- Email: [email protected]
Prevalence and psychosocial correlates of self-reported past suicide attempts among bariatric surgery candidates
Surgery for Obesity and Related Diseases 6 (2010) 702–706
Integrated health article
Prevalence and psychosocial correlates of self-reported past suicide attempts among bariatric surgery candidates Amy K. Windover, Ph.D.a,b,*, Julie Merrell, Ph.D.a, Kathleen Ashton, Ph.D.a,b, Leslie J. Heinberg, Ph.D.a,b a Cleveland Clinic, Cleveland, Ohio Cleveland Clinic Lerner College of Medicine, Case Western Reserve University, Cleveland, Ohio Received April 2, 2010; accepted August 30, 2010
b
Abstract
Background: Although research has been limited, suicidal behavior has commonly been identified as a contraindication for bariatric surgery. The present study aimed to determine the prevalence and correlates of past suicide attempts in a bariatric surgery population at an academic medical center. Methods: A retrospective chart review, including the demographic and psychosocial variables, was conducted of 1020 consecutive bariatric surgery candidates presenting during a 32-month period. Results: Of the 1020 patients, 115 (11.2%) self-reported ⱖ1 previous suicide attempt. The patients with a positive suicide history were significantly younger (mean 42.9 ⫾ 11.0 years), less educated (mean 13.4 ⫾ 2.4 years), had a greater body mass index (mean 52.3 ⫾ 11.6 kg/m2), and were more predominantly single (32.2% versus 20.9%), female (90.4% versus 74.8%), and receiving disability (45.2% versus 21.8%) compared with patients without a suicide history. A positive suicide history was also significantly associated with a history of psychiatric hospitalization, outpatient psychotherapy and/or psychotropic medication, sexual abuse, and substance abuse. Conclusion: Assessing suicide history is an important aspect of the bariatric preoperative assessment. Additional research is needed to evaluate the effects of suicide history on the postoperative outcomes and adherence. (Surg Obes Relat Dis 2010;6:702–706.) © 2010 American Society for Metabolic and Bariatric Surgery. All rights reserved.
Keywords:
Suicide history; Past suicide attempts; Bariatric surgery
Bariatric surgery is currently the most effective longterm treatment of severe obesity [1,2]. Benefits include sustained weight loss, remediation of medical co-morbidities, improved depression [3], and improved quality of life [1]. Long-term mortality, as well as disease-specific mortality for coronary artery disease, cancer, and diabetes, is also significantly reduced after gastric bypass surgery [4]. Despite these benefits, the postoperative suicide rates have been speculated to be much greater than those for the population at large. For example, 1 study reported 3 suicides of 462 gastric bypass patients [5]. In a review of 4 studies,
*Correspondence: Amy K. Windover, Ph.D., Bariatrics and Metabolic Institute, Cleveland Clinic, 9500 Euclid Avenue, M66, Cleveland, OH 44195. E-mail: [email protected]
8 suicides were identified of 1785 patients who underwent bariatric surgery [6]. In contrast, another study noted only 1 suicide death of 1040 postoperative gastric bypass patients [7]. A study of the death rates and cause of death for all Pennsylvania residents who had undergone bariatric surgery from 1995 to 2004 (n ⫽ 16,683) revealed 16 suicides [8]. When averaged, these studies have suggested a rate of 140 suicide deaths per 100,000 compared with a populationwide incidence of 14.37 suicide deaths per 100,000 [9]. More concerning, compared with a control group matched on age, gender, and body mass index (BMI), 15 suicides were identified in a bariatric surgery group (260 per 100,000) versus 5 (9 per 100,000) in a control group [4]. Although not included in these numbers, that study also identified a greater number of other accidental deaths and deaths by poisoning of undetermined intent in the bariatric
1550-7289/10/$ – see front matter © 2010 American Society for Metabolic and Bariatric Surgery. All rights reserved. doi:10.1016/j.soard.2010.08.014
A. K. Windover et al. / Surgery for Obesity and Related Diseases 6 (2010) 702–706
surgery versus control group [4]. It is possible that at least some of these accidental deaths and deaths by poisoning could have been undeclared suicides. Such research highlights the potential concern for the long-term risk of suicide among bariatric surgery patients. Although causative factors have yet to be identified, severe obesity has been associated with an increased risk of psychopathology, including depression [10,11] and suicide attempts [12]. Psychopathology and/or a history of attempted suicide, in turn, have been associated with greater risk of suicide [13]. Consequently, active or recent suicidal ideation and behavior have been commonly identified as psychological contraindications for bariatric surgery [14], because suicidal ideation and behavior are indicative of current psychiatric instability and suggest the need for emergent psychiatric care. Furthermore, such instability might make the behavioral changes and adherence more difficult to achieve. Although a history of suicidal ideation and behavior is often a part of preoperative psychological assessment [15,16], research exploring the history of attempted suicide, and its correlates and effect on weight loss outcomes in bariatric populations have been limited. To the best of our knowledge, only 2 studies have explored the prevalence of past suicide attempts among bariatric surgery patients. The first study found that 14 women who had undergone gastric bypass surgery had significantly more past suicide attempts compared with a control group of 14 morbidly obese women seeking nonoperative weight management [17]. However, 15 years later, the second study examined the lifetime prevalence of attempted suicide and self-injurious behavior in a sample of 121 bariatric surgery candidates and found 11 (9.1%) to selfreport a history of attempted suicide [18]. In contrast to the bariatric published data, studies exploring the prevalence of past suicide attempts have been more prevalent for other moderate-to-high risk surgeries, such as organ transplantation [19 –21]. A history of attempted suicide has been significantly associated with a shorter time to infection or rejection in cardiac transplant patients [20]. It has also been identified as a predictor of attenuated survival time in this population [20]. Population-based studies of nonfatal self harm have revealed a greater prevalence of suicide attempt histories among younger, previously married, white women [22]. Depression, other mental disorders, alcohol abuse, and other substance abuse have also been found to be more prevalent among those with a history of attempted suicide compared with the adult population at large [22]. The presence of similar demographic and clinical factors within weight loss surgery populations has not been previously examined. Thus, the present study examined correlates of past suicidal behavior. On the basis of epidemiologic studies of suicide, we hypothesized that those with a history of suicide would be more likely to be women, have
703
a history of psychiatric treatment, and be younger than those without this history.
Methods Participants Data from the psychological evaluation were retrospectively reviewed from the medical charts of 1020 consecutive patients evaluated at the Cleveland Clinic Bariatric and Metabolic Institute for bariatric surgery from January 2006 to August 2008. The 1020 participants were primarily women (76.6%), white (69.1%), married (47.0%), and employed (53.3%). The age range was 18 –77 years (mean 46.3 ⫾ 11.3). Their education level was 7–24 years (mean 13.8 ⫾ 2.5). The presenting BMI was 31.5–99.3 kg/m2 (mean 50.3 ⫾ 11.1). Procedures The data reviewed included past suicide attempts, demographic information, preoperative BMI, history of physical and sexual abuse, history of substance abuse/dependence, and current or historical mental health treatment, including inpatient hospitalization, psychotropic medication management, and psychotherapy. These data were obtained during semistructured preoperative psychological and behavioral assessment interviews using a template form in the electronic medical record. During the assessment of past psychiatric history, all patients were queried whether they had ever tried to kill themselves or hurt themselves in some other manner. When answered in the affirmative, follow-up questions were used to assess the number of attempts, method, and subsequent treatment. However, for the present study, patients were simply dichotomized into either the presence or absence of a suicide history. A history of suicidal ideation, plan, or intent was not considered to be positive. Only those patients who endorsed a specific attempt were dichotomized as having positive suicide history (SH⫹). Objective psychological testing (Symptom Checklist-90-R) was also administered as a part of the preoperative evaluation; however, it was not used to determine the suicide history [23]. Additionally, if patients were currently in treatment with a mental health professional or had past inpatient psychiatric hospitalization within the past 5 years, mental health records were obtained before surgical clearance. The medical chart review and patient self-report noted historical and current psychotropic medications that were confirmed at each subsequent visit by bariatric nurses. Patients reporting a psychiatric hospitalization or suicide attempt in the past year were programmatically required to demonstrate 12 months of mental health intervention and emotional stability before surgery. The institutional review board of the Cleveland Clinic approved the use of the patient data.
704
A. K. Windover et al. / Surgery for Obesity and Related Diseases 6 (2010) 702–706
Statistical analysis The data were analyzed using the Statistical Package for Social Sciences, version 17.0, (SPSS, Chicago, IL). For the purposes of the present study, a SH⫹ was defined as having ⱖ1 self-reported suicide attempt. The prevalence of suicide history in our population was determined. Demographic differences among patients with a SH⫹ (n ⫽ 115) versus patients without a suicide history (SH⫺) (n ⫽ 905) were evaluated using independent samples t tests for continuous variables and chi-square analyses for categorical variables. The results were considered significant at P ⱕ.05, unless otherwise indicated. Binary logistic regression analysis was conducted to determine the effect of a SH on the clinical variables of interest. Any sociodemographic differences between the groups were entered in the first step of the regression equation, followed by baseline BMI, and, finally, the psychosocial variables: a history of psychiatric hospitalization, a history of outpatient psychotherapy and/or psychotropic medication management, current psychotropic medication use, history of sexual abuse, history of physical abuse, and history of tobacco use. Owing to the multiple comparisons, a Bonferroni correction was applied, making the ␣ value more conservative (␣ ⫽ .004). Results Of the 1020 patients evaluated for 32 months, 115 (11.2%) reported a history of ⱖ1 suicide attempts. The SH⫹ bariatric surgery candidates were significantly younger, less educated, had a greater BMI, and were more predominantly single, female, and receiving disability compared with the SH⫺ patients. The percentages, mean values, standard deviations, and significance levels comparing the SH⫹ and SH⫺ groups are summarized in Table 1. Binary logistic regression analysis was used on a subset of 980 patients without missing data to determine the clinical correlates of SH⫹ bariatric surgery candidates, independent of sociodemographic and BMI differences. Step 1 included the following demographic variables: gender, marital status, job status, age, and education. Step 2 consisted of the preoperative BMI on program entry. Step 3 included clinical variables of interest: a history of psychiatric hospitalization, history of outpatient psychotherapy and/or psychotropic medication management, current psychotropic medication use, history of sexual abuse, history of physical abuse, and history of tobacco use. As previously noted, a SH⫹ was significantly correlated with job status but only showed a trend toward significance for age and gender after controlling for multiple comparisons. The significance levels of each step and individual predictive  weights are listed in Table 2. A significant difference was not found between the 2 groups for the baseline BMI [chi-square (1, n ⫽ 980) ⫽ 2.8, P ⬎.09].
Table 1 Comparison of sociodemographic characteristics Variable Gender Female Male Ethnicity White Black Other Marital status Married Never married Divorced Other Employment status Employed Disability Unemployed Other Surgery type Roux-en-Y LapBand Sleeve Other Age (yr) Education (yr) Preoperative BMI (kg/m2)
SH⫹ (n ⫽ 115)
SH⫺ (n ⫽ 905)
104 (90.4) 11 (9.6)
677 (74.8) 228 (25.2)
76 (66.7) 29 (25.4) 9 (7.9)
630 (70.2) 217 (24.2) 50 (5.6)
34 (29.6) 37 (32.2) 30 (26.1) 14 (12.2)
446 (49.3) 189 (20.9) 170 (18.8) 99 (11.0)
37 (32.5) 52 (45.6) 16 (14.0) 9 (7.9)
508 (56.3) 197 (21.8) 103 (11.4) 94 (10.4)
33 (71.7) 4 (8.7) 7 (15.2) 2 (4.4) 43 ⫾ 11 13.4 ⫾ 2.4 52.3 ⫾ 11.6
263 (65.9) 71 (17.8) 48 (12.0) 17 (4.3) 46.8 ⫾ 11.3 13.9 ⫾ 2.5 50 ⫾ 11
df
Chi-square 2
13.9*
2
3.0
3
22.3†
3
45.8*
3
6.5
1018 994 1015
⫺3.4† ⫺1.9‡ 2.2§
SH ⫽ suicide history; df ⫽ degrees of freedom; BMI ⫽ body mass index. Data presented as numbers, with percentages in parentheses, or mean ⫾ standard deviation. * P ⬍.0001, 2-tailed. † P ⬍.001, 2-tailed. ‡ Marginally significant at P ⬍.052. § P ⬍.05, 2-tailed.
After controlling for the significant demographic variables, a history of psychiatric hospitalization and a history of outpatient psychotherapy and/or psychotropic medication management also correlated significantly with SH⫹ [chisquare (7, n ⫽ 980) ⫽ 175, P ⬍.001; Table 2]. Trends toward significance were noted between groups for a history of substance abuse and a history of sexual abuse (Table 2). When entered with the other covariates, no significant differences between groups were found for marital status, education, preoperative BMI, current psychotropic medication use, a positive physical abuse history, and a positive history of tobacco use. Discussion The prevalence of SH⫹ in our sample was 73 times that of the population at large (age range 18 –77) [22]. Although somewhat greater, this corresponded to the findings of the limited research available demonstrating a greater prevalence of SH⫹ in bariatric populations compared with the
A. K. Windover et al. / Surgery for Obesity and Related Diseases 6 (2010) 702–706
705
Table 2 Summary of binary logistic regression analyses for sociodemographic and clinical predictors of Patients with positive suicide history (n ⫽ 980) Model Binary logistic regression 1 Step 1 Gender Marital status Job status Age Education Step 2 BMI on program entry Step 3 History of Psychiatric hospitalization Outpatient mental health treatment Sexual abuse Physical abuse Active psychiatric medication Tobacco use Substance abuse/dependence
Chi-square 50.2*
df

SE
Exp()
⫺1† .1 ⫺.2† .02† .1
.3 .1 .1 .01 .04
.3 1.1 .7 1 11
⫺.01
.009
2.3* 1.1* .5† .2 ⫺.002 ⫺.1 ⫺.6†
.3 .4 .3 .3 .3 .2 .3
5
2.8
1
175*
7
.1
10.4 3.1 1.7 .5 .9 .8 .5
SE ⫽ standard error; other abbreviations as in Table 1. * P ⬍.001. † P ⬍.05, trend toward significant.
population base rates [4 – 6]. Also consonant with these studies, a greater prevalence of SH⫹ was found among those with greater preoperative BMIs [24]. Theories for the relationship between extreme obesity and psychiatric issues have included the negative effects of social stigma on mood and the increased appetite, decreased physical activity, and the use of weight gain-promoting medication common in depression [25]. Consistent with epidemiologic studies, we found a greater prevalence of SH⫹ among younger patients, women, and primarily single/never married, but also divorced, patients. No significant differences were found for race/ethnicity. Additionally, we noted a significant relationship between SH⫹ and receiving disability; however, having less education was only marginally significant. Significant correlations found among the SH⫹ patients and a history of psychiatric hospitalization and outpatient psychotherapy and/or psychotropic medication management are congruent with epidemiologic studies reporting a greater prevalence of SH⫹ among those diagnosed with depression and other mental disorders. The increased rates of psychiatric hospitalization and outpatient mental health treatment were not surprising, because often those who have attempted suicide require medical attention and might be appropriately referred for ongoing mental health treatment at that time. The severe degree of emotional distress associated with SH⫹ also often impairs an individual’s interpersonal and occupational functioning. Thus, this might account for the relationship between SH⫹ and those receiving disability. Such data on the prevalence and psychosocial correlates supports the common practice of assessing history of attempted suicide, as well as other psychosocial variables.
The routine assessment of SH⫹ both pre- and postoperatively might inform the treatment team regarding the need for closer monitoring and the potential benefit of evidencebased psychotherapy. Given the evaluative nature of the assessment, it is quite likely that some patients chose not to disclose a past suicide attempt and were incorrectly dichotomized as SH⫺. However, it is unlikely that patients were incorrectly dichotomized as SH⫹ when they did not have a true history. Thus, the reported results might not be a true representation of all SH⫹ patients, and future research should include assessments for suicide history that are separate from the evaluative process. A second limitation of the present study was that our relatively small sample of SH⫹ patients reduced the power of our analyses. Previous research has identified a greater prevalence of suicide attempts after bariatric surgery [17,18]. Future research should examine the predictive utility of self-reported past suicide attempts on both the short- and long-term bariatric surgery outcomes, including suicide attempts after surgery. Conclusion The results of the present study have identified a greater prevalence of SH⫹ among bariatric surgery candidates compared with the population at large. The sociodemographic and clinical correlates of SH⫹ among bariatric surgery candidates include younger age, female gender, single, and receiving disability, as well as having a history of psychiatric hospitalization, outpatient psychotherapy and/or psychotropic medication management, sexual abuse, and substance abuse/dependence. Additional research is
706
A. K. Windover et al. / Surgery for Obesity and Related Diseases 6 (2010) 702–706
needed to examine the predictive utility of SH⫹ on bariatric surgery outcomes.
[12]
Disclosures [13]
The authors have no commercial associations that might be a conflict of interest in relation to this article. [14]
References [15] [1] Buchwald H. Consensus conference statement: bariatric surgery for morbid obesity: health implications for patients, health professionals, and third-party payers. Surg Obes Relat Dis 2005;1:371– 81. [2] Elder KA, Wolfe BM. Bariatric surgery: a review of procedures and outcomes. Gastroenterology 2007;132:2253–71. [3] Dymek M, le Grange D, Neven K, Alverdy J. Quality of life and psychosocial adjustment in patients after Roux-en-Y gastric bypass: a brief report. Obes Surg 2001;11:32–9. [4] Adams T, Gress R, Smith S, et al. Long-term mortality after gastric bypass surgery. N Engl J Med 2007;357:753– 61. [5] Waters GS, Pories, WJ, Swanson MS, et al. Long-term studies of mental health after the Greenville gastric bypass operation for morbid obesity. Am J Surg 1991;161:154 –7. [6] Hsu G, Benotti P, Dwyer J, et al. Nonsurgical factors that influence the outcome of bariatric surgery: a review. Psychosom Med 1998;60: 338 – 46. [7] Higa K, Boone K, Ho T. Complications of the laparoscopic Rouxen-Y gastric bypass: 1,040 patients—what have we learned? Obes Surg 2000;10:509 –13. [8] Omalu BI, Ives DG, Buhari AM, et al. Death rates and causes of death after bariatric surgery for Pennsylvania residents, 1995 to 2004. Arch Surg 2007;142:923–28. [9] Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Web-based injury statistics query and reporting system (WISQARS). Available at: www.cdc.gov/ncipc/ wisqars. Accessed September 3, 2009. [10] Mitchell JE, deZwaan M. Bariatric surgery: a guide for mental health professionals. New York: Routledge Taylor & Francis Group; 2005. p. 1–158. [11] Petry NM, Barry D, Pietrzak RH, Wagner JA. Overweight and obesity are associated with psychiatric disorders: results from the Na-
[16] [17] [18]
[19]
[20]
[21] [22]
[23]
[24]
[25]
tional Epidemiological Survey on alcohol and related conditions. Psychosom Med 2008;70:288 –97. Dong C, Li WD, Li D, Price R. Extreme obesity is associated with attempted suicides: results from a family study. Int J Obes 2000;30: 388 –90. Jacobs D, Brewer M. APA practice guideline: provides recommendations for assessing and treating patients with suicidal behaviors. Psychiatric Ann 2004;34:373– 80. Bauchowitz AU, Gonder-Frederick LA, Olbrisch ME, et al. Psychosocial evaluation of bariatric surgery candidates: a survey of present practices. Psychosom Med 2005;67:825–3. LeMont D, Moorehead M, Parish M, Reto C, Ritz S. Suggestions for the pre-surgical psychological assessment of bariatric surgery candidates. Durham, NC: Duke University Medical Center; 2004. Powers P, Rosemurgy A, Coovert D, Boyd F. Psychological sequelae of bariatric surgery: a pilot study. Psychosomatics 1988;29:283– 88. Rosen L, Aniskiewicz A. Psychosocial functioning of two groups of morbidly obese patients. Int J Obes 1983;7:53–9. Sansone R, Wiederman MW, Schumacher D, Routsong-Weichers L. The prevalence of self-harm behaviors among a sample of gastric surgery candidates. J Psychosom Res 2008;65:441– 44. Levenson J, Olbrisch M. Psychosocial evaluation of organ transplant candidates: a comprehensive survey of process, criteria, and outcomes in heart, liver, and kidney transplantation. Psychosomatics 1993;34:314 –23. Owen J, Bonds C, Wellisch D. Psychiatric evaluations for heart transplant candidates: predicting post-transplant hospitalizations, rejection episodes, and survival. Psychosomatics 2006;47:213–22. Reither A, Mahler E. Suicide in liver transplant patients. Psychosomatics 1994;35:574 –77. National Institute of Mental Health. Suicide in the U.S.: statistics and prevention. 2007. Available at: www.nimh.nih.gov/health/publications/ suicide-in-the-us-statistics-and-prevention. Accessed July 20, 2008. Derogatis, LR. Symptom Checklist-90-R Administration, Scoring, and Procedures Manual. Minneapolis: National Computer Systems; 1994. p. 1–123. Mather A, Cox B, Enns M, Sareen J. Associations of obesity with psychiatric disorders and suicidal behaviors in a nationally representative sample. J Psychosom Res 2009;66:277– 85. Simon GE, Von Korff M, Saunders K, et al. Association between obesity and psychiatric disorders in the US adult population. Arch Gen Psychiatry 2006;63:824 –30.
Integrated health article
Prevalence and psychosocial correlates of self-reported past suicide attempts among bariatric surgery candidates Amy K. Windover, Ph.D.a,b,*, Julie Merrell, Ph.D.a, Kathleen Ashton, Ph.D.a,b, Leslie J. Heinberg, Ph.D.a,b a Cleveland Clinic, Cleveland, Ohio Cleveland Clinic Lerner College of Medicine, Case Western Reserve University, Cleveland, Ohio Received April 2, 2010; accepted August 30, 2010
b
Abstract
Background: Although research has been limited, suicidal behavior has commonly been identified as a contraindication for bariatric surgery. The present study aimed to determine the prevalence and correlates of past suicide attempts in a bariatric surgery population at an academic medical center. Methods: A retrospective chart review, including the demographic and psychosocial variables, was conducted of 1020 consecutive bariatric surgery candidates presenting during a 32-month period. Results: Of the 1020 patients, 115 (11.2%) self-reported ⱖ1 previous suicide attempt. The patients with a positive suicide history were significantly younger (mean 42.9 ⫾ 11.0 years), less educated (mean 13.4 ⫾ 2.4 years), had a greater body mass index (mean 52.3 ⫾ 11.6 kg/m2), and were more predominantly single (32.2% versus 20.9%), female (90.4% versus 74.8%), and receiving disability (45.2% versus 21.8%) compared with patients without a suicide history. A positive suicide history was also significantly associated with a history of psychiatric hospitalization, outpatient psychotherapy and/or psychotropic medication, sexual abuse, and substance abuse. Conclusion: Assessing suicide history is an important aspect of the bariatric preoperative assessment. Additional research is needed to evaluate the effects of suicide history on the postoperative outcomes and adherence. (Surg Obes Relat Dis 2010;6:702–706.) © 2010 American Society for Metabolic and Bariatric Surgery. All rights reserved.
Keywords:
Suicide history; Past suicide attempts; Bariatric surgery
Bariatric surgery is currently the most effective longterm treatment of severe obesity [1,2]. Benefits include sustained weight loss, remediation of medical co-morbidities, improved depression [3], and improved quality of life [1]. Long-term mortality, as well as disease-specific mortality for coronary artery disease, cancer, and diabetes, is also significantly reduced after gastric bypass surgery [4]. Despite these benefits, the postoperative suicide rates have been speculated to be much greater than those for the population at large. For example, 1 study reported 3 suicides of 462 gastric bypass patients [5]. In a review of 4 studies,
*Correspondence: Amy K. Windover, Ph.D., Bariatrics and Metabolic Institute, Cleveland Clinic, 9500 Euclid Avenue, M66, Cleveland, OH 44195. E-mail: [email protected]
8 suicides were identified of 1785 patients who underwent bariatric surgery [6]. In contrast, another study noted only 1 suicide death of 1040 postoperative gastric bypass patients [7]. A study of the death rates and cause of death for all Pennsylvania residents who had undergone bariatric surgery from 1995 to 2004 (n ⫽ 16,683) revealed 16 suicides [8]. When averaged, these studies have suggested a rate of 140 suicide deaths per 100,000 compared with a populationwide incidence of 14.37 suicide deaths per 100,000 [9]. More concerning, compared with a control group matched on age, gender, and body mass index (BMI), 15 suicides were identified in a bariatric surgery group (260 per 100,000) versus 5 (9 per 100,000) in a control group [4]. Although not included in these numbers, that study also identified a greater number of other accidental deaths and deaths by poisoning of undetermined intent in the bariatric
1550-7289/10/$ – see front matter © 2010 American Society for Metabolic and Bariatric Surgery. All rights reserved. doi:10.1016/j.soard.2010.08.014
A. K. Windover et al. / Surgery for Obesity and Related Diseases 6 (2010) 702–706
surgery versus control group [4]. It is possible that at least some of these accidental deaths and deaths by poisoning could have been undeclared suicides. Such research highlights the potential concern for the long-term risk of suicide among bariatric surgery patients. Although causative factors have yet to be identified, severe obesity has been associated with an increased risk of psychopathology, including depression [10,11] and suicide attempts [12]. Psychopathology and/or a history of attempted suicide, in turn, have been associated with greater risk of suicide [13]. Consequently, active or recent suicidal ideation and behavior have been commonly identified as psychological contraindications for bariatric surgery [14], because suicidal ideation and behavior are indicative of current psychiatric instability and suggest the need for emergent psychiatric care. Furthermore, such instability might make the behavioral changes and adherence more difficult to achieve. Although a history of suicidal ideation and behavior is often a part of preoperative psychological assessment [15,16], research exploring the history of attempted suicide, and its correlates and effect on weight loss outcomes in bariatric populations have been limited. To the best of our knowledge, only 2 studies have explored the prevalence of past suicide attempts among bariatric surgery patients. The first study found that 14 women who had undergone gastric bypass surgery had significantly more past suicide attempts compared with a control group of 14 morbidly obese women seeking nonoperative weight management [17]. However, 15 years later, the second study examined the lifetime prevalence of attempted suicide and self-injurious behavior in a sample of 121 bariatric surgery candidates and found 11 (9.1%) to selfreport a history of attempted suicide [18]. In contrast to the bariatric published data, studies exploring the prevalence of past suicide attempts have been more prevalent for other moderate-to-high risk surgeries, such as organ transplantation [19 –21]. A history of attempted suicide has been significantly associated with a shorter time to infection or rejection in cardiac transplant patients [20]. It has also been identified as a predictor of attenuated survival time in this population [20]. Population-based studies of nonfatal self harm have revealed a greater prevalence of suicide attempt histories among younger, previously married, white women [22]. Depression, other mental disorders, alcohol abuse, and other substance abuse have also been found to be more prevalent among those with a history of attempted suicide compared with the adult population at large [22]. The presence of similar demographic and clinical factors within weight loss surgery populations has not been previously examined. Thus, the present study examined correlates of past suicidal behavior. On the basis of epidemiologic studies of suicide, we hypothesized that those with a history of suicide would be more likely to be women, have
703
a history of psychiatric treatment, and be younger than those without this history.
Methods Participants Data from the psychological evaluation were retrospectively reviewed from the medical charts of 1020 consecutive patients evaluated at the Cleveland Clinic Bariatric and Metabolic Institute for bariatric surgery from January 2006 to August 2008. The 1020 participants were primarily women (76.6%), white (69.1%), married (47.0%), and employed (53.3%). The age range was 18 –77 years (mean 46.3 ⫾ 11.3). Their education level was 7–24 years (mean 13.8 ⫾ 2.5). The presenting BMI was 31.5–99.3 kg/m2 (mean 50.3 ⫾ 11.1). Procedures The data reviewed included past suicide attempts, demographic information, preoperative BMI, history of physical and sexual abuse, history of substance abuse/dependence, and current or historical mental health treatment, including inpatient hospitalization, psychotropic medication management, and psychotherapy. These data were obtained during semistructured preoperative psychological and behavioral assessment interviews using a template form in the electronic medical record. During the assessment of past psychiatric history, all patients were queried whether they had ever tried to kill themselves or hurt themselves in some other manner. When answered in the affirmative, follow-up questions were used to assess the number of attempts, method, and subsequent treatment. However, for the present study, patients were simply dichotomized into either the presence or absence of a suicide history. A history of suicidal ideation, plan, or intent was not considered to be positive. Only those patients who endorsed a specific attempt were dichotomized as having positive suicide history (SH⫹). Objective psychological testing (Symptom Checklist-90-R) was also administered as a part of the preoperative evaluation; however, it was not used to determine the suicide history [23]. Additionally, if patients were currently in treatment with a mental health professional or had past inpatient psychiatric hospitalization within the past 5 years, mental health records were obtained before surgical clearance. The medical chart review and patient self-report noted historical and current psychotropic medications that were confirmed at each subsequent visit by bariatric nurses. Patients reporting a psychiatric hospitalization or suicide attempt in the past year were programmatically required to demonstrate 12 months of mental health intervention and emotional stability before surgery. The institutional review board of the Cleveland Clinic approved the use of the patient data.
704
A. K. Windover et al. / Surgery for Obesity and Related Diseases 6 (2010) 702–706
Statistical analysis The data were analyzed using the Statistical Package for Social Sciences, version 17.0, (SPSS, Chicago, IL). For the purposes of the present study, a SH⫹ was defined as having ⱖ1 self-reported suicide attempt. The prevalence of suicide history in our population was determined. Demographic differences among patients with a SH⫹ (n ⫽ 115) versus patients without a suicide history (SH⫺) (n ⫽ 905) were evaluated using independent samples t tests for continuous variables and chi-square analyses for categorical variables. The results were considered significant at P ⱕ.05, unless otherwise indicated. Binary logistic regression analysis was conducted to determine the effect of a SH on the clinical variables of interest. Any sociodemographic differences between the groups were entered in the first step of the regression equation, followed by baseline BMI, and, finally, the psychosocial variables: a history of psychiatric hospitalization, a history of outpatient psychotherapy and/or psychotropic medication management, current psychotropic medication use, history of sexual abuse, history of physical abuse, and history of tobacco use. Owing to the multiple comparisons, a Bonferroni correction was applied, making the ␣ value more conservative (␣ ⫽ .004). Results Of the 1020 patients evaluated for 32 months, 115 (11.2%) reported a history of ⱖ1 suicide attempts. The SH⫹ bariatric surgery candidates were significantly younger, less educated, had a greater BMI, and were more predominantly single, female, and receiving disability compared with the SH⫺ patients. The percentages, mean values, standard deviations, and significance levels comparing the SH⫹ and SH⫺ groups are summarized in Table 1. Binary logistic regression analysis was used on a subset of 980 patients without missing data to determine the clinical correlates of SH⫹ bariatric surgery candidates, independent of sociodemographic and BMI differences. Step 1 included the following demographic variables: gender, marital status, job status, age, and education. Step 2 consisted of the preoperative BMI on program entry. Step 3 included clinical variables of interest: a history of psychiatric hospitalization, history of outpatient psychotherapy and/or psychotropic medication management, current psychotropic medication use, history of sexual abuse, history of physical abuse, and history of tobacco use. As previously noted, a SH⫹ was significantly correlated with job status but only showed a trend toward significance for age and gender after controlling for multiple comparisons. The significance levels of each step and individual predictive  weights are listed in Table 2. A significant difference was not found between the 2 groups for the baseline BMI [chi-square (1, n ⫽ 980) ⫽ 2.8, P ⬎.09].
Table 1 Comparison of sociodemographic characteristics Variable Gender Female Male Ethnicity White Black Other Marital status Married Never married Divorced Other Employment status Employed Disability Unemployed Other Surgery type Roux-en-Y LapBand Sleeve Other Age (yr) Education (yr) Preoperative BMI (kg/m2)
SH⫹ (n ⫽ 115)
SH⫺ (n ⫽ 905)
104 (90.4) 11 (9.6)
677 (74.8) 228 (25.2)
76 (66.7) 29 (25.4) 9 (7.9)
630 (70.2) 217 (24.2) 50 (5.6)
34 (29.6) 37 (32.2) 30 (26.1) 14 (12.2)
446 (49.3) 189 (20.9) 170 (18.8) 99 (11.0)
37 (32.5) 52 (45.6) 16 (14.0) 9 (7.9)
508 (56.3) 197 (21.8) 103 (11.4) 94 (10.4)
33 (71.7) 4 (8.7) 7 (15.2) 2 (4.4) 43 ⫾ 11 13.4 ⫾ 2.4 52.3 ⫾ 11.6
263 (65.9) 71 (17.8) 48 (12.0) 17 (4.3) 46.8 ⫾ 11.3 13.9 ⫾ 2.5 50 ⫾ 11
df
Chi-square 2
13.9*
2
3.0
3
22.3†
3
45.8*
3
6.5
1018 994 1015
⫺3.4† ⫺1.9‡ 2.2§
SH ⫽ suicide history; df ⫽ degrees of freedom; BMI ⫽ body mass index. Data presented as numbers, with percentages in parentheses, or mean ⫾ standard deviation. * P ⬍.0001, 2-tailed. † P ⬍.001, 2-tailed. ‡ Marginally significant at P ⬍.052. § P ⬍.05, 2-tailed.
After controlling for the significant demographic variables, a history of psychiatric hospitalization and a history of outpatient psychotherapy and/or psychotropic medication management also correlated significantly with SH⫹ [chisquare (7, n ⫽ 980) ⫽ 175, P ⬍.001; Table 2]. Trends toward significance were noted between groups for a history of substance abuse and a history of sexual abuse (Table 2). When entered with the other covariates, no significant differences between groups were found for marital status, education, preoperative BMI, current psychotropic medication use, a positive physical abuse history, and a positive history of tobacco use. Discussion The prevalence of SH⫹ in our sample was 73 times that of the population at large (age range 18 –77) [22]. Although somewhat greater, this corresponded to the findings of the limited research available demonstrating a greater prevalence of SH⫹ in bariatric populations compared with the
A. K. Windover et al. / Surgery for Obesity and Related Diseases 6 (2010) 702–706
705
Table 2 Summary of binary logistic regression analyses for sociodemographic and clinical predictors of Patients with positive suicide history (n ⫽ 980) Model Binary logistic regression 1 Step 1 Gender Marital status Job status Age Education Step 2 BMI on program entry Step 3 History of Psychiatric hospitalization Outpatient mental health treatment Sexual abuse Physical abuse Active psychiatric medication Tobacco use Substance abuse/dependence
Chi-square 50.2*
df

SE
Exp()
⫺1† .1 ⫺.2† .02† .1
.3 .1 .1 .01 .04
.3 1.1 .7 1 11
⫺.01
.009
2.3* 1.1* .5† .2 ⫺.002 ⫺.1 ⫺.6†
.3 .4 .3 .3 .3 .2 .3
5
2.8
1
175*
7
.1
10.4 3.1 1.7 .5 .9 .8 .5
SE ⫽ standard error; other abbreviations as in Table 1. * P ⬍.001. † P ⬍.05, trend toward significant.
population base rates [4 – 6]. Also consonant with these studies, a greater prevalence of SH⫹ was found among those with greater preoperative BMIs [24]. Theories for the relationship between extreme obesity and psychiatric issues have included the negative effects of social stigma on mood and the increased appetite, decreased physical activity, and the use of weight gain-promoting medication common in depression [25]. Consistent with epidemiologic studies, we found a greater prevalence of SH⫹ among younger patients, women, and primarily single/never married, but also divorced, patients. No significant differences were found for race/ethnicity. Additionally, we noted a significant relationship between SH⫹ and receiving disability; however, having less education was only marginally significant. Significant correlations found among the SH⫹ patients and a history of psychiatric hospitalization and outpatient psychotherapy and/or psychotropic medication management are congruent with epidemiologic studies reporting a greater prevalence of SH⫹ among those diagnosed with depression and other mental disorders. The increased rates of psychiatric hospitalization and outpatient mental health treatment were not surprising, because often those who have attempted suicide require medical attention and might be appropriately referred for ongoing mental health treatment at that time. The severe degree of emotional distress associated with SH⫹ also often impairs an individual’s interpersonal and occupational functioning. Thus, this might account for the relationship between SH⫹ and those receiving disability. Such data on the prevalence and psychosocial correlates supports the common practice of assessing history of attempted suicide, as well as other psychosocial variables.
The routine assessment of SH⫹ both pre- and postoperatively might inform the treatment team regarding the need for closer monitoring and the potential benefit of evidencebased psychotherapy. Given the evaluative nature of the assessment, it is quite likely that some patients chose not to disclose a past suicide attempt and were incorrectly dichotomized as SH⫺. However, it is unlikely that patients were incorrectly dichotomized as SH⫹ when they did not have a true history. Thus, the reported results might not be a true representation of all SH⫹ patients, and future research should include assessments for suicide history that are separate from the evaluative process. A second limitation of the present study was that our relatively small sample of SH⫹ patients reduced the power of our analyses. Previous research has identified a greater prevalence of suicide attempts after bariatric surgery [17,18]. Future research should examine the predictive utility of self-reported past suicide attempts on both the short- and long-term bariatric surgery outcomes, including suicide attempts after surgery. Conclusion The results of the present study have identified a greater prevalence of SH⫹ among bariatric surgery candidates compared with the population at large. The sociodemographic and clinical correlates of SH⫹ among bariatric surgery candidates include younger age, female gender, single, and receiving disability, as well as having a history of psychiatric hospitalization, outpatient psychotherapy and/or psychotropic medication management, sexual abuse, and substance abuse/dependence. Additional research is
706
A. K. Windover et al. / Surgery for Obesity and Related Diseases 6 (2010) 702–706
needed to examine the predictive utility of SH⫹ on bariatric surgery outcomes.
[12]
Disclosures [13]
The authors have no commercial associations that might be a conflict of interest in relation to this article. [14]
References [15] [1] Buchwald H. Consensus conference statement: bariatric surgery for morbid obesity: health implications for patients, health professionals, and third-party payers. Surg Obes Relat Dis 2005;1:371– 81. [2] Elder KA, Wolfe BM. Bariatric surgery: a review of procedures and outcomes. Gastroenterology 2007;132:2253–71. [3] Dymek M, le Grange D, Neven K, Alverdy J. Quality of life and psychosocial adjustment in patients after Roux-en-Y gastric bypass: a brief report. Obes Surg 2001;11:32–9. [4] Adams T, Gress R, Smith S, et al. Long-term mortality after gastric bypass surgery. N Engl J Med 2007;357:753– 61. [5] Waters GS, Pories, WJ, Swanson MS, et al. Long-term studies of mental health after the Greenville gastric bypass operation for morbid obesity. Am J Surg 1991;161:154 –7. [6] Hsu G, Benotti P, Dwyer J, et al. Nonsurgical factors that influence the outcome of bariatric surgery: a review. Psychosom Med 1998;60: 338 – 46. [7] Higa K, Boone K, Ho T. Complications of the laparoscopic Rouxen-Y gastric bypass: 1,040 patients—what have we learned? Obes Surg 2000;10:509 –13. [8] Omalu BI, Ives DG, Buhari AM, et al. Death rates and causes of death after bariatric surgery for Pennsylvania residents, 1995 to 2004. Arch Surg 2007;142:923–28. [9] Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Web-based injury statistics query and reporting system (WISQARS). Available at: www.cdc.gov/ncipc/ wisqars. Accessed September 3, 2009. [10] Mitchell JE, deZwaan M. Bariatric surgery: a guide for mental health professionals. New York: Routledge Taylor & Francis Group; 2005. p. 1–158. [11] Petry NM, Barry D, Pietrzak RH, Wagner JA. Overweight and obesity are associated with psychiatric disorders: results from the Na-
[16] [17] [18]
[19]
[20]
[21] [22]
[23]
[24]
[25]
tional Epidemiological Survey on alcohol and related conditions. Psychosom Med 2008;70:288 –97. Dong C, Li WD, Li D, Price R. Extreme obesity is associated with attempted suicides: results from a family study. Int J Obes 2000;30: 388 –90. Jacobs D, Brewer M. APA practice guideline: provides recommendations for assessing and treating patients with suicidal behaviors. Psychiatric Ann 2004;34:373– 80. Bauchowitz AU, Gonder-Frederick LA, Olbrisch ME, et al. Psychosocial evaluation of bariatric surgery candidates: a survey of present practices. Psychosom Med 2005;67:825–3. LeMont D, Moorehead M, Parish M, Reto C, Ritz S. Suggestions for the pre-surgical psychological assessment of bariatric surgery candidates. Durham, NC: Duke University Medical Center; 2004. Powers P, Rosemurgy A, Coovert D, Boyd F. Psychological sequelae of bariatric surgery: a pilot study. Psychosomatics 1988;29:283– 88. Rosen L, Aniskiewicz A. Psychosocial functioning of two groups of morbidly obese patients. Int J Obes 1983;7:53–9. Sansone R, Wiederman MW, Schumacher D, Routsong-Weichers L. The prevalence of self-harm behaviors among a sample of gastric surgery candidates. J Psychosom Res 2008;65:441– 44. Levenson J, Olbrisch M. Psychosocial evaluation of organ transplant candidates: a comprehensive survey of process, criteria, and outcomes in heart, liver, and kidney transplantation. Psychosomatics 1993;34:314 –23. Owen J, Bonds C, Wellisch D. Psychiatric evaluations for heart transplant candidates: predicting post-transplant hospitalizations, rejection episodes, and survival. Psychosomatics 2006;47:213–22. Reither A, Mahler E. Suicide in liver transplant patients. Psychosomatics 1994;35:574 –77. National Institute of Mental Health. Suicide in the U.S.: statistics and prevention. 2007. Available at: www.nimh.nih.gov/health/publications/ suicide-in-the-us-statistics-and-prevention. Accessed July 20, 2008. Derogatis, LR. Symptom Checklist-90-R Administration, Scoring, and Procedures Manual. Minneapolis: National Computer Systems; 1994. p. 1–123. Mather A, Cox B, Enns M, Sareen J. Associations of obesity with psychiatric disorders and suicidal behaviors in a nationally representative sample. J Psychosom Res 2009;66:277– 85. Simon GE, Von Korff M, Saunders K, et al. Association between obesity and psychiatric disorders in the US adult population. Arch Gen Psychiatry 2006;63:824 –30.