- Email: [email protected]
Abnormal temperament in patients with morbid obesity seeking surgical treatment
Journal of Affective Disorders 118 (2009) 155–160
Contents lists available at ScienceDirect
Journal of Affective Disorders j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j a d
Research report
Abnormal temperament in patients with morbid obesity seeking surgical treatment Benedikt Amann a,b,⁎, Roland Mergl d, Carla Torrent c, Giulio Perugi e, Frank Padberg b, Nadja El-Gjamal b, Gregor Laakmann b a b c d e
Benito Menni, CASM, Research Unit, CIBERSAM, Sant Boi de Llobregat, Barcelona, Spain Psychiatry Consultation-Liaison Service, University Hospital Grosshadern, Ludwig-Maximilians-University, Munich, Germany Bipolar Disorders Program, Clinical Institute of Neuroscience, CIBERSAM, University Hospital Clinic, IDIBAPS, University of Barcelona, Barcelona, Spain Department of Psychiatry, University of Leipzig, Germany Department of Psychiatry, University of Pisa, Pisa, Italy
a r t i c l e
i n f o
Article history: Received 20 November 2008 Received in revised form 14 January 2009 Accepted 14 January 2009 Available online 25 February 2009 Keywords: Temperament Affective disorder Depression Mania Subthreshold Obesity
a b s t r a c t Background: Obesity and its related disorders are growing epidemic across the world. Research on links between the bipolar spectrum and obesity has proliferated in the last few years. As some forms of abnormal temperament are considered as subtypes of the soft bipolar spectrum, we aimed to evaluate abnormal temperaments in morbidly obese patients. Methods: Using a short version of the Temperament Evaluation of Memphis, Pisa, Paris and San Diego, we investigated abnormal depressive, cyclothymic, hyperthymic, irritable or anxious temperament in 213 patients with morbid obesity compared to a control group of 90 patients admitted prior to organ transplantation. Additionally, the Beck-Depression Inventory (BDI) and the Self-Report Manic Inventory (SRMI) were applied to assess current mood status. Results: The obese group showed statistically significantly more psychiatric comorbidities compared to the control group. Abnormal temperaments were significantly more often observed in patients with morbid obesity rather than in controls. Cyclothymic, irritable and anxious temperaments showed specificity to obesity. Obese patients had significantly higher scores on the BDI, while no difference for SRMI scores was found among the whole groups. All temperaments were positively correlated with BDI and SRMI in the obese group. Limitations: The control group was not matched for demographic characteristics. Conclusions: Our results need replication but indicate an affective overlap in the form of abnormal temperament and depressive symptoms in obese patients, whereas mood swings should be evaluated and early mood stabilization considered for patients with significant weight gain to prevent obesity or to reduce already existing overweight. Studies of mood stabilizers and prospective observations would shed further insight on this complex interface of a major clinical and public health issue. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The prevalence of obesity has increased markedly during the past few decades, and this disorder is now responsible for
⁎ Corresponding author. Benito Menni, CASM, Research Unit, Dr. Antoni Pujadas 38, 08830 Sant Boi de Llobregat, Spain. Tel.: +34 936529999; fax: +34 936400268. E-mail address: [email protected] (B. Amann). 0165-0327/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2009.01.020
more health care expenditures than any other medical condition. At times, obesity has been considered as “an epidemic phenomenon”, showing a current prevalence of 10–35% in countries with western-style economies (Seidell and Flegal, 1997; Wilborn et al., 2005; Bray and Bellanger, 2006). Therapeutic interventions include dietary programs, cognitive–behavioural therapy, appetite inhibiting drugs, physical exercise and surgical interventions such as gastric banding. However, decades of obesity research have been
156
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
unable to suggest sufficient etiological explanations and to provide generally accepted strategies for safe and long-lasting weight loss. Suggested pathogenic factors include among other things: increased dietary energy intake and decreased energy expenditure, genetic susceptibility, as well as endocrinological and behavioral alterations. Comorbid psychopathology as a factor in the weight gain process, and its therapeutic implications, has been largely ignored for years despite clear evidence of an association (Pickering et al., 2007). Recent research in psychiatry has suggested a positive relationship between obesity and mood disorders, especially for major depression and bipolar disorder (Roberts et al., 2000; Onyike et al., 2003; Pickering et al., 2007). In the latter disorder, findings also indicate that eating disorder symptoms are prevalent and that the bipolar patients are, on average, more obese than the normal controls, including drug-naïve patients (Elmslie et al., 2000; McElroy et al., 2004; Maina et al., 2008; Wildes et al., 2008; Fiedorowicz et al., 2008). As the boundaries of bipolarity have expanded in recent years with studies supporting the need of widening the concept of the bipolar spectrum, a recent investigation has also suggested high prevalence rates of minor forms of bipolar disorder, especially hypomanic symptoms, among obese patients seeking surgical treatment (Alciati et al., 2007). Other studies have revealed temperamental abnormalities, e.g. greater novelty seeking in obese patients, using the Temperament and Character Inventory of Cloninger (Sullivan et al., 2007) or abnormal cyclothymic temperament in patients with bulimia (Perugi et al, 2006). In modern psychiatry, temperament is considered to be an enduring aspect of personality, which captures the subject's best stable behaviour, predicts mood changes and is strongly influenced by genetic constitution (Kagan, 1994; von Zerssen and Akiskal, 1998; Cloninger et al., 2006). Emil Kraepelin (1909–1915) considered that abnormal depressive, manic, irritable or cyclothymic temperament not only are presented as basic affective dispositions, but also as subclinical types of manic-depressive illness. Later, Akiskal et al. changed the manic temperament to hyperthymic and added a generalized anxious temperament (Akiskal and Akiskal, 1992; Akiskal, 1996; Akiskal and Pinto, 2000). The conceptualisation of these five temperaments led to the development of an operational instrument, i.e. the Temperament Evaluation of Memphis, Pisa, Paris and San Diego (TEMPS), as well as a self-rating questionnaire (TEMPSA) with 109 (for men) and 110 (for women) items (Akiskal et al, 2005b). Mendlowicz et al. (2005) showed that relatives of patients with bipolar disorder scored higher in the cyclothymic and anxious temperament domains compared to normal controls, but less in comparison to euthymic bipolar patients. Subsequently, Akiskal et al. (2006) suggested to classify the major depressive patients with cyclothymic temperament as part of the bipolar spectrum, labeled as “bipolar disorder II1/2”. Using the same methodological procedure in a recent publication as in our obese cohort, we observed a correlation between abnormal temperament, especially the cyclothymic temperament, and somatization disorder. We suggested that somatization disorder should be included in the soft bipolar spectrum and its treatment options reconsidered (Amann et al., in press). The aim of the present study was (1) to identify differences in abnormal temperament in 213 obese and 90
control subjects, (2) to correlate abnormal temperament with acute affective states, and (3) to evaluate if abnormal temperament, as part of the soft bipolar spectrum, should be added to the list of potential risk factors for obesity. 2. Methods The psychiatric consultation–liaison (C–L) service at the University Hospital Munich-Grosshadern responds mainly to clinical or medicolegal psychiatric problems in medical or surgical services. Referrals include, among others, patients with adjustment disorders related to their somatic disease, somatoform disorders or somatic patients with comorbid psychiatric illness. The service also carries out specific evaluations, such as psychiatric assessment prior to gastric banding in obese patients or prior to transplantation. Between 2005 and 2008 a total of 213 obese patients were referred to the psychiatric C–L service and were consecutively evaluated for suitability to surgical intervention, mainly gastric banding. Apart from demographic data, a complete somatic and psychiatric history was recorded by an experienced psychiatrist (B.A., N. E-G., G.L.) using our own systematic inquiry schedule listing all relevant psychiatric and somatic diseases. Psychiatric diagnoses were established based on an extensive diagnostic interview using ICD-10 (WHO, 1992) and DSM-IV criteria (American Psychiatric Association, 1994). Information was collected directly from patients and confirmed by family members — if present — during the interview. All patients were asked to complete the following selfrating battery which forms part of a standard protocol in the C–L service: the Beck Depression Inventory (BDI; Beck et al., 1961; German version: Hautzinger et al., 1992), the SelfReport Manic Inventory (SRMI; Shugar et al., 1992, German translation by Braunig et al., 1996), and the brief TEMPS-M (Erfurth et al., 2005a,b) which consists of 35 questions with 7 each dedicated to one of the five temperaments. The brief TEMPS-M derives from a scale developed by Akiskal as an operational instrument, i.e. the Temperament Evaluation of Memphis, Pisa, Paris and San Diego (TEMPS), to evaluate cyclothymic, depressive, irritable, hyperthymic and anxious temperaments (Akiskal et al, 2005b). The English (Akiskal et al., 2005a,b), the French (Hantouche et al., 2001; Akiskal et al., 2005c), the Spanish (Sanchez-Moreno et al., 2005) and the German versions (Blöink et al., 2005) show robust psychometric characteristics and strong validity. In the validation study for this self-report questionnaire, cut-off scores were investigated using the equivalent of ±2 S.D. to calculate the percentage of individuals who would meet the criteria of an abnormal temperament (Erfurth et al., 2005a,b); this was also applied in our investigation. A sample of 90 patients who underwent the same diagnostic process as the obese sample served as the control group. They were also consecutively admitted to the C–L service for routine psychiatric assessment regarding suitability prior to liver (n = 40), lung (n = 28), heart (n = 17), kidney (n = 3) or combined (heart and lung) (n = 2) transplantation. 2.1. Statistical analyses Statistical analyses were performed using the statistics software SPSS™ (Statistical package for the Social Sciences,
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
SPSS Inc, Chicago, Illinois, version 15.0). Comparison of clinical and sociodemographic characterisitics across the two groups was carried out using t-tests for independent sample comparison (in the case of metric variables like age) and Mann-Whitney U-tests (in the case of ordinal variables like the total score of the BDI). χ2-tests were applied in order to compare the obese and the control group regarding categorical variables. In the case of two-by-two tables and expected numbers b5, Fisher's exact test was used. Regarding the temperaments, since multiple dependent variables were used, we first performed a multivariate analysis of covariance (MANCOVA) with age and sex as covariates and group as a main factor. Then, group differences between the obese and the control group were tested with one-way analysis of covariance (ANCOVA). Spearman-Brown correlation coefficients were used to analyze which abnormal temperament was related to BDI- or SMRI-scores. The significance level was set at p b 0.05. 3. Results No significant differences in marital status and educational level were found. Other demographical variables, including sex and age, differed within the groups: The average age in the obese group was 39.6 years (S.D. = 11.8 years; range 17–67 years) and in the control group 50.9 years (S.D. = 9.4 years; range 21–68 years) ( t = − 8.84; df = 1,208.328; p b 0.001). As expected, we found a statistically significant overrepresentation of females in the obese group, with 156 females (73.2%) in the obese compared to 26 (28.9%) in the control group (χ2 = 51.88; df = 1,1; p b 0.001). Obese patients had higher weight parameters than controls (M = 135.3, S.D. = 30 kg vs. M = 75.4, S.D. = 15.6 kg) and BMI (M = 46.4, S.D. = 9.3 vs. M = 25, S.D. = 4.5), respectively (weight: t = 22.74; df = 1288.858; p b 0.001; BMI: t = 27.03; df = 1295.411; p b 0.001). Sixty-one of the 213 obese patients (28.6%) had a family history of obesity based on enquiry during the interview and confirmed in interview with the family where possible. Group differences in the rate of psychiatric comorbidities were statistically significant (χ2 = 14.07; df = 1; p = 0.0002): 104 of 213 (48.8%) in the obese patients reported having been diagnosed with mental diseases compared to 23 of 90 (25.6%) transplantation candidates. Fifty-five (25.8%) obese patients suffered from depression, twenty-six from adjustment disorder (12.2%), seven (3.3%) from a combination of both, four (1.9%) from personality disorder, four (1.9%) from depression and personality disorder, three (1.4%) from depression and anxiety, two (0.9%) from bipolar disorder and one (0.5%) from schizophrenia. In the transplantation group 23 (25.6%) reported from psychiatric illness in their history with 11 patients (12.2%) having suffered from depression, 6 (6.7%) from adjustment disorder, 4 (4.4%) from a combination of the two, and one each (1.1%) from personality disorder and anxiety disorder. Report of a family history of mental illnesses was positive in 22 of 213 (10.3%) of the obese patients, with 12 (5.6%) members having a history of affective disorders, 6 (2.8%) having addiction and one (0.5%) each having suicide, schizophrenia, OCD, addiction and somatization disorder, and addiction and suicide. In the control group 6 family
157
members (6.7%) suffered from addiction, 1 (1.1%) from schizophrenia and one family member committed suicide. With 8 of 90 (8.9%) patients with a positive family history, the difference was not statistically significant (χ2 = 0.15; df = 1; p b 0.70). One hundred thirty-eight patients (64.8%) in the obese group in comparison to 33 patients (36.7%) in the control group were rated as showing at least one abnormal temperament (χ2 = 20.35; df = 1; p b 0.0001). Of these 138 obese patients, 85 (62%) had one abnormal temperament, 29 (21%) had two, 19 (14%) had three, 4 had (3%) four and 1 (0.7%) had all five abnormal temperaments. Out of 33 patients with at least one abnormal temperament in the control group, 28 (85%) had one, 3 (9%) had two, and 2 (6%) had three abnormal temperaments. The two groups differed significantly regarding the mean number of abnormal temperaments (obese group: M = 1.04; S.D. = 1.05; control group: M = 0.44; S.D. = 0.67; Z = − 4.997; p b 0.001). This group difference remains statistically significant if the effects of age and sex are controlled for (F = 10.16; df = 1299; p = 0.002). MANCOVA (with age and sex as covariates, temperament scores as dependent variables and group (obese, control as main factor) revealed a significant overall difference between obese subjects and controls regarding the temperament (Wilks Lambda: F = 4.72; df = 5295; p b 0.001). Post-hoc ANCOVAs demonstrated significantly higher temperament scores in obese patients (as compared to controls) r e g a r d i n g c yc l ot hy m i c t e m p e ra m e n t ( F = 13 . 3 9 ; df = 1464.557; p b 0.001), irritable temperament (F = 4.46; df = 1126.445; p = 0.035) and anxious temperament (F = 4.47; df = 1127.521; p = 0.035), but not depressive and hyperthymic temperament (due to significant effects of the covariate “sex” on these variables). In total, 43 patients in the obese group were classified as abnormal in the depressive, 43 in the cyclothymic, 58 in the hyperthymic, 46 in the anxious and 32 in the irritable temperament. In the control group 7 patients exhibited a
Fig. 1. Comparison of the percentage of total numbers of abnormal temperaments in obese (n = 213) and control group (n = 90). One patient can have more than one abnormal temperament.
158
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
Table 1 Demographic and clinical characteristics of 213 obese patients and 90 controls (⁎mean, S.D., range).
Age (years) ⁎ Sex (female), numbers total, (%) Weight⁎ Body mass index⁎ Psychiatric comorbidities, total, (%) Family history, total, (%) Affective disorders Addiction Others Family history obesity, total, (%) SRMI (N 14:abnormal)⁎ BDI (N18:abnormal)⁎ Occupation total, (%) Active work Without occupation Pensioner Student Marital status total (%) Living together/married Single Divorced/separated Widowed
Obese group (n = 213)
Control group (n = 90)
39.6 (11.8;17–67) 156 (73.2)
50.9 (9.4;21–68) 26 (28.9)
135.3 (30; 75–235) 46.4 (9.3; 28–81) 104 (48.8)
75.4 (15.6; 41–135) 25 (4.5; 15–39) 23 (25.6)
22 (10.3) 12 6 4 61 (28.6)
8 (8.9) 0 6 2 0
5.2 (5.7;0–42) 17 (10.8;0– 50)
3.9 (4.3;0–23) 10.6 (7.7; 0–35)
141 (69.8) 34 (16.8) 18 (8.9) 9 (4.5)
38 (42.2) 16 (17.8) 35 (38.9) 1 (1.1)
136 (64.1) 57 (26.9) 16 (7.5) 3 (1.4)
66 (73.3) 14 (15.6) 8 (8.9) 2(2.2)
depressive, 4 a cyclothymic, 22 a hypertymic, 5 an anxious, 2 an irritable temperament (Fig. 1). Current depressive symptoms as measured by BDI ( N 18: abnormal) were higher in the obese group (M = 17; S.D. = 10.8; max. = 50, min. = 0) versus the control group (M = 10.6; S.D. = 7.7; max. = 35, min. = 0) and statistically significantly different (Z = − 5.03; p b 0.001) (Table 1). 87 patients (40.8%) in the obese group vs. 13 (14.4%) in the control group scored more than 18 in the BDI and fulfilled therefore a depressive state. 16 of 213 (7.5%) had an abnormal elevated mood in the SMRI (N14: abnormal) in the obese group in contrast to 3 of 90 (3.3%) in the control group. Median values of the SMRI between the groups were not statistically significantly different (Z = −1.12; p = 0.26). Interestingly, all abnormal temperaments were significantly correlated with the BDI and SRMI scores in the obese groups (Table 2). In the control group BDI and SMRI scores were positively correlated with depressive, cyclothymic,
irritable and anxious temperament, while, on the other hand, no correlation was observed between BDI and SMRI and between the hyperthymic temperament. Spearman-Brown correlation coefficients were used to analyze if temperaments were related to an increasing BMI which was not statistically significant in none of the groups. We also addressed the question of whether affective temperaments were mainly found in patients with comorbid depression in their history (in the obese and control group). Therefore, in both groups patients with and without comorbid depression were compared regarding the rate of patients with abnormal temperament scores. In the obese group we found statistically significantly more frequent comorbid depression associated with depressive temperament (χ2 = 7.24; df = 1; p = 0.007) (32.7% versus 15.8%). In the control group, subjects with comorbid depression had also significantly more often an abnormal depressive temperament (27.3%) than subjects without this condition (Fisher's exact test: p = 0.037; two-sided test). 4. Discussion This study represents the first use of the brief TEMPS-M to evaluate abnormal temperaments in obese persons in the general population. Our results indicate that patients with obesity show a higher rate of pathological temperament defined by the briefTEMPS-M compared to a control group of patients undergoing routine psychiatric assessment prior to transplantation. The prevalence of almost 65% abnormal temperaments in obese patients is clearly higher than one would expect in the general population. There, prevalence rates vary between 12.9% (Turkey; Vahip et al., 2005), 16.4% (Hungary; Rozsa et al., 2006) and 19.7% (Germany, as assessed using TEMPS-M; Erfurth et al., 2005b). Obese patients scored significantly higher in the cycloythmic, irritable and anxious temperaments compared to the control group. Presuming that cyclothymic temperament is part of the soft bipolar spectrum, the results fit in with previous studies which suggested a high prevalence of bipolar symptoms in an obese population (Alciati et al., 2007). The lack of significant group differences in the depressive and hyperthymic temperament may have been due to significant sex differences, as suggested by significant effects of the covariate “sex” in the ANCOVA. Additionally, the control group of transplantation candidates also showed a high percentage of abnormal temperaments and a peak in the hyperthymic trait was found. One explanation for this could
Table 2 Spearman-Brown correlations between temperamental traits and severity of affective disorders in 213 obese patients and 90 controls. Temperamental traits Depressive
Cyclothymic
Hyperthymic
Irritable
Anxious
rho = 0.73 p b 0.001*** rho = 0.50 p b 0.001***
rho = − 0.42 p b 0.001*** rho = − 0.17 p = 0.011*
rho = 0.24 p b 0.001** rho = 0.26 p b 0.001***
rho = 0.54 p b 0.001*** rho = 0.29 p b 0.001***
rho = 0.58 p = 0.001*** rho = 0.31 p = 0.003**
rho = − 0.15 p = 0.15 rho = − 0.05 p = 0.63
rho = 0.24 p = 0.03* rho = 0.25 p = 0.02*
rho = 0.44 p b 0.001*** rho = 0.35 p = 0.001***
Obese patients (n = 213) BDI total score SRMI total score
rho = 0.69 p b 0.001*** rho = 0.29 p b 0.001*** Control group (n = 90)
BDI total score SRMI total score
rho = 0.48 p b 0.001*** rho = 0.33 p = 0.002**
Notes: + p b 0.10; * p b 0.05; ** p b 0.01; *** p b 0.001. BDI: Beck Depression Inventory (Beck et al., 1961; German version: Hautzinger et al., 1992). SRMI: Self-Report Manic Inventory (Shugar et al., 1992, German translation by Braunig et al., 1996).
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
be a possible selection bias for the control group with transplantation candidates having a high rate of hyperthymia due to drug use or other risky behaviors favoring the illnesses that motivated transplantation (Perretta et al., 1998). Furthermore, there is controversy over the specificity of the hyperthymic trait in general: on the one hand, an investigation by Kesebir et colleagues (2005) showed that both bipolar patients and their relatives had a significantly higher prevalence of hyperthymic temperament compared to controls. However, on the other, another study found that hyperthymic scores were higher in normal controls than in bipolar patients and their healthy relatives (Mendlowicz et al., 2005). Another argument against the validity of the hyperthymic temperament construct derives from genetic investigations: The s allele of the 5HTTLPR polymorphism in the serotonin transporter gene was found to be associated with higher TEMPS scores of depressive, anxious, irritable, and particularly cyclothymic temperaments, but not with higher scores of hyperthymic temperament (Gonda et al., 2006). If the hyperthymic temperament due to doubtful validity is subtracted from both groups, in the obese group 77 (53%) of 144 patients and in the control group 10 (15%) of 63 would present with an abnormal temperament which would still be significantly different and might also be more a more reliable estimate of the rate of abnormal temperament in the control group and in the general population. As mentioned before, the association between obesity and cyclothymic traits is in line with previous evidence of high rates of hypomanic symptoms among obese patients seeking surgical treatment (Alciati et al., 2007). Furthermore, cyclothymic temperament has also been associated with eating problems in young women from a student population not psychiatrically ill (Signoretta et al., 2005) and with bulimia and binge eating disorders in atypical depressive patients (Perugi et al., 2006). It has been hypothesized that cyclothymic temperament and related sensation seeking and self-stimulating behaviour might involve any type of potentially addictive substance or activity, such as food, alcohol, drugs, physical exercise, work, travelling, the internet, or sex (Perugi and Akiskal, 2002). Relationships between obesity and addiction have also been reported in clinical (McIntyre et al., 2007) and biological studies (Volkow and Wise, 2005). In the present sample, in addition to a high prevalence of abnormal temperament, differences in acute depressive states, measured by BDI, were also detected. Obese patients were more often depressed compared to the control group, a finding which is in line with the previous literature (Roberts et al., 2000; Onyike et al., 2003; Pickering et al., 2007). Using a cut-off point of N18 in the BDI, 40% of obese patients were considered as having a depressive episode. As expected, a higher prevalence of obesity in women was noted in our sample which is in accordance with previous studies (e.g. Wilborn et al., 2005). Our results here are in line with investigations suggesting a higher prevalence of mental disorders in obesity; however, they contradict other studies which have found a high incidence of mental disorders in the families of obese patients (Black et al., 1992; Roberts et al., 2000; Onyike et al., 2003; Pickering et al., 2007). Along with our psychopathological findings, the high comorbidity of psychiatric diseases in obese patients of other studies might contribute to diagnostic discussions. Our findings might also
159
help obesity change its position in current diagnostic classifications, such as the ICD 10, so that it moves closer towards mental illness rather than solely being subsumed under ‘Endocrine, nutritional and metabolic diseases’. A limitation of our study is that the control group was not matched by demographic characteristics. This control group was chosen as it was a well-characterized sample of our psychiatric C–L service with an emphasis on a somatic disease which was being considered for surgical intervention. On these grounds, the sample of transplantation candidates was considered suitable to be compared with a group hypothesised to show psychiatric abnormality, despite the practical impossibility of matching for age and sex. Higher group differences would be expected, if psychiatrically healthy controls would have been studied. However, such an approach would have been associated with the problems of a “hypernormal” control group which we wanted to avoid in the present study (Buckley et al., 1992). Our results need replication and they make no claim to explain the complex etiology of obesity. However, they have the potential to add an important predisposing factor for the development of this hard-to-treat disease. Our results, together with other similar findings, also suggest that there should be a re-appraisal of the relationship between obesity and mental illness, as has been undertaken with other eating disorders, such as anorexia and bulimia. Furthermore, we would recommend psychiatric evaluation of patients in early stages of weight gain, focusing on affectivity with its spectrum, such as depressive and hypomanic symptoms and abnormal temperament. If pathological, an antidepressive or mood-stabilizing treatment should be considered. Studies of mood stabilizers and prospective observations could shed further insight on this complex interface of major clinical and public health issues. Role of funding source The study was supported by a grant from the Stanley Medical Research Centre, and a grant from the Instituto de Salud Carlos III, Centro de Investigación, Biomédica de Red de Salud Mental, CIBERSAM. Conflict of Interest All authors exclude any conflict of interest.
Acknowledgements The study was supported by a grant from the Stanley Medical Research Centre, and a grant from the Instituto de Salud Carlos III, Centro de Investigación, Biomédica de Red de Salud Mental, CIBERSAM. The first author, BA, receives two grants as principal investigator from the Instituto de Salud Carlos III (FIS CP06/00359 and PI071278). We also thank to Mrs Angela Poeller, Christa Fahmueller-Klose and Diana Wittmann (all Hospital Grosshadern) for their overall great help in the study and to Prof Peter McKenna for his help in revising the manuscript. References Amann, B., Padberg, F., Mergl, R., Naber, D., Baghai, T., Reimers, K., El-Giamal, N., Laakamann, G., in press. An investigation of temperamental traits in patients with somatoform disorder: do they belong to the affective spectrum? Psychosomatics, in press. American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders (4th ed; DSM-IV). APA, Washington DC.
160
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
Akiskal, H.S., 1996. The temperamental foundations of affective disorders. In: Mundt, C., Hahlweg, K., Fiedler, P. (Eds.), Interpersonal Factors in the Origin and Course of Affective Disorders. Gaskell, London, pp. 3–30. Akiskal, H.S., Akiskal, K., 1992. Cyclothymic, hyperthymic and depressive temperaments as subaffective variants of mood disorders. In: Tasman, A., Riba, M.B. (Eds.), Annual Review, vol. 11. American Psychiatric Press, Washington DC, pp. 43–62. Akiskal, H.S., Pinto, O., 2000. Soft bipolar spectrum: footnotes to Kraepelin on the interface of hypomania, temperament and depression. In: Marneros, A., Angst, J. (Eds.), Bipolar Disorders: 100 Years after Manic-depressive Insanity. Kluwer Academic, Dordrecht, pp. 37–62. Akiskal, H.S., Akiskal, K.K., Haykal, R.F., Manning, J.S., Connor, P.D., 2005a. TEMPS-A: progress towards validation of a self-rated clinical version of the Temperament Evaluation of the Memphis, Pisa, Paris, and San Diego Autoquestionnaire. J. Affect. Disord. 85 (1–2), 3–16. Akiskal, H.S., Mendlowicz, M.V., Jean-Louis, G., Rapaport, M.H., Kelsoe, J.R., Gillin, J.C., Smith, T.L., 2005b. TEMPS-A: validation of a short version of a self-rated instrument designed to measure variations in temperament. J. Affect. Disord. 85 (1–2), 45–52. Akiskal, H.S., Akiskal, K., Allilaire, J.F., Azorin, J.M., Bourgeois, M.L., Sechter, D., Fraud, J.P., Chatenet-Duchene, L., Lancrenon, S., Perugi, G., Hantouche, E.G., 2005c. Validating affective temperaments in their subaffective and socially positive attributes: psychometric, clinical and familial data from a French national study. J. Affect. Disord. 85 (1–2), 29–36. Akiskal, H.S., Akiskal, K.K., Lancrenon, S., Hantouche, E., 2006. Validating the soft bipolar spectrum in the French National EPIDEP Study: the prominence of BP-II 1/2. J Affect. Disord. 96 (3), 207–213. Alciati, A., D'Ambrosio, A., Foschi, D., Corsi, F., Mellado, C., Angst, J., 2007. Bipolar spectrum disorders in severely obese patients seeking surgical treatment. J. Affect. Disord. 101 (1–3), 131–138. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J., Erbaugh, J., 1961. An inventory for measuring depression. Arch. Gen. Psychiatry 4, 561–571. Black, D.W., Goldstein, R.B., Mason, E.E., Bell, S.E., Blum, N., 1992. Depression and other mental disorders in the relatives of morbidly obese patients. J. Affect. Disord. 25 (2), 91–95. Bloink, R., Brieger, P., Akiskal, H.S., Marneros, A., 2005. Factorial structure and internal consistency of the German TEMPS-A scale: validation against the NEO-FFI questionnaire. J. Affect. Disord. 85 (1–2), 77–83. Braunig, P., Shugar, G., Kruger, S., 1996. An investigation of the Self-Report Manic Inventory as a diagnostic and severity scale for mania. Compr. Psychiatry 37 (1), 52–55. Bray, G.A., Bellanger, T., 2006. Epidemiology, trends, and morbidities of obesity and the metabolic syndrome. Endocrine 29 (1), 109–117. Buckley, P., O'Callaghan, E., Larkin, C., Waddington, J.L., 1992. Schizophrenia research: the problem of controls. Biol. Psychiatry 32 (3), 215–217 1. Cloninger, C.R., Svrakic, D.M., Przybeck, T.R., 2006. Can personality assessment predict future depression? A twelve-month follow-up of 631 subjects. J. Affect. Disord. 92 (1), 35–44. Elmslie, J.L., Silverstone, J.T., Mann, J.I., Williams, S.M., Romans, S.E., 2000. Prevalence of overweight and obesity in bipolar patients. J. Clin. Psychiatry 61 (3), 179–184. Erfurth, A., Gerlach, A.L., Hellweg, I., Boenigk, I., Michael, N., Akiskal, H.S., 2005a. Studies on a German (Munster) version of the temperament auto-questionnaire TEMPS-A: construction and validation of the briefTEMPS-M. J. Affect. Disord. 85 (1–2), 53–69. Erfurth, A., Gerlach, A.L., Michael, N., Boenigk, I., Hellweg, I., Signoretta, S., Akiskal, K., Akiskal, H.S., 2005b. Distribution and gender effects of the subscales of a German version of the temperament autoquestionnaire briefTEMPS-M in a university student population. J. Affect. Disord. 85 (1–2), 71–76. Fiedorowicz, J.G., Palagummi, N.M., Forman-Hoffman, V.L., Miller del, D., Haynes, W.G., 2008. Elevated prevalence of obesity, metabolic syndrome, and cardiovascular risk factors in bipolar disorder. Ann. Clin. Psychiatry 20 (3), 131–137. Gonda, X., Rihmer, Z., Zsombok, T., Bagdy, G., Akiskal, K.K., Akiskal, H.S., 2006. The 5HTTLPR polymorphism of the serotonin transporter gene is associated with affective temperaments as measured by TEMPS-A. J. Affect. Disord. 91 (2–3), 125–131. Hantouche, E.G., Kochmann, F., Akiskal, H.S., 2001. Evaluation des temperaments affectifs: version complete des outils dáutoevaluation. Encèphale 28 (sp III), 24–30. Hautzinger, M., Bailer, M., Worall, H., Keller, F., 1992. Das Beck Depressionsinventar – BDI. Huber, Bern. Kagan, J., 1994. Galen's Prophecy: Temperament in Human Nature. Basic Books, New York. Kesebir, S., Vahip, S., Akdeniz, F., Yuncu, Z., Alkan, M., Akiskal, H., 2005. Affective temperaments as measured by TEMPS-A in patients with
bipolar I disorder and their first-degree relatives: a controlled study. J. Affect. Disord. 85 (1–2), 127–133. Kraepelin, E., 1909–1915. Psychiatrie Ein Lehrbuch für Studierende und Ärzte. (8. Auflage). Johann Ambrosius Barth, Leipzig. Maina, G., Salvi, V., Vitalucci, A., D'Ambrosio, V., Bogetto, F., 2008. Prevalence and correlates of overweight in drug-naive patients with bipolar disorder. J. Affect. Disord. 110 (1–2), 149–155. Mendlowicz, M.V., Jean-Louis, G., Kelsoe, J.R., Akiskal, H.S., 2005. A comparison of recovered bipolar patients, healthy relatives of bipolar probands, and normal controls using the short TEMPS-A. J. Affect. Disord. 85 (1–2), 147–151. McElroy, S.L., Kotwal, R., Malhotra, S., Nelson, E.B., Keck, P.E., Nemeroff, C.B., 2004. Are mood disorders and obesity related? A review for the mental health professional. J. Clin. Psychiatry 65 (5), 634–651. McIntyre, R.S., McElroy, S.L., Konarski, J.Z., Soczynska, J.K., Bottas, A., Castel, S., Wilkins, K., Kennedy, S.H., 2007. Substance use disorders and overweight/obesity in bipolar I disorder: preliminary evidence for competing addictions. J. Clin. Psychiatry 68 (9), 1352–1357. Onyike, C.U., Crum, R.M., Lee, H.B., Lyketsos, C.G., Eaton, W.W., 2003. Is obesity associated with major depression? Results from the Third National Health and Nutrition Examination Survey. Am. J. Epidemiol. 158 (12), 1139–1147 15. Perretta, P., Akiskal, H.S., Nisita, C., Zaccagnini, E., Lorenzetti, C., Della Santa, M., Cassano, G.B., 1998. The high prevalence of bipolar II and associated cyclothymic and hyperthymic temperaments in HIV patients. J. Affect. Disord. 50, 215–224. Perugi, G., Akiskal, H.S., 2002. The soft bipolar spectrum redefined: focus on the cyclothymic, anxious-sensitive, impulse-dyscontrol, and bingeeating connection in bipolar II and related conditions. Psychiatr. Clin. North Am. 25 (4), 713–737. Perugi, G., Toni, C., Passino, M.C., Akiskal, K.K., Kaprinis, S., Akiskal, H.S., 2006. Bulimia nervosa in atypical depression: the mediating role of cyclothymic temperament. J. Affect. Disord. 92, 91–97. Pickering, R.P., Grant, B.F., Chou, S.P., Compton, W.M., 2007. Are overweight, obesity, and extreme obesity associated with psychopathology? Results from the national epidemiologic survey on alcohol and related conditions. J. Clin. Psychiatry 68 (7), 998–1009. Roberts, R.E., Kaplan, G.A., Shema, S.J., Strawbridge, W.J., 2000. Are the obese at greater risk for depression? Am. J. Epidemiol. 152 (2), 163–170 15. Rozsa, S., Rihmer, A., Ko, N., Gonda, X., Szili, I., Szadoczky, E., Pestality, P., Rihmer, Z., 2006. Affective temperaments: psychometric properties of the Hungarian TEMPS-A. Psychiatr. Hung. 21 (2), 147–160. Sanchez-Moreno, J., Barrantes-Vidal, N., Vieta, E., Martinez-Aran, A., SaizRuiz, J., Montes, J.M., Akiskal, K., Akiskal, H.S., 2005. Process of adaptation to Spanish of the Temperament Evaluation of Memphis, Pisa, Paris and San Diego Scale. Self applied version (TEMPS-A). Actas Esp. Psiquiatr. 33 (5), 325–330. Seidell, J.C., Flegal, K.M., 1997. Assessing obesity: classification and epidemiology. Br Med Bull. 53 (2), 238–252. Shugar, G., Schertzer, S., Toner, B.B., Di Gasbarro, I., 1992. Development, use, and factor analysis of a self-report inventory for mania. Compr. Psychiatry 33 (5), 325–331. Signoretta, S., Maremmani, I., Liguori, A., Perugi, G., Akiskal, H.S., 2005. Affective temperament traits measured by TEMPS-I and emotionalbehavioral problems in clinically-well children, adolescents, and young adults. J. Affect. Disord. 85 (1–2), 169–180. Sullivan, S., Cloninger, C.R., Przybeck, T.R., Klein, S., 2007. Personality characteristics in obesity and relationship with successful weight loss. Int. J. Obes (Lond) 31 (4), 669–674. von Zerssen, D., Akiskal, H.S., 1998. Personality factors in affective disorders: historical developments and current issues with special reference to the concepts of temperament and character. J. Affect. Disord. 51, 1–5. WHO, 1992. The ICD-10 Classification of Mental and Behavioural Disorders. WHO, Geneva. Wilborn, C., Beckham, J., Campbell, B., Harvey, T., Galbreath, M., La Bounty, P., Nassar, E., Wismann, J., Kreider, R., 2005. Obesity: prevalence, theories, medical consequences, management, and research directions. J. Int. Soc. Sports Nutr. 9 (2), 4–31. Wildes, J.E., Marcus, M.D., Fagiolini, A., 2008. Prevalence and correlates of eating disorder co-morbidity in patients with bipolar disorder. Psychiatry Res. 161 (1), 51–58 30. Vahip, S., Kesebir, S., Alkan, M., Yazici, O., Akiskal, K.K., Akiskal, H.S., 2005. Affective temperaments in clinically-well subjects in Turkey: initial psychometric data on the TEMPS-A. J. Affect. Disord. 85 (1–2), 113–125. Volkow, N.D., Wise, R.A., 2005. How can drug addiction help us understand obesity? Nat. Neurosci. 8 (5), 555–560.
Contents lists available at ScienceDirect
Journal of Affective Disorders j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j a d
Research report
Abnormal temperament in patients with morbid obesity seeking surgical treatment Benedikt Amann a,b,⁎, Roland Mergl d, Carla Torrent c, Giulio Perugi e, Frank Padberg b, Nadja El-Gjamal b, Gregor Laakmann b a b c d e
Benito Menni, CASM, Research Unit, CIBERSAM, Sant Boi de Llobregat, Barcelona, Spain Psychiatry Consultation-Liaison Service, University Hospital Grosshadern, Ludwig-Maximilians-University, Munich, Germany Bipolar Disorders Program, Clinical Institute of Neuroscience, CIBERSAM, University Hospital Clinic, IDIBAPS, University of Barcelona, Barcelona, Spain Department of Psychiatry, University of Leipzig, Germany Department of Psychiatry, University of Pisa, Pisa, Italy
a r t i c l e
i n f o
Article history: Received 20 November 2008 Received in revised form 14 January 2009 Accepted 14 January 2009 Available online 25 February 2009 Keywords: Temperament Affective disorder Depression Mania Subthreshold Obesity
a b s t r a c t Background: Obesity and its related disorders are growing epidemic across the world. Research on links between the bipolar spectrum and obesity has proliferated in the last few years. As some forms of abnormal temperament are considered as subtypes of the soft bipolar spectrum, we aimed to evaluate abnormal temperaments in morbidly obese patients. Methods: Using a short version of the Temperament Evaluation of Memphis, Pisa, Paris and San Diego, we investigated abnormal depressive, cyclothymic, hyperthymic, irritable or anxious temperament in 213 patients with morbid obesity compared to a control group of 90 patients admitted prior to organ transplantation. Additionally, the Beck-Depression Inventory (BDI) and the Self-Report Manic Inventory (SRMI) were applied to assess current mood status. Results: The obese group showed statistically significantly more psychiatric comorbidities compared to the control group. Abnormal temperaments were significantly more often observed in patients with morbid obesity rather than in controls. Cyclothymic, irritable and anxious temperaments showed specificity to obesity. Obese patients had significantly higher scores on the BDI, while no difference for SRMI scores was found among the whole groups. All temperaments were positively correlated with BDI and SRMI in the obese group. Limitations: The control group was not matched for demographic characteristics. Conclusions: Our results need replication but indicate an affective overlap in the form of abnormal temperament and depressive symptoms in obese patients, whereas mood swings should be evaluated and early mood stabilization considered for patients with significant weight gain to prevent obesity or to reduce already existing overweight. Studies of mood stabilizers and prospective observations would shed further insight on this complex interface of a major clinical and public health issue. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The prevalence of obesity has increased markedly during the past few decades, and this disorder is now responsible for
⁎ Corresponding author. Benito Menni, CASM, Research Unit, Dr. Antoni Pujadas 38, 08830 Sant Boi de Llobregat, Spain. Tel.: +34 936529999; fax: +34 936400268. E-mail address: [email protected] (B. Amann). 0165-0327/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2009.01.020
more health care expenditures than any other medical condition. At times, obesity has been considered as “an epidemic phenomenon”, showing a current prevalence of 10–35% in countries with western-style economies (Seidell and Flegal, 1997; Wilborn et al., 2005; Bray and Bellanger, 2006). Therapeutic interventions include dietary programs, cognitive–behavioural therapy, appetite inhibiting drugs, physical exercise and surgical interventions such as gastric banding. However, decades of obesity research have been
156
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
unable to suggest sufficient etiological explanations and to provide generally accepted strategies for safe and long-lasting weight loss. Suggested pathogenic factors include among other things: increased dietary energy intake and decreased energy expenditure, genetic susceptibility, as well as endocrinological and behavioral alterations. Comorbid psychopathology as a factor in the weight gain process, and its therapeutic implications, has been largely ignored for years despite clear evidence of an association (Pickering et al., 2007). Recent research in psychiatry has suggested a positive relationship between obesity and mood disorders, especially for major depression and bipolar disorder (Roberts et al., 2000; Onyike et al., 2003; Pickering et al., 2007). In the latter disorder, findings also indicate that eating disorder symptoms are prevalent and that the bipolar patients are, on average, more obese than the normal controls, including drug-naïve patients (Elmslie et al., 2000; McElroy et al., 2004; Maina et al., 2008; Wildes et al., 2008; Fiedorowicz et al., 2008). As the boundaries of bipolarity have expanded in recent years with studies supporting the need of widening the concept of the bipolar spectrum, a recent investigation has also suggested high prevalence rates of minor forms of bipolar disorder, especially hypomanic symptoms, among obese patients seeking surgical treatment (Alciati et al., 2007). Other studies have revealed temperamental abnormalities, e.g. greater novelty seeking in obese patients, using the Temperament and Character Inventory of Cloninger (Sullivan et al., 2007) or abnormal cyclothymic temperament in patients with bulimia (Perugi et al, 2006). In modern psychiatry, temperament is considered to be an enduring aspect of personality, which captures the subject's best stable behaviour, predicts mood changes and is strongly influenced by genetic constitution (Kagan, 1994; von Zerssen and Akiskal, 1998; Cloninger et al., 2006). Emil Kraepelin (1909–1915) considered that abnormal depressive, manic, irritable or cyclothymic temperament not only are presented as basic affective dispositions, but also as subclinical types of manic-depressive illness. Later, Akiskal et al. changed the manic temperament to hyperthymic and added a generalized anxious temperament (Akiskal and Akiskal, 1992; Akiskal, 1996; Akiskal and Pinto, 2000). The conceptualisation of these five temperaments led to the development of an operational instrument, i.e. the Temperament Evaluation of Memphis, Pisa, Paris and San Diego (TEMPS), as well as a self-rating questionnaire (TEMPSA) with 109 (for men) and 110 (for women) items (Akiskal et al, 2005b). Mendlowicz et al. (2005) showed that relatives of patients with bipolar disorder scored higher in the cyclothymic and anxious temperament domains compared to normal controls, but less in comparison to euthymic bipolar patients. Subsequently, Akiskal et al. (2006) suggested to classify the major depressive patients with cyclothymic temperament as part of the bipolar spectrum, labeled as “bipolar disorder II1/2”. Using the same methodological procedure in a recent publication as in our obese cohort, we observed a correlation between abnormal temperament, especially the cyclothymic temperament, and somatization disorder. We suggested that somatization disorder should be included in the soft bipolar spectrum and its treatment options reconsidered (Amann et al., in press). The aim of the present study was (1) to identify differences in abnormal temperament in 213 obese and 90
control subjects, (2) to correlate abnormal temperament with acute affective states, and (3) to evaluate if abnormal temperament, as part of the soft bipolar spectrum, should be added to the list of potential risk factors for obesity. 2. Methods The psychiatric consultation–liaison (C–L) service at the University Hospital Munich-Grosshadern responds mainly to clinical or medicolegal psychiatric problems in medical or surgical services. Referrals include, among others, patients with adjustment disorders related to their somatic disease, somatoform disorders or somatic patients with comorbid psychiatric illness. The service also carries out specific evaluations, such as psychiatric assessment prior to gastric banding in obese patients or prior to transplantation. Between 2005 and 2008 a total of 213 obese patients were referred to the psychiatric C–L service and were consecutively evaluated for suitability to surgical intervention, mainly gastric banding. Apart from demographic data, a complete somatic and psychiatric history was recorded by an experienced psychiatrist (B.A., N. E-G., G.L.) using our own systematic inquiry schedule listing all relevant psychiatric and somatic diseases. Psychiatric diagnoses were established based on an extensive diagnostic interview using ICD-10 (WHO, 1992) and DSM-IV criteria (American Psychiatric Association, 1994). Information was collected directly from patients and confirmed by family members — if present — during the interview. All patients were asked to complete the following selfrating battery which forms part of a standard protocol in the C–L service: the Beck Depression Inventory (BDI; Beck et al., 1961; German version: Hautzinger et al., 1992), the SelfReport Manic Inventory (SRMI; Shugar et al., 1992, German translation by Braunig et al., 1996), and the brief TEMPS-M (Erfurth et al., 2005a,b) which consists of 35 questions with 7 each dedicated to one of the five temperaments. The brief TEMPS-M derives from a scale developed by Akiskal as an operational instrument, i.e. the Temperament Evaluation of Memphis, Pisa, Paris and San Diego (TEMPS), to evaluate cyclothymic, depressive, irritable, hyperthymic and anxious temperaments (Akiskal et al, 2005b). The English (Akiskal et al., 2005a,b), the French (Hantouche et al., 2001; Akiskal et al., 2005c), the Spanish (Sanchez-Moreno et al., 2005) and the German versions (Blöink et al., 2005) show robust psychometric characteristics and strong validity. In the validation study for this self-report questionnaire, cut-off scores were investigated using the equivalent of ±2 S.D. to calculate the percentage of individuals who would meet the criteria of an abnormal temperament (Erfurth et al., 2005a,b); this was also applied in our investigation. A sample of 90 patients who underwent the same diagnostic process as the obese sample served as the control group. They were also consecutively admitted to the C–L service for routine psychiatric assessment regarding suitability prior to liver (n = 40), lung (n = 28), heart (n = 17), kidney (n = 3) or combined (heart and lung) (n = 2) transplantation. 2.1. Statistical analyses Statistical analyses were performed using the statistics software SPSS™ (Statistical package for the Social Sciences,
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
SPSS Inc, Chicago, Illinois, version 15.0). Comparison of clinical and sociodemographic characterisitics across the two groups was carried out using t-tests for independent sample comparison (in the case of metric variables like age) and Mann-Whitney U-tests (in the case of ordinal variables like the total score of the BDI). χ2-tests were applied in order to compare the obese and the control group regarding categorical variables. In the case of two-by-two tables and expected numbers b5, Fisher's exact test was used. Regarding the temperaments, since multiple dependent variables were used, we first performed a multivariate analysis of covariance (MANCOVA) with age and sex as covariates and group as a main factor. Then, group differences between the obese and the control group were tested with one-way analysis of covariance (ANCOVA). Spearman-Brown correlation coefficients were used to analyze which abnormal temperament was related to BDI- or SMRI-scores. The significance level was set at p b 0.05. 3. Results No significant differences in marital status and educational level were found. Other demographical variables, including sex and age, differed within the groups: The average age in the obese group was 39.6 years (S.D. = 11.8 years; range 17–67 years) and in the control group 50.9 years (S.D. = 9.4 years; range 21–68 years) ( t = − 8.84; df = 1,208.328; p b 0.001). As expected, we found a statistically significant overrepresentation of females in the obese group, with 156 females (73.2%) in the obese compared to 26 (28.9%) in the control group (χ2 = 51.88; df = 1,1; p b 0.001). Obese patients had higher weight parameters than controls (M = 135.3, S.D. = 30 kg vs. M = 75.4, S.D. = 15.6 kg) and BMI (M = 46.4, S.D. = 9.3 vs. M = 25, S.D. = 4.5), respectively (weight: t = 22.74; df = 1288.858; p b 0.001; BMI: t = 27.03; df = 1295.411; p b 0.001). Sixty-one of the 213 obese patients (28.6%) had a family history of obesity based on enquiry during the interview and confirmed in interview with the family where possible. Group differences in the rate of psychiatric comorbidities were statistically significant (χ2 = 14.07; df = 1; p = 0.0002): 104 of 213 (48.8%) in the obese patients reported having been diagnosed with mental diseases compared to 23 of 90 (25.6%) transplantation candidates. Fifty-five (25.8%) obese patients suffered from depression, twenty-six from adjustment disorder (12.2%), seven (3.3%) from a combination of both, four (1.9%) from personality disorder, four (1.9%) from depression and personality disorder, three (1.4%) from depression and anxiety, two (0.9%) from bipolar disorder and one (0.5%) from schizophrenia. In the transplantation group 23 (25.6%) reported from psychiatric illness in their history with 11 patients (12.2%) having suffered from depression, 6 (6.7%) from adjustment disorder, 4 (4.4%) from a combination of the two, and one each (1.1%) from personality disorder and anxiety disorder. Report of a family history of mental illnesses was positive in 22 of 213 (10.3%) of the obese patients, with 12 (5.6%) members having a history of affective disorders, 6 (2.8%) having addiction and one (0.5%) each having suicide, schizophrenia, OCD, addiction and somatization disorder, and addiction and suicide. In the control group 6 family
157
members (6.7%) suffered from addiction, 1 (1.1%) from schizophrenia and one family member committed suicide. With 8 of 90 (8.9%) patients with a positive family history, the difference was not statistically significant (χ2 = 0.15; df = 1; p b 0.70). One hundred thirty-eight patients (64.8%) in the obese group in comparison to 33 patients (36.7%) in the control group were rated as showing at least one abnormal temperament (χ2 = 20.35; df = 1; p b 0.0001). Of these 138 obese patients, 85 (62%) had one abnormal temperament, 29 (21%) had two, 19 (14%) had three, 4 had (3%) four and 1 (0.7%) had all five abnormal temperaments. Out of 33 patients with at least one abnormal temperament in the control group, 28 (85%) had one, 3 (9%) had two, and 2 (6%) had three abnormal temperaments. The two groups differed significantly regarding the mean number of abnormal temperaments (obese group: M = 1.04; S.D. = 1.05; control group: M = 0.44; S.D. = 0.67; Z = − 4.997; p b 0.001). This group difference remains statistically significant if the effects of age and sex are controlled for (F = 10.16; df = 1299; p = 0.002). MANCOVA (with age and sex as covariates, temperament scores as dependent variables and group (obese, control as main factor) revealed a significant overall difference between obese subjects and controls regarding the temperament (Wilks Lambda: F = 4.72; df = 5295; p b 0.001). Post-hoc ANCOVAs demonstrated significantly higher temperament scores in obese patients (as compared to controls) r e g a r d i n g c yc l ot hy m i c t e m p e ra m e n t ( F = 13 . 3 9 ; df = 1464.557; p b 0.001), irritable temperament (F = 4.46; df = 1126.445; p = 0.035) and anxious temperament (F = 4.47; df = 1127.521; p = 0.035), but not depressive and hyperthymic temperament (due to significant effects of the covariate “sex” on these variables). In total, 43 patients in the obese group were classified as abnormal in the depressive, 43 in the cyclothymic, 58 in the hyperthymic, 46 in the anxious and 32 in the irritable temperament. In the control group 7 patients exhibited a
Fig. 1. Comparison of the percentage of total numbers of abnormal temperaments in obese (n = 213) and control group (n = 90). One patient can have more than one abnormal temperament.
158
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
Table 1 Demographic and clinical characteristics of 213 obese patients and 90 controls (⁎mean, S.D., range).
Age (years) ⁎ Sex (female), numbers total, (%) Weight⁎ Body mass index⁎ Psychiatric comorbidities, total, (%) Family history, total, (%) Affective disorders Addiction Others Family history obesity, total, (%) SRMI (N 14:abnormal)⁎ BDI (N18:abnormal)⁎ Occupation total, (%) Active work Without occupation Pensioner Student Marital status total (%) Living together/married Single Divorced/separated Widowed
Obese group (n = 213)
Control group (n = 90)
39.6 (11.8;17–67) 156 (73.2)
50.9 (9.4;21–68) 26 (28.9)
135.3 (30; 75–235) 46.4 (9.3; 28–81) 104 (48.8)
75.4 (15.6; 41–135) 25 (4.5; 15–39) 23 (25.6)
22 (10.3) 12 6 4 61 (28.6)
8 (8.9) 0 6 2 0
5.2 (5.7;0–42) 17 (10.8;0– 50)
3.9 (4.3;0–23) 10.6 (7.7; 0–35)
141 (69.8) 34 (16.8) 18 (8.9) 9 (4.5)
38 (42.2) 16 (17.8) 35 (38.9) 1 (1.1)
136 (64.1) 57 (26.9) 16 (7.5) 3 (1.4)
66 (73.3) 14 (15.6) 8 (8.9) 2(2.2)
depressive, 4 a cyclothymic, 22 a hypertymic, 5 an anxious, 2 an irritable temperament (Fig. 1). Current depressive symptoms as measured by BDI ( N 18: abnormal) were higher in the obese group (M = 17; S.D. = 10.8; max. = 50, min. = 0) versus the control group (M = 10.6; S.D. = 7.7; max. = 35, min. = 0) and statistically significantly different (Z = − 5.03; p b 0.001) (Table 1). 87 patients (40.8%) in the obese group vs. 13 (14.4%) in the control group scored more than 18 in the BDI and fulfilled therefore a depressive state. 16 of 213 (7.5%) had an abnormal elevated mood in the SMRI (N14: abnormal) in the obese group in contrast to 3 of 90 (3.3%) in the control group. Median values of the SMRI between the groups were not statistically significantly different (Z = −1.12; p = 0.26). Interestingly, all abnormal temperaments were significantly correlated with the BDI and SRMI scores in the obese groups (Table 2). In the control group BDI and SMRI scores were positively correlated with depressive, cyclothymic,
irritable and anxious temperament, while, on the other hand, no correlation was observed between BDI and SMRI and between the hyperthymic temperament. Spearman-Brown correlation coefficients were used to analyze if temperaments were related to an increasing BMI which was not statistically significant in none of the groups. We also addressed the question of whether affective temperaments were mainly found in patients with comorbid depression in their history (in the obese and control group). Therefore, in both groups patients with and without comorbid depression were compared regarding the rate of patients with abnormal temperament scores. In the obese group we found statistically significantly more frequent comorbid depression associated with depressive temperament (χ2 = 7.24; df = 1; p = 0.007) (32.7% versus 15.8%). In the control group, subjects with comorbid depression had also significantly more often an abnormal depressive temperament (27.3%) than subjects without this condition (Fisher's exact test: p = 0.037; two-sided test). 4. Discussion This study represents the first use of the brief TEMPS-M to evaluate abnormal temperaments in obese persons in the general population. Our results indicate that patients with obesity show a higher rate of pathological temperament defined by the briefTEMPS-M compared to a control group of patients undergoing routine psychiatric assessment prior to transplantation. The prevalence of almost 65% abnormal temperaments in obese patients is clearly higher than one would expect in the general population. There, prevalence rates vary between 12.9% (Turkey; Vahip et al., 2005), 16.4% (Hungary; Rozsa et al., 2006) and 19.7% (Germany, as assessed using TEMPS-M; Erfurth et al., 2005b). Obese patients scored significantly higher in the cycloythmic, irritable and anxious temperaments compared to the control group. Presuming that cyclothymic temperament is part of the soft bipolar spectrum, the results fit in with previous studies which suggested a high prevalence of bipolar symptoms in an obese population (Alciati et al., 2007). The lack of significant group differences in the depressive and hyperthymic temperament may have been due to significant sex differences, as suggested by significant effects of the covariate “sex” in the ANCOVA. Additionally, the control group of transplantation candidates also showed a high percentage of abnormal temperaments and a peak in the hyperthymic trait was found. One explanation for this could
Table 2 Spearman-Brown correlations between temperamental traits and severity of affective disorders in 213 obese patients and 90 controls. Temperamental traits Depressive
Cyclothymic
Hyperthymic
Irritable
Anxious
rho = 0.73 p b 0.001*** rho = 0.50 p b 0.001***
rho = − 0.42 p b 0.001*** rho = − 0.17 p = 0.011*
rho = 0.24 p b 0.001** rho = 0.26 p b 0.001***
rho = 0.54 p b 0.001*** rho = 0.29 p b 0.001***
rho = 0.58 p = 0.001*** rho = 0.31 p = 0.003**
rho = − 0.15 p = 0.15 rho = − 0.05 p = 0.63
rho = 0.24 p = 0.03* rho = 0.25 p = 0.02*
rho = 0.44 p b 0.001*** rho = 0.35 p = 0.001***
Obese patients (n = 213) BDI total score SRMI total score
rho = 0.69 p b 0.001*** rho = 0.29 p b 0.001*** Control group (n = 90)
BDI total score SRMI total score
rho = 0.48 p b 0.001*** rho = 0.33 p = 0.002**
Notes: + p b 0.10; * p b 0.05; ** p b 0.01; *** p b 0.001. BDI: Beck Depression Inventory (Beck et al., 1961; German version: Hautzinger et al., 1992). SRMI: Self-Report Manic Inventory (Shugar et al., 1992, German translation by Braunig et al., 1996).
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
be a possible selection bias for the control group with transplantation candidates having a high rate of hyperthymia due to drug use or other risky behaviors favoring the illnesses that motivated transplantation (Perretta et al., 1998). Furthermore, there is controversy over the specificity of the hyperthymic trait in general: on the one hand, an investigation by Kesebir et colleagues (2005) showed that both bipolar patients and their relatives had a significantly higher prevalence of hyperthymic temperament compared to controls. However, on the other, another study found that hyperthymic scores were higher in normal controls than in bipolar patients and their healthy relatives (Mendlowicz et al., 2005). Another argument against the validity of the hyperthymic temperament construct derives from genetic investigations: The s allele of the 5HTTLPR polymorphism in the serotonin transporter gene was found to be associated with higher TEMPS scores of depressive, anxious, irritable, and particularly cyclothymic temperaments, but not with higher scores of hyperthymic temperament (Gonda et al., 2006). If the hyperthymic temperament due to doubtful validity is subtracted from both groups, in the obese group 77 (53%) of 144 patients and in the control group 10 (15%) of 63 would present with an abnormal temperament which would still be significantly different and might also be more a more reliable estimate of the rate of abnormal temperament in the control group and in the general population. As mentioned before, the association between obesity and cyclothymic traits is in line with previous evidence of high rates of hypomanic symptoms among obese patients seeking surgical treatment (Alciati et al., 2007). Furthermore, cyclothymic temperament has also been associated with eating problems in young women from a student population not psychiatrically ill (Signoretta et al., 2005) and with bulimia and binge eating disorders in atypical depressive patients (Perugi et al., 2006). It has been hypothesized that cyclothymic temperament and related sensation seeking and self-stimulating behaviour might involve any type of potentially addictive substance or activity, such as food, alcohol, drugs, physical exercise, work, travelling, the internet, or sex (Perugi and Akiskal, 2002). Relationships between obesity and addiction have also been reported in clinical (McIntyre et al., 2007) and biological studies (Volkow and Wise, 2005). In the present sample, in addition to a high prevalence of abnormal temperament, differences in acute depressive states, measured by BDI, were also detected. Obese patients were more often depressed compared to the control group, a finding which is in line with the previous literature (Roberts et al., 2000; Onyike et al., 2003; Pickering et al., 2007). Using a cut-off point of N18 in the BDI, 40% of obese patients were considered as having a depressive episode. As expected, a higher prevalence of obesity in women was noted in our sample which is in accordance with previous studies (e.g. Wilborn et al., 2005). Our results here are in line with investigations suggesting a higher prevalence of mental disorders in obesity; however, they contradict other studies which have found a high incidence of mental disorders in the families of obese patients (Black et al., 1992; Roberts et al., 2000; Onyike et al., 2003; Pickering et al., 2007). Along with our psychopathological findings, the high comorbidity of psychiatric diseases in obese patients of other studies might contribute to diagnostic discussions. Our findings might also
159
help obesity change its position in current diagnostic classifications, such as the ICD 10, so that it moves closer towards mental illness rather than solely being subsumed under ‘Endocrine, nutritional and metabolic diseases’. A limitation of our study is that the control group was not matched by demographic characteristics. This control group was chosen as it was a well-characterized sample of our psychiatric C–L service with an emphasis on a somatic disease which was being considered for surgical intervention. On these grounds, the sample of transplantation candidates was considered suitable to be compared with a group hypothesised to show psychiatric abnormality, despite the practical impossibility of matching for age and sex. Higher group differences would be expected, if psychiatrically healthy controls would have been studied. However, such an approach would have been associated with the problems of a “hypernormal” control group which we wanted to avoid in the present study (Buckley et al., 1992). Our results need replication and they make no claim to explain the complex etiology of obesity. However, they have the potential to add an important predisposing factor for the development of this hard-to-treat disease. Our results, together with other similar findings, also suggest that there should be a re-appraisal of the relationship between obesity and mental illness, as has been undertaken with other eating disorders, such as anorexia and bulimia. Furthermore, we would recommend psychiatric evaluation of patients in early stages of weight gain, focusing on affectivity with its spectrum, such as depressive and hypomanic symptoms and abnormal temperament. If pathological, an antidepressive or mood-stabilizing treatment should be considered. Studies of mood stabilizers and prospective observations could shed further insight on this complex interface of major clinical and public health issues. Role of funding source The study was supported by a grant from the Stanley Medical Research Centre, and a grant from the Instituto de Salud Carlos III, Centro de Investigación, Biomédica de Red de Salud Mental, CIBERSAM. Conflict of Interest All authors exclude any conflict of interest.
Acknowledgements The study was supported by a grant from the Stanley Medical Research Centre, and a grant from the Instituto de Salud Carlos III, Centro de Investigación, Biomédica de Red de Salud Mental, CIBERSAM. The first author, BA, receives two grants as principal investigator from the Instituto de Salud Carlos III (FIS CP06/00359 and PI071278). We also thank to Mrs Angela Poeller, Christa Fahmueller-Klose and Diana Wittmann (all Hospital Grosshadern) for their overall great help in the study and to Prof Peter McKenna for his help in revising the manuscript. References Amann, B., Padberg, F., Mergl, R., Naber, D., Baghai, T., Reimers, K., El-Giamal, N., Laakamann, G., in press. An investigation of temperamental traits in patients with somatoform disorder: do they belong to the affective spectrum? Psychosomatics, in press. American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders (4th ed; DSM-IV). APA, Washington DC.
160
B. Amann et al. / Journal of Affective Disorders 118 (2009) 155–160
Akiskal, H.S., 1996. The temperamental foundations of affective disorders. In: Mundt, C., Hahlweg, K., Fiedler, P. (Eds.), Interpersonal Factors in the Origin and Course of Affective Disorders. Gaskell, London, pp. 3–30. Akiskal, H.S., Akiskal, K., 1992. Cyclothymic, hyperthymic and depressive temperaments as subaffective variants of mood disorders. In: Tasman, A., Riba, M.B. (Eds.), Annual Review, vol. 11. American Psychiatric Press, Washington DC, pp. 43–62. Akiskal, H.S., Pinto, O., 2000. Soft bipolar spectrum: footnotes to Kraepelin on the interface of hypomania, temperament and depression. In: Marneros, A., Angst, J. (Eds.), Bipolar Disorders: 100 Years after Manic-depressive Insanity. Kluwer Academic, Dordrecht, pp. 37–62. Akiskal, H.S., Akiskal, K.K., Haykal, R.F., Manning, J.S., Connor, P.D., 2005a. TEMPS-A: progress towards validation of a self-rated clinical version of the Temperament Evaluation of the Memphis, Pisa, Paris, and San Diego Autoquestionnaire. J. Affect. Disord. 85 (1–2), 3–16. Akiskal, H.S., Mendlowicz, M.V., Jean-Louis, G., Rapaport, M.H., Kelsoe, J.R., Gillin, J.C., Smith, T.L., 2005b. TEMPS-A: validation of a short version of a self-rated instrument designed to measure variations in temperament. J. Affect. Disord. 85 (1–2), 45–52. Akiskal, H.S., Akiskal, K., Allilaire, J.F., Azorin, J.M., Bourgeois, M.L., Sechter, D., Fraud, J.P., Chatenet-Duchene, L., Lancrenon, S., Perugi, G., Hantouche, E.G., 2005c. Validating affective temperaments in their subaffective and socially positive attributes: psychometric, clinical and familial data from a French national study. J. Affect. Disord. 85 (1–2), 29–36. Akiskal, H.S., Akiskal, K.K., Lancrenon, S., Hantouche, E., 2006. Validating the soft bipolar spectrum in the French National EPIDEP Study: the prominence of BP-II 1/2. J Affect. Disord. 96 (3), 207–213. Alciati, A., D'Ambrosio, A., Foschi, D., Corsi, F., Mellado, C., Angst, J., 2007. Bipolar spectrum disorders in severely obese patients seeking surgical treatment. J. Affect. Disord. 101 (1–3), 131–138. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J., Erbaugh, J., 1961. An inventory for measuring depression. Arch. Gen. Psychiatry 4, 561–571. Black, D.W., Goldstein, R.B., Mason, E.E., Bell, S.E., Blum, N., 1992. Depression and other mental disorders in the relatives of morbidly obese patients. J. Affect. Disord. 25 (2), 91–95. Bloink, R., Brieger, P., Akiskal, H.S., Marneros, A., 2005. Factorial structure and internal consistency of the German TEMPS-A scale: validation against the NEO-FFI questionnaire. J. Affect. Disord. 85 (1–2), 77–83. Braunig, P., Shugar, G., Kruger, S., 1996. An investigation of the Self-Report Manic Inventory as a diagnostic and severity scale for mania. Compr. Psychiatry 37 (1), 52–55. Bray, G.A., Bellanger, T., 2006. Epidemiology, trends, and morbidities of obesity and the metabolic syndrome. Endocrine 29 (1), 109–117. Buckley, P., O'Callaghan, E., Larkin, C., Waddington, J.L., 1992. Schizophrenia research: the problem of controls. Biol. Psychiatry 32 (3), 215–217 1. Cloninger, C.R., Svrakic, D.M., Przybeck, T.R., 2006. Can personality assessment predict future depression? A twelve-month follow-up of 631 subjects. J. Affect. Disord. 92 (1), 35–44. Elmslie, J.L., Silverstone, J.T., Mann, J.I., Williams, S.M., Romans, S.E., 2000. Prevalence of overweight and obesity in bipolar patients. J. Clin. Psychiatry 61 (3), 179–184. Erfurth, A., Gerlach, A.L., Hellweg, I., Boenigk, I., Michael, N., Akiskal, H.S., 2005a. Studies on a German (Munster) version of the temperament auto-questionnaire TEMPS-A: construction and validation of the briefTEMPS-M. J. Affect. Disord. 85 (1–2), 53–69. Erfurth, A., Gerlach, A.L., Michael, N., Boenigk, I., Hellweg, I., Signoretta, S., Akiskal, K., Akiskal, H.S., 2005b. Distribution and gender effects of the subscales of a German version of the temperament autoquestionnaire briefTEMPS-M in a university student population. J. Affect. Disord. 85 (1–2), 71–76. Fiedorowicz, J.G., Palagummi, N.M., Forman-Hoffman, V.L., Miller del, D., Haynes, W.G., 2008. Elevated prevalence of obesity, metabolic syndrome, and cardiovascular risk factors in bipolar disorder. Ann. Clin. Psychiatry 20 (3), 131–137. Gonda, X., Rihmer, Z., Zsombok, T., Bagdy, G., Akiskal, K.K., Akiskal, H.S., 2006. The 5HTTLPR polymorphism of the serotonin transporter gene is associated with affective temperaments as measured by TEMPS-A. J. Affect. Disord. 91 (2–3), 125–131. Hantouche, E.G., Kochmann, F., Akiskal, H.S., 2001. Evaluation des temperaments affectifs: version complete des outils dáutoevaluation. Encèphale 28 (sp III), 24–30. Hautzinger, M., Bailer, M., Worall, H., Keller, F., 1992. Das Beck Depressionsinventar – BDI. Huber, Bern. Kagan, J., 1994. Galen's Prophecy: Temperament in Human Nature. Basic Books, New York. Kesebir, S., Vahip, S., Akdeniz, F., Yuncu, Z., Alkan, M., Akiskal, H., 2005. Affective temperaments as measured by TEMPS-A in patients with
bipolar I disorder and their first-degree relatives: a controlled study. J. Affect. Disord. 85 (1–2), 127–133. Kraepelin, E., 1909–1915. Psychiatrie Ein Lehrbuch für Studierende und Ärzte. (8. Auflage). Johann Ambrosius Barth, Leipzig. Maina, G., Salvi, V., Vitalucci, A., D'Ambrosio, V., Bogetto, F., 2008. Prevalence and correlates of overweight in drug-naive patients with bipolar disorder. J. Affect. Disord. 110 (1–2), 149–155. Mendlowicz, M.V., Jean-Louis, G., Kelsoe, J.R., Akiskal, H.S., 2005. A comparison of recovered bipolar patients, healthy relatives of bipolar probands, and normal controls using the short TEMPS-A. J. Affect. Disord. 85 (1–2), 147–151. McElroy, S.L., Kotwal, R., Malhotra, S., Nelson, E.B., Keck, P.E., Nemeroff, C.B., 2004. Are mood disorders and obesity related? A review for the mental health professional. J. Clin. Psychiatry 65 (5), 634–651. McIntyre, R.S., McElroy, S.L., Konarski, J.Z., Soczynska, J.K., Bottas, A., Castel, S., Wilkins, K., Kennedy, S.H., 2007. Substance use disorders and overweight/obesity in bipolar I disorder: preliminary evidence for competing addictions. J. Clin. Psychiatry 68 (9), 1352–1357. Onyike, C.U., Crum, R.M., Lee, H.B., Lyketsos, C.G., Eaton, W.W., 2003. Is obesity associated with major depression? Results from the Third National Health and Nutrition Examination Survey. Am. J. Epidemiol. 158 (12), 1139–1147 15. Perretta, P., Akiskal, H.S., Nisita, C., Zaccagnini, E., Lorenzetti, C., Della Santa, M., Cassano, G.B., 1998. The high prevalence of bipolar II and associated cyclothymic and hyperthymic temperaments in HIV patients. J. Affect. Disord. 50, 215–224. Perugi, G., Akiskal, H.S., 2002. The soft bipolar spectrum redefined: focus on the cyclothymic, anxious-sensitive, impulse-dyscontrol, and bingeeating connection in bipolar II and related conditions. Psychiatr. Clin. North Am. 25 (4), 713–737. Perugi, G., Toni, C., Passino, M.C., Akiskal, K.K., Kaprinis, S., Akiskal, H.S., 2006. Bulimia nervosa in atypical depression: the mediating role of cyclothymic temperament. J. Affect. Disord. 92, 91–97. Pickering, R.P., Grant, B.F., Chou, S.P., Compton, W.M., 2007. Are overweight, obesity, and extreme obesity associated with psychopathology? Results from the national epidemiologic survey on alcohol and related conditions. J. Clin. Psychiatry 68 (7), 998–1009. Roberts, R.E., Kaplan, G.A., Shema, S.J., Strawbridge, W.J., 2000. Are the obese at greater risk for depression? Am. J. Epidemiol. 152 (2), 163–170 15. Rozsa, S., Rihmer, A., Ko, N., Gonda, X., Szili, I., Szadoczky, E., Pestality, P., Rihmer, Z., 2006. Affective temperaments: psychometric properties of the Hungarian TEMPS-A. Psychiatr. Hung. 21 (2), 147–160. Sanchez-Moreno, J., Barrantes-Vidal, N., Vieta, E., Martinez-Aran, A., SaizRuiz, J., Montes, J.M., Akiskal, K., Akiskal, H.S., 2005. Process of adaptation to Spanish of the Temperament Evaluation of Memphis, Pisa, Paris and San Diego Scale. Self applied version (TEMPS-A). Actas Esp. Psiquiatr. 33 (5), 325–330. Seidell, J.C., Flegal, K.M., 1997. Assessing obesity: classification and epidemiology. Br Med Bull. 53 (2), 238–252. Shugar, G., Schertzer, S., Toner, B.B., Di Gasbarro, I., 1992. Development, use, and factor analysis of a self-report inventory for mania. Compr. Psychiatry 33 (5), 325–331. Signoretta, S., Maremmani, I., Liguori, A., Perugi, G., Akiskal, H.S., 2005. Affective temperament traits measured by TEMPS-I and emotionalbehavioral problems in clinically-well children, adolescents, and young adults. J. Affect. Disord. 85 (1–2), 169–180. Sullivan, S., Cloninger, C.R., Przybeck, T.R., Klein, S., 2007. Personality characteristics in obesity and relationship with successful weight loss. Int. J. Obes (Lond) 31 (4), 669–674. von Zerssen, D., Akiskal, H.S., 1998. Personality factors in affective disorders: historical developments and current issues with special reference to the concepts of temperament and character. J. Affect. Disord. 51, 1–5. WHO, 1992. The ICD-10 Classification of Mental and Behavioural Disorders. WHO, Geneva. Wilborn, C., Beckham, J., Campbell, B., Harvey, T., Galbreath, M., La Bounty, P., Nassar, E., Wismann, J., Kreider, R., 2005. Obesity: prevalence, theories, medical consequences, management, and research directions. J. Int. Soc. Sports Nutr. 9 (2), 4–31. Wildes, J.E., Marcus, M.D., Fagiolini, A., 2008. Prevalence and correlates of eating disorder co-morbidity in patients with bipolar disorder. Psychiatry Res. 161 (1), 51–58 30. Vahip, S., Kesebir, S., Alkan, M., Yazici, O., Akiskal, K.K., Akiskal, H.S., 2005. Affective temperaments in clinically-well subjects in Turkey: initial psychometric data on the TEMPS-A. J. Affect. Disord. 85 (1–2), 113–125. Volkow, N.D., Wise, R.A., 2005. How can drug addiction help us understand obesity? Nat. Neurosci. 8 (5), 555–560.