- Email: [email protected]
Using the presurgical psychological evaluation to predict 5-year weight loss outcomes in bariatric surgery patients
Author’s Accepted Manuscript Using the pre-surgical psychological evaluation to predict 5-year Weight loss outcomes in bariatric surgery patients Ryan J. Marek, Yossef S. Ben-Porath, Manfred van Dulmen, Kathleen Ashton, Leslie J. Heinberg www.elsevier.com/locate/buildenv
PII: DOI: Reference:
S1550-7289(16)30811-5 http://dx.doi.org/10.1016/j.soard.2016.11.008 SOARD2826
To appear in: Surgery for Obesity and Related Diseases Cite this article as: Ryan J. Marek, Yossef S. Ben-Porath, Manfred van Dulmen, Kathleen Ashton and Leslie J. Heinberg, Using the pre-surgical psychological evaluation to predict 5-year Weight loss outcomes in bariatric surgery patients, Surgery for Obesity and Related Diseases, http://dx.doi.org/10.1016/j.soard.2016.11.008 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 Using the Pre-Surgical Psychological Evaluation to Predict 5-Year Weight Loss Outcomes in Bariatric Surgery Patients
Ryan J. Marek, M.A., Yossef S. Ben-Porath, Ph.D., Manfred van Dulmen. Ph.D. Department of Psychological Sciences Kent State University Kent, OH 44242
Kathleen Ashton, Ph.D., Leslie J. Heinberg, Ph.D. Cleveland Clinic Lerner College of Medicine 9500 Euclid Avenue/M61 Cleveland, OH 44195 United States of America
Conflict of Interest: Yossef Ben-Porath is a paid consultant to the MMPI-2-RF publisher, the University of Minnesota and Distributor, Pearson. As co-author of the MMPI-2-RF, he receives royalties on sales of the test. Disclosure: Portions of this manuscript were accepted to the ASMBS Masters in Behavioral Health Course at Obesity Week in New Orleans, LA (November, 2016) and were used to satisfy oral and written requirements for the first author’s dissertation at Kent State University. Correspondence: All correspondence should be addressed to: Ryan J. Marek, Department of Psychological Sciences, Kent State University, Kent, OH 44242. Email: [email protected]
2 Abstract Background: Psychosocial factors contribute to poorer weight loss outcomes following bariatric surgery; however, findings on associations between preoperative psychiatric diagnoses, psychological testing, and weight loss are inconsistent. Objectives: Examine associations between pre-surgical psychiatric diagnoses derived from a semi-structured clinical interview and test scores from the Minnesota Multiphasic PersonalityInventory – 2 – Restructured Form (MMPI-2-RF) and 5-year Body Mass Index (BMI) outcomes. Setting: Cleveland Clinic Bariatric and Metabolic Institute Methods: 446 consecutively, consented patients who underwent a Roux-en-Y Gastric Bypass (RYGB) at least 5-years prior were included in the study. A majority were women (74.2%) and Caucasian (66.2%). Patients’ mean pre-surgical BMI was 49.14 kg/m2 [Standard Deviation (SD) = 9.50 kg/m2]. Psychiatric diagnoses were obtained from a pre-surgical, semi-structured clinical interview and all participants were administered the MMPI-2-RF at their pre-surgical evaluations. BMIs were collected at 4 post-operative time points across a 5-year trajectory. This prospective design utilized latent growth curve modeling. Results: Older individuals evidenced a slower rate of BMI reduction over time. A pre-surgical diagnosis of Binge Eating Disorder predicted higher BMIs at the 5-year outcome. Scores on MMPI-2-RF measures of emotional and behavioral dysfunction domains incrementally predicted poorer weight loss outcomes. Conclusions: Pre-operative indicators of psychopathology, notably indicators that are dimensional in nature, are important in predicting post-operative outcomes. Closer follow-up with patients who evidence pre-surgical psychological factors, both before and after surgery, may help improve outcomes.
3 Keywords: Obesity, Bariatric Surgery, Assessment, Weight Loss, Outcome, Psychology,
Introduction Bariatric surgery results in significant weight loss and resolution of comorbid conditions for patients with severe obesity, demonstrating good efficacy as far as ten years following the procedure.1,2 However, weight loss begins to plateau around 12 months following surgery, and a subset of patients begin to show weight regain from that point forward (though some patients begin regaining weight around the 6-month outcome).3-7 Pre-surgical behavioral and psychosocial factors, including internalizing (i.e., anhedonia, anxiety, low frustration tolerance) and externalizing (i.e., impulsivity, sensation seeking, substance use) psychopathology, maladaptive eating behaviors, and poor behavioral adherence have been linked to poorer weight loss outcomes following bariatric surgery.8-17 Due in part to this association, a number of obesity-related societies recommend conducting pre-surgical psychological evaluations as part of patients’ pre-surgical regimen.5,18 A primary goal of these evaluations is to identify psychosocial factors that could impede outcomes and recommend appropriate pre-surgical and/or postoperative interventions to help the patient achieve and maintain optimal results. Despite numerous studies demonstrating evidence that pre-surgical psychopathology predicts poorer weight loss outcomes, 8-13 much of the literature on the role of pre-surgical internalizing and externalizing psychopathology in predicting poorer weight loss outcomes is mixed. For example, some studies reported that pre-surgical measures of depression predicted poorer short-term outcomes16,17,19 whereas others found no association between measures of
4 depression and short-term outcomes.12,20 Some studies indicate that pre-surgical externalizing behaviors, such as binge and graze eating are predictive of poorer weight loss outcomes,10 whereas other studies do not support those associations.21 A meta-analysis11 reported that presurgical variables, such as previous weight loss attempts, binge/sweet/maladaptive eating habits, depression, anxiety, history of sexual abuse, self-esteem, past/current alcohol abuse/use, and other psychiatric disorders do not predict poorer outcomes. Another integrative review of the literature8 implied that reliance on the current diagnostic system in addition to methodological differences across studies likely contributed to the inconsistent findings reported. For example, some depression self-report measures have inadequate construct validity when differentiating between distress and low positive emotionality (a core feature of those diagnosed with Major Depression Disorder). Likewise, the categorical-polythetic approach to classifying mental illness used worldwide assumes that psychopathology can be classified into distinct categories – where meeting a diagnostic threshold is a matter of having a sufficient number of symptoms associated with the diagnosis. However, the diagnostic system is polythetic in that different combinations of symptoms lead to the same diagnosis. The authors8 encouraged use of theoretically-derived, dimensional assessments of the hierarchical model of psychopathology (which are congruent with constructs outlined in RDoC) to better establish consistency when reporting associations with outcome criteria. Use of a conceptually-grounded instrument such as the Minnesota Multiphasic Personality Inventory – 2 – Restructured Form (MMPI-2-RF), 22,23 which is dimensional in nature and structured in congruence with the hierarchical model of psychopathology will likely offer more consistent prediction coefficients.
5 We sought to further establish the utility of pre-surgical psychological evaluations by examining mid-term (5-year) weight loss results in a sample of Roux-en-Y (RYGB) patients. The multi-method assessment practice conducted at the hospital made it possible to explore associations between findings in a medical chart review, a semi-structured clinical interview, and internalizing and externalizing scale scores from the MMPI-2-RF with 5-year BMI outcomes. Based on prior literature, no specific psychiatric diagnosis consistently predicts poorer weight loss outcomes; 8-17 therefore, all diagnoses and chart review information (e.g., history of abuse, number of psychiatric medications, etc) were tested for exploratory purposes. Lastly, MMPI-2RF scale assessing internalizing and externalizing dysfunction were added to the prediction model to test the incremental contribution of an objective personality assessment instrument that adheres to contemporary dimensional conceptualizations of psychopathology in the pre-surgical psychological evaluation. Materials and Methods Participants This study was prospective in nature, relying on an ongoing large-scale data collection. Data from 451 consecutively consented patients who underwent RYGB at least 5-years prior and resided in Northeast Ohio were available for the study. Prior to the pre-surgical evaluation, patients were asked to consent to have their medical and psychological data and used for research purposes. No incentives to participate were offered. If patients visited the medical center over the course of the 5-years, their weights were accessible from their electronic medical records. Of these individuals, 5 (0.11%) were excluded from further analyses because they produced an invalid MMPI-2-RF protocol based on guidelines outlined in the MMPI-2-RF Technical Manual23: Cannot Say > 18, Variable Response Inconsistency > 80, True Response
6 Inconsistency > 80, Infrequent Responses > 120, & Infrequent Psychopathology Responses > 100. There were no statistically significant differences on demographic or psychological variables between those who produced valid or invalid protocols, though in larger, better powered samples of bariatric surgery patients, lower education levels have been associated with invalid protocols in this setting.24 Of the 446 participants retained for analyses, 74.2% were women. The sample included 66.2% Caucasians, 19.7% African Americans, and 14.1% other ethnicities. The mean age was 46.75 years [Standard Deviation (SD) = 11.63; Range 18 – 74]. Patients’ mean pre-surgical BMI was 49.14 kg/m2 (SD = 9.50 kg/m2). The hospital’s Institutional Review Board approved the study.
Measures Semi-Structured Psycho-Diagnostic Interview. A doctoral level psychologist or a supervised post-doctoral fellow administered a semi-structured interview that is mandated by the Cleveland Clinic at initial intake. Information obtained during the interview included: an evaluation of Diagnostic and Statistical Manual – Fourth Edition – Text Revision (DSM-IVTR)25 diagnoses and mental health history/treatment, current mental health diagnoses/treatment, past and current substance use/abuse/dependence, physical and sexual abuse history, and history of past suicide attempts. It is important to note that this semistructured interview has not been validated against other published structured interviews. However, the interview template requires that clinicians assess for symptoms associated with all depressive, anxiety, psychotic bipolar, eating, and substance use-related disorders. BMI was recorded from the electronic medical record at the time of their evaluation, whereas postoperative BMIs were gathered within a 1-month range of the post-operative time point.
7 Although reliability data (e.g., inter-rater reliability) were not available for the interview, the information obtained has demonstrated good convergent and discriminant validity with psychopathology self-report measures.26 These data were coded through a retrospective chart review by trained research assistants and were then double entered to maintain consistency and reduce typographical error. Inter-rater reliability (Kappa statistics) between coders was .96 (Range = .81-1.00). Minnesota Multiphasic Personality Inventory – 2 – Restructured Form (MMPI-2RF).22,23 The MMPI-2-RF consists of 338 true-false items that are scored on 9 measures of protocol validity and 42 substantive scales that assess psychological constructs. The instrument has been recommended as an objective psychological measure in bariatric surgery evaluations because of the scale scores’ psychometric properties, replicated normative data, and associations with outcomes in bariatric surgery settings.27
Procedure A doctoral level psychologist or a supervised post-doctoral fellow administered a semistructured interview that is given at the Cleveland Clinic at initial intake. Information obtained during the interview included: an evaluation of Diagnostic and Statistical Manual – Fourth Edition – Text Revision (DSM-IV-TR)25 diagnoses (DSM-5 criteria was used for BED) and mental health history/treatment, current mental health diagnoses/treatment, past and current substance use/abuse/dependence, physical and sexual abuse history, and history of past suicide attempts. It is important to note that this semi-structured interview has not been validated against other published structured interviews. Like a SCID, it uses initial prompt queries for the following disorders: eating disorders, depression, bipolar, psychosis, generalized anxiety, panic, obsessive-compulsive, substance abuse/dependence and Post-Traumatic Stress Disorder. DSM-5
8 criteria for binge eating disorder with all symptoms (whether or not initial queries are positive) are elicited for all patients. Similarly, the research criteria for Night Eating Syndrome28 are all queried. Statistical Analyses All analyses were computed in Mplus 7.3.29 Latent growth curve analyses (LGCAs) were estimated using BMIs across time.30 In the current investigation, the intercept was specified to reflect differences at the 5-year post-operative time point (termed “5-Year BMIs”) whereas the slope (termed “BMI reduction over time”) reflected the rate of change from the pre-surgical BMIs through the 5-year BMIs. Conditional LGCA’s were then used to test whether pre-surgical variables predicted 5-Year BMIs and BMI reduction over time. Models were estimated using Robust Maximum Likelihood, as this method is preferred when variables that are non-normally distributed are in the analysis models. Full Information Maximum Likelihood (FIML) with age included in the covariance matrix was used to handle missing data across time.31 The following recommended goodness-of-fit indexes were used to evaluate the adequacy of the models tested:32 the Root Mean Square Error of Approximation (RMSEA), the Comparative Fit Index (CFI), the Standardized Root Mean Square (SRMR), and the Bayesian Information Criterion (BIC). Additionally, Chi Square (χ2) difference testing33 was used to compare models, with a statistically significant chi-square value indicating that the more complex model fits the data better than a less restrictive one. Standardized Beta Weights (β) were reported for path coefficients. Prediction analyses were conducted hierarchically. In other words, BED and MDD were entered into the model first. Following, other psychiatric diagnoses and medical chart review information were tested. Next, hypothesized MMPI-2-RF Substantive Scale Scores were entered based on scale set.
9 Results Descriptive Statistics and Preliminary Analyses Descriptive statistics for the variables used in subsequent analyses are reported in Table 1, which contains means and standard deviations for continuous variables, prevalence percentage for categorical variables, and percent missing for all variables used in the subsequent analyses. Importantly, 38.1% of the sample engaged in an under-reporting response style on the MMPI-2RF [Uncommon Virtues > 65T or Adjustment Validity > 60T]. Although a portion of these individuals’ scores may reflect traditional upbringing or being psychologically well-adjusted, 33% scored at or above the highest recommended interpretative T-Score cut-offs on these scales [Uncommon Virtues > 80T or Adjustment Validity > 70T], which implied that under-reporting was very likely in this sample. BMI means and standard deviations across time as well as the percentage of missing data were similar to weight loss outcomes reported in other longitudinal RYGB studies.3,6 Correlations between demographic or psychological variables and number of time points with missing data over time indicated that younger individuals had higher amounts of missing data over time (r = -.18, p < .001). No other statistically significant correlation coefficients were observed between demographic or psychological variables and attrition over time. Latent Growth Curve Modeling The first LGCA tested the appropriateness of a linear model (i.e., weight loss was a continuous, negative linear trajectory). Model fit statistics of linear and non-linear models are located in Supplemental Table 1. The best overall model fit for BMI change over the 5-year trajectory was a non-linear model with correlated error variances [χ2(9) = 19.15, p = .02, RMSEA = .05, CFI = .96, SRMR = .07, BIC = 13706.88]. The mean of the slope was negative (-
10 10.40 kg/m2) indicating that on average patients reduced 10.40 kg/m2 of their BMI over the 5year time span. Variance statistics around the 5-Year BMI estimate (Variance = 49.78, p < .001) and BMI-Reduction over time (Variance = 12.96, p = .031) were both statically significant, indicating considerable variability in patients’ BMIs both at the 5-year post-operative time point and the rate at which patients lose weight over time. The correlation between pre-surgical BMIs and BMI-reduction over time was statistically significant (r = -.62, p < .001), indicating that patients with higher pre-surgical BMIs evidenced a faster rate of change than did patients with lower starting BMIs. All subsequent analyses controlled for this trend. The non-linear model was then estimated using age as a predictor of 5-Year BMIs and BMI-Reduction over time. Model fit improved [χ2(7) = 12.36, p = .089, RMSEA = .041, CFI = .98, SRMR = .07, BIC = 10245.35] and demonstrated an incremental fit for the data [Δχ2(2) = 7.48, p = .024, ΔBIC = 3461.53]. Moreover, younger individuals had statistically higher BMIs at the pre-surgical evaluation than did older patients in this sample (β = -.09, p = .049), though there were no differences in age and BMI by the five-year outcome (β = -.01, p = .807). However, BMI reduction across time varied as a function of age, such that older adults tended to have a slower rate of BMI reduction across time than did younger adults (β = .20, p <.001). A depiction of this structural equation model is located in Supplemental Figure 1. Prediction Analyses Conditional LCGA analyses were conducted to estimate whether pre-surgical psychopathology predicted 5-year post-surgical BMIs as well as the rate of BMI-change over the 5-year trajectory. Medical chart review information was tested first, followed by diagnoses, and then MMPI-2-RF scales by scale set. Variables were retained in the model if they significantly predicted 5-year BMIs or BMI reduction over time. History of suicide attempts, number of
11 psychotropic medications, history of outpatient therapy, history of physical and sexual abuse, and history of smoking cigarettes were not predictive of 5-year BMIs or BMI reduction over time. Among psychiatric diagnoses, a pre-surgical diagnosis of BED predicted higher BMIs at the 5-year outcome (β = .16, p = .008). No other diagnoses were predictive of 5-year BMIs or BMI reduction over time. Patients who scored higher on the MMPI-2-RF scales Behavioral/Externalizing Dysfunction (β = .11, p = .030), Low Positive Emotions (β = .13, p = .032), and Hypomanic Activation (β = .13, p = .028) also evidenced higher BMIs at the 5-year outcome after controlling for age and BED. In regards to predicting variability in BMI-reduction over time, patients who scored higher on MMPI-2-RF scales Hypomanic Activation (β = .24, p = .002), Anger Proneness (β = .16, p = .004), and Activation (β = .17, p = .036) prior to surgery evidenced a slower rate of BMI-reduction over 5-years after controlling for age, BED, and the correlation between pre-surgical BMI and BMI reduction over time.
Discussion The goal of the current investigation was to examine whether categorical psychiatric diagnoses predicted BMI trajectories over a five-year post-bariatric surgery course (above and beyond known factors such as age) and whether, in turn, dimensional facets of psychopathology added predictive utility. It was hypothesized that dimensional measures of psychopathology, as measured on the MMPI-2-RF, would be incrementally predictive of 5-year weight loss outcomes above and beyond psychiatric diagnoses and demographic variables. As expected, older patients and those who obtained surgery at a lower BMI lost weight at a slower rate. Although many psychiatric diagnoses did not predict outcome, those who met criteria for BED prior to surgery began evidencing higher BMIs by 5-years post-surgery. Also, as hypothesized, scale scores such
12 as Low Positive Emotions (a measure of patients’ difficulty feeling positive emotions, lacking interest and energy, and displaying vegetative symptoms associated with Major Depressive Disorder) and Behavioral/Externalizing Dysfunction (a broad measure of disinhibition, hypomania, substance use, aggression, etc.) predicted higher BMIs 5-years post-surgery after controlling for various demographic variables and psychiatric diagnoses in the model. Taken together, the current investigation provided support for the utility of using empirically-grounded, objective psychological assessments during the pre-surgical evaluation. Results indicated that BMI-reduction over time is non-linear in that most individuals tended to evidence a decrease in BMI up to the one-year follow-up and then began regaining a small amount of weight by the 5-year outcome. Fit indices improved when non-linear trajectories were modeled as a function of age. The model in this investigation implied that older adults tended to have a slower rate of BMI-reduction over time. This finding is comparable to past research suggesting that, although older adults benefit from bariatric surgery (e.g., improved quality of life, reduction of medical comorbities, etc), they tend not to lose as much weight as younger adults. 34-36 Research implies that physical frailty and resistance to changing lifelong diet and activity habits are associated with a slower rate of BMI-change over time.37 Helping older adults find and engage in physical activities (such as water aerobics) as well as helping them identify the benefits of engaging in exercise may help accelerate their weight loss trajectories after bariatric surgery. In the current investigation, a pre-surgical diagnosis of BED was the only psychiatric diagnosis that was predictive of higher 5-year BMIs. This finding is consistent with other studies indicating that a pre-surgical diagnosis of BED is associated with poorer weight loss outcomes.3840
Moreover, many patients who meet diagnostic criteria for BED prior to surgery are more likely
13 to engage in graze eating behaviors and report having a sense of loss of control over eating, which in turn, can influence post-operative weight loss.41-43 However, other studies have reported no association between BED and weight loss after bariatric surgery.16,44 It is important to note that all patients diagnosed with BED had to complete binge eating treatment and show a positive benefit prior to surgery. Therefore, findings in the current investigation are likely underestimates of the association between untreated pre-surgical BED and medium-term outcomes. Moreover, patients are not able to necessarily consume large amounts of food, but may have consumed higher calorie/surgery/fatty foods in smaller amounts. Therefore, emphasis on conducting postoperative assessments of broader dietary adherence criteria (including grazing and feelings of loss of control over eating) and how it moderates weight loss outcomes is recommended. Studies that have examined the ability of post-surgical variables (such as dietary adherence, loss of control, executive functioning) to predict poorer outcomes have collectively found that factors associated with externalizing psychological dysfunction were related to poorer weight loss outcomes in this population. 14,45-48 In regards to weight, patients who scored higher on Behavioral/Externalizing Dysfunction scales on the MMPI-2-RF evidenced poorer one-year weight loss outcomes in a sample of gastric bypass patients.12 The current investigation extends these findings to the 5-year post-surgical bariatric surgery pattients. These results indicate that patients who evidence externalizing psychopathology prior to surgery may benefit from interventions designed to assist them gain better impulse-control. However, individuals with externalizing disorders tend not to respond as well to treatment.49 Therefore, maintaining postsurgical contact with patients who evidence greater pre-surgical externalizing psychopathology may lead to imrpoved ouctomes for these individuals.
14 In the internalizing domain, MMPI-2-RF scale scores on Low Positive Emotions were associated with higher BMIs at the 5-year follow-up, whereas higher scores on Anger-Proneness were associated with a slower rate of BMI-reduction over time. Low Positive Emotions encompasses the inability to experience positive emotions, anhedonia, pessimism, and introversion, and is a good predictor of Major Depressive Disorder.22,50,51 If a patient is experiencing symptoms such as anhedonia, they may be less likely to engage in physical activity. Individuals more prone to low positive emotionality are also less likely to enagage in the treatment process,22 and may similarly fail to engage in activities that promote weight loss maintenace. Helping patients gain insight into the benefits of exercise both prior to and following surgery lead to their becoming more active and achieving and maintaining good weight loss outcomes. In regards to Anger Proneness, patients with higher scores tend to experience irritability and low frustration tolerance. These patients may become discouraged at various postoperative timepoints and pre- and post-surgical treatment for patients who exhibit poor frustration tolerance should focus on achiegving and maintaining small goals throughout their surgical recovery. In an integrative review of the literature, Marek, Ben-Porath, and Heinberg8 discuss the limitations of relying on atheoretical diagnosistic systems and measures to predict outcomes in this setting. The authors review the problem of comorbitiy and research on the dimensional hierarchical model of psychopathology, which is similar to the constructs in the Research Domain Criteria (RDoC) outlined by the National Institute of Mental Health (NIMH)52 The current investigation is important because we would have been erroneously concluded that internalizing psychopathology was not associated with poorer weight loss outcomes had traditional psychiatric diagnoses been the sole outcome predictors. A strength to the MMPI-2-RF
15 is that it is conceptually and empirically linked to the hierarchical model of psychopathology. The incremental predictive utility obtained with this instrument in the current investigation lends support both to use of the MMPI-2-RF as part of a presurgical evaluation and the utility of the hierarchical model in short and longer-term weight-loss prediction.12,19 The current investigation had some limitations. BMI was the only dependent variable available, thus limiting the ability to predict how other revelavant outcomes such as postoperative psychopathology, eating behaviors, and quality of life affect weight loss across time. Furthermore, data on other factors that may have contributed to higher BMIs by the 5-year outcome, such as whether additional non-bariatric procedures had impacted their weight or whether they were taking medications that may affect their weight were not collected. Moreover, data on medical diagnoses (such as Type 2 Diabetes) were also not available. Another limitation is that the current study was composed of a sample who tended to under-report psychopatology and behaviors prior to surgery. If a patient engages in an under-reporting response style on the MMPI-2-RF, they do so across the evaluation, as evidenced in other samples.53 When underreporting is evidenced, prediction coefficients are attenuated and there is currently no validated method to correct for range restriction due to under-reporting in structural equation modeling. The sample was composed of individuals who, if presented with psychosocial risk factors, had to adhere to pre-surgical treatment and show a positive treatment response. Therefore, prediction coefficients reported in the current investigation are likely underestimates. Follow up studies are needed to better examine how pre-surgical treatment changes mediate outcomes. Lastly, percent excess weight loss (%EWL) is reported in Table 1, but was not included in the latent growth curve models for a few reasons. Modeling longitudinal latent growth curve analyses using %EWL would mean that there is no variability at the starting time point. A latent model with an
16 indicator that has zero variability cannot be modeled. Therefore, the latent growth curve would need to start at the 3-month time point and the correlation between the intercept and the slope would be quite different as compared to a correlation between a baseline measurement and its rate of change over time. Nonetheless, the current investigation had several strengths. The current investigation included examining outcome at the 5-year post-operative time point. Many outcome studies have focused on shorter-term outcome time points when using pre-surgical psychological assessments to predict weight loss outcomes. Lastly, Full Information Maximum Likelihood was used to handle missing data in the analyses which helped reduce bias in the predication coefficients. Future studies should integraeg post-operative psychological assessements to better account for how these psychological construct moderate weight loss. Conclusions Overall, bariatric surgery weight loss outcomes as a result of a RYGB indicate that most patients evidence a similar rate of BMI-reduction up until the one-year, post-surgical follow-up. From that point to the five-year post-surgical, most patients evidence a small amount of weight gain. BMI reduction across time was slower in older patients and patients with a lower BMI at the time of the evaluation. Additional variability in five-year BMI outcomes was accounted for by a pre-surgical diagnosis of BED and higher scores on some Internalizing and Externalizing scales of the MMPI-2-RF. Patients who were more impulsive, showed lower frustration tolerance, or were more anhedonic, pessmistic, and introverted evidenced higher BMIs at the 5year outcome than patients who did not evidence these symptoms/behaviors. Clinicians who observe symptoms, features, or behaviors in line with these constructs should reinforce the importance of follow-up visits, focusing on enhancing patient's self-control as a target for
17 intervention, and more closely monitor patients who may be less engaged or easily discouraged throughout the post-surgical process.
References
1. 2. 3. 4. 5.
6. 7.
8. 9.
10.
11. 12.
13.
14.
Brethauer SA, Chand B, Schauer PR. Risks and benefits of bariatric surgery: current evidence. Cleveland Clinic Journal of Medicine. 2006;73(11):993-1007. Valezi AC, de Almeida Menezes M, Mali J, Jr. Weight Loss Outcome After Roux-en-Y Gastric Bypass: 10 Years of Follow-up. Obes Surg. 2013;23(8):1290-1293. Magro DO, Geloneze B, Delfini R, Pareja BC, Callejas F, Pareja JC. Long-term weight regain after gastric bypass: a 5-year prospective study. Obes Surg. 2008;18(6):648-651. O'Brien PE, Dixon JB, Brown W. Obesity is a surgical disease: Overview of obesity and bariatric surgery. Anz J Surg. 2004;74(4):200-204. Mechanick JI, Youdim A, Jones DB, et al. Clinical practice guidelines for the perioperative nutritional, metabolic, and nonsurgical support of the bariatric surgery patient--2013 update: cosponsored by American Association of Clinical Endocrinologists, the Obesity Society, and American Society for Metabolic & Bariatric Surgery. Surg Obes Relat Dis. 2013;9(2):159-191. Sjostrom L. Bariatric surgery and reduction in morbidity and mortality: experiences from the SOS study. Int J Obes. 2008;32:S93-97. Courcoulas AP, Christian NJ, Belle SH, et al. Weight change and health outcomes at 3 years after bariatric surgery among individuals with severe obesity. JAMA : the journal of the American Medical Association. 2013;310(22):2416-2425. Marek RJ, Ben-Porath YS, Heinberg LJ. Understanding the role of psychopathology in bariatric surgery outcomes. Obes Rev. 2016;17(2):126-141. Meany G, Conceicao E, Mitchell JE. Binge eating, binge eating disorder and loss of control eating: effects on weight outcomes after bariatric surgery. Eur Eat Disord Rev. 2014;22(2):87-91. Conceição EM, Utzinger LM, Pisetsky EM. Eating disorders and problematic eating behaviours before and after bariatric surgery: characterization, assessment and association with treatment outcomes. Eur Eat Disord Rev. 2015;23(6):417-425. Livhits M, Mercado C, Yermilov I, et al. Preoperative predictors of weight loss following bariatric surgery: systematic review. Obes Surg. 2012;22(1):70-89. Marek RJ, Tarescavage AM, Ben-Porath YS, Ashton K, Merrell Rish J, Heinberg LJ. Using presurgical psychological testing to predict 1-year appointment adherence and weight loss in bariatric surgery patients: predictive validity and methodological considerations. Surg Obes Relat Dis. 2015;11:1171-1181. Belanger SB, Wechsler FS, Nademin ME, Virden III TB. Predicting outcome of gastric bypass surgery utilizing personality scale elevations, psychosocial factors, and diagnostic group membership. Obes Surg. 2010;20(10):1361-1371. Conceição EM, Mitchell JE, Engel SG, Machado PP, Lancaster K, Wonderlich SA. What is “grazing”? Reviewing its definition, frequency, clinical characteristics, and impact on
18
15. 16.
17.
18. 19.
20.
21.
22.
23.
24.
25. 26.
27.
28. 29. 30.
bariatric surgery outcomes, and proposing a standardized definition. Surg Obes Relat Dis. 2014;10(5):973-982. Odom J, Zalesin KC, Washington TL, et al. Behavioral predictors of weight regain after bariatric surgery. Obes Surg. 2010;20(3):349-356. 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(4):544-549. Semanscin-Doerr DA, Windover A, Ashton K, Heinberg LJ. Mood disorders in laparoscopic sleeve gastrectomy patients: does it affect early weight loss? Surg Obes Relat Dis. 2010;6(2):191-196. Sogg S, Lauretti J, West-Smith L. Recommendations for the Pre-Surgical Psychosocial Evaluation of Bariatric Surgery Patients. Surg Obes Relat Dis. 2016. Marek RJ, Ben-Porath YS, Merrell J, Ashton K, Heinberg LJ. Predicting one and three month postoperative Somatic Concerns, Psychological Distress, and Maladaptive Eating Behaviors in bariatric surgery candidates with the Minnesota Multiphasic Personality Inventory-2 Restructured Form (MMPI-2-RF). Obes Surg. 2014;24(4):631-639. 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):61-68. Wadden TA, Faulconbridge LF, Jones‐ Corneille LR, et al. Binge eating disorder and the outcome of bariatric surgery at one year: a prospective, observational study. Obesity. 2011;19(6):1220-1228. Ben-Porath YS, Tellegen A. The Minnesota Multiphasic Personality Inventory-2 Restructured Form (MMPI-2-RF): Manual for administration, scoring, and interpretation. . Minneapolis, Minnesota: University of Minnesota Press; 2008/2011. Tellegen A, Ben-Porath YS. The Minnesota Multiphasic Personality Inventory-2 Restructured Form (MMPI-2-RF): Technical Manual. Minneapolis. Minneapolis, Minnesota: University of Minnesota Press; 2008/2011. Tarescavage AM, Windover A, Ben-Porath YS, et al. Use of the MMPI-2-RF suicidal/death ideation and substance abuse scales in screening bariatric surgery candidates. Psychol Assess. 2013;25(4):1384-1389. Association AP. Diagnostic and statistical manual of mental disorders. 4th ed., text rev. ed. Washington, D.C.: American Psychiatric Publishing; 2000. Marek RJ, Ben-Porath YS, Windover A, et al. Assessing psychosocial functioning of bariatric surgery candidates with the Minnesota multiphasic personality inventory-2 restructured form (MMPI-2-RF). Obes Surg. 2013;23(11):1864-1873. Marek RJ, Heinberg LJ, Lavery M, Merrell Rish J, Ashton K. A review of psychological assessment instruments for use in bariatric surgery evaluations. Psychol Assess. 2016;28(9):1142-1157. Allison KC, Lundgren JD, O'Reardon JP, et al. Proposed diagnostic criteria for night eating syndrome. The International journal of eating disorders. 2010;43(3):241-247. Muthén LK, Muthén BO. Mplus User's Guide. Seventh ed. Los Angeles, CA: Muthén & Muthén; 1998-2012. Duncan TE, Duncan SC. An introduction to latent growth curve modeling. Behav Ther. 2004;35(2):333-363.
19 31.
32. 33. 34. 35.
36. 37. 38.
39.
40. 41.
42.
43.
44.
45.
46.
47.
Enders CK. The impact of nonnormality on full information maximum-likelihood estimation for structural equation models with missing data. Psychol Methods. 2001;6(4):352. Hu Lt, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Modeling.1999;6(1):1-55. Satorra A. Scaled and adjusted restricted tests in multi-sample analysis of moment structures. Springer; 2000. Sugerman HJ, DeMaria EJ, Kellum JM, Sugerman EL, Meador JG, Wolfe LG. Effects of Bariatric Surgery in Older Patients. Annals of Surgery. 2004;240(2):243-247. Villareal DT, Apovian CM, Kushner RF, Klein S. Obesity in older adults: technical review and position statement of the American Society for Nutrition and NAASO, The Obesity Society. Obesity research. 2005;13(11):1849-1863. Villareal DT, Chode S, Parimi N, et al. Weight loss, exercise, or both and physical function in obese older adults. N Engl J Med. 2011;364(13):1218-1229. Elia M. Obesity in the elderly. Obes Res. 2001;9 Suppl 4(S11):244S-248S. Dymek MP, le Grange D, Neven K, Alverdy J. Quality of Life and Psychosocial Adjustment in Patients after Roux-en-Y Gastric Report Bypass: A Brief Report. Obes Surg. 2001;11(1):32-39. Green AEC, Dymek-Valentine M, Pytluk S, le Grange D, Alverdy J. Psychosocial outcome of gastric bypass surgery for patients with and without binge eating. Obes Surg. 2004;14(7):975-985. Sallet PC, Sallet JA, Dixon JB, et al. Eating behavior as a prognostic factor for weight loss after gastric bypass. Obes Surg. 2007;17(4):445-451. de Zwaan M, Hilbert A, Swan-Kremeier L, et al. Comprehensive interview assessment of eating behavior 18-35 months after gastric bypass surgery for morbid obesity. Surg Obes Relat Dis. 2010;6(1):79-85. Conceicao E, Mitchell JE, Vaz AR, et al. The presence of maladaptive eating behaviors after bariatric surgery in a cross sectional study: importance of picking or nibbling on weight regain. Eat Behav. 2014;15(4):558-562. Conceição EM, Mitchell JE, Engel SG, Machado PP, Lancaster K, Wonderlich SA. What is “grazing”? Reviewing its definition, frequency, clinical characteristics, and impact on bariatric surgery outcomes, and proposing a standardized definition. Surg Obes Relat Dis. 2014 Oct 31;10(5):973-82. White MA, Masheb RM, Rothschild BS, Burke-Martindale CH, Grilo CM. The prognostic significance of regular binge eating in extremely obese gastric bypass patients: 12-month postoperative outcomes. J Clin Psychol. 2006;67(12):1928-1935. El Chaar M, McDeavitt K, Richardson S, Gersin KS, Kuwada TS, Stefanidis D. Does patient compliance with preoperative bariatric office visits affect postoperative excess weight loss? Surg Obes Relat Dis. 2011;7(6):743-748. Gould JC, Beverstein G, Reinhardt S, Garren MJ. Impact of routine and long-term follow-up on weight loss after laparoscopic gastric bypass. Surg Obes Relat Dis. 2007;3(6):627-630; discussion 630. Poole NA, Al Atar A, Kuhanendran D, et al. Compliance with surgical after-care following bariatric surgery for morbid obesity: a retrospective study. Obes Surg. 2005;15(2):261-265.
20 48.
49. 50.
51. 52.
53.
Spitznagel MB, Alosco M, Galioto R, et al. The role of cognitive function in postoperative weight loss outcomes: 36-month follow-up. Obes Surg. 2014;24(7):10781084. Roth A, Fonagy P. What works for whom?: a critical review of psychotherapy research. Guilford Press; 2006. Sellbom M, Ben-Porath YS, Bagby RM. On the hierarchical structure of mood and anxiety disorders: confirmatory evidence and elaboration of a model of temperament markers. Journal of abnormal psychology. 2008;117(3):576-590. Watson D, Tellegen A. Toward a consensual structure of mood. Psychol Bull. 1985;98(2):219. Insel T, Cuthbert B, Garvey M, et al. Research domain criteria (RDoC): toward a new classification framework for research on mental disorders. Am J Psychiat. 2010;167(7):748-751. Forbey JD, Lee TT, Ben-Porath YS, Arbisi PA, Gartland D. Associations between MMPI-2-RF validity scale scores and extra-test measures of personality and psychopathology. Assessment. 2013;20(4):448-461.
21 Table 1 Descriptive Statistics Psychosocial Measures Medical Chart Review History of Suicide Attempts Number of Psychotropic Medications History of Outpatient Therapy History of Smoking Cigarettes History of Physical Abuse History of Sexual Abuse Psychiatric Diagnoses Major Depression Disorder Depression NOS Bipolar I or II Disorder Panic Disorder Generalized Anxiety Disorder Anxiety Disorder NOS Substance Use Disorder Binge Eating Disorder MMPI-2-RF Scales Emotional/Internalizing Dysfunction Behavioral/Externalizing Dysfunction Demoralization Low Positive Emotions Antisocial Behaviors Dysfunctional Negative Emotions Hypomanic Activation Suicidal/Death Ideation Helplessness/Hopelessness Self-Doubt Inefficacy Stress/Worry Anxiety Anger Proneness Behavior-Restricting Fears Multiple-Specific Fears Juvenile Conduct Problems Substance Use Disorder Aggression Activation Aggressiveness-Revised Disconstraint-Revised Negative Emotionality/Neuroticism Revised
Running Foot: 5-Year Weight Loss Outcomes
M
SD
Prevalence (%)
Missing (%)
1.3 -
.91 -
6.3 47.1 41.0 14.1 15.0
0.0 0.0 0.0 0.0 0.0 0.0
-
-
8.3 21.3 3.8 4.1 1.8 10.0 9.6 20.4
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
49 44 50 51 45 45 42 48 46 51 47 49 48 47 48 50 48 43 44 43 49 44 48
10 8 10 10 8 9 8 7 8 10 10 10 9 9 9 9 10 5 8 9 8 8 10
-
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
22 Psychosocial Measures M SD Prevalence (%) Missing (%) Introversion/Low Positive Emotionality – 52 10 0.0 Revised BMIs Pre-Surgical BMI 49.1 9.5 0.0 3-Month Post-Surgical BMI 38.6 7.6 15.5 6-Month Post-Surgical BMI 34.6 7.2 24.7 1-Year Post-Surgical BMI 32.4 7.6 28.3 5-Year Post-Surgical BMI 34.7 7.6 52.2 Note: %Excess Weight Loss at the Post-Operative Outcomes were: 3-Months (M = 34.9, SD = 11.8), 6-Months (M = 48.1, SD = 15.0), 1-Year (M = 59.9, SD = 17.9), 5-Years (M = 47.5, SD = 21.8); NOS (Not Otherwise Specified);
Running Foot: 5-Year Weight Loss Outcomes
PII: DOI: Reference:
S1550-7289(16)30811-5 http://dx.doi.org/10.1016/j.soard.2016.11.008 SOARD2826
To appear in: Surgery for Obesity and Related Diseases Cite this article as: Ryan J. Marek, Yossef S. Ben-Porath, Manfred van Dulmen, Kathleen Ashton and Leslie J. Heinberg, Using the pre-surgical psychological evaluation to predict 5-year Weight loss outcomes in bariatric surgery patients, Surgery for Obesity and Related Diseases, http://dx.doi.org/10.1016/j.soard.2016.11.008 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 Using the Pre-Surgical Psychological Evaluation to Predict 5-Year Weight Loss Outcomes in Bariatric Surgery Patients
Ryan J. Marek, M.A., Yossef S. Ben-Porath, Ph.D., Manfred van Dulmen. Ph.D. Department of Psychological Sciences Kent State University Kent, OH 44242
Kathleen Ashton, Ph.D., Leslie J. Heinberg, Ph.D. Cleveland Clinic Lerner College of Medicine 9500 Euclid Avenue/M61 Cleveland, OH 44195 United States of America
Conflict of Interest: Yossef Ben-Porath is a paid consultant to the MMPI-2-RF publisher, the University of Minnesota and Distributor, Pearson. As co-author of the MMPI-2-RF, he receives royalties on sales of the test. Disclosure: Portions of this manuscript were accepted to the ASMBS Masters in Behavioral Health Course at Obesity Week in New Orleans, LA (November, 2016) and were used to satisfy oral and written requirements for the first author’s dissertation at Kent State University. Correspondence: All correspondence should be addressed to: Ryan J. Marek, Department of Psychological Sciences, Kent State University, Kent, OH 44242. Email: [email protected]
2 Abstract Background: Psychosocial factors contribute to poorer weight loss outcomes following bariatric surgery; however, findings on associations between preoperative psychiatric diagnoses, psychological testing, and weight loss are inconsistent. Objectives: Examine associations between pre-surgical psychiatric diagnoses derived from a semi-structured clinical interview and test scores from the Minnesota Multiphasic PersonalityInventory – 2 – Restructured Form (MMPI-2-RF) and 5-year Body Mass Index (BMI) outcomes. Setting: Cleveland Clinic Bariatric and Metabolic Institute Methods: 446 consecutively, consented patients who underwent a Roux-en-Y Gastric Bypass (RYGB) at least 5-years prior were included in the study. A majority were women (74.2%) and Caucasian (66.2%). Patients’ mean pre-surgical BMI was 49.14 kg/m2 [Standard Deviation (SD) = 9.50 kg/m2]. Psychiatric diagnoses were obtained from a pre-surgical, semi-structured clinical interview and all participants were administered the MMPI-2-RF at their pre-surgical evaluations. BMIs were collected at 4 post-operative time points across a 5-year trajectory. This prospective design utilized latent growth curve modeling. Results: Older individuals evidenced a slower rate of BMI reduction over time. A pre-surgical diagnosis of Binge Eating Disorder predicted higher BMIs at the 5-year outcome. Scores on MMPI-2-RF measures of emotional and behavioral dysfunction domains incrementally predicted poorer weight loss outcomes. Conclusions: Pre-operative indicators of psychopathology, notably indicators that are dimensional in nature, are important in predicting post-operative outcomes. Closer follow-up with patients who evidence pre-surgical psychological factors, both before and after surgery, may help improve outcomes.
3 Keywords: Obesity, Bariatric Surgery, Assessment, Weight Loss, Outcome, Psychology,
Introduction Bariatric surgery results in significant weight loss and resolution of comorbid conditions for patients with severe obesity, demonstrating good efficacy as far as ten years following the procedure.1,2 However, weight loss begins to plateau around 12 months following surgery, and a subset of patients begin to show weight regain from that point forward (though some patients begin regaining weight around the 6-month outcome).3-7 Pre-surgical behavioral and psychosocial factors, including internalizing (i.e., anhedonia, anxiety, low frustration tolerance) and externalizing (i.e., impulsivity, sensation seeking, substance use) psychopathology, maladaptive eating behaviors, and poor behavioral adherence have been linked to poorer weight loss outcomes following bariatric surgery.8-17 Due in part to this association, a number of obesity-related societies recommend conducting pre-surgical psychological evaluations as part of patients’ pre-surgical regimen.5,18 A primary goal of these evaluations is to identify psychosocial factors that could impede outcomes and recommend appropriate pre-surgical and/or postoperative interventions to help the patient achieve and maintain optimal results. Despite numerous studies demonstrating evidence that pre-surgical psychopathology predicts poorer weight loss outcomes, 8-13 much of the literature on the role of pre-surgical internalizing and externalizing psychopathology in predicting poorer weight loss outcomes is mixed. For example, some studies reported that pre-surgical measures of depression predicted poorer short-term outcomes16,17,19 whereas others found no association between measures of
4 depression and short-term outcomes.12,20 Some studies indicate that pre-surgical externalizing behaviors, such as binge and graze eating are predictive of poorer weight loss outcomes,10 whereas other studies do not support those associations.21 A meta-analysis11 reported that presurgical variables, such as previous weight loss attempts, binge/sweet/maladaptive eating habits, depression, anxiety, history of sexual abuse, self-esteem, past/current alcohol abuse/use, and other psychiatric disorders do not predict poorer outcomes. Another integrative review of the literature8 implied that reliance on the current diagnostic system in addition to methodological differences across studies likely contributed to the inconsistent findings reported. For example, some depression self-report measures have inadequate construct validity when differentiating between distress and low positive emotionality (a core feature of those diagnosed with Major Depression Disorder). Likewise, the categorical-polythetic approach to classifying mental illness used worldwide assumes that psychopathology can be classified into distinct categories – where meeting a diagnostic threshold is a matter of having a sufficient number of symptoms associated with the diagnosis. However, the diagnostic system is polythetic in that different combinations of symptoms lead to the same diagnosis. The authors8 encouraged use of theoretically-derived, dimensional assessments of the hierarchical model of psychopathology (which are congruent with constructs outlined in RDoC) to better establish consistency when reporting associations with outcome criteria. Use of a conceptually-grounded instrument such as the Minnesota Multiphasic Personality Inventory – 2 – Restructured Form (MMPI-2-RF), 22,23 which is dimensional in nature and structured in congruence with the hierarchical model of psychopathology will likely offer more consistent prediction coefficients.
5 We sought to further establish the utility of pre-surgical psychological evaluations by examining mid-term (5-year) weight loss results in a sample of Roux-en-Y (RYGB) patients. The multi-method assessment practice conducted at the hospital made it possible to explore associations between findings in a medical chart review, a semi-structured clinical interview, and internalizing and externalizing scale scores from the MMPI-2-RF with 5-year BMI outcomes. Based on prior literature, no specific psychiatric diagnosis consistently predicts poorer weight loss outcomes; 8-17 therefore, all diagnoses and chart review information (e.g., history of abuse, number of psychiatric medications, etc) were tested for exploratory purposes. Lastly, MMPI-2RF scale assessing internalizing and externalizing dysfunction were added to the prediction model to test the incremental contribution of an objective personality assessment instrument that adheres to contemporary dimensional conceptualizations of psychopathology in the pre-surgical psychological evaluation. Materials and Methods Participants This study was prospective in nature, relying on an ongoing large-scale data collection. Data from 451 consecutively consented patients who underwent RYGB at least 5-years prior and resided in Northeast Ohio were available for the study. Prior to the pre-surgical evaluation, patients were asked to consent to have their medical and psychological data and used for research purposes. No incentives to participate were offered. If patients visited the medical center over the course of the 5-years, their weights were accessible from their electronic medical records. Of these individuals, 5 (0.11%) were excluded from further analyses because they produced an invalid MMPI-2-RF protocol based on guidelines outlined in the MMPI-2-RF Technical Manual23: Cannot Say > 18, Variable Response Inconsistency > 80, True Response
6 Inconsistency > 80, Infrequent Responses > 120, & Infrequent Psychopathology Responses > 100. There were no statistically significant differences on demographic or psychological variables between those who produced valid or invalid protocols, though in larger, better powered samples of bariatric surgery patients, lower education levels have been associated with invalid protocols in this setting.24 Of the 446 participants retained for analyses, 74.2% were women. The sample included 66.2% Caucasians, 19.7% African Americans, and 14.1% other ethnicities. The mean age was 46.75 years [Standard Deviation (SD) = 11.63; Range 18 – 74]. Patients’ mean pre-surgical BMI was 49.14 kg/m2 (SD = 9.50 kg/m2). The hospital’s Institutional Review Board approved the study.
Measures Semi-Structured Psycho-Diagnostic Interview. A doctoral level psychologist or a supervised post-doctoral fellow administered a semi-structured interview that is mandated by the Cleveland Clinic at initial intake. Information obtained during the interview included: an evaluation of Diagnostic and Statistical Manual – Fourth Edition – Text Revision (DSM-IVTR)25 diagnoses and mental health history/treatment, current mental health diagnoses/treatment, past and current substance use/abuse/dependence, physical and sexual abuse history, and history of past suicide attempts. It is important to note that this semistructured interview has not been validated against other published structured interviews. However, the interview template requires that clinicians assess for symptoms associated with all depressive, anxiety, psychotic bipolar, eating, and substance use-related disorders. BMI was recorded from the electronic medical record at the time of their evaluation, whereas postoperative BMIs were gathered within a 1-month range of the post-operative time point.
7 Although reliability data (e.g., inter-rater reliability) were not available for the interview, the information obtained has demonstrated good convergent and discriminant validity with psychopathology self-report measures.26 These data were coded through a retrospective chart review by trained research assistants and were then double entered to maintain consistency and reduce typographical error. Inter-rater reliability (Kappa statistics) between coders was .96 (Range = .81-1.00). Minnesota Multiphasic Personality Inventory – 2 – Restructured Form (MMPI-2RF).22,23 The MMPI-2-RF consists of 338 true-false items that are scored on 9 measures of protocol validity and 42 substantive scales that assess psychological constructs. The instrument has been recommended as an objective psychological measure in bariatric surgery evaluations because of the scale scores’ psychometric properties, replicated normative data, and associations with outcomes in bariatric surgery settings.27
Procedure A doctoral level psychologist or a supervised post-doctoral fellow administered a semistructured interview that is given at the Cleveland Clinic at initial intake. Information obtained during the interview included: an evaluation of Diagnostic and Statistical Manual – Fourth Edition – Text Revision (DSM-IV-TR)25 diagnoses (DSM-5 criteria was used for BED) and mental health history/treatment, current mental health diagnoses/treatment, past and current substance use/abuse/dependence, physical and sexual abuse history, and history of past suicide attempts. It is important to note that this semi-structured interview has not been validated against other published structured interviews. Like a SCID, it uses initial prompt queries for the following disorders: eating disorders, depression, bipolar, psychosis, generalized anxiety, panic, obsessive-compulsive, substance abuse/dependence and Post-Traumatic Stress Disorder. DSM-5
8 criteria for binge eating disorder with all symptoms (whether or not initial queries are positive) are elicited for all patients. Similarly, the research criteria for Night Eating Syndrome28 are all queried. Statistical Analyses All analyses were computed in Mplus 7.3.29 Latent growth curve analyses (LGCAs) were estimated using BMIs across time.30 In the current investigation, the intercept was specified to reflect differences at the 5-year post-operative time point (termed “5-Year BMIs”) whereas the slope (termed “BMI reduction over time”) reflected the rate of change from the pre-surgical BMIs through the 5-year BMIs. Conditional LGCA’s were then used to test whether pre-surgical variables predicted 5-Year BMIs and BMI reduction over time. Models were estimated using Robust Maximum Likelihood, as this method is preferred when variables that are non-normally distributed are in the analysis models. Full Information Maximum Likelihood (FIML) with age included in the covariance matrix was used to handle missing data across time.31 The following recommended goodness-of-fit indexes were used to evaluate the adequacy of the models tested:32 the Root Mean Square Error of Approximation (RMSEA), the Comparative Fit Index (CFI), the Standardized Root Mean Square (SRMR), and the Bayesian Information Criterion (BIC). Additionally, Chi Square (χ2) difference testing33 was used to compare models, with a statistically significant chi-square value indicating that the more complex model fits the data better than a less restrictive one. Standardized Beta Weights (β) were reported for path coefficients. Prediction analyses were conducted hierarchically. In other words, BED and MDD were entered into the model first. Following, other psychiatric diagnoses and medical chart review information were tested. Next, hypothesized MMPI-2-RF Substantive Scale Scores were entered based on scale set.
9 Results Descriptive Statistics and Preliminary Analyses Descriptive statistics for the variables used in subsequent analyses are reported in Table 1, which contains means and standard deviations for continuous variables, prevalence percentage for categorical variables, and percent missing for all variables used in the subsequent analyses. Importantly, 38.1% of the sample engaged in an under-reporting response style on the MMPI-2RF [Uncommon Virtues > 65T or Adjustment Validity > 60T]. Although a portion of these individuals’ scores may reflect traditional upbringing or being psychologically well-adjusted, 33% scored at or above the highest recommended interpretative T-Score cut-offs on these scales [Uncommon Virtues > 80T or Adjustment Validity > 70T], which implied that under-reporting was very likely in this sample. BMI means and standard deviations across time as well as the percentage of missing data were similar to weight loss outcomes reported in other longitudinal RYGB studies.3,6 Correlations between demographic or psychological variables and number of time points with missing data over time indicated that younger individuals had higher amounts of missing data over time (r = -.18, p < .001). No other statistically significant correlation coefficients were observed between demographic or psychological variables and attrition over time. Latent Growth Curve Modeling The first LGCA tested the appropriateness of a linear model (i.e., weight loss was a continuous, negative linear trajectory). Model fit statistics of linear and non-linear models are located in Supplemental Table 1. The best overall model fit for BMI change over the 5-year trajectory was a non-linear model with correlated error variances [χ2(9) = 19.15, p = .02, RMSEA = .05, CFI = .96, SRMR = .07, BIC = 13706.88]. The mean of the slope was negative (-
10 10.40 kg/m2) indicating that on average patients reduced 10.40 kg/m2 of their BMI over the 5year time span. Variance statistics around the 5-Year BMI estimate (Variance = 49.78, p < .001) and BMI-Reduction over time (Variance = 12.96, p = .031) were both statically significant, indicating considerable variability in patients’ BMIs both at the 5-year post-operative time point and the rate at which patients lose weight over time. The correlation between pre-surgical BMIs and BMI-reduction over time was statistically significant (r = -.62, p < .001), indicating that patients with higher pre-surgical BMIs evidenced a faster rate of change than did patients with lower starting BMIs. All subsequent analyses controlled for this trend. The non-linear model was then estimated using age as a predictor of 5-Year BMIs and BMI-Reduction over time. Model fit improved [χ2(7) = 12.36, p = .089, RMSEA = .041, CFI = .98, SRMR = .07, BIC = 10245.35] and demonstrated an incremental fit for the data [Δχ2(2) = 7.48, p = .024, ΔBIC = 3461.53]. Moreover, younger individuals had statistically higher BMIs at the pre-surgical evaluation than did older patients in this sample (β = -.09, p = .049), though there were no differences in age and BMI by the five-year outcome (β = -.01, p = .807). However, BMI reduction across time varied as a function of age, such that older adults tended to have a slower rate of BMI reduction across time than did younger adults (β = .20, p <.001). A depiction of this structural equation model is located in Supplemental Figure 1. Prediction Analyses Conditional LCGA analyses were conducted to estimate whether pre-surgical psychopathology predicted 5-year post-surgical BMIs as well as the rate of BMI-change over the 5-year trajectory. Medical chart review information was tested first, followed by diagnoses, and then MMPI-2-RF scales by scale set. Variables were retained in the model if they significantly predicted 5-year BMIs or BMI reduction over time. History of suicide attempts, number of
11 psychotropic medications, history of outpatient therapy, history of physical and sexual abuse, and history of smoking cigarettes were not predictive of 5-year BMIs or BMI reduction over time. Among psychiatric diagnoses, a pre-surgical diagnosis of BED predicted higher BMIs at the 5-year outcome (β = .16, p = .008). No other diagnoses were predictive of 5-year BMIs or BMI reduction over time. Patients who scored higher on the MMPI-2-RF scales Behavioral/Externalizing Dysfunction (β = .11, p = .030), Low Positive Emotions (β = .13, p = .032), and Hypomanic Activation (β = .13, p = .028) also evidenced higher BMIs at the 5-year outcome after controlling for age and BED. In regards to predicting variability in BMI-reduction over time, patients who scored higher on MMPI-2-RF scales Hypomanic Activation (β = .24, p = .002), Anger Proneness (β = .16, p = .004), and Activation (β = .17, p = .036) prior to surgery evidenced a slower rate of BMI-reduction over 5-years after controlling for age, BED, and the correlation between pre-surgical BMI and BMI reduction over time.
Discussion The goal of the current investigation was to examine whether categorical psychiatric diagnoses predicted BMI trajectories over a five-year post-bariatric surgery course (above and beyond known factors such as age) and whether, in turn, dimensional facets of psychopathology added predictive utility. It was hypothesized that dimensional measures of psychopathology, as measured on the MMPI-2-RF, would be incrementally predictive of 5-year weight loss outcomes above and beyond psychiatric diagnoses and demographic variables. As expected, older patients and those who obtained surgery at a lower BMI lost weight at a slower rate. Although many psychiatric diagnoses did not predict outcome, those who met criteria for BED prior to surgery began evidencing higher BMIs by 5-years post-surgery. Also, as hypothesized, scale scores such
12 as Low Positive Emotions (a measure of patients’ difficulty feeling positive emotions, lacking interest and energy, and displaying vegetative symptoms associated with Major Depressive Disorder) and Behavioral/Externalizing Dysfunction (a broad measure of disinhibition, hypomania, substance use, aggression, etc.) predicted higher BMIs 5-years post-surgery after controlling for various demographic variables and psychiatric diagnoses in the model. Taken together, the current investigation provided support for the utility of using empirically-grounded, objective psychological assessments during the pre-surgical evaluation. Results indicated that BMI-reduction over time is non-linear in that most individuals tended to evidence a decrease in BMI up to the one-year follow-up and then began regaining a small amount of weight by the 5-year outcome. Fit indices improved when non-linear trajectories were modeled as a function of age. The model in this investigation implied that older adults tended to have a slower rate of BMI-reduction over time. This finding is comparable to past research suggesting that, although older adults benefit from bariatric surgery (e.g., improved quality of life, reduction of medical comorbities, etc), they tend not to lose as much weight as younger adults. 34-36 Research implies that physical frailty and resistance to changing lifelong diet and activity habits are associated with a slower rate of BMI-change over time.37 Helping older adults find and engage in physical activities (such as water aerobics) as well as helping them identify the benefits of engaging in exercise may help accelerate their weight loss trajectories after bariatric surgery. In the current investigation, a pre-surgical diagnosis of BED was the only psychiatric diagnosis that was predictive of higher 5-year BMIs. This finding is consistent with other studies indicating that a pre-surgical diagnosis of BED is associated with poorer weight loss outcomes.3840
Moreover, many patients who meet diagnostic criteria for BED prior to surgery are more likely
13 to engage in graze eating behaviors and report having a sense of loss of control over eating, which in turn, can influence post-operative weight loss.41-43 However, other studies have reported no association between BED and weight loss after bariatric surgery.16,44 It is important to note that all patients diagnosed with BED had to complete binge eating treatment and show a positive benefit prior to surgery. Therefore, findings in the current investigation are likely underestimates of the association between untreated pre-surgical BED and medium-term outcomes. Moreover, patients are not able to necessarily consume large amounts of food, but may have consumed higher calorie/surgery/fatty foods in smaller amounts. Therefore, emphasis on conducting postoperative assessments of broader dietary adherence criteria (including grazing and feelings of loss of control over eating) and how it moderates weight loss outcomes is recommended. Studies that have examined the ability of post-surgical variables (such as dietary adherence, loss of control, executive functioning) to predict poorer outcomes have collectively found that factors associated with externalizing psychological dysfunction were related to poorer weight loss outcomes in this population. 14,45-48 In regards to weight, patients who scored higher on Behavioral/Externalizing Dysfunction scales on the MMPI-2-RF evidenced poorer one-year weight loss outcomes in a sample of gastric bypass patients.12 The current investigation extends these findings to the 5-year post-surgical bariatric surgery pattients. These results indicate that patients who evidence externalizing psychopathology prior to surgery may benefit from interventions designed to assist them gain better impulse-control. However, individuals with externalizing disorders tend not to respond as well to treatment.49 Therefore, maintaining postsurgical contact with patients who evidence greater pre-surgical externalizing psychopathology may lead to imrpoved ouctomes for these individuals.
14 In the internalizing domain, MMPI-2-RF scale scores on Low Positive Emotions were associated with higher BMIs at the 5-year follow-up, whereas higher scores on Anger-Proneness were associated with a slower rate of BMI-reduction over time. Low Positive Emotions encompasses the inability to experience positive emotions, anhedonia, pessimism, and introversion, and is a good predictor of Major Depressive Disorder.22,50,51 If a patient is experiencing symptoms such as anhedonia, they may be less likely to engage in physical activity. Individuals more prone to low positive emotionality are also less likely to enagage in the treatment process,22 and may similarly fail to engage in activities that promote weight loss maintenace. Helping patients gain insight into the benefits of exercise both prior to and following surgery lead to their becoming more active and achieving and maintaining good weight loss outcomes. In regards to Anger Proneness, patients with higher scores tend to experience irritability and low frustration tolerance. These patients may become discouraged at various postoperative timepoints and pre- and post-surgical treatment for patients who exhibit poor frustration tolerance should focus on achiegving and maintaining small goals throughout their surgical recovery. In an integrative review of the literature, Marek, Ben-Porath, and Heinberg8 discuss the limitations of relying on atheoretical diagnosistic systems and measures to predict outcomes in this setting. The authors review the problem of comorbitiy and research on the dimensional hierarchical model of psychopathology, which is similar to the constructs in the Research Domain Criteria (RDoC) outlined by the National Institute of Mental Health (NIMH)52 The current investigation is important because we would have been erroneously concluded that internalizing psychopathology was not associated with poorer weight loss outcomes had traditional psychiatric diagnoses been the sole outcome predictors. A strength to the MMPI-2-RF
15 is that it is conceptually and empirically linked to the hierarchical model of psychopathology. The incremental predictive utility obtained with this instrument in the current investigation lends support both to use of the MMPI-2-RF as part of a presurgical evaluation and the utility of the hierarchical model in short and longer-term weight-loss prediction.12,19 The current investigation had some limitations. BMI was the only dependent variable available, thus limiting the ability to predict how other revelavant outcomes such as postoperative psychopathology, eating behaviors, and quality of life affect weight loss across time. Furthermore, data on other factors that may have contributed to higher BMIs by the 5-year outcome, such as whether additional non-bariatric procedures had impacted their weight or whether they were taking medications that may affect their weight were not collected. Moreover, data on medical diagnoses (such as Type 2 Diabetes) were also not available. Another limitation is that the current study was composed of a sample who tended to under-report psychopatology and behaviors prior to surgery. If a patient engages in an under-reporting response style on the MMPI-2-RF, they do so across the evaluation, as evidenced in other samples.53 When underreporting is evidenced, prediction coefficients are attenuated and there is currently no validated method to correct for range restriction due to under-reporting in structural equation modeling. The sample was composed of individuals who, if presented with psychosocial risk factors, had to adhere to pre-surgical treatment and show a positive treatment response. Therefore, prediction coefficients reported in the current investigation are likely underestimates. Follow up studies are needed to better examine how pre-surgical treatment changes mediate outcomes. Lastly, percent excess weight loss (%EWL) is reported in Table 1, but was not included in the latent growth curve models for a few reasons. Modeling longitudinal latent growth curve analyses using %EWL would mean that there is no variability at the starting time point. A latent model with an
16 indicator that has zero variability cannot be modeled. Therefore, the latent growth curve would need to start at the 3-month time point and the correlation between the intercept and the slope would be quite different as compared to a correlation between a baseline measurement and its rate of change over time. Nonetheless, the current investigation had several strengths. The current investigation included examining outcome at the 5-year post-operative time point. Many outcome studies have focused on shorter-term outcome time points when using pre-surgical psychological assessments to predict weight loss outcomes. Lastly, Full Information Maximum Likelihood was used to handle missing data in the analyses which helped reduce bias in the predication coefficients. Future studies should integraeg post-operative psychological assessements to better account for how these psychological construct moderate weight loss. Conclusions Overall, bariatric surgery weight loss outcomes as a result of a RYGB indicate that most patients evidence a similar rate of BMI-reduction up until the one-year, post-surgical follow-up. From that point to the five-year post-surgical, most patients evidence a small amount of weight gain. BMI reduction across time was slower in older patients and patients with a lower BMI at the time of the evaluation. Additional variability in five-year BMI outcomes was accounted for by a pre-surgical diagnosis of BED and higher scores on some Internalizing and Externalizing scales of the MMPI-2-RF. Patients who were more impulsive, showed lower frustration tolerance, or were more anhedonic, pessmistic, and introverted evidenced higher BMIs at the 5year outcome than patients who did not evidence these symptoms/behaviors. Clinicians who observe symptoms, features, or behaviors in line with these constructs should reinforce the importance of follow-up visits, focusing on enhancing patient's self-control as a target for
17 intervention, and more closely monitor patients who may be less engaged or easily discouraged throughout the post-surgical process.
References
1. 2. 3. 4. 5.
6. 7.
8. 9.
10.
11. 12.
13.
14.
Brethauer SA, Chand B, Schauer PR. Risks and benefits of bariatric surgery: current evidence. Cleveland Clinic Journal of Medicine. 2006;73(11):993-1007. Valezi AC, de Almeida Menezes M, Mali J, Jr. Weight Loss Outcome After Roux-en-Y Gastric Bypass: 10 Years of Follow-up. Obes Surg. 2013;23(8):1290-1293. Magro DO, Geloneze B, Delfini R, Pareja BC, Callejas F, Pareja JC. Long-term weight regain after gastric bypass: a 5-year prospective study. Obes Surg. 2008;18(6):648-651. O'Brien PE, Dixon JB, Brown W. Obesity is a surgical disease: Overview of obesity and bariatric surgery. Anz J Surg. 2004;74(4):200-204. Mechanick JI, Youdim A, Jones DB, et al. Clinical practice guidelines for the perioperative nutritional, metabolic, and nonsurgical support of the bariatric surgery patient--2013 update: cosponsored by American Association of Clinical Endocrinologists, the Obesity Society, and American Society for Metabolic & Bariatric Surgery. Surg Obes Relat Dis. 2013;9(2):159-191. Sjostrom L. Bariatric surgery and reduction in morbidity and mortality: experiences from the SOS study. Int J Obes. 2008;32:S93-97. Courcoulas AP, Christian NJ, Belle SH, et al. Weight change and health outcomes at 3 years after bariatric surgery among individuals with severe obesity. JAMA : the journal of the American Medical Association. 2013;310(22):2416-2425. Marek RJ, Ben-Porath YS, Heinberg LJ. Understanding the role of psychopathology in bariatric surgery outcomes. Obes Rev. 2016;17(2):126-141. Meany G, Conceicao E, Mitchell JE. Binge eating, binge eating disorder and loss of control eating: effects on weight outcomes after bariatric surgery. Eur Eat Disord Rev. 2014;22(2):87-91. Conceição EM, Utzinger LM, Pisetsky EM. Eating disorders and problematic eating behaviours before and after bariatric surgery: characterization, assessment and association with treatment outcomes. Eur Eat Disord Rev. 2015;23(6):417-425. Livhits M, Mercado C, Yermilov I, et al. Preoperative predictors of weight loss following bariatric surgery: systematic review. Obes Surg. 2012;22(1):70-89. Marek RJ, Tarescavage AM, Ben-Porath YS, Ashton K, Merrell Rish J, Heinberg LJ. Using presurgical psychological testing to predict 1-year appointment adherence and weight loss in bariatric surgery patients: predictive validity and methodological considerations. Surg Obes Relat Dis. 2015;11:1171-1181. Belanger SB, Wechsler FS, Nademin ME, Virden III TB. Predicting outcome of gastric bypass surgery utilizing personality scale elevations, psychosocial factors, and diagnostic group membership. Obes Surg. 2010;20(10):1361-1371. Conceição EM, Mitchell JE, Engel SG, Machado PP, Lancaster K, Wonderlich SA. What is “grazing”? Reviewing its definition, frequency, clinical characteristics, and impact on
18
15. 16.
17.
18. 19.
20.
21.
22.
23.
24.
25. 26.
27.
28. 29. 30.
bariatric surgery outcomes, and proposing a standardized definition. Surg Obes Relat Dis. 2014;10(5):973-982. Odom J, Zalesin KC, Washington TL, et al. Behavioral predictors of weight regain after bariatric surgery. Obes Surg. 2010;20(3):349-356. 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(4):544-549. Semanscin-Doerr DA, Windover A, Ashton K, Heinberg LJ. Mood disorders in laparoscopic sleeve gastrectomy patients: does it affect early weight loss? Surg Obes Relat Dis. 2010;6(2):191-196. Sogg S, Lauretti J, West-Smith L. Recommendations for the Pre-Surgical Psychosocial Evaluation of Bariatric Surgery Patients. Surg Obes Relat Dis. 2016. Marek RJ, Ben-Porath YS, Merrell J, Ashton K, Heinberg LJ. Predicting one and three month postoperative Somatic Concerns, Psychological Distress, and Maladaptive Eating Behaviors in bariatric surgery candidates with the Minnesota Multiphasic Personality Inventory-2 Restructured Form (MMPI-2-RF). Obes Surg. 2014;24(4):631-639. 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):61-68. Wadden TA, Faulconbridge LF, Jones‐ Corneille LR, et al. Binge eating disorder and the outcome of bariatric surgery at one year: a prospective, observational study. Obesity. 2011;19(6):1220-1228. Ben-Porath YS, Tellegen A. The Minnesota Multiphasic Personality Inventory-2 Restructured Form (MMPI-2-RF): Manual for administration, scoring, and interpretation. . Minneapolis, Minnesota: University of Minnesota Press; 2008/2011. Tellegen A, Ben-Porath YS. The Minnesota Multiphasic Personality Inventory-2 Restructured Form (MMPI-2-RF): Technical Manual. Minneapolis. Minneapolis, Minnesota: University of Minnesota Press; 2008/2011. Tarescavage AM, Windover A, Ben-Porath YS, et al. Use of the MMPI-2-RF suicidal/death ideation and substance abuse scales in screening bariatric surgery candidates. Psychol Assess. 2013;25(4):1384-1389. Association AP. Diagnostic and statistical manual of mental disorders. 4th ed., text rev. ed. Washington, D.C.: American Psychiatric Publishing; 2000. Marek RJ, Ben-Porath YS, Windover A, et al. Assessing psychosocial functioning of bariatric surgery candidates with the Minnesota multiphasic personality inventory-2 restructured form (MMPI-2-RF). Obes Surg. 2013;23(11):1864-1873. Marek RJ, Heinberg LJ, Lavery M, Merrell Rish J, Ashton K. A review of psychological assessment instruments for use in bariatric surgery evaluations. Psychol Assess. 2016;28(9):1142-1157. Allison KC, Lundgren JD, O'Reardon JP, et al. Proposed diagnostic criteria for night eating syndrome. The International journal of eating disorders. 2010;43(3):241-247. Muthén LK, Muthén BO. Mplus User's Guide. Seventh ed. Los Angeles, CA: Muthén & Muthén; 1998-2012. Duncan TE, Duncan SC. An introduction to latent growth curve modeling. Behav Ther. 2004;35(2):333-363.
19 31.
32. 33. 34. 35.
36. 37. 38.
39.
40. 41.
42.
43.
44.
45.
46.
47.
Enders CK. The impact of nonnormality on full information maximum-likelihood estimation for structural equation models with missing data. Psychol Methods. 2001;6(4):352. Hu Lt, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Modeling.1999;6(1):1-55. Satorra A. Scaled and adjusted restricted tests in multi-sample analysis of moment structures. Springer; 2000. Sugerman HJ, DeMaria EJ, Kellum JM, Sugerman EL, Meador JG, Wolfe LG. Effects of Bariatric Surgery in Older Patients. Annals of Surgery. 2004;240(2):243-247. Villareal DT, Apovian CM, Kushner RF, Klein S. Obesity in older adults: technical review and position statement of the American Society for Nutrition and NAASO, The Obesity Society. Obesity research. 2005;13(11):1849-1863. Villareal DT, Chode S, Parimi N, et al. Weight loss, exercise, or both and physical function in obese older adults. N Engl J Med. 2011;364(13):1218-1229. Elia M. Obesity in the elderly. Obes Res. 2001;9 Suppl 4(S11):244S-248S. Dymek MP, le Grange D, Neven K, Alverdy J. Quality of Life and Psychosocial Adjustment in Patients after Roux-en-Y Gastric Report Bypass: A Brief Report. Obes Surg. 2001;11(1):32-39. Green AEC, Dymek-Valentine M, Pytluk S, le Grange D, Alverdy J. Psychosocial outcome of gastric bypass surgery for patients with and without binge eating. Obes Surg. 2004;14(7):975-985. Sallet PC, Sallet JA, Dixon JB, et al. Eating behavior as a prognostic factor for weight loss after gastric bypass. Obes Surg. 2007;17(4):445-451. de Zwaan M, Hilbert A, Swan-Kremeier L, et al. Comprehensive interview assessment of eating behavior 18-35 months after gastric bypass surgery for morbid obesity. Surg Obes Relat Dis. 2010;6(1):79-85. Conceicao E, Mitchell JE, Vaz AR, et al. The presence of maladaptive eating behaviors after bariatric surgery in a cross sectional study: importance of picking or nibbling on weight regain. Eat Behav. 2014;15(4):558-562. Conceição EM, Mitchell JE, Engel SG, Machado PP, Lancaster K, Wonderlich SA. What is “grazing”? Reviewing its definition, frequency, clinical characteristics, and impact on bariatric surgery outcomes, and proposing a standardized definition. Surg Obes Relat Dis. 2014 Oct 31;10(5):973-82. White MA, Masheb RM, Rothschild BS, Burke-Martindale CH, Grilo CM. The prognostic significance of regular binge eating in extremely obese gastric bypass patients: 12-month postoperative outcomes. J Clin Psychol. 2006;67(12):1928-1935. El Chaar M, McDeavitt K, Richardson S, Gersin KS, Kuwada TS, Stefanidis D. Does patient compliance with preoperative bariatric office visits affect postoperative excess weight loss? Surg Obes Relat Dis. 2011;7(6):743-748. Gould JC, Beverstein G, Reinhardt S, Garren MJ. Impact of routine and long-term follow-up on weight loss after laparoscopic gastric bypass. Surg Obes Relat Dis. 2007;3(6):627-630; discussion 630. Poole NA, Al Atar A, Kuhanendran D, et al. Compliance with surgical after-care following bariatric surgery for morbid obesity: a retrospective study. Obes Surg. 2005;15(2):261-265.
20 48.
49. 50.
51. 52.
53.
Spitznagel MB, Alosco M, Galioto R, et al. The role of cognitive function in postoperative weight loss outcomes: 36-month follow-up. Obes Surg. 2014;24(7):10781084. Roth A, Fonagy P. What works for whom?: a critical review of psychotherapy research. Guilford Press; 2006. Sellbom M, Ben-Porath YS, Bagby RM. On the hierarchical structure of mood and anxiety disorders: confirmatory evidence and elaboration of a model of temperament markers. Journal of abnormal psychology. 2008;117(3):576-590. Watson D, Tellegen A. Toward a consensual structure of mood. Psychol Bull. 1985;98(2):219. Insel T, Cuthbert B, Garvey M, et al. Research domain criteria (RDoC): toward a new classification framework for research on mental disorders. Am J Psychiat. 2010;167(7):748-751. Forbey JD, Lee TT, Ben-Porath YS, Arbisi PA, Gartland D. Associations between MMPI-2-RF validity scale scores and extra-test measures of personality and psychopathology. Assessment. 2013;20(4):448-461.
21 Table 1 Descriptive Statistics Psychosocial Measures Medical Chart Review History of Suicide Attempts Number of Psychotropic Medications History of Outpatient Therapy History of Smoking Cigarettes History of Physical Abuse History of Sexual Abuse Psychiatric Diagnoses Major Depression Disorder Depression NOS Bipolar I or II Disorder Panic Disorder Generalized Anxiety Disorder Anxiety Disorder NOS Substance Use Disorder Binge Eating Disorder MMPI-2-RF Scales Emotional/Internalizing Dysfunction Behavioral/Externalizing Dysfunction Demoralization Low Positive Emotions Antisocial Behaviors Dysfunctional Negative Emotions Hypomanic Activation Suicidal/Death Ideation Helplessness/Hopelessness Self-Doubt Inefficacy Stress/Worry Anxiety Anger Proneness Behavior-Restricting Fears Multiple-Specific Fears Juvenile Conduct Problems Substance Use Disorder Aggression Activation Aggressiveness-Revised Disconstraint-Revised Negative Emotionality/Neuroticism Revised
Running Foot: 5-Year Weight Loss Outcomes
M
SD
Prevalence (%)
Missing (%)
1.3 -
.91 -
6.3 47.1 41.0 14.1 15.0
0.0 0.0 0.0 0.0 0.0 0.0
-
-
8.3 21.3 3.8 4.1 1.8 10.0 9.6 20.4
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
49 44 50 51 45 45 42 48 46 51 47 49 48 47 48 50 48 43 44 43 49 44 48
10 8 10 10 8 9 8 7 8 10 10 10 9 9 9 9 10 5 8 9 8 8 10
-
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
22 Psychosocial Measures M SD Prevalence (%) Missing (%) Introversion/Low Positive Emotionality – 52 10 0.0 Revised BMIs Pre-Surgical BMI 49.1 9.5 0.0 3-Month Post-Surgical BMI 38.6 7.6 15.5 6-Month Post-Surgical BMI 34.6 7.2 24.7 1-Year Post-Surgical BMI 32.4 7.6 28.3 5-Year Post-Surgical BMI 34.7 7.6 52.2 Note: %Excess Weight Loss at the Post-Operative Outcomes were: 3-Months (M = 34.9, SD = 11.8), 6-Months (M = 48.1, SD = 15.0), 1-Year (M = 59.9, SD = 17.9), 5-Years (M = 47.5, SD = 21.8); NOS (Not Otherwise Specified);
Running Foot: 5-Year Weight Loss Outcomes