Content uploaded by David Mahony
Author content
All content in this area was uploaded by David Mahony
Content may be subject to copyright.
UNCORRECTED PROOF
3RESEARCH ARTICLE
4Psychological Gender Differences in Bariatric
5Surgery Candidates
6David Mahony
7Received: 22 May 2007 / Accepted: 29 June 2007
8#Springer Science + Business Media B.V. 2007
11 Abstract
12 Background Over 177,000 bariatric surgeries were per-
13 formed in 2006. Most patients are required to receive
14 presurgical psychological clearance, although there are no
15 empirically validated psycho-surgical risk factors. In an
16 effort to establish normative data on suspected risk factors,
17 the present study was conducted to determine if males and
18 females differ on psycho-surgical risk factors.
19 Methods Subjects consisted of 361 consecutive bariatric
20 surgery candidates undergoing a psychological evaluation
21 in a private practice setting. They were administered the
22 PsyBari, a test that detects and measures psycho-surgical
23 risk factors, and the Beck Depression Inventory (BDI-2).
24 Results The results indicate that males have significantly
25 higher BMIs than females ( p= 0.035). Females have tried
26 significantly more diets than males ( p< 0.000). Females are
27 significantly more likely to report a history of depression
28 than males ( p<0.000). Females received significantly
29 higher scores on the PsyBari Depression Index than males
30 (p<0.000.). Females received significantly higher BDI-2
31 scores than males ( p< 0.001). Females are significantly
32 more likely to report a history of anxiety than males ( p=
33 0.004). Females received significantly higher scores on the
34 PsyBari Social Anxiety Index than males ( p= 0.038).
35 Conclusion The results indicate that males and females
36 differ significantly on suspected psycho-surgical risk
37 factors. Assessments of bariatric surgery candidates should
38 recognize that males and females have different baselines
39 for psycho-surgical risk factors. Further research on
40bariatric surgery candidates should report results separated
41by gender.
42Keywords Psychological test .Bariatric surgery .
43Depression .Anxiety
44Introduction
45Over the past decade, bariatric surgery has become a
46popular treatment for morbidly obese patients. In 2006
47alone, an estimated 177,000 patients received this treatment
48[1]. The procedure requires patients to complete a battery of
49presurgical exams, and most surgeons require a psycholog-
50ical evaluation. Formal referral questions for the psycho-
51logical evaluation are provided to psychologists by surgical
52organizations in an effort to help them identify patients who
53are considered at risk for postsurgical psychological or
54behavioral problems (PSPBPs) [1]. These standard referral
55questions are not based on any empirical evidence and may
56or may not be effective in determining which patients are at
57risk for PSPBPs.
58Since the requirement for psychological evaluations on
59bariatric surgery candidates (BSC) has been established, a
60great deal of research has been conducted in an effort to
61identify psychosocial risk factors [2,3]. These efforts have
62begun to tease apart suspected risk factors, although initial
63findings suggest that factors previously considered to place
64a patient at risk are not necessarily problematic.
65For example, it was initially thought that BSCs who
66scored high on depression scales would be at risk for
67PSPBPs because symptoms of depression would interfere
68with postsurgical compliance, motivation, and coping
69skills. Studies have found that, although a large percentage
70of BSCs experience presurgical depression, these symptoms
OBES SURG
DOI 10.1007/s11695-007-9245-5
D. Mahony (*)
Department of Psychiatry, Lutheran Medical Center,
150 55 St. Suite 2-45, Brooklyn,
New York, NY 11220, USA
e-mail: dmahony@lmcmc.com
Q1
JrnlID 11695_ArtID 9245_Proof# 1 - 17/12/2007
UNCORRECTED PROOF
71 usually improve postsurgically without any PSPBPs [4,5].
72 In fact, a study done by Averbukh et al. [6] found a
73 negative correlation between presurgical depression scores
74 and postsurgical weight loss. Instead of depression inter-
75 fering with postsurgical functioning, there may be a
76 rebound effect where higher levels of presurgical depres-
77 sion predict better postsurgical functioning.
78 Similarly, a history of sexual abuse was considered to be
79 a risk factor, especially for female patients. The belief was
80 that these patients want to avoid sexual attention because it
81 reminds them of past sexual trauma. So, they use their
82 weight as a means to protect themselves from sexual
83 attention. Once they lose weight, they will receive more
84 sexual attention and experience increased anxiety due to the
85 reminder of the past sexual trauma. In spite of these
86 assumptions, studies have found no increased anxiety or
87 PSPBPs for this population [7,8].
88 These findings, and others, leave the psychological
89 examiner in the position of having to grant or deny surgical
90 clearance without having much empirical criteria as to which
91 patients will experience PSPBPs. In fact, the assessment
92 practices of mental health professionals vary widely for
93 bariatric surgery, with no established standards [9,10].
94 In an effort to address this problem and standardize the
95 bariatric surgery psychological evaluation, the PsyBari was
96 created [11]. The PsyBari is a psychological test that
97 measures variables considered to be important in identify-
98 ing patients at risk for PSPBPs. It is a 217-item paper-and-
99 pencil test that scores BSCs on multiple indices related to
100 psychological and behavioral variables considered to be
101 important in bariatric surgery psychological evaluations.
102 This includes a depression and social anxiety index. Using
103 the PsyBari, a wide range of potential psychosocial risk
104 factors can be assessed rapidly. The examiner can review
105 individual items and standardized T scores on each index.
106 As part of the initial standardization process, presurgical
107 norms have to be established for the PsyBari. This includes
108 determining in what ways BSCs are a heterogeneous
109 population. Previous research on the psychosocial variables
110 considered to be important in bariatric surgery often group
111 the patients together into one homogenous group [5,13].
112 This is in contrast to a wealth of research that exists on
113 psychosocial variables involved in nonsurgical obese pa-
114 tients, i.e., obese patients who are not considering bariatric
115 surgery. These researchers have shown that nonsurgical
116 obese patients are a heterogeneous population, with one of
117 the most frequently cited differences being gender. In fact,
118 the data differentiating nonsurgical obese males and
119 females is so extensive that researchers usually present
120 and discuss these populations separately [14,15]. For
121 example, researchers have consistently found that nonsur-
122 gical obese females report higher levels of depression and
123 social anxiety than nonsurgical obese males [14,16].
124It would be meaningful to know if BSCs also differ on
125variables such as gender. This would allow examiners to
126evaluate patients’psychosocial variables more accurately.
127Put another way, it is important to know what role gender
128plays in presurgical psychological evaluations. We cannot
129simply extrapolate findings of nonsurgical obese popula-
130tions onto BSCs because these two populations differ in
131important ways. For example, BSCs have to have a BMI of
13235 or above to be eligible for bariatric surgery, whereas
133subjects in nonsurgical obesity research may have lower
134BMIs.
135The present study is focused on determining if gender
136subgroups exist within the BSC population as they do with
137nonsurgical obese populations. Differentiating BSCs on
138psychological and behavioral variables is important because
139it may indicate that they have different norms and different
140risk factors for bariatric surgery.
141In specific, the present study hypothesizes that, consis-
142tent with the findings in nonsurgical obesity research,
143female BSCs will report higher levels of depression and
144social anxiety than male BSCs. This is predicted to be
145independent of BMI, and it is hypothesized that male BSCs
146will report higher BMIs than female BSCs. The study limits
147the number of PsyBari indices to those that have been
148found to show gender differences in nonsurgical obese
149populations in an effort to limit the likelihood of a type I
150error. The present study also hypothesizes that female BSCs
151will have more experience dieting and more experience
152with psychological and psychiatric treatment. These varia-
153bles were included because their importance is in assessing
154dieting experience and willingness to acknowledge and
155address psychological problems.
156Materials and Methods
157A power analysis was calculated using an effect size of 0.5,
158alpha level set at 0.05, and power set at 0.95, resulting in a
159sample size of 176 subjects [17]. A total of 361 records
160were available from subjects who underwent psychological
161evaluations for bariatric surgery in a private practice setting
162from August 2002 to September 2006. This included 244
163(67.6%) females and 117 (32.4%) males. Ages ranged from
16419 to 70, with a mean age of 41.7 (SD = 10.8). Subjects
165identified their ethnic group as Caucasian (64.6%), African-
166American (13.2%), Hispanic (9.9%), Asian (1.1%), Native-
167American (0.5%), and other (4.7%).
168Subjects were referred from two local bariatric surgery
169programs. They were administered the PsyBari, a test that
170measures weight-related psychological and behavioral
171variables considered to be relevant in predicting postsurgi-
172cal performance, and the Beck Depression Inventory 2
173(BDI-2) [18]. Subjects then completed a 1-h semistructured
JrnlID 11695_ArtID 9245_Proof# 1 - 17/12/2007
OBES SURG
UNCORRECTED PROOF
174 interview. The results of the PsyBari and BDI-2, along with
175 the interview results, determined whether or not the subject
176 received psychological clearance.
177 The data were analyzed for normality using the Shapiro–
178 Wilk test [19]. Most of the data were positively skewed and
179 did not meet criteria for normal distribution. This is
180 consistent with BSCs’tendency to minimize psychological
181 symptoms. Due to these results, nonparametric tests, the
182 Mann–Whitney and Chi Square, were used. The alpha rate
183 was set at 0.05.
184 Results
185 Regarding weight and eating habits, men reported signifi-
186 cantly higher BMIs (M=49.20, SD= 7.84) than women
187 (M=47.23, SD= 7.32), U= 10,927.00, p= 0.035 (two-tailed).
188 Women have tried significantly more diets (M=7.75,
189 SD=4.08) than men (M=5.48, SD 3.18), U=8,483.00,
190 p<0.000 (two-tailed).
191 For reported levels of depression, women were signifi-
192 cantly more likely to report a history of depression than men
193 (45.7% for women vs. 17.4% for men), X
2
(4, N=347)=
194 26.624, p<0.000. On the PsyBari Depression Index,
195 women received significantly higher scores (M= 10.91,
196 SD=5.15) than men (M=8.01, SD = 5.10), U=4,447.500,
197 p<0.000 (two tailed). On the BDI-2, women also received
198 significantly higher scores (M=13.73, SD=9.94) than men
199 (M=9.74, SD= 7.72), U= 8,211.000, p<0.001 (two tailed).
200 For reported levels of anxiety, women were more likely
201 to acknowledge a history of anxiety than men (23.2% of
202 women vs. 10.4% of men), X
2
(4, N=348)= 8.133, p=
203 0.004. On the PsyBari Social Anxiety Index, women scored
204 significantly higher (M=6.77, SD=4.52) than men (M=
205 5.58, SD=4.39), U= 5,959.000, p= 0.038.
206 In regards to mental health treatment, women were
207 significantly more likely to acknowledge a history of
208 psychotherapy than men (36.5% of women vs. 15.9% of
209 men) X
2
(4, N=346)=15.373, p<0.000. Women were
210 significantly more like to acknowledge taking psychiatric
211 medicine in the past than men (33% of women vs. 14.2% of
212 men) X
2
(4, N=346)= 13.813, p< 0.000.
213 Discussion
214 The results show that, similar to nonsurgical obese popu-
215 lations, BSCs should not be considered a homogenous
216 population. BSCs differ significantly, based on gender, on
217 many psychosocial variables routinely measured in psycho-
218 logical assessments. Specifically, female BSCs were signif-
219 icantly more likely to acknowledge a history of depression
220 and social anxiety. Female BSCs received significantly
221higher scores on the PsyBari depression and social anxiety
222indices. Female BSCs also received significantly higher
223BDI-2 scores. Female BSCs admitted to receiving psycho-
224logical and psychiatric treatment more than men. Female
225BSCs also had significantly more diet history than male
226BSCs. These differences are independent of weight because
227it was found that male BSCs have significantly higher BMIs
228than female BSCs.
229The results show that gender is an important variable
230when assessing BSCs. Males and females have different
231baselines on psychosocial variables, and individual BSCs
232should be compared to gender-based norms. Future research
233can determine if these gender differences persist after
234bariatric surgery and weight loss. Males and females may
235have different PSPBPs; different eating habits; different
236social adjustments; and, overall, different reactions to the
237surgery. In other words, males may have different risk factors
238than females. These differences need to be taken into account
239during presurgical psychological evaluations.
240As more information is obtained on BSCs, future research
241will be needed to determine if separate norms should be
242established for other psychosocial variables such as SES,
243race/ethnicity, and age. Establishment of these norms will
244assist future attempts at determining and measuring variables
245that place BSCs at high risk for PSPBPs.
246References
2471. American Society for Bariatric Surgery. American Society for
248Bariatric Surgery web site. www.asbs.org (2007). Accessed 2
249April 2007.
2502. Wadden TA, Sarwer DB, Womble LG, et al. Psychosocial aspects
251of obesity and obesity surgery. Obes Surg 2001;81:1001–24.
2523. Greenburg I. Psychological aspects of bariatric surgery. Nutr Clin
253Pract 2003;18:124–30.
2544. Nickel C, Widermann C, Harms D, et al. Patients with extreme
255obesity: change in mental symptoms three years after gastric
256banding. Int J Psychiatry Med 2005;35:109–22.
2575. Vallis MT, Butler GS, Perey B, et al. The role of psychological
258functioning in morbid obesity and its treatment with gastroplasty.
259Obes Surg 2001;11:716–25.
2606. Averbukh Y, Heshka S, El-Shoreya H, et al. Depression score
261predicts weight loss following roux-en y gastric bypass. Obes
262Surg 2003;13:833–6.
2637. Larsen JK, Geenan R. Childhood sexual abuse is not associated
264with a poor outcome after gastric banding for severe obesity. Obes
265Surg 2005;15:534–7.
2668. Buser A, Dymek-Valentine M, Hilburger J, et al. Outcome
267following gastric bypass surgery: impact of past sexual abuse.
268Obes Surg 2004;14:170–4.
2699. Fabricatore AN, Crerand CE, Wadden TA, et al. How do mental
270health professionals evaluate candidates for bariatric surgery?
271Survey results. Obes Surg 2006;16:567–73.
27210. Bauchowitz AU, Gonder-Frederick LA, Olbrisch ME, et al.
273Psychosocial evaluation of bariatric surgery candidates: a survey
274of present practices. Psychosom Med 2006;67:825–32.
27511. Mahony D. The PsyBari. A psychological evaluation for bariatric
276surgery candidates 2007. www.psybari.com. Accessed 21 May 2007.
JrnlID 11695_ArtID 9245_Proof# 1 - 17/12/2007
OBES SURG
UNCORRECTED PROOF
277 12. Saunders R. Binge eating in gastric bypass patients before surgery.
278 Obes Surg 1999;9:72–6.
279 13. Van Gemert WG, Adang EM, Greve JWM, et al. Quality of life
280 assessment of morbidly obese patients: effect of weight-reducing
281 surgery. Am J Clin Nutr 1998;76:197–201.
282 14. Heo M, Pietrobelli A, Fontaine KR, et al. Depressive mood and
283 obesity in US adults: comparison and moderation by sex, age, and
284 race. Int J Obes 2006;30:513–9.
285 15. Linde JA, Jeffery RW, Levy RL, et al. Binge eating disorder,
286 weight control self-efficacy, and depression in overweight men
287 and women. Int J Obes 2004;2:418–25.
28816. Rosenberger PH, Henderson KE, Grilo CM. Psychiatric disorder
289comorbidity and association with eating disorders in bariatric
290surgery patients: a cross-sectional study using structured inter-
291view-based diagnoses. J Clin Psychiatry 2006;67:1080–5.
29217. Erdfelder E, Faul F, Buchner A. GPOWER: a general power
293analysis program. Behav Res Methods Instrum Comput
2941996;28:1–11.
29518. Beck AT, Steer RA, Brown GK. Manual for the beck depression
296inventory-II. San Antonio: Psychological Corporation; 1996.
29719. SPSS. Statistical Package for Social Sciences, version 12.0.1.
298Chicago: SPSS Inc; 2003.
Q2
JrnlID 11695_ArtID 9245_Proof# 1 - 17/12/2007
OBES SURG