bypass (34%), sleeve gastrectomy (42%), and adjustable gastric banding (14%). Although bariatric surgeries improve medical comorbidities, many patients rely on these surgeries to produce the behavioral changes needed to maintain weight loss [12]. On the contrary, Odom et al. reported that certain behavioral factors such as increased food urges, concern
641
over addictive behaviors, and a decreased sense of wellbeing are strong predictors of weight regain after bariatric surgery [13]. Because of the significant interplay between depression and obesity [14], presurgical psychological evaluation of patients planning to undergo bariatric surgery is an appropriate prerequisite. Our objective was to analyze the effect of weight loss after different bariatric procedures on the severity of depression as measured with BDI scores. Furthermore, to evaluate the effect of BDI change on 6–12-month weight loss and the effect of psychosocial status on BDI scores. Of the 318 patients enrolled in the study only 137 patients met the inclusion criteria. This is partly due to mandatory completion of preoperative BDI form for psychological evaluation and clearance, an important component for surgical consideration. While the follow-up completion of the BDI was on volunteer basis, there was a significant drop off of patients filling out the BDI form despite follow-up. Exceptions being a small percentage of patients seeking revisional surgery or with significant weight regain were strongly recommended to be followed up with psychologist for behavioral modification and were monitored closely with BDI forms. One of the limitations of the study is a lack of comparison of outcomes with patients whose BDI scores were lacking or with a group that did not get follow-up, which could potentially effect key conclusions of this study. Another limitation of the study is globalization of the time periods. Usual protocol at our center for follow-up includes one week follow-up after discharge, followed by 1 month and every 3 months during the first year. Patients had annual follow-ups, subsequent to their first year for sleeve gastrectomy and Roux-en-Y gastric bypass. Patients with adjustable gastric banding were followed up during the time of band adjustment. Despite this regiment there is overlap of patient appointments as such the study was divided into presurgical or baseline outcomes measured on average 4.6 months before surgery, postsurgical divided into 2 groups of 0–6 months as “up to 6 months” and as “6–12 months,” and for followup beyond 6 months to the end of the study period which was 1 year from the time of bariatric surgery. The prevalence of depression in adults in the United States between 2006 and 2008 was about 9.1%. Depression was more common among individuals with chronic diseases (obesity, cardiovascular disease, type 2 diabetes) and among individuals with unhealthy behaviors (smoking, physical inactivity). Depression is more prevalent in women, those previously married (divorced, widowed, or separated), those with education level less than equivalent to high school, and those unable to work or unemployed individuals [15]. Models to evaluate the association between important demographic characteristics and psychosocial variables such as gender, marital status, being employed, history of abuse, and previous psychological history suggested that
642
S. Ayloo et al. / Surgery for Obesity and Related Diseases 11 (2015) 637–642
individuals who were employed and who had no previous psychological history had a statistically significant improvement in BDI scores, reflecting an improvement of depressive symptoms. This observation suggests that these psychosocial factors may play a critical role in the bariatric population for the overall postoperative well-being of the patient. In addition, evaluation of the effect of the change in BDI after more than 6 months and BMI indicated no statistical correlation. One of the limitations of this study has been to group all the affective disorders together without differentiating depressive disorders from anxiety disorders that might potentially have an effect on the change of BDI and BMI In reviewing our series of patients, 84% of patients were women. Patient demographic characteristics including age, gender, race, marital status, education, and employment status; psychosocial characteristics of family support, history of psychiatric disorder, treatment with psychiatric medications, and history of physical, sexual, or mental abuse; and medical co-morbidities were not significantly different among the different surgical groups, thus minimizing the effect of these variables on the BDI scores. The preoperative BDI scores for different procedures were not significantly different, although the preoperative BMI was significantly different. This was expected because the gastric sleeve procedure was originally intended for patients with a higher BMI. When assessing the effect of weight on BDI, the absolute change was considered to minimize the difference in the preoperative BMI among the procedures [16,17,18]. Conclusions Bariatric surgery leads to significant improvement in weight loss and BDI scores. Weight loss was greatest after gastric bypass, followed by sleeve gastrectomy and adjustable gastric banding. The improvement in BDI scores was independent of the magnitude of weight loss, which was true for each of the procedures analyzed. Patients who were employed with no prior psychological history experienced a significant improvement in BDI scores. No correlation was found in improvement in BDI up to 6 months to weight loss at 6–12 months. We believe that the change in mood affected by bariatric surgery is due to the interplay of multiple factors including initial weight loss. The data suggest that psychosocial factors may play a more substantial role in the change in mood than was previously recognized. Disclosures The authors have no funding information to disclose; all authors declare no conflict of interest.
References [1] Roberts RE, Kaplan GA, Shema SJ, Strawbridge WJ. Are the obese at greater risk for depression? Am J Epidemiol 2000;152(2):163–70 [2] Luppino FS, de Wit LM, Bouvy PF, et al. Overweight, obesity, and depression: a systematic review and meta-analysis of longitudinal studies. Arch Gen Psychiatry 2010;67:220–9. [3] Muller A, Mitchell JE, Sondag C, de Zwaan M. Psychiatric aspects of bariatric surgery. Curr Psychiatry Rep 2013;15:397. [4] Marcus MD, Kalarchian MA, Courcoulas AP. Psychiatric evaluation and follow-up of bariatric surgery patients. Am J Psychiatry 2009;166:285–91. [5] Kalarchian MA, Marcus MD, Levine MD, et al. Psychiatric disorders among bariatric surgery candidates: relationship to obesity and functional health status. Am J Psychiatry 2007;164:328–34. (quiz 74). [6] Kalarchian MA, Marcus MD, Levine MD, Soulakova JN, Courcoulas AP, Wisinski MS. Relationship of psychiatric disorders to 6-month outcomes after gastric bypass. Surg Obes Relat Dis 2008;4:544–9. [7] Bauchowitz AU, Gonder-Frederick LA, Olbrisch ME, et al. Psychosocial evaluation of bariatric surgery candidates: a survey of present practices. Psychosom Med 2005;67:825–32. [8] Hayden MJ, Dixon JB, Dixon ME, Shea TL, O'Brien PE. Characterization of the improvement in depressive symptoms following bariatric surgery. Obes Surg 2011;21:328–35. [9] Dixon JB, Dixon ME, O'Brien PE. Depression in association with severe obesity: changes with weight loss. Arch Intern Med 2003;163: 2058–65. [10] Sanchez-Santos R, Del Barrio MJ, Gonzalez C, et al. Long-term health-related quality of life following gastric bypass: influence of depression. Obes Surg 2006;16:580–5. [11] ASMBS. Estimate of bariatric surgery numbers 2014. [page on the internet]. Gainseville: American Society for Metabolica and Bariatric Surgery; [cited November 4, 2014]. Available from: http://asmbs.org/ 2014/03/estimate-of-bariatric-surgery-numbers/. [12] Thonney B, Pataky Z, Badel S, Bobbioni-Harsch E, Golay A. The relationship between weight loss and psychosocial functioning among bariatric surgery patients. Am J Surg 2010;199:183–8. [13] Odom J, Zalesin KC, Washington TL, et al. Behavioral predictors of weight regain after bariatric surgery. Obes Surg. 2010;20:349–56. [14] de Zwaan M, Enderle J, Wagner S, et al. Anxiety and depression in bariatric surgery patients: a prospective, follow-up study using structured clinical interviews. J Affect Disord 2011;133(1–2):61–8. [15] Centers for Disease Control and Prevention. Current depression among adults—United States, 2006 and 2008. Morbid Mortal Weekly Rep [page on the internet]. [cited October 1, 2010]. Available from: http://www.cdc.gov/features/dsdepression/revised_table_estimates_ for_depression_mmwr_erratum_feb-2011.pdf. [16] Averbukh Y, Heshka S, El-Shoreya H, et al. Depression score predicts weight loss following Roux-en-Y gastric bypass. Obes Surg 2003;13:833–6. [17] Schowalter M, Benecke A, Lager C, et al. Changes in depression following gastric banding: a 5- to 7-year prospective study. Obes Surg 2008;18:314–20. [18] Munoz DJ, Chen E, Fischer S, et al. Considerations for the use of the Beck Depression Inventory in the assessment of weight-loss surgery seeking patients. Obes Surg 2007;17:1097–101.
Integrated health original article
Correlation between the Beck Depression Inventory and bariatric surgical procedures Subhashini Ayloo, M.D.a,*, Kara Thompson, M.D.b, Nabajit Choudhury, M.D.a, Raiyah Sheriffdeen, M.D.a a
Division of General Minimally Invasive and Robotic Surgery, Department of Surgery, University of Illinois at Chicago, Chicago, Illinois b Department of Medicine, Research Office, Dalhousie University, Halifax, Nova Scotia Received August 4, 2014; accepted November 10, 2014
Abstract
Background: The Beck Depression Inventory (BDI) is a psychosocial screen for depression in obese patients seeking bariatric surgery. Gastric bypass improves postsurgical BDI scores due to weight loss, which predicts future weight loss. The effect of different bariatric procedures with differences in weight loss on BDI scores is unknown. Objective: To evaluate the relationship between different bariatric procedures and changes in the BDI scores, adjusting for the initial BDI score, and to consider the impact of psychosocial variables. The secondary objective was to assess the relationship between changes in BDI scores and weight loss at 6 to 12 months. Setting: University Hospital, United States. Methods: Bariatric surgical patients were prospectively enrolled and retrospectively reviewed. We assessed changes in BDI after adjusting for the presurgical BDI and analyzed the relationship between patient demographic characteristics/psychological disorders and changes in BDI. Results: We enrolled 137 patients who underwent a gastric band procedure, sleeve gastrectomy, or gastric bypass. We found a significant decrease in BMI and BDI scores across the full sample. Unlike BDI, change in BMI varied with procedure. Normalizing for baseline BDI, change in BDI did not significantly correlate with change in BMI. Patients who were employed and those without psychiatric history experienced even greater improvement in BDI scores. No statistically significant correlation was found between the change in BDI and weight loss at 6–12 months. Conclusions: BDI scores were independent of the type of bariatric procedure and the amount of weight loss. Advantageous psychosocial parameters were associated with greater improvement in BDI scores. (Surg Obes Relat Dis 2015;11:637–642.) r 2015 American Society for Metabolic and Bariatric Surgery. All rights reserved.
Keywords:
Beck Depression Inventory; Psychosocial; Weight loss; Depression; Bariatric surgery
Depression is a highly prevalent co-morbidity occurring in patients with severe obesity [1]. Several studies have reported an association between obesity and mild to moderate depressive symptoms compared with nondepressed controls [2–4]. Depression in bariatric patients is of specific interest, * Correspondence: Subhashini Ayloo, MD FACS, Division of General Minimally Invasive and Robotic Surgery, 840 S. Wood St., Mail Code 958, Suite 435 E, Chicago, IL 60612. E-mail: [email protected]
because the presence of depressive symptoms before surgery has been associated with poorer outcomes after surgery, including regaining of lost weight [5,6]. A prerequisite for bariatric surgery includes a presurgical evaluation of the patient’s psychosocial status, most often by a mental health professional. Most programs use symptom inventories such as the Beck Depression Inventory (BDI) as the most common assessment tool [7]. These assessments also serve as an indicator for further thorough psychological consultation and for comparison of mood
http://dx.doi.org/10.1016/j.soard.2014.11.005 1550-7289/r 2015 American Society for Metabolic and Bariatric Surgery. All rights reserved.
638
S. Ayloo et al. / Surgery for Obesity and Related Diseases 11 (2015) 637–642
changes in the postsurgical period, including either an improvement in or an increased likelihood of feelings of depression. Previous studies have shown a trend toward improvement in symptoms of depression after bariatric surgery that may be correlated with weight loss [8–10]. The objective of this study was to analyze the effect of bariatric surgery, patient demographic characteristics, and clinical characteristics on postoperative BDI scores in urban, severely obese patients who underwent bariatric surgery at a tertiary center. Patients were further stratified by the type of bariatric surgery, including gastric banding, sleeve gastrectomy, and Rouxen-Y gastric bypass, to determine if a correlation was present between the type of procedure and weight loss and the BDI scores. Methods BDI is a multiple-choice questionnaire composed of 21 items. The questions assess symptoms of irritability, feelings of guilt, suicidal ideation, fatigue, and weight loss. Scores of 0–9 are considered to indicate minimal depression, 10–19 mild depression, 19–29 moderate depression, and Z30 severe depression. Sample and procedures A total of 318 patients with available presurgical BDI scores who underwent bariatric surgery between January 2006 and December 2012 were retrospectively reviewed for inclusion in the study. All surgeries were performed by a single surgeon at a university-based tertiary care center that offered Roux-en-Y gastric bypass, sleeve gastrectomy, and gastric banding. Only patients who completed postsurgical follow-up and who completed a postsurgical BDI were included in the study. Of the 318 patients, 137 met these requirements. After Institutional Review Board approval, the charts for the 137 patients were reviewed for age, gender, height, weight, presurgical body mass index (BMI), BDI scores, and postsurgical BMI and BDI scores. Presurgical or baseline data was collected about 4.6 months before surgery. About one half of the sample had their psychiatric assessment performed between 3–6 months before surgery. Postsurgical collection of data are classified into 2 groups: 1 week to 6 months follow-up after surgery as “up to 6 months,” and 46 months until the end of study period (i.e., 12 months from the bariatric procedure) as “6–12 months,” with an assumption that significant weight changes from bariatric procedures usually take about 6 months. The timeline for the collection of data was categorized into 0– 3 months, 3–6 months and 6–12 months. Ideal measure for BMI in up to 6 month group was taken between 3–6 months postsurgery. If that measure was not available, the 0–3 months measure was used. For nearly three quarters of the
sample, the BMI was measured from 3–6 months. Additionally, the type and number of co-morbidities and the psychiatric history, including a history of psychotropic drug use, were noted. Finally, charts were reviewed for social factors including the level of education completed, selfreported perception of family support, marital status, history of physical, mental, or sexual abuse, and employment status. Patients were then stratified by the type of surgery, which included Roux-en-Y gastric bypass, sleeve gastrectomy, and adjustable gastric banding. Statistical analysis Summary and descriptive statistics were generated for each patient’s baseline clinical characteristics. The Wald test and analysis of variance (ANOVA) were used to assess baseline imbalances between patients in the 3 surgical groups. The primary outcome, the change in BDI score from pre- to postsurgery, was analyzed using the paired t test. Analysis of covariance (ANCOVA) was used to assess the absolute change in BDI among the surgical groups after adjusting for the baseline BDI score. The least square mean differences with standard errors (SE) are reported. The relationship between marital status, employment, history of abuse, and psychiatric diagnosis and the change in the BDI score was investigated with exploratory analysis using ANCOVA. Sensitivity analysis was conducted using nonparametric rank order ANCOVA models. The Pearson correlation coefficient was calculated to assess the relationship between the change in BDI score and the change in BMI up to 6 months. Sensitivity analysis was performed carrying forward the BMI measurement up to 6 months for patients with missing data and the BMI measurement at 6– 12 months. A 2-sided P value of less than .05 was considered statistically significant. Statistical analysis was performed with SAS statistical software version 9.3 (SAS Institute, Cary, NC). Results Patient demographic characteristics The patient population in this study included 115 women and 22 men with a mean (⫾ SD) age of 41.7 ⫾ 10.4 years, height of 165.7 ⫾ 10.3 cm, weight of 132.0 ⫾ 24.8 kg, presurgical BMI of 47.7 ⫾ 7.7 kg/m2, and postsurgical BMI of 42.2 ⫾ 7.7 kg/m2. Seventy-eight (57%) patients were African American, 44 (32%) patients were married, 64 (47%) patients were employed, and 55 (40%) patients had completed high school or the equivalent at the time of surgery. Patient demographic characteristics were stratified by the type of bariatric surgery and are presented in Table 1. Of the 137 patients, 58 patients underwent gastric banding, 51 underwent sleeve gastrectomy, and 28 underwent gastric bypass. The Wald test and ANOVA were used to assess the
Depression and Different Bariatric Procedures / Surgery for Obesity and Related Diseases 11 (2015) 637–642
639
Table 1 Patient demographic characteristics Variables
Gender (female) (%) Age (yr) Height (cm) Weight (kg) Race (%) African-American Caucasian Others (Hispanic, Asian) Marital status (%) Married Single, divorced, or widowed Education (%) Less than high school High school or equivalent College or higher Employed (%) Perception of positive family support History of physical, sexual, or mental abuse History of psychiatric diagnosis (%) History of psychiatric medications (%) Medical co-morbidities (%) Hypertension Type 2 diabetes mellitus Obstructive sleep apnea Asthma Dyslipidemia Gastroesophageal reflux disease Hypothyroidism Average number of co-morbidities Body mass index (kg/m2) Presurgery Beck Depression Inventory score (mean) Presurgery
All surgeries
Gastric band
Sleeve gastrectomy
Gastric bypass
n ¼ 137
n ¼ 58
n ¼ 51
n ¼ 28
83.9 (n ¼ 115) 41.65 (⫾10.3) 165.73 (þ10.3) 131.97 (þ24.8)
96 (n ¼ 49) 44.0 (⫾ 11.1) 166.6 (⫾ 9.4) 132.3 (⫾ 28.1)
82 (n ¼ 42) 39.6 (⫾ 9.4) 166.8 (⫾ 9.2) 138.3 (⫾21.3)
86 (n ¼ 24) 40.7 (⫾ 10.0) 161.9 (⫾ 13.3) 119.7 (⫾ 19.0)
57 (n ¼ 78) 25 (n ¼ 34) 18 (n ¼ 25)
66 (n ¼ 38) 21 (n ¼ 12) 13.7 (n ¼ 8)
47 (n ¼ 24) 35 (n ¼ 18) 17.6 (n ¼ 9)
57 (n ¼ 16) 14 (n ¼ 4) 29 (n ¼ 8)
NS
32 (n ¼ 44) 68 (n ¼ 93)
31 (n ¼ 18) 69 (n ¼ 40)
33 (n ¼ 17) 67 (n ¼ 34)
32 (n ¼ 9) 68 (n ¼ 19)
.96
NS
(n (n (n (n (n (n (n (n
¼ ¼ ¼ ¼ ¼ ¼ ¼ ¼
18) 55) 63) 64) 127) 24) 50) 28)
14 45 40 48 93 22 34 21
51 (n 33 (n 34 (n 28 (n 27 (n 20 (n 11 (n 2.7
¼ ¼ ¼ ¼ ¼ ¼ ¼
70) 45) 47) 39) 37) 28) 15)
¼ ¼ ¼ ¼ ¼ ¼ ¼ ¼
P
.91 .09 .004
(n (n (n (n (n (n (n (n
¼ ¼ ¼ ¼ ¼ ¼ ¼ ¼
7) 20) 24) 25) 47) 9) 20) 11)
11 (n ¼ 3) 32 (n ¼ 9) 57 (n ¼ 16) 39 (n ¼ 11) 93 (n ¼ 26) 7 (n ¼ 2) 36 (n ¼ 10) 18 (n ¼ 5)
.67 .98 .24 .87 .92
57 (n ¼ 33) 33 (n ¼ 19) 34 (n ¼ 20) 28 (n ¼ 16) 22 (n ¼ 13) 19 (n ¼ 11) 5 (n ¼ 3) 2.8
41 (n 35 (n 41 (n 33 (n 25 (n 20 (n 20 (n 2.6
¼ ¼ ¼ ¼ ¼ ¼ ¼
21) 18) 21) 17) 13) 10) 10)
57 (n ¼ 16) 29 (n ¼ 8) 21 (n ¼ 6) 21 (n ¼ 6) 39 (n ¼ 11) 25 (n ¼ 7) 7 (n ¼ 2) 2.7
.20 .83 .22 .53 .27 .80 .06 .96
47.7 (⫾7.7)
47.3 (⫾ 7.8)
49.9 (⫾ 7.8)
44.5 (⫾ 6.0)
.017
11.3 (⫾ 8.8)
9.9 (⫾ 7.8)
11.5 (⫾ 8.9)
13.7 (⫾ 1.1)
.25
13 40 46 47 93 18 36 20
association between surgical procedures and demographic variables including age, gender, race, marital status, education, employment, co-morbidities, family support, history of physical/sexual/mental abuse, history of psychiatric disorder, or psychiatric medications. No statistically significant differences were found among the groups. Medical and psychiatric variables Table 1 presents data on co-morbidities present at the time of surgery. Although patients had a wide range of medical histories, the mean number of co-morbidities did not vary significantly among the different surgical groups (2.7 ⫾ 1.9; P ¼ .9636). The most common affective disorder for all patients in this data set was depression (n ¼ 31, 23%), followed by anxiety (n ¼ 11, 8%) which was clinically diagnosed by psychologist dedicated in assessing bariatric surgical patients. The majority of patients, however, did not have clinical psychiatric diagnoses. Patients who underwent gastric banding, sleeve gastrectomy and gastric bypass were
(n (n (n (n (n (n (n (n
8) 26) 23) 28) 54) 13) 20) 12)
14 39 47 49 92 18 39 22
each prescribed a mean of .4, .4, and .3 psychotropic medications, respectively. A mean of .4 psychotropic medications was prescribed for all patients combined. We found a statistically significant difference in the baseline mean preoperative BMI among the different surgical procedures (P ¼ .017) with the highest preoperative BMI in the sleeve gastrectomy group (mean 49.9 ⫾ 7.8, SD). No significant difference was noted in the mean preoperative BDI scores among the different groups (P ¼ .1693). At 6 months postsurgery, we observed significant (P o .0001) improvement in BDI scores from 11.3 ⫾ 8.8 (SD) to 7.2 ⫾ 8.3 (SD) with a mean difference of 4.11 ⫾ 11.69 (SD) and a significant drop in BMI from 47.7 ⫾7.7 to 42.2 ⫾ 7.73 with a mean difference of 5.60 ⫾ 3.80 (SD) (P o .0001). ANCOVA was performed to look at the relationship between different surgical procedures and the change in BDI after adjusting for preoperative BDI (Table 2). Gastric bypass produced the largest decrease in the least squares (LS) adjusted mean BDI change at 5.8 ⫾ 1.58 (SE), followed by sleeve
640
Outcome
Gastric band Baseline
Sleeve gastrectomy Up to 6 mo
Mean (SD) Number of patients Weight (kg) Least square difference (95% versus sleeve gastrectomy Least square difference (95% versus lap band BMI Least square difference (95% versus sleeve gastrectomy Least square difference (95% versus lap band BDI Least square difference (95% versus sleeve gastrectomy Least square difference (95% versus lap band
58 132.3 (28.0) CI) CI)
Change from baseline *Least square mean (SE)
57 57 122.9 (26.9) -8.8 (1.2) 10.22 (6.51, 13.92)
Up to 6 mo Mean (SD)
51 138.3 (21.3)
49 119.5 (21.3) -
Change from baseline *Least square mean (SE) 49 -19.0 (1.3)
Baseline
47.3 (7.8)
28 119.7 (18.9)
Change from baseline *Least square mean (SE)
28 28 97.1 (21.2) -23.5 (1.8) -4.49 (-9.15, .17) -14.71 (-19.14, -10.28)
44.2 (7.8) -3.2 (.4) 3.70 (2.51, 4.88)
CI)
49.8 (7.8)
42.9 (7.2) -
-6.9 (.4)
44.5 (6.0)
9.9 (7.8)
Up to 6 mo
Mean (SD)
-
CI)
CI)
Baseline
Gastric bypass
8.59 (9.9) 2.33(-.83, 5.49)
CI)
BMI ¼ body mass index; BDI ¼ Beck Depression Inventory.
-2.5 (1.0)
-4.98 (-6.38, -3.58) 11.4 (8.8)
-
36.7 (5.6) -8.1 (.5) -1.28 (-2.76, .19)
6.3 (7.3) -
-4.9 (1.1)
13.7 (10.1)
5.6 (5.9) -5.8 (1.5) -0.90(-4.79, 2.97) -3.24(-7.07, .59)
S. Ayloo et al. / Surgery for Obesity and Related Diseases 11 (2015) 637–642
Table 2 Changes in Beck Depression Inventory from baseline among different surgical procedures
Depression and Different Bariatric Procedures / Surgery for Obesity and Related Diseases 11 (2015) 637–642
gastrectomy at 4.91 ⫾ 1.16 (SE) and gastric banding at 2.57 ⫾ 1.09 (SE). We found no statistically significant difference in BDI change among surgical procedures when adjusting for presurgical BDI scores (P ¼ .1739). These results were similar when the model included the change in BMI (P ¼ .4061). There was a statistically significant difference for change in weight between surgical procedures after adjusting for preoperative weight (P o .0001). Gastric bypass had a LS mean weight loss difference of -14.71, 95% CI (-19.14, -10.28) greater compared to gastric band and a LS mean weight loss difference of -4.49 (95% CI -9.15, .17) greater compared to sleeve gastrectomy. Gastric band had 10.22 (95% CI 6.51, 13.92) less weight loss compared to sleeve gastrectomy. As well there was a statistically significant difference for change in BMI between surgical procedures after adjusting for preoperative BMI (P o .0001). Results for BMI are shown in Table 2. Parametric ANCOVA analyses were performed to evaluate potentially influential demographic variables for association with the change in BDI, adjusting for initial BDI. Employment and psychiatric history were statistically significant. Results were consistent with those from the nonparametric analysis. Patients who were employed had a greater adjusted LS mean decrease in BDI of 5.77 ⫾ 1.03 (SE) compared to those not employed, with a mean decrease change in BDI of 2.64 (SE, .96; P ¼ .0284). Patients with no psychiatric diagnosis had a greater decrease in adjusted LS mean BDI change 5.50 (SE, .88) compared to those with a psychiatric diagnosis who had a decrease in BDI change of 1.68 (SE, 1.16; P ¼ .0101) after adjusting for initial BDI score. The correlation between BDI and BMI was examined with the Pearson correlation coefficient considering the difference between baseline and 6 to 12 months after surgery (n ¼ 94). No statistically significant association was identified between the BDI change and the BMI change (P ¼ .1327). A sensitivity analysis was then performed in which the value up to 6 months after surgery was carried forward to include the entire sample. Results were consistent with the original analysis (P ¼ .8790) and were also consistent with those from nonparametric analysis. Discussion According to the American Society for Metabolic and Bariatric Surgery, 179,000 bariatric procedures were performed in 2013 [11]. The most commonly performed bariatric operations are Roux-en-Y gastric bypass (34%), sleeve gastrectomy (42%), and adjustable gastric banding (14%). Although bariatric surgeries improve medical comorbidities, many patients rely on these surgeries to produce the behavioral changes needed to maintain weight loss [12]. On the contrary, Odom et al. reported that certain behavioral factors such as increased food urges, concern
641
over addictive behaviors, and a decreased sense of wellbeing are strong predictors of weight regain after bariatric surgery [13]. Because of the significant interplay between depression and obesity [14], presurgical psychological evaluation of patients planning to undergo bariatric surgery is an appropriate prerequisite. Our objective was to analyze the effect of weight loss after different bariatric procedures on the severity of depression as measured with BDI scores. Furthermore, to evaluate the effect of BDI change on 6–12-month weight loss and the effect of psychosocial status on BDI scores. Of the 318 patients enrolled in the study only 137 patients met the inclusion criteria. This is partly due to mandatory completion of preoperative BDI form for psychological evaluation and clearance, an important component for surgical consideration. While the follow-up completion of the BDI was on volunteer basis, there was a significant drop off of patients filling out the BDI form despite follow-up. Exceptions being a small percentage of patients seeking revisional surgery or with significant weight regain were strongly recommended to be followed up with psychologist for behavioral modification and were monitored closely with BDI forms. One of the limitations of the study is a lack of comparison of outcomes with patients whose BDI scores were lacking or with a group that did not get follow-up, which could potentially effect key conclusions of this study. Another limitation of the study is globalization of the time periods. Usual protocol at our center for follow-up includes one week follow-up after discharge, followed by 1 month and every 3 months during the first year. Patients had annual follow-ups, subsequent to their first year for sleeve gastrectomy and Roux-en-Y gastric bypass. Patients with adjustable gastric banding were followed up during the time of band adjustment. Despite this regiment there is overlap of patient appointments as such the study was divided into presurgical or baseline outcomes measured on average 4.6 months before surgery, postsurgical divided into 2 groups of 0–6 months as “up to 6 months” and as “6–12 months,” and for followup beyond 6 months to the end of the study period which was 1 year from the time of bariatric surgery. The prevalence of depression in adults in the United States between 2006 and 2008 was about 9.1%. Depression was more common among individuals with chronic diseases (obesity, cardiovascular disease, type 2 diabetes) and among individuals with unhealthy behaviors (smoking, physical inactivity). Depression is more prevalent in women, those previously married (divorced, widowed, or separated), those with education level less than equivalent to high school, and those unable to work or unemployed individuals [15]. Models to evaluate the association between important demographic characteristics and psychosocial variables such as gender, marital status, being employed, history of abuse, and previous psychological history suggested that
642
S. Ayloo et al. / Surgery for Obesity and Related Diseases 11 (2015) 637–642
individuals who were employed and who had no previous psychological history had a statistically significant improvement in BDI scores, reflecting an improvement of depressive symptoms. This observation suggests that these psychosocial factors may play a critical role in the bariatric population for the overall postoperative well-being of the patient. In addition, evaluation of the effect of the change in BDI after more than 6 months and BMI indicated no statistical correlation. One of the limitations of this study has been to group all the affective disorders together without differentiating depressive disorders from anxiety disorders that might potentially have an effect on the change of BDI and BMI In reviewing our series of patients, 84% of patients were women. Patient demographic characteristics including age, gender, race, marital status, education, and employment status; psychosocial characteristics of family support, history of psychiatric disorder, treatment with psychiatric medications, and history of physical, sexual, or mental abuse; and medical co-morbidities were not significantly different among the different surgical groups, thus minimizing the effect of these variables on the BDI scores. The preoperative BDI scores for different procedures were not significantly different, although the preoperative BMI was significantly different. This was expected because the gastric sleeve procedure was originally intended for patients with a higher BMI. When assessing the effect of weight on BDI, the absolute change was considered to minimize the difference in the preoperative BMI among the procedures [16,17,18]. Conclusions Bariatric surgery leads to significant improvement in weight loss and BDI scores. Weight loss was greatest after gastric bypass, followed by sleeve gastrectomy and adjustable gastric banding. The improvement in BDI scores was independent of the magnitude of weight loss, which was true for each of the procedures analyzed. Patients who were employed with no prior psychological history experienced a significant improvement in BDI scores. No correlation was found in improvement in BDI up to 6 months to weight loss at 6–12 months. We believe that the change in mood affected by bariatric surgery is due to the interplay of multiple factors including initial weight loss. The data suggest that psychosocial factors may play a more substantial role in the change in mood than was previously recognized. Disclosures The authors have no funding information to disclose; all authors declare no conflict of interest.
References [1] Roberts RE, Kaplan GA, Shema SJ, Strawbridge WJ. Are the obese at greater risk for depression? Am J Epidemiol 2000;152(2):163–70 [2] Luppino FS, de Wit LM, Bouvy PF, et al. Overweight, obesity, and depression: a systematic review and meta-analysis of longitudinal studies. Arch Gen Psychiatry 2010;67:220–9. [3] Muller A, Mitchell JE, Sondag C, de Zwaan M. Psychiatric aspects of bariatric surgery. Curr Psychiatry Rep 2013;15:397. [4] Marcus MD, Kalarchian MA, Courcoulas AP. Psychiatric evaluation and follow-up of bariatric surgery patients. Am J Psychiatry 2009;166:285–91. [5] Kalarchian MA, Marcus MD, Levine MD, et al. Psychiatric disorders among bariatric surgery candidates: relationship to obesity and functional health status. Am J Psychiatry 2007;164:328–34. (quiz 74). [6] Kalarchian MA, Marcus MD, Levine MD, Soulakova JN, Courcoulas AP, Wisinski MS. Relationship of psychiatric disorders to 6-month outcomes after gastric bypass. Surg Obes Relat Dis 2008;4:544–9. [7] Bauchowitz AU, Gonder-Frederick LA, Olbrisch ME, et al. Psychosocial evaluation of bariatric surgery candidates: a survey of present practices. Psychosom Med 2005;67:825–32. [8] Hayden MJ, Dixon JB, Dixon ME, Shea TL, O'Brien PE. Characterization of the improvement in depressive symptoms following bariatric surgery. Obes Surg 2011;21:328–35. [9] Dixon JB, Dixon ME, O'Brien PE. Depression in association with severe obesity: changes with weight loss. Arch Intern Med 2003;163: 2058–65. [10] Sanchez-Santos R, Del Barrio MJ, Gonzalez C, et al. Long-term health-related quality of life following gastric bypass: influence of depression. Obes Surg 2006;16:580–5. [11] ASMBS. Estimate of bariatric surgery numbers 2014. [page on the internet]. Gainseville: American Society for Metabolica and Bariatric Surgery; [cited November 4, 2014]. Available from: http://asmbs.org/ 2014/03/estimate-of-bariatric-surgery-numbers/. [12] Thonney B, Pataky Z, Badel S, Bobbioni-Harsch E, Golay A. The relationship between weight loss and psychosocial functioning among bariatric surgery patients. Am J Surg 2010;199:183–8. [13] Odom J, Zalesin KC, Washington TL, et al. Behavioral predictors of weight regain after bariatric surgery. Obes Surg. 2010;20:349–56. [14] de Zwaan M, Enderle J, Wagner S, et al. Anxiety and depression in bariatric surgery patients: a prospective, follow-up study using structured clinical interviews. J Affect Disord 2011;133(1–2):61–8. [15] Centers for Disease Control and Prevention. Current depression among adults—United States, 2006 and 2008. Morbid Mortal Weekly Rep [page on the internet]. [cited October 1, 2010]. Available from: http://www.cdc.gov/features/dsdepression/revised_table_estimates_ for_depression_mmwr_erratum_feb-2011.pdf. [16] Averbukh Y, Heshka S, El-Shoreya H, et al. Depression score predicts weight loss following Roux-en-Y gastric bypass. Obes Surg 2003;13:833–6. [17] Schowalter M, Benecke A, Lager C, et al. Changes in depression following gastric banding: a 5- to 7-year prospective study. Obes Surg 2008;18:314–20. [18] Munoz DJ, Chen E, Fischer S, et al. Considerations for the use of the Beck Depression Inventory in the assessment of weight-loss surgery seeking patients. Obes Surg 2007;17:1097–101.
