EMPIRICAL ARTICLE
Neuropsychological Function in Patients with Anorexia Nervosa or Bulimia Nervosa Siri Weider, Cand. Psychol1,2* Marit Sæbï Indredavik, MD, PhD3,4 Stian Lydersen, PhD3 Knut Hestad, PhD1,5
ABSTRACT Objective: This study explored the neuropsychological performance of patients diagnosed with anorexia nervosa (AN) or bulimia nervosa (BN) compared with healthy controls (HCs). An additional aim was to investigate the effect of several possible mediators on the association between eating disorders (EDs) and cognitive function.
mass index (lowest lifetime BMI) and depressive symptoms explained all findings in the BN group. Although this adjustment reduced the difference between the AN and HC groups, the AN group still performed worse than the HCs regarding verbal learning and memory, visual learning and memory, visuospatial ability, working memory, and executive functioning.
Method: Forty patients with AN, 39 patients with BN, and 40 HCs who were comparable in age and education were consecutively recruited to complete a standardized neuropsychological test battery covering the following cognitive domains: verbal learning and memory, visual learning and memory, speed of information processing, visuospatial ability, working memory, executive function, verbal fluency, attention/vigilance, and motor function.
Discussion: Patients with EDs scored below the HCs on several cognitive function measures, this difference being most pronounced for the AN group. The nadir BMI and depressive symptoms had strong mediating effects. Longitudinal studies are needed to identify the importance of weight restoration and treatment of depressive symptoms in the prevention of a possible cognitive decline. C 2014 Wiley Periodicals, Inc. V
Results: The AN group scored significantly below the HCs on eight of the nine measured cognitive domains. The BN group also showed inferior performance on six cognitive domains. After adjusting for possible mediators, the nadir body
Accepted 24 March 2014 Additional Supporting Information may be found in the online version of this article. Supported by The National Program for Integrated Clinical Specialist and PhD Training for Psychologists, Norway. [This program is a joint cooperation between the Universities of Bergen, Oslo, and Tromsï, the Norwegian University of Science and Technology (Trondheim), the Regional Health Authorities, and the Norwegian Psychological Association. The program is funded jointly by the Ministry of Education and Research and the Ministry of Health and Care Services.] *Correspondence to: Siri Weider, Department of Psychology, Norwegian University of Science and Technology, Dragvoll, 7491 Trondheim, Norway. E-mail: [email protected] 1 Department of Psychology, Norwegian University of Science and Technology, Trondheim, Norway 2 Specialised Unit for Eating Disorder Patients, Department of Psychiatry, Levanger Hospital, Health Trust Nord-Trïndelag, Levanger, Norway 3 Regional Centre for Child and Youth Mental Health and Child Welfare, Norwegian University of Science and Technology, Trondheim, Norway 4 Department of Child and Adolescent Psychiatry, St. Olav’s University Hospital, Trondheim, Norway 5 Division of Mental Health, Innlandet Hospital Trust, Hamar, Norway Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eat.22283 C 2014 Wiley Periodicals, Inc. V
International Journal of Eating Disorders 00:00 00–00 2014
Keywords: anorexia nervosa; bulimia nervosa; neuropsychology (Int J Eat Disord 2014; 00:000–000)
Introduction Several studies have reported neuropsychological impairments in patients with eating disorders (EDs). Reviews have identified difficulties in attention, executive function, learning, memory, verbal functioning, and visuospatial ability.1–3 These findings are, however, most consistent in patients with anorexia nervosa (AN), where a specific pattern of neuropsychological difficulties has been suggested, encompassing deficits in central coherence4,5 and in aspects of executive functioning associated with frontal brain function, especially set-shifting.6 There is insufficient research on neuropsychological function in patients with bulimia nervosa (BN),2,7 and existing studies are hampered by small sample sizes.7 In addition, most earlier studies have been domain-specific, studying only a limited area of cognition. At present, there is no evidence of a specific cognitive profile matching that in AN.7 Most studies on EDs have focused on either patients with AN or BN, and few studies have 1
WEIDER ET AL.
investigated the relationship between cognitive function-associated problems in these closely related disorders. The cognitive deficits observed in the ED population have been explained either by malnutrition, assuming that the cognitive problems will improve after refeeding,8 as persistent problems after weight gain caused by the detrimental effects of previous starvation and malnutrition,9 or as predisposing traits with increased risk of developing an ED.10,11 Genetic markers as well as epigenetic influences contributing to development of EDs are currently being sought, suggesting that the observed cognitive difficulties may exist before disease progression and thus contribute to disease development. Twin studies support genetic vulnerability to EDs,12 and findings of altered brain chemistry and structure in patients strongly suggest a neurobiological component of ED etiology.13 How ED-related aspects may themselves affect cognitive functioning is not clear.14 Studies of patients with AN have generally failed to find associations between cognitive difficulties and the current body mass index (BMI; kg/m2).9,15 Additionally, the effect of the nadir (lowest lifetime) BMI on cognitive functioning in EDs is unclear, although researchers have found this factor to be important in regard to enduring decrease of cerebral gray matter in patients recovered from EDs.16 Findings indicate, however, that the nadir BMI does not affect performance in central coherence,17 set-shifting,17 or attention to details.18 Patients with EDs have a high prevalence of comorbid disorders. In a large study on female inpatients with EDs, the most frequent comorbidities were mood disorders (mainly unipolar depression), found in 94% of patients.19 Depression also negatively affects cognitive functioning in various domains.20 The present understanding of the effect of such comorbid psychiatric disorders on cognitive function in EDs is incomplete.21 Although some researchers have failed to find any effect of depression as a confounding variable,15,22,23 others have found that depression plays an important role in ED-associated neuropsychological difficulties.14
Aims of the Study
The aim of this study was to examine the neuropsychological profiles of patients with AN or BN compared with healthy controls (HCs) using a broad battery of neuropsychological tests. We also sought to test the possible mediating effects of IQ, psychotropic medication, and several disease2
specific factors such as BMI, the nadir BMI, and depressive symptoms.
Method Participants A total of 95 patients aged between 17 and 63 years were consecutively invited to participate in this study on cognitive function in EDs. The patients were recruited from two special inpatient ED units at Levanger Hospital between September 2008 and April 2010 and between August 2011 and February 2013. A total of nine patients refused to participate or discontinued testing, leaving 86 patients. The inclusion criteria were (a) being admitted for treatment for AN or BN according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision (DSM-IV-TR)24; (b) having an ED as the primary diagnosis; (c) speaking Norwegian as a first language; and (d) being somatically able to participate in the study. The exclusion criteria were confirmed brain damage (n 5 1; cerebral infarction); psychosis; diabetes (n 5 1); neurological disease; and neuropsychiatric disorders, such as attention-deficit hyperactivity disorder (ADHD; n 5 3), Tourette’s syndrome, autism spectrum disorder, and chronic fatigue syndrome (n 5 2), leaving a total of 79 enrolled participants resulting in a participation rate of 89.8%. The five patients with AN who refused to participate or discontinued the assessment were older than the remaining patients with AN (M 5 43.6 years vs. M 5 27.5 years, respectively) and had a lower BMI (M 5 15.6 vs. M 5 16.2). There were no differences in demographics between the four patients with BN who refused to participate or discontinued the assessment and the remaining patients with BN. A total of 40 patients (38 females and two males) were diagnosed with AN. Among these patients, seven no longer met the BMI criteria for AN (18.5. It is also worth noting that all of these patients fulfilled the criteria for AN according to the new DSM-V. In total, 27 of the patients with AN (67.7%) were diagnosed with restrictive AN, whereas the remaining 13 (32.3%) were diagnosed with binge-purge AN. Additionally, 39 patients (37 females and two males) fulfilled all criteria for BN. Among these, five had previously (>1 year prior) been diagnosed with AN. None of them had been previously hospitalized for AN. Several of the patients suffered from comorbid diagnoses. The most prevalent comorbid diagnoses in the AN group were depression (n 5 17) and post-traumatic stress disorder (PTSD; n 5 5), followed by bipolar II disorder (n 5 2), obsessive compulsive disorder (OCD; n 5 3), and International Journal of Eating Disorders 00:00 00–00 2014
NEUROPSYCHOLOGICAL PERFORMANCE OF PATIENTS WITH AN OR BN
generalized anxiety disorder (GAD; n 5 3). Depression was also the most prevalent comorbid diagnosis in the BN group (n 5 10), followed by PTSD (n 5 4), GAD (n 5 2), OCD (n 5 1), and bipolar disorder (n 5 2). The patients were at different stages in their treatment. However, all patients were recruited in connection with receiving inpatient care. To avoid the effects of acute starvation, no patients were recruited before Day 10 of their admission. In certain instances, because of somatic complications or patients being too severely emaciated to participate, testing was postponed until the patients were more nutritionally and medically stable. At the time of assessment, 36 of the patients with AN (90.0%) were receiving inpatient treatment. Additionally, one patient was in day treatment, and three were outpatients. Among the patients with BN, 32 (82.1%) were receiving inpatient treatment. Another patient was in day treatment, and six were outpatients. The outpatients were between sequential admissions (n 5 6) or had recently been discharged from the units (n 5 3) but still fulfilled all diagnostic criteria for their disorders. Thirteen patients with AN (32.5%) and 14 patients with BN (35.9%) were hospitalized for the first time for ED treatment. The remaining patients were previously hospitalized at least once. The types and frequencies of medication in the AN group were as follows: antidepressants (selective serotonin reuptake inhibitors) (n 5 25), antipsychotics (n 5 15), anxiolytics (n 5 7), hypnotics (n 5 7), thyroid hormones (n 5 3), and antiepileptics (n 5 3, either as a mood stabilizer or as a treatment for peripheral neuropathic pain). Medication use in the BN group was as follows: antidepressants (n 5 26), hypnotics (n 5 7), antipsychotics (n 5 5), anxiolytics (n 5 3), thyroid hormones (n 5 1), and antiepileptics (n 5 1, as a mood stabilizer) (see Supporting Information Table S1 for compound names and average doses). Several of the patients were polymedicated, and the proportion of individuals being medicated in each patient group was 82.5% for AN and 84.6% for BN. On the basis of the skewed sex distribution and a hypothesized higher-than-average educational level in the patient groups, we chose to include an HC group matched for sex, and comparable in age and education, consisting of 40 healthy individuals. The HCs were recruited from different educational facilities, including adult education participants at the Ole Vig Upper Secondary School in Stjïrdal, Sïr-Trïndelag University College, Nord-Trïndelag University College, the Norwegian University of Science and Technology, and the Folkeuniversitetet Adult Education Association, and by word of mouth (n 5 3). All HCs spoke Norwegian as a first language. The exclusion criteria were the same as for the patient groups, with the following additional criteria: (a) a lifetime history of EDs or eating problems, (b) currently on a diet, (c) having a BMI 26 kg/m2, and International Journal of Eating Disorders 00:00 00–00 2014
(d) having a known psychiatric diagnosis. Five participants were excluded based on the exclusion criteria. Comparability of the groups was assured by an individual pairing of one patient with AN with one patient with BN and a subsequent search for an HC of a similar age (63 years), the same sex and a similar educational level. The close match of the patient pairs was possible due to the homogeneity of age and educational level of this ED population. There was, however, greater variability in age and the level of education in the AN group than in the BN group. In an attempt to find an HC comparable with the patients at the extremes, the HCs were, in some cases, a closer match to the patient with AN in the pair than to the patient with BN. This study was approved by the Regional Committee for Medical and Health Ethics of Central Norway (reference 4.2007.2229). All participants gave written informed consent in accordance with the Declaration of Helsinki. Materials Clinical Assessment. All diagnoses were drawn by a specialist on EDs (medical doctor or clinical psychologist) and validated at the day of testing using the Mini International Neuropsychiatric Interview (MINI)25 based on criteria from the DSM-IV-TR.24 All participants completed the Beck Depression Inventory-II (BDI-II),26 as a measure of depressive symptoms, and the Eating Disorder Inventory-2 (EDI-2),27 as a measure of severity of the ED. Cognitive Assessment. The patients completed a comprehensive battery of internationally well-known neuropsychological tests covering a broad age spectrum that measure numerous cognitive domains (see Table 1). A number of these tests were chosen based on earlier research that proved them sensitive to cognitive difficulties in EDs. The level of general intelligence was assessed by the Wechsler Adult Intelligence Scale, 3rd Edition (WAIS-III).29 Based on the knowledge that the Working Memory Index and the Processing Speed Index, as calculated according to the WAIS-III manual, are indexes sensitive for measuring current pathology,30 the General Ability Index (GAI)30 was used as a measure of premorbid intelligence. The GAI is composed of the raw scores from the following subtests: Picture Completion, Matrix Reasoning, Block Design, Vocabulary, Similarities, and Information. Raw scores on the subtests for the Working Memory Index and the Processing Speed Index contributed to the summary scores for various cognitive domains. The WAIS-III profiles of the groups (with nearly identical participants) are presented elsewhere.31 Statistical Analysis The data were analyzed using SPSS 19.0. One-way, between-groups analysis of variance with Sidak post hoc tests were used for comparisons of demographic and
3
WEIDER ET AL. TABLE 1.
Neuropsychological test battery (subtests in brackets)
Cognitive Domain
Test Variables
Verbal learning and memory
California Verbal Learning Test-II (total recall 1–5, short-delay free recall, long-delay free recall), Wechsler Memory Scale-Revised (WMS-R) (Logical Memory I and II) WMS-R (Visual Memory I and II), Rey-Osterrieth Complex Figure Test (ROCF) (immediate recall trial, delayed recall trial) Trail Making Test, Part A, Wechsler Adult Intelligence Scale-III (WAIS-III) (Digit Symbol, Symbol Search), Color-Word Interference Test (CWIT)a (color naming, word reading) WAIS-III (Block Design, Matrix Reasoning), ROCF (copy) Paced Auditory Serial Addition Test (PASAT) 3, PASAT 2, WAIS-III (Letter Number Sequencing, Digit Span), WMS-R (Spatial Span) Category Test Computer Version with 108 cards, Wisconsin Card Sorting Test-64 Card Version (total errors, perseverative responses), Tower Testa (total achievement score), Trail Making Test, Part B, CWIT (inhibition, inhibition/switching) Verbal Fluency Testa (letter, category, switching) Conners’ Continuous Performance Test-II (omissions, commissions, hit RT, hit RT SE, detectability, hit RT by block, hit RT by block SE, hit RT by ISI, hit RT by ISI SE) Grooved Pegboard Test DH, Grooved Pegboard Test NDH, Grip strength DH, Grip strength NDH
Visual learning and memory Speed of information processing Visuospatial ability Working memory Executive function Verbal fluency Attention and vigilance Motor function a
Note: From Delis-Kaplan Executive Function System; DH, dominant hand; NDH, nondominant hand. See Strauss et al.28 for test references.
TABLE 2. Demographic and clinical characteristics of groups with anorexia nervosa and bulimia nervosa compared with healthy controls
Age BMI Nadir BMI Number of years of education Number of years with ED Father’s education status Mother’s education status BDI-II EDI-2 IQ (GAI)
AN (n 5 40)
BN (n 5 39)
HCs (n 5 40)
Mean (SD)
Mean (SD)
Mean (SD)
F
p-value
27.53 (9.724) 16.24 (1.925)a 13.64 (2.434)a 13.53 (2.219) 11.91 (9.520)c 12.93 (3.518) 12.93 (2.956) 30.35 (13.981)c 92.68 (40.335)c 106.53 (19.010)
27.54 (8.741) 21.76 (3.980) 16.89 (2.562)b 12.54 (1.819) 10.79 (7.200)c 12.46 (2.732) 12.08 (2.860) 23.61 (13.176)c 91.34 (46.117) 110.08 (13.519)
27.45 (10.008) 22.41 (1.744) 20.75 (1.648) 13.10 (1.809) N/A 12.68 (3.033) 13.08 (2.759) 5.55 (6.756) 16.70 (11.735) 114.78 (15.040)
0.001 61.293 100.297 2.517 N/A 0.219 1.394 47.685 58.492 2.654
0.999
Neuropsychological Function in Patients with Anorexia Nervosa or Bulimia Nervosa Siri Weider, Cand. Psychol1,2* Marit Sæbï Indredavik, MD, PhD3,4 Stian Lydersen, PhD3 Knut Hestad, PhD1,5
ABSTRACT Objective: This study explored the neuropsychological performance of patients diagnosed with anorexia nervosa (AN) or bulimia nervosa (BN) compared with healthy controls (HCs). An additional aim was to investigate the effect of several possible mediators on the association between eating disorders (EDs) and cognitive function.
mass index (lowest lifetime BMI) and depressive symptoms explained all findings in the BN group. Although this adjustment reduced the difference between the AN and HC groups, the AN group still performed worse than the HCs regarding verbal learning and memory, visual learning and memory, visuospatial ability, working memory, and executive functioning.
Method: Forty patients with AN, 39 patients with BN, and 40 HCs who were comparable in age and education were consecutively recruited to complete a standardized neuropsychological test battery covering the following cognitive domains: verbal learning and memory, visual learning and memory, speed of information processing, visuospatial ability, working memory, executive function, verbal fluency, attention/vigilance, and motor function.
Discussion: Patients with EDs scored below the HCs on several cognitive function measures, this difference being most pronounced for the AN group. The nadir BMI and depressive symptoms had strong mediating effects. Longitudinal studies are needed to identify the importance of weight restoration and treatment of depressive symptoms in the prevention of a possible cognitive decline. C 2014 Wiley Periodicals, Inc. V
Results: The AN group scored significantly below the HCs on eight of the nine measured cognitive domains. The BN group also showed inferior performance on six cognitive domains. After adjusting for possible mediators, the nadir body
Accepted 24 March 2014 Additional Supporting Information may be found in the online version of this article. Supported by The National Program for Integrated Clinical Specialist and PhD Training for Psychologists, Norway. [This program is a joint cooperation between the Universities of Bergen, Oslo, and Tromsï, the Norwegian University of Science and Technology (Trondheim), the Regional Health Authorities, and the Norwegian Psychological Association. The program is funded jointly by the Ministry of Education and Research and the Ministry of Health and Care Services.] *Correspondence to: Siri Weider, Department of Psychology, Norwegian University of Science and Technology, Dragvoll, 7491 Trondheim, Norway. E-mail: [email protected] 1 Department of Psychology, Norwegian University of Science and Technology, Trondheim, Norway 2 Specialised Unit for Eating Disorder Patients, Department of Psychiatry, Levanger Hospital, Health Trust Nord-Trïndelag, Levanger, Norway 3 Regional Centre for Child and Youth Mental Health and Child Welfare, Norwegian University of Science and Technology, Trondheim, Norway 4 Department of Child and Adolescent Psychiatry, St. Olav’s University Hospital, Trondheim, Norway 5 Division of Mental Health, Innlandet Hospital Trust, Hamar, Norway Published online 00 Month 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/eat.22283 C 2014 Wiley Periodicals, Inc. V
International Journal of Eating Disorders 00:00 00–00 2014
Keywords: anorexia nervosa; bulimia nervosa; neuropsychology (Int J Eat Disord 2014; 00:000–000)
Introduction Several studies have reported neuropsychological impairments in patients with eating disorders (EDs). Reviews have identified difficulties in attention, executive function, learning, memory, verbal functioning, and visuospatial ability.1–3 These findings are, however, most consistent in patients with anorexia nervosa (AN), where a specific pattern of neuropsychological difficulties has been suggested, encompassing deficits in central coherence4,5 and in aspects of executive functioning associated with frontal brain function, especially set-shifting.6 There is insufficient research on neuropsychological function in patients with bulimia nervosa (BN),2,7 and existing studies are hampered by small sample sizes.7 In addition, most earlier studies have been domain-specific, studying only a limited area of cognition. At present, there is no evidence of a specific cognitive profile matching that in AN.7 Most studies on EDs have focused on either patients with AN or BN, and few studies have 1
WEIDER ET AL.
investigated the relationship between cognitive function-associated problems in these closely related disorders. The cognitive deficits observed in the ED population have been explained either by malnutrition, assuming that the cognitive problems will improve after refeeding,8 as persistent problems after weight gain caused by the detrimental effects of previous starvation and malnutrition,9 or as predisposing traits with increased risk of developing an ED.10,11 Genetic markers as well as epigenetic influences contributing to development of EDs are currently being sought, suggesting that the observed cognitive difficulties may exist before disease progression and thus contribute to disease development. Twin studies support genetic vulnerability to EDs,12 and findings of altered brain chemistry and structure in patients strongly suggest a neurobiological component of ED etiology.13 How ED-related aspects may themselves affect cognitive functioning is not clear.14 Studies of patients with AN have generally failed to find associations between cognitive difficulties and the current body mass index (BMI; kg/m2).9,15 Additionally, the effect of the nadir (lowest lifetime) BMI on cognitive functioning in EDs is unclear, although researchers have found this factor to be important in regard to enduring decrease of cerebral gray matter in patients recovered from EDs.16 Findings indicate, however, that the nadir BMI does not affect performance in central coherence,17 set-shifting,17 or attention to details.18 Patients with EDs have a high prevalence of comorbid disorders. In a large study on female inpatients with EDs, the most frequent comorbidities were mood disorders (mainly unipolar depression), found in 94% of patients.19 Depression also negatively affects cognitive functioning in various domains.20 The present understanding of the effect of such comorbid psychiatric disorders on cognitive function in EDs is incomplete.21 Although some researchers have failed to find any effect of depression as a confounding variable,15,22,23 others have found that depression plays an important role in ED-associated neuropsychological difficulties.14
Aims of the Study
The aim of this study was to examine the neuropsychological profiles of patients with AN or BN compared with healthy controls (HCs) using a broad battery of neuropsychological tests. We also sought to test the possible mediating effects of IQ, psychotropic medication, and several disease2
specific factors such as BMI, the nadir BMI, and depressive symptoms.
Method Participants A total of 95 patients aged between 17 and 63 years were consecutively invited to participate in this study on cognitive function in EDs. The patients were recruited from two special inpatient ED units at Levanger Hospital between September 2008 and April 2010 and between August 2011 and February 2013. A total of nine patients refused to participate or discontinued testing, leaving 86 patients. The inclusion criteria were (a) being admitted for treatment for AN or BN according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Text Revision (DSM-IV-TR)24; (b) having an ED as the primary diagnosis; (c) speaking Norwegian as a first language; and (d) being somatically able to participate in the study. The exclusion criteria were confirmed brain damage (n 5 1; cerebral infarction); psychosis; diabetes (n 5 1); neurological disease; and neuropsychiatric disorders, such as attention-deficit hyperactivity disorder (ADHD; n 5 3), Tourette’s syndrome, autism spectrum disorder, and chronic fatigue syndrome (n 5 2), leaving a total of 79 enrolled participants resulting in a participation rate of 89.8%. The five patients with AN who refused to participate or discontinued the assessment were older than the remaining patients with AN (M 5 43.6 years vs. M 5 27.5 years, respectively) and had a lower BMI (M 5 15.6 vs. M 5 16.2). There were no differences in demographics between the four patients with BN who refused to participate or discontinued the assessment and the remaining patients with BN. A total of 40 patients (38 females and two males) were diagnosed with AN. Among these patients, seven no longer met the BMI criteria for AN (18.5. It is also worth noting that all of these patients fulfilled the criteria for AN according to the new DSM-V. In total, 27 of the patients with AN (67.7%) were diagnosed with restrictive AN, whereas the remaining 13 (32.3%) were diagnosed with binge-purge AN. Additionally, 39 patients (37 females and two males) fulfilled all criteria for BN. Among these, five had previously (>1 year prior) been diagnosed with AN. None of them had been previously hospitalized for AN. Several of the patients suffered from comorbid diagnoses. The most prevalent comorbid diagnoses in the AN group were depression (n 5 17) and post-traumatic stress disorder (PTSD; n 5 5), followed by bipolar II disorder (n 5 2), obsessive compulsive disorder (OCD; n 5 3), and International Journal of Eating Disorders 00:00 00–00 2014
NEUROPSYCHOLOGICAL PERFORMANCE OF PATIENTS WITH AN OR BN
generalized anxiety disorder (GAD; n 5 3). Depression was also the most prevalent comorbid diagnosis in the BN group (n 5 10), followed by PTSD (n 5 4), GAD (n 5 2), OCD (n 5 1), and bipolar disorder (n 5 2). The patients were at different stages in their treatment. However, all patients were recruited in connection with receiving inpatient care. To avoid the effects of acute starvation, no patients were recruited before Day 10 of their admission. In certain instances, because of somatic complications or patients being too severely emaciated to participate, testing was postponed until the patients were more nutritionally and medically stable. At the time of assessment, 36 of the patients with AN (90.0%) were receiving inpatient treatment. Additionally, one patient was in day treatment, and three were outpatients. Among the patients with BN, 32 (82.1%) were receiving inpatient treatment. Another patient was in day treatment, and six were outpatients. The outpatients were between sequential admissions (n 5 6) or had recently been discharged from the units (n 5 3) but still fulfilled all diagnostic criteria for their disorders. Thirteen patients with AN (32.5%) and 14 patients with BN (35.9%) were hospitalized for the first time for ED treatment. The remaining patients were previously hospitalized at least once. The types and frequencies of medication in the AN group were as follows: antidepressants (selective serotonin reuptake inhibitors) (n 5 25), antipsychotics (n 5 15), anxiolytics (n 5 7), hypnotics (n 5 7), thyroid hormones (n 5 3), and antiepileptics (n 5 3, either as a mood stabilizer or as a treatment for peripheral neuropathic pain). Medication use in the BN group was as follows: antidepressants (n 5 26), hypnotics (n 5 7), antipsychotics (n 5 5), anxiolytics (n 5 3), thyroid hormones (n 5 1), and antiepileptics (n 5 1, as a mood stabilizer) (see Supporting Information Table S1 for compound names and average doses). Several of the patients were polymedicated, and the proportion of individuals being medicated in each patient group was 82.5% for AN and 84.6% for BN. On the basis of the skewed sex distribution and a hypothesized higher-than-average educational level in the patient groups, we chose to include an HC group matched for sex, and comparable in age and education, consisting of 40 healthy individuals. The HCs were recruited from different educational facilities, including adult education participants at the Ole Vig Upper Secondary School in Stjïrdal, Sïr-Trïndelag University College, Nord-Trïndelag University College, the Norwegian University of Science and Technology, and the Folkeuniversitetet Adult Education Association, and by word of mouth (n 5 3). All HCs spoke Norwegian as a first language. The exclusion criteria were the same as for the patient groups, with the following additional criteria: (a) a lifetime history of EDs or eating problems, (b) currently on a diet, (c) having a BMI 26 kg/m2, and International Journal of Eating Disorders 00:00 00–00 2014
(d) having a known psychiatric diagnosis. Five participants were excluded based on the exclusion criteria. Comparability of the groups was assured by an individual pairing of one patient with AN with one patient with BN and a subsequent search for an HC of a similar age (63 years), the same sex and a similar educational level. The close match of the patient pairs was possible due to the homogeneity of age and educational level of this ED population. There was, however, greater variability in age and the level of education in the AN group than in the BN group. In an attempt to find an HC comparable with the patients at the extremes, the HCs were, in some cases, a closer match to the patient with AN in the pair than to the patient with BN. This study was approved by the Regional Committee for Medical and Health Ethics of Central Norway (reference 4.2007.2229). All participants gave written informed consent in accordance with the Declaration of Helsinki. Materials Clinical Assessment. All diagnoses were drawn by a specialist on EDs (medical doctor or clinical psychologist) and validated at the day of testing using the Mini International Neuropsychiatric Interview (MINI)25 based on criteria from the DSM-IV-TR.24 All participants completed the Beck Depression Inventory-II (BDI-II),26 as a measure of depressive symptoms, and the Eating Disorder Inventory-2 (EDI-2),27 as a measure of severity of the ED. Cognitive Assessment. The patients completed a comprehensive battery of internationally well-known neuropsychological tests covering a broad age spectrum that measure numerous cognitive domains (see Table 1). A number of these tests were chosen based on earlier research that proved them sensitive to cognitive difficulties in EDs. The level of general intelligence was assessed by the Wechsler Adult Intelligence Scale, 3rd Edition (WAIS-III).29 Based on the knowledge that the Working Memory Index and the Processing Speed Index, as calculated according to the WAIS-III manual, are indexes sensitive for measuring current pathology,30 the General Ability Index (GAI)30 was used as a measure of premorbid intelligence. The GAI is composed of the raw scores from the following subtests: Picture Completion, Matrix Reasoning, Block Design, Vocabulary, Similarities, and Information. Raw scores on the subtests for the Working Memory Index and the Processing Speed Index contributed to the summary scores for various cognitive domains. The WAIS-III profiles of the groups (with nearly identical participants) are presented elsewhere.31 Statistical Analysis The data were analyzed using SPSS 19.0. One-way, between-groups analysis of variance with Sidak post hoc tests were used for comparisons of demographic and
3
WEIDER ET AL. TABLE 1.
Neuropsychological test battery (subtests in brackets)
Cognitive Domain
Test Variables
Verbal learning and memory
California Verbal Learning Test-II (total recall 1–5, short-delay free recall, long-delay free recall), Wechsler Memory Scale-Revised (WMS-R) (Logical Memory I and II) WMS-R (Visual Memory I and II), Rey-Osterrieth Complex Figure Test (ROCF) (immediate recall trial, delayed recall trial) Trail Making Test, Part A, Wechsler Adult Intelligence Scale-III (WAIS-III) (Digit Symbol, Symbol Search), Color-Word Interference Test (CWIT)a (color naming, word reading) WAIS-III (Block Design, Matrix Reasoning), ROCF (copy) Paced Auditory Serial Addition Test (PASAT) 3, PASAT 2, WAIS-III (Letter Number Sequencing, Digit Span), WMS-R (Spatial Span) Category Test Computer Version with 108 cards, Wisconsin Card Sorting Test-64 Card Version (total errors, perseverative responses), Tower Testa (total achievement score), Trail Making Test, Part B, CWIT (inhibition, inhibition/switching) Verbal Fluency Testa (letter, category, switching) Conners’ Continuous Performance Test-II (omissions, commissions, hit RT, hit RT SE, detectability, hit RT by block, hit RT by block SE, hit RT by ISI, hit RT by ISI SE) Grooved Pegboard Test DH, Grooved Pegboard Test NDH, Grip strength DH, Grip strength NDH
Visual learning and memory Speed of information processing Visuospatial ability Working memory Executive function Verbal fluency Attention and vigilance Motor function a
Note: From Delis-Kaplan Executive Function System; DH, dominant hand; NDH, nondominant hand. See Strauss et al.28 for test references.
TABLE 2. Demographic and clinical characteristics of groups with anorexia nervosa and bulimia nervosa compared with healthy controls
Age BMI Nadir BMI Number of years of education Number of years with ED Father’s education status Mother’s education status BDI-II EDI-2 IQ (GAI)
AN (n 5 40)
BN (n 5 39)
HCs (n 5 40)
Mean (SD)
Mean (SD)
Mean (SD)
F
p-value
27.53 (9.724) 16.24 (1.925)a 13.64 (2.434)a 13.53 (2.219) 11.91 (9.520)c 12.93 (3.518) 12.93 (2.956) 30.35 (13.981)c 92.68 (40.335)c 106.53 (19.010)
27.54 (8.741) 21.76 (3.980) 16.89 (2.562)b 12.54 (1.819) 10.79 (7.200)c 12.46 (2.732) 12.08 (2.860) 23.61 (13.176)c 91.34 (46.117) 110.08 (13.519)
27.45 (10.008) 22.41 (1.744) 20.75 (1.648) 13.10 (1.809) N/A 12.68 (3.033) 13.08 (2.759) 5.55 (6.756) 16.70 (11.735) 114.78 (15.040)
0.001 61.293 100.297 2.517 N/A 0.219 1.394 47.685 58.492 2.654
0.999
