Reduced insight in bipolar I disorder: Neurofunctional and neurostructural correlates

Journal of Affective Disorders 116 (2009) 56–63

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

Reduced insight in bipolar I disorder: Neurofunctional and neurostructural correlates A preliminary study M. Varga a,⁎, A. Babovic b, K. Flekkoy c, U. Ronneberg d, N.I. Landro e, A.S. David f, S. Opjordsmoen g a

Department of Acute Psychiatric Emergency Ward, Aker University Hospital, Oslo, Norway Department of Nuclear Medicine, Rikshospitalet-Radiumhospitalet, Oslo, Norway c Department of Geriatric Medicine, Ulleval University Hospital/Dept. of Psychology, University of Oslo, Oslo, Norway d The Norwegian Board of Health, Oslo, Norway e Center for the Study of Human Cognition, Dept. of Psychology, University of Oslo, Oslo, Norway f Division of Psychological Medicine, Institute of Psychiatry, King's College, London, UK g Division of Psychiatry, Ulleval University Hospital/Faculty of Medicine, Institute of Psychiatry, University of Oslo, Oslo, Norway b

a r t i c l e

i n f o

Article history: Received 25 August 2008 Received in revised form 10 November 2008 Accepted 10 November 2008 Available online 5 December 2008 Keywords: Bipolar disorder Insight CT SPECT

a b s t r a c t Background: To correlate measures of insight for own psychopathology to structural and functional brain imaging findings in 21 patients with DSM-IV bipolar I disorder. Methods: Insight was assessed using the Scale to Assess Unawareness of Mental Disorder (SUMD). Resting single photon emission computed tomography (SPECT) and computed tomography (CT) was conducted in patients and 21 normal comparison subjects matched for age, gender and handedness. Results: Reduced general insight and symptom awareness, but not symptom attribution, were significantly related to cortical and subcortical atrophy, respectively. No correlations between SPECT and insight measures were identified. Limitations: Limited sample size and the use of resting state SPECT. Conclusions: General and symptom awareness were related to measures of brain atrophy but not to neurofunctioning as measured by SPECT. Future research should consider the structure and function of specific cortical regions, including the frontal and parietal cortices. © 2008 Elsevier B.V. All rights reserved.

1. Introduction The neurological condition of anosognosia is characterised by a seeming lack of awareness of disabilities caused by brain injury or disease (Babinski, 1914). Lewis (1934) postulated that a disturbance of cerebral function may be implicated in loss of insight in both neurological disorders and functional psychoses. It has been demonstrated that 50–80% of patients with schizophrenia show poor insight into their illness (Amador

⁎ Corresponding author. Dept. of Acute Psychiatric Emergency Ward, Aker University Hospital, N-0514 Oslo, Norway. Tel.: +47 23033396, +47 91346194; fax: +47 23033399. E-mail address: [email protected] (M. Varga). 0165-0327/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2008.11.005

and Gorman, 1998). Based on DSM-IV, one of the major differences between schizophrenia and other psychotic disorders is the lack of insight in schizophrenia. However, insight function is significantly affected in bipolar patients as well, especially during the manic phase but also to some extent in the remitted state (Dell'Osso et al., 2002; Ghaemi and Rosenquist, 2004; Varga et al., 2006). Little is known about the neurobiological underpinnings of insight deficits in general, and the pathophysiology of bipolar disorder is incompletely understood. In schizophrenia, the potential relationship between prefrontal function and insight has been examined in a number of neuropsychological and structural imaging studies (i.e., Larøi et al., 2000; Flashman et al., 2001; Shad et al., 2004). As in schizophrenia, neuropsychological studies indicate that a selective prefrontal

M. Varga et al. / Journal of Affective Disorders 116 (2009) 56–63

(and possibly parietal) dysfunction is present in bipolar disorder (i.e., Goldberg et al., 1995; Olley et al., 2005; Varga et al., 2006), but this does not necessarily imply that a selective structural–biochemical abnormality is present. In a review of CT and MRI studies in the major psychoses, Raz and Raz (1990) found that patients with schizophrenia and affective psychosis did not differ in terms of ventricular enlargement or abnormal widening of the cortical fissures. Bearden et al. (2001) reviewed neuroimaging studies of bipolar patients and found that one of the most prominent findings was the presence of white matter lesions (hyperintensities), located most frequently in the frontal lobes and basal ganglia. According to Strakowski et al. (2005), abnormalities in prefrontal cortical areas, striatum and amygdala seem to exist early in the course of the illness and could possibly predate illness onset. In contrast, abnormalities in the cerebellar vermis, lateral ventricles and other prefrontal regions (e.g., left inferior) appear to develop with repeated affective episodes, and may represent the effects of illness progression and associated factors (Strakowski et al., 2005). Functional neuroimaging studies have demonstrated distinct alterations of cerebral blood flow during different states of bipolar illness (for reviews see Bearden et al., 2001; Strakowski et al., 2005). For example, Benabarre et al. (2004) found in a recent single photon emission computed tomography (SPECT) study regional decreased uptake of 99mTc-HMPAO in the frontal region and the basal ganglia in 17 bipolar I depressed patients. However, Tutus et al. (1998) found increased rCBF in the left frontal lobe in unmedicated unipolar depressed patients as compared to bipolar patients; a difference which disappeared during remission. There were no significant differences in rCBF between bipolar patients and healthy control subjects. PET studies have generally shown overall increased cortical metabolism when patients are in the manic state and decreased cortical metabolism in the same patients in the depressive state (O'Connell,1995; Sarikaya et al.,1999; Friedman et al., 2006 for a review). Studies of euthymic patients suggest a combination of increased activity in anterior cingulate regions and decreased activity in other prefrontal areas, consistent with the presence of mild symptoms of both mania and depression (Friedman et al., 2006). However, the cause of the regional cerebral blood flow (rCBF) deficits is unknown. The aim of the present study was to examine the possible relationship between reduced insight and neurofunctional and neurostructural measures in patients with bipolar I disorder, and to establish a relationship, as not yet formulated, between these markers and the psychopathology of the disorder itself. Neuropsychological tests were included to assess the neurocognitive correlates of insight failure, as already published in Varga et al., 2006. To the best of our knowledge, this is the first study where structural and functional neuroimaging measures have been combined with neuropsychological assessments to investigate structural and functional correlates of compromised insight in bipolar disorder. 2. Materials and methods 2.1. Subjects Twenty-one bipolar I patients, recruited from in- and outpatient wards at Ulleval University Hospital, participated in the

57

study. This group is a sub-set of our larger study group described in Varga et al. (2006). The Structured Clinical Interview for DSMIV (SCID; 1994) was administered to verify the primary axis I diagnosis. A consensus group of two experienced psychiatrists was established (U.R.; S.O.), performing independent ratings based on the SCID protocols with excellent agreement. All patients were less than 60 years of age. The exclusion criteria were as follows: concomitant neurological disorder and/or brain organic conditions, chronic or long-term substance abuse, and electroconvulsive therapy within the last 3 months. Treatment given was part of a standard, clinical regime, and no attempts were made to control for specific medication or nonmedication effects. At evaluation, all but two patients (90.5%) were taking mood-stabilisers (lithium n = 15; valproic acid n =3; carbamazepine n = 1). Three patients (14.3%) were using antipsychotics (olanzapine n = 2; chlorprothixene n = 1), 14.3% antidepressants (citalopram n= 3), and 9.5% hypnotics (promethazin n= 1; zopiclone n = 1). In 19 of the 21 patients, the duration of stable treatment was 6 months or more. Most patients were in a phase of full (n =12) or partial (n = 1) remission from their illness. Full remission requires a period of at least 2 months with no significant symptoms of mania or depression (APA, 1994). For partial remission, (a) the symptoms of a manic or depressive episode are either still present but full criteria are no longer met, or (b) there are no longer any significant symptoms of a manic or depressive episode, but the period of remission has been less than 2 months. One patient (4.8%) was manic at the time of examination, 19% were mildly depressed, 9.5% moderately depressed and 4.8% severely depressed (for the seven depressed patients, the mean MADRS score was 15.4; range 8–28). There were 10 men and 11 women, mean age 41 years (SD =7.8) (see Table 1 for details). As control subjects, 21 healthy volunteers (10 men and 11 women) were recruited among acquaintances and personnel employed in various departments at Ulleval University Hospital; mean age 40 years (SD =11.1). Inclusion criteria were no significant mental illness, no previous or present neurological disorders, no major head trauma, no alcohol and/or drug abuse, and no history of serious mental disorder among first-degree relatives. 2.2. Psychiatric scales and neurocognitive tests A detailed description of the psychometric scales used in the present study has been presented elsewhere (Varga et al., 2006). These included: Global Assessment of Functioning — split version (GAF), Brief Psychiatric Rating Scale (BPRS), Montgomery–Asberg Depression Rating Scale (MADRS), Mania Rating Scale (MRS from SADS-C), Strauss–Carpenter Scale (SCLFS) and Clinical Global Impressions Scale (CGI). Degree of insight into own mental disorder and relevant symptoms was assessed with the Scale to Assess Unawareness of Mental Disorder (SUMD; Amador et al., 1993). The scale has been used in large samples in the DSM-IV field trials and is perhaps the most common scale in current use. The SUMD consists of 3 primary items (SUMD General, Awareness and Misattribution), scored on a Likert-type scale from 1 (good insight) to 5 (no insight). The patients were categorised as having generally “preserved” or “impaired” insight based on a threshold mean score of ≤3.0. The threshold score is identical to those used in other studies (e.g., Larøi et al., 2000; Varga et al., 2006). For the purpose of the present study, data analysis

58

M. Varga et al. / Journal of Affective Disorders 116 (2009) 56–63

Table 1 Demographic, clinical and neuropsychological variables for the patient group and normal controls (mean; SD). Bipolar total (n = 21)

Bipolar1 remitted (n = 12)

Bipolar2 symptomatic (n = 9)

Controls (n = 21)

Demographics Sex (M/F) Age (years) Education (years) Overall IQ

10/11 41.0 (7.8) 14.1 (3.2) 9.84 (2.5)

6/6 40.5 (7.4) 14.1 (3.4) 10.5 (2.0)

4/5 42.2 (8.8) 14.2 (3.0) 8.89 (2.8)

10/11 40.0 (11.1) 14.7 (2.4) 12.6 (1.8)***

Clinical variables Age at onset (years) Illness duration (years) Number of admissions (n) GAF-S GAF-F BPRS total score MADRS MRS from SADS-C CGI total SCLFS total SUMD General (mean score) SUMD Awareness (sum score) SUMD Misattribution (sum score)

25.7 (9.2) 15.6 (6.9) 4.10 (5.6) 61.8 (10.6) 56.8 (13.1) 23.4 (7.2) 6.33 (7.9) 2.62 (11.3) 3.67 (2.0) 14.4 (4.0) 1.35 (0.65) 7.95 (9.06) 6.48 (4.98)

27.8 (9.9) 12.8 (5.7) 2.25 (1.6) 68.5 (5.5) 65.3 (8.1) 19.1 (2.6) 1.58 (1.8) 0.25 (0.9) 2.42 (1.0) 17.2 (2.4) 1.28 (0.4) 4.17 (4.2) 6.08 (6.3)

22.9 (7.9) 19.3 (6.8) 6.56 (8.0) 52.8 (9.1)*** 45.6 (9.4)*** 29.2 (7.3)** 12.7 (8.6)** 5.78 (17.3) 5.33 (1.8)*** 10.8 (2.7)*** 1.44 (0.9) 13.0 (11.4)* 7.00 (2.6)

Neuropsychology a Information (WAIS) Similarities (WAIS) Digit Span total (WAIS) Block Design (WAIS) WCST categories achieved

51.0 (12.3) 54.8 (8.6) 46.4 (10.9) 46.5 (10.4) 4.14 (2.2)

52.2 (13.2) 53.9 (8.9) 51.1 (10.2) 47.5 (8.4) 4.75 (1.5)

49.3 (11.6) 55.9 (8.6) 40.0 (8.7)* 45.2 (12.9) 3.33 (2.7)

57.9 (9.8)* 63.5 (7.9)*** 54.9 (8.9)** 54.4 (8.8)* 4.81 (1.7)

Bipolar1 = in full remission; Bipolar2 = symptomatic (manic, depressed or in partial remission). Levels of significance: *p b 0.05, **p b 0.01, ***p b 0.001. Overall IQ: mean of WAIS measures (raw-scores). a All neuropsychological values are presented as T-scores, apart from WCST categories achieved (number of categories completed).

focused on current awareness and attribution of symptoms at the time of examination. Information regarding relevant symptoms for each patient was retrieved from patient records, the patient's therapists, personnel with day-to-day contact with the patient and from the psychiatrist who assessed the patient on the SCID, GAF, MRS from SADS-C and MADRS. Items in the SUMD referring to irrelevant symptoms for each patient were omitted. The SUMD General score was calculated using the mean sum score across the 3 constituting items, while a sum score of all relevant items/symptoms was derived for the Awareness of symptom and Misattribution of symptom subscales. All patients were assured that information from the SUMD interview was anonymous and therefore would not affect their treatment. The neuropsychological test battery used is described in Varga et al. (2006). Executive functioning, memory, attention and general intelligence were compared between patients and control subjects. The overall IQ estimate used refers to the mean of the WAIS-III subtests Information and Similarities. All patients were tested as outpatients.

turer's instructions (GE Health Care) and lasted approximately 30 min. During the image acquisition period, which was carried out in a quiet room with dampened lightening, the subjects had their eyes closed and arms in a comfortable position to avoid movement artefacts. The evaluations were analysed visually and independently by two experienced nuclear medicine physicians blind for clinical diagnosis. Semiquantitative estimation was done for each subject on a scale from 1 to 5; 1 being no pathology and 5 representing definite pathology. The uptake in cerebellum was considered the normal reference region. When cerebellar uptake was compromised, occipital uptake was used. A relative blood flow ratio (rCBF) was calculated for each region of interest using the average tissue uptake in the cortical region divided by the uptake in cerebellum or the occipital region. Regions of interest (ROIs) were the frontal, parietal, temporal and occipital areas, right and left side. Patients and controls were assessed according to the same criteria. Kappa-values for SPECT examinations ranged from 0.77 (left parietal perfusion) to 0.96 (general perfusion).

2.3. Single-photon emission computed tomography

2.4. Computed tomography

Technetium-99m hexamethylpropylene amine oxime (HMPAO) brain single-photon emission computed tomography (SPECT) was performed in the 21 bipolar patients and the 21 healthy controls, using a double head rotating gamma camera (GR Health Care) in order to examine possible alterations in cerebral blood perfusion. SPECT acquisition of the images was initiated 15 min after the i.v. administration of 99mTc-HMPAO (Ceretec, approximately 700 Mbq) according to the manufac-

All CT examinations were carried out as a routine brain examination with a GE 9800 Quick Scanner (General Electric, Milwaukee, Wi.). The scans were obtained transaxially without angulation at 120 kW and with a scan time of 3 s. The slice thickness through the posterior fossa and basal ganglia was 5 mm using 170 mAs, thereafter 10 mm using 140 mAs to the vertex. The CT scans were assessed independently by two experienced senior neuroradiologists blind to clinical diagnosis.

M. Varga et al. / Journal of Affective Disorders 116 (2009) 56–63

Based on a visual impression, the scans were rated manually according to subcortical and cortical atrophy on a four-point scale ranging from normal, via slight and moderate atrophy, to severe atrophy. Assessment of atrophy was conducted bilaterally along general, frontal, parietal, temporal and occipital dimensions. The inter-rater reliability measure (Kappa) for subcortical atrophy was 0.48, and 0.90 for cortical atrophy when rated according to normal versus pathological results. Because of the relatively low Kappa-value for subcortical atrophy, the most stringent criteria for atrophy were utilised in the further analyses of the data. 2.5. Procedures All patients underwent neuroimaging assessments consisting of SPECT and CT to assess brain function and structure, respectively. The insight and psychiatric assessments and the neuropsychological testing were performed according to standardised procedures without knowledge of the SPECT or CT results. All subjects gave written informed consent before participating in the study. The regional Ethics Committee approved the project. 2.6. Statistical analyses Statistical analyses were performed using SPSS for Windows. Analysis of variance (ANOVA), independent samples ttests (two-tailed), and Pearson's and Spearman's correlations were utilised depending on the distribution of data. Given the exploratory nature of the study, Bonferroni adjustments were not performed (see Perneger, 1998). 3. Results 3.1. Demographic, clinical, and neurocognitive characteristics Patients and controls were carefully matched on age, gender, handedness and years of education (Table 1) but differed significantly in terms of overall intellectual functioning (p b 0.001). However, most patients had at least a college education and 16 were currently employed. Most patients were in a phase of full or partial remission from their illness, although

59

clinical information revealed that some of them still showed subtle social and occupational dysfunctions (e.g., social withdrawal, unemployment/not being able to keep a job). There were significant group differences between remitted and nonremitted/symptomatic patients on most clinical characteristics except for age at illness onset, duration of illness, number of admissions and the mania scores. Neuropsychological performance was significantly impaired in the patients compared to the controls on all cognitive tasks except the Wisconsin Card Sorting Test (WCST; Varga et al., 2006). Remitted patients scored significantly better only on WAIS Digit Span (p b 0.05) compared to non-remitted patients. Thirteen patients (61.9%) were rated as having complete overall insight into their mental disorder as measured by SUMD General (score ≤3), whereas 38.1% failed (score ≥3) to some extent to recognise one or more of the three general aspects of their illness (i.e., the presence of a mental illness, benefits achieved with medication, and the social consequences of the disorder). Fifteen patients (71.4%) were rated as being largely unaware of the presence of at least one relevant symptom (score ≥3) (mostly of a depressive nature, e.g., anhedonia, social isolation etc.). Five of the six individuals who showed high to moderate levels of symptom awareness attributed these symptoms to their mental disorder. There were no significant differences in insight between remitted and symptomatic patients. 3.2. Neuroimaging A summary of SPECT and CT results can be seen in Table 2. Overall, in ten patients (47.6%) there was slight to moderate reduced global cerebral blood flow as assessed by SPECT, particularly in the frontal regions (33.3%). However, the relative 99mTc-HMPAO activity in the selected brain regions (see Materials and methods) revealed no significant differences between patients with bipolar I disorder and healthy controls. Pathological CT-findings were identified in 66.7% of the patients. Of these, 50% showed indications of frontal atrophy, 25% had general atrophy, 16.7% parietal atrophy, and 8.3% temporal atrophy. Enlargement of the third and anterior lateral ventricles were observed in 47.6% of the cases (subcortical atrophy). No relationship was found for age and SPECT measures, probably reflecting the use of age-related rCBF in areas of reference.

Table 2 CT (n = 20) and SPECT (n = 21) findings. Atrophy

Bipolar total (n = 21)

Bipolar1 remitted (n = 12)

Bipolar2 symptomatic (n = 9)

Controls (n = 21)

CT no cortical atrophy Slight cortical atrophy – Frontal – Parietal – Temporal – General Moderate cortical atrophy – Frontal CT no subcortical atrophy Slight subcortical atrophy Moderate subcortical atrophy SPECT normal Slightly abnormal Moderately abnormal

8 (40%) 10 (50%) 4 (20%) 2 (10%) 1 (5%) 3 (15%) 2 (10%) 2 (10%) 10 (50%) 9 (45%) 1 (5%) 11 (52.4%) 8 (38.1%) 2 (9.52%)

5 (25%) 5 (25%) 3 (15%) 1 (5%) 1 (5%) 0 1 (5%) 1 (5%) 6 (30%) 5 (25%) 0 8 (38.1%) 2 (9.52%) 2 (9.52%)

3 5 1 1 0 3 1 1 4 4 1 3 6 0

13 (61.9%) 8 (38.1%) 0

Bipolar1 = in full remission; Bipolar2 = symptomatic (manic, depressed or in partial remission). CT-data: one patient missing.

(15%) (25%) (5%) (5%) (15%) (5%) (5%) (20%) (20%) (5%) (14.3%) (28.6%)

60

M. Varga et al. / Journal of Affective Disorders 116 (2009) 56–63

Table 3 Correlations between SUMD subscores and neuroimaging measures: Spearman.

Table 5 Correlations between clinical variables and neuroimaging measures (Spearman).

SUMD

CT SPECT

Cortical atrophy Subcortical atrophy General perfusion

SPECT

General

Awareness

Misattribution

0.24 0.46* − 0.003

0.45* 0.38 − 0.10

− 0.01 0.12 − 0.21

Levels of significance: *p b 0.05.

3.3. Insight and neuroimaging Table 3 presents Spearman's correlations between the different SUMD subscales and neuroimaging measures. Significant correlations emerged between subcortical atrophy (increased lateral and third ventricles) and general illness awareness (SUMD General; p b 0.05). Cortical atrophy was moderately correlated with the Awareness subscale of the SUMD (p b 0.05). No significant correlations were found between SUMD Misattribution and CT findings. There were no significant correlations between any of the SUMD subscales and SPECT measures. 3.4. Insight, clinical variables, neurocognition and neuroimaging The associations between clinical variables and SUMD and neuroimaging measures are presented in Tables 4 and 5. Regarding neurocognition, there was no relationship between WCST and SUMD, but sub-tests of the WAIS-III did correlate with different dimensions of the SUMD (Similarities and General awareness, Digit Span and Symptom awareness, Block Design and Misattribution; see Table 4). Overall, higher levels of functioning and lower levels of psychopathology were associated Table 4 Correlations between clinical and neuropsychological variables, and insight (Pearson). SUMD General

Awareness

Misattribution

Clinical variables GAF-S GAF-F BPRS (total) Anxiety–depression Anergia Thought disturbance Activation Hostility–suspiciousness MADRS (total) MRS from SADS-C (total) CGI (total) SCLFS (total) Number of admissions Illness duration Age at onset Age (years)

− 0.35 − 0.21 0.39 − 0.14 − 0.21 0.78** 0.22 0.81** − 0.24 0.81** 0.45* − 0.27 − 0.09 − 0.20 0.10 − 0.05

− 0.58* − 0.48* 0.74* 0.23 0.06 0.88** 0.62** 0.84** 0.03 0.84* 0.68** − 0.54* 0.18 − 0.01 0.30 0.35

− 0.09 − 0.07 0.07 0.08 0.18 − 0.01 0.11 − 0.07 0.11 − 0.07 0.17 − 0.11 0.13 0.13 0.13 0.27

Neuropsychology Information (WAIS) Similarities (WAIS) Digit Span (WAIS) Block Design (WAIS) WCST categories achieved

− 0.11 − 0.48* − 0.29 − 0.15 − 0.26

− 0.08 − 0.27 − 0.48* − 0.02 − 0.43

Levels of significance: *p b 0.05, **p b 0.01 (two-tailed).

0.25 0.06 0.08 0.54* 0.11

GAF-S GAF-F BPRS (total) Illness duration Age at onset Age (years)

CT

CT

General perfusion

Cortical

Subcortical

− 0.20 − 0.13 − 0.06 0.29 − 0.12 0.02

− 0.26 − 0.29 0.07 − 0.22 0.50* 0.58**

− 0.36 − 0.23 0.06 − 0.05 0.44 0.54*

Levels of significance: *p b 0.05, **p b 0.01 (two-tailed). SPECT General perfusion: based on scale from 1 to 5. CT Cortical and Subcortical: scale from 0 to 3.

with preserved general and symptom awareness. Symptoms of mania, thought disturbance and hostility–suspiciousness were significantly correlated with reduced general and symptom awareness. There were no associations between any of the clinical and neurocognitive variables and the SUMD Misattribution score, except for a positive correlation between Misattribution and WAIS Block Design (a primarily parietal measure). There were no correlations between clinical variables and the rCBF indices (Table 5). However, the patients' age was significantly related to both cortical (p b 0.01) and subcortical (p b 0.05) atrophy, and age at illness onset to cortical atrophy (p b 0.05). No significant correlations were found between neurocognitive and neuroimaging measures. 4. Discussion In the present study, lack of insight for own psychopathology in a sample of primarily remitted bipolar I patients was not found to be associated with any specific functional brain deficits as assessed by SPECT. On the other hand, poor general and symptom awareness (SUMD General and Awareness subscale scores) was significantly correlated to subcortical and cortical atrophy as assessed by CT. Using a similar methodology, Larøi et al. (2000) found a significant correlation between cortical atrophy and a general SUMD-score in patients with chronic schizophrenia. No significant correlations emerged between SUMD Misattribution and any of the neuroimaging measures. No correlations were found between neurocognition and neuroimaging measures, precluding a neurocognitive interpretation of neuroimaging markers. This is of importance, since the neuropsychological test battery used in this study covered areas of neurocognition known to be linked to insight deficits (e.g., Lysaker et al., 1998; Amador and David, 2004) as well as neurocognitive functions at risk for being compromised in schizophrenia and bipolar disorder (especially prefrontal executive tests and visuospatial functions). Expansion of the battery would have been desirable, especially for executive measures (e.g., STROOP), but was decided against in order to preserve test-motivation in the patients. Given the general nature of atrophy identified in the present study, the results are consistent with a relationship between reduced insight and both prefrontal and parietal dysfunction. No correlations were found for neuropsychological and neuroimaging measures, probably reflecting the insensitivity of unstimulated SPECT. The lack of correlations between neurocognition and neuroimaging precludes a neuroanatomical identification of

M. Varga et al. / Journal of Affective Disorders 116 (2009) 56–63

neurocognition. The results do, however, support our earlier neuropsychological study on a larger group of bipolar I patients, showing impaired insight to be related to frontotemporal (WAIS Similarities and Digit Span; Grooved Pegboard) and parietal (WAIS Block Design) markers (Varga et al., 2006). With the exception of WAIS Block Design, Misattribution was found to be unrelated to all neurocognitive measures in our larger study, which is consistent with the CT scan results. Misattribution appears therefore not to be determined by cognitive impairment but by other self-referential processes. In this context it is interesting that deviant self-understanding is positively related to WAIS Block Design, perhaps suggesting a right hemisphere parietal involvement. Sulcal prominence (cortical atrophy) was found in 57% of the patients. In accordance with previous studies (e.g., Nasrallah et al., 1982; Dougherty and Rauch, 1997; Roy et al., 1998; Stoll et al., 2000), ventricular enlargement was found in about half of the patients. The results suggest that lack of awareness in bipolar patients may be understood as having at least in part a structural basis. Decreased rCBF was indicated in about half of the patients. This is in contradiction to the findings of Tutus et al. (1998), possibly reflecting the difference in medication status of the patients in the two studies. However, there were no significant differences between patients and controls on rCBF measures (Table 2). In the patients, the presence or absence of blood flow abnormalities was not related to particular symptoms or cognitive profile. This finding is consonant with Tutus et al.'s (1998) results. Although Olley et al. (2005) argue that patients who are in a remitted state may be experiencing manic or depressive symptoms at a subclinical level, the similarities between the groups in this study might be due to a mostly similar psychological state at the time of investigation. In a study looking at the relationship between neuropsychological functioning and cerebral blood flow in bipolar patients (n = 30; manic, hypomanic, depressed and euthymic states), Benabarre et al. (2005) found a laterality difference in the striatal areas between the patients and healthy controls. The authors argue that this may well relate to problems in self-evaluation and insight. We were unable to confirm this pattern in our sample of 21 cases, most of whom were in a euthymic state. It could be argued that the SPECT method is not sensitive enough to detect more subtle functional changes present in the bipolar population. Cognitive dysfunction (in terms of hypo- or hyperfrontality) is most evident when patients are executing cognitive tasks thought to be dependent upon the integrity of different brain areas. Thus, it is possible that SPECT could have yielded indications of brain dysfunction, and similarly a relation to insight, if the patients were engaged in performing specific cognitive tasks such as a NBack paradigm tapping executive working memory functions. Another possibility is that deeper structures are involved (e.g., gyrus cinguli; Shad et al., 2006a), not directly accessible by SPECT. There is also the possibility that no single brain region is consistently related to the symptom unawareness in bipolar patients. Other reasons for absence of associations could include the small sample size and the clinical state (e.g., severity of psychopathology and medication status). In both bipolar disorder and schizophrenia, poor insight has been related to both clinical symptoms and neurocognitive impairment, but results are inconsistent (for details, see Rossell

61

et al., 2003; Amador and David, 2004; Aleman et al., 2006; Varga et al., 2006). Evidence supporting frontal cortical involvement in mediating insight in bipolar disorder is primarily based on cognitive test data. However, recently identified neuroanatomical correlates of poor insight in schizophrenia show some resemblance with anosognosia, which is associated with lesions in the right parietal lobe and bilateral frontal lobes, and give support to the cognitive findings (e.g., Stuss and Benson, 1986; Shad et al., 2006a; Shad et al., 2006b). Unawareness of illness in schizophrenia has also been shown to be significantly associated with smaller brain size (Flashman et al., 2000), frontal atrophy (Larøi et al., 2000; Flashman et al., 2001; Shad et al., 2006a), and increased ventricle–brain ratios (Takai et al., 1992). Flashman et al. (2001), Shad et al. (2006a) and most recently Sapara et al. (2007) also report that different subregions of the frontal lobes may correlate with different dimensions of unawareness. Evidence for a link between the parietal lobes and reduced insight has been equivocal (Barr, 1998), although McEvoy et al. (1996) and Larøi et al. (2000) found a statistically significant association in a group of patients with schizophrenia. The same lines of inquiry are worth pursuing in future research in bipolar disorder. It is important to bear in mind that the frontal lobes are not functionally homogeneous. Different subregions may correlate with awareness. Similarly, modern insight research looks at the various insight dimensions, and future research should utilize these dimensions and examine whether they individually relate to symptoms, cognitive deficits or neuroanatomical measures. For the first time in bipolar subjects, the present study examined the relationship between structural and functional neuroimaging data and dimensions of insight using a comprehensive insight scale. However, the present results are preliminary. The small sample size prevents us from controlling statistically for possible confounding factors such as illness duration and age by using multivariate analyses. The effects of long-term use of psychotropic medications and illness chronicity on insight functioning and brain regional volumes may limit the interpretations of the results. Brain structure and rCBF were rated by experienced neuroradiologists and nuclear medicine physicians blind for clinical diagnosis. Still, the analyses were semiqualitative and may have lacked in precision, especially for subcortical structures. Finally, the SUMD, originally designed to measure level of insight in psychotic patients with schizophrenia, may not be ideal for use in remitted or low-symptomatic bipolar patients as several items might be irrelevant and therefore have to be omitted. The present insight assessment is limited by mainly depressive symptoms. However, the General subscale represents dimensions of insight applicable for any psychiatric symptomatology, including mood symptoms. In this respect, the SUMD can be adapted to non-psychotic samples. Several insight instruments exist, which may be considered in future research. The SUMD is, however, the most widely used insight scale in psychosis research and is unique in its detailed assessment of patients' awareness of, and attribution for, a wide range of signs and symptoms. The Schedule for the Assessment of Insight — Expanded Version (SAI-E; Kemp and David, 1997) has been applied widely in both Western and non-Western countries, but might be more suitable for first-onset psychosis. The Birchwood Insight Scale (Birchwood et al., 1994) is a selfreport scale also widely used in psychosis research, and has the advantage of being time-saving and less costly than clinical

62

M. Varga et al. / Journal of Affective Disorders 116 (2009) 56–63

interviews such as the SUMD. Jonsdottir et al. (2008), however, warrant cautious use of the scale across different diagnostic groups, as the psychometric properties of the scale were found to be poorer in bipolar disorders than in schizophrenia. Beck's Cognitive Insight Scale (Beck et al., 2004), designed for assessment of self-reflection on patients' anomalous experiences and interpretations of own beliefs, has been reported by Engh et al. (2007) to be applicable both for schizophrenia, bipolar disorder and control subjects (when items referring to psychotic experiences are omitted). In spite of its clear limitations, this study should encourage research to more fully explore specific brain regions that may be involved in mediating insight deficits, the prefrontal and parietal cortex. At the time of execution of this study, the present imaging techniques were the only one available. In future research, imaging techniques such as functional MRI and Proton magnetic resonance spectroscopy (1HMRS) might be the methods of choice to identify neurobiological correlates of insight dimensions in bipolar disorder because of the possibilities they offer for examining the biochemistry of various brain regions. An increased understanding of the psychobiology of insight may facilitate the development of cognitive remediation strategies and/or novel psychopharmacological interventions to enhance insight and improve clinical outcome. Role of the funding source Funding sources were the Norwegian Research Council for Science and the Humanities, and the Norwegian Foundation for Health and Rehabilitation. The support served scientific purposes only as part of the first author's doctoral thesis. No commercial interests were involved. Conflict of interest None.

Acknowledgements The study was supported by grants from the Norwegian Research Council for Science and the Humanities, and the Norwegian Foundation for Health and Rehabilitation. Thanks are extended to Dr. Monika Haakonsen, Dept. of Neuroradiology, Ulleval University Hospital, Oslo, Norway; Dr. Reidar Dullerud, Lovisenberg Hospital, Oslo, Norway; and Dr. Per Mathisen, Sørlandet Hospital HF, Kristiansand, Norway.

References

Aleman, A., Agrawal, N., Morgan, K.D., David, A.S., 2006. Insight in psychosis and neuropsychological function: meta-analysis. Br. J. Psychiatry 189, 204–212. Amador, X.F., Gorman, J.M., 1998. Psychopathologic domains and insight in schizophrenia. Psychiatr. Clin. North Am. 21, 27–42. Amador, X.F., David, A.S. (Eds.), 2004. Insight and Psychosis, 2nd ed. Oxford University Press, New York. Amador, X.F., Strauss, D.H., Yale, S.A., Flaum, M.M., Endicott, J., Gorman, J.M., 1993. Assessment of insight in psychosis. Am. J. Psychiatry 150, 873–879. American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental Disorders, 4th Edition. American Psychiatric Press, Washington DC. (DSM-IV). Babinski, M.J.,1914. Contribution à l'étude des troubles mentaux dans l'hémiplégie organique cérébrale. Rev. Neurol. (Paris) 1, 845–848. Barr, W.B., 1998. Neurobehavioral disorders of awareness and their relevance to schizophrenia, In: Amador, X.F., David, A.S. (Eds.), Insight and Psychosis, 1st ed. Oxford University Press, New York, pp. 107–141. Bearden, C.E., Hoffman, K.M., Cannon, T.D., 2001. The neuropsychology and neuroanatomy of bipolar affective disorder: a critical review. Bipolar Disord. 3, 106–150.

Beck, A.T., Baruch, E., Balter, J.M., Steer, R.A., Warman, D.M., 2004. A new instrument for measuring insight: the Beck Cognitive Insight Scale. Schizophr. Res. 68, 319–329. Benabarre, A., Vieta, E., Martin, F., Lomena, F., Martinez-Aran, A., Colom, F., Corbella, B., Gasto, C., Valdes, M., 2004. Clinical value of 99mTc-HMPAO SPECT in depressed bipolar I patients. Psychiatry Res. 132, 285–289. Benabarre, A., Vieta, E., Martinez-Aran, A., Garcia-Garcia, M., Martin, F., Lomena, F., Torrent, C., Sanchez-Moreno, J., Colom, F., Reinares, M., Brugue, E., Valdes, M., 2005. Neuropsychological disturbances and cerebral blood flow in bipolar disorder. Aust. N. Z. J. Psychiatry 39, 227–234. Birchwood, M., Smith, J., Drury, V., Healy, J., Macmillan, F., Slade, M., 1994. A self-report insight scale for psychosis: reliability, validity and sensitivity to change. Acta Psychiatr. Scand. 89, 62–67. Dell'Osso, L., Pini, S., Cassano, G.B., Mastrocinque, C., Seckinger, R.A., Saettoni, M., Papasogli, A., Yale, S.A., Amador, X.F., 2002. Insight into illness in patients with mania, mixed mania, bipolar depression and major depression with psychotic features. Bipolar Disord. 4, 315–322. Dougherty, D., Rauch, S.L., 1997. Neuroimaging and neurobiological models of depression. Harv. Rev. Psychiatr. 5, 138–159. Engh, J.A., Friis, S., Birkenaes, A., Jonsdottir, H., Ringen, P.A., Ruud, T., Sundet, K.S., Opjordsmoen, S., Andreassen, O.A., 2007. Measuring cognitive insight in schizophrenia and bipolar disorder: a comparative study. BMC Psychiatry 11 7:71. Flashman, L.A., McAllister, T.W., Andreasen, N.C., Saykin, A.J., 2000. Smaller brain size associated with unawareness of illness in patients with schizophrenia. Am. J. Psychiatry 157, 1167–1169. Flashman, L.A., McAllister, T.W., Johnson, S.C., Rick, J.H., Green, R.L., Saykin, A.J., 2001. Specific frontal lobe subregions correlated with unawareness of illness in schizophrenia: a preliminary study. J Neuropsychiatry Clin Neurosci 13, 255–257. Friedman, J.N., Hurley, R.A., Taber, K.H., 2006. Bipolar disorder: imaging state versus trait. J. Neuropsychiat. Clin. Neurosci. 18, 296–301. Ghaemi, S.N., Rosenquist, K.J., 2004. Is insight in mania state-dependent? A meta-analysis. J. Nerv. Ment. Dis. 192, 771–775. Goldberg, J.F., Harrow, M., Grossman, L.S., 1995. Course and outcome in bipolar affective disorder: a longitudinal follow-up study. Am. J. Psychiatry 152, 379–384. Jonsdottir, H., Engh, J.A., Friis, S., Birkenaes, A., Ringen, P.A., Vaskinn, A., Sundet, K., Opjordsmoen, S., Andreassen, O.A., 2008. Measurement of insight in patients with bipolar disorder: are self-rated scales developed for patients with schizophrenia applicable? J. Nerv. Ment. Dis. 196, 333–335. Kemp, R., David, A., 1997. Insight and compliance. In: Blackwell, B. (Ed.), Treatment Compliance and The Therapeutic Alliance in Serious Mental Illness. Harwood Academic Publishers, pp. 61–84. Larøi, F., Fannemel, M., Rønneberg, U., Flekkøy, K., Opjordsmoen, S., Dullerud, R., Haakonsen, M., 2000. Unawareness of illness in chronic schizophrenia and its relationship to structural brain measures and neuropsychological tests. Psychiatry Res Neuroimage Section 100, 49–58. Lewis, A., 1934. The psychopathology of insight. Br. J. Med. Psychol. 14, 332–348. Lysaker, P.H., Bell, M.D., Bryson, G., Kaplan, E., 1998. Neurocognitive function and insight in schizophrenia: support for an association with impairments in executive function but not with impairments in global function. Acta Psychiatr. Scand. 97, 297–301. McEvoy, J.P., Hartman, M., Gottlieb, D., Goodwin, S., Apperson, L.J., Wilson, W., 1996. Common sense, insight, and neuropsychological test performance in schizophrenic patients. Schizophr. Bull. 22, 635–641. Nasrallah, H.A., McCalley-Whitters, M., Jacoby, C.G., 1982. Cerebral ventricular enlargement in young manic males: a controlled CT study. J. Affect. Disord. 4, 15–19. O'Connell, R.A., 1995. Psychiatric disorders, In: Van Heertum, R.L., Tikofsky, R.S. (Eds.), Cerebral SPECT Imaging, 2nd edn. Raven Press Ltd, New York. Olley, A., Mahli, G.S., Mitchell, P.B., Batchelor, J., Lagopoulos, J., Austin, M.P., 2005. When euthymia is just not enough: the neuropsychology of bipolar disorder. J. Nerv. Ment. Dis. 193, 323–330. Perneger, T.V., 1998. What's wrong with Bonferroni adjustments. Br. Med. J. 18, 1236–1238. Raz, S., Raz, N., 1990. Structural brain abnormalities in the major psychoses: a quantitative review of the evidence from computerized imaging. Psychol. Bull. 108, 93–108. Rossell, S.L., Coakes, J., Shapleske, J., Woodruff, P.W., David, A.S., 2003. Insight: its relationship with cognitive function, brain volume and symptoms in schizophrenia. Psychol. Med. 33, 111–119. Roy, P.D., Zipursky, R.B., Saint-Cyr, J.A., Bury, A., Langevin, R., Seeman, M.V., 1998. Temporal horn enlargement is present in schizophrenia and bipolar disorder. Biol. Psychiatry 44, 418–422. Sapara, A., Cooke, M., Fannon, D., Francis, A., Buchanan, R.W., Anilkumar, A.P., Barkataki, I., Aasen, I., Kuipers, E., Kumari, V., 2007. Prefrontal cortex and insight in schizophrenia: a volumetric MRI study. Schizophr. Res. 89, 22–34. Sarikaya, A., Karasin, E., Cermik, T.F., Abay, E., Berkarda, S., 1999. Evaluation of dysthymic disorder with technetium-99 m hexamethylpropylene amine

M. Varga et al. / Journal of Affective Disorders 116 (2009) 56–63 oxime brain single-photon emission tomography. Eur. J. Nucl. Med. 26, 260–264. Shad, M.U., Muddasani, S., Sahni, S.D., Keshavan, M.S., 2004. Insight and prefrontal cortex in first-episode schizophrenia. NeuroImage 22,1315–1320. Shad, M.U., Muddasani, S., Keshavan, M.S., 2006a. Prefrontal subregions and dimensions of insight in first-episode schizophrenia — a pilot study. Psychiatry Res. 146, 35–42. Shad, M.U., Tamminga, C.A., Cullum, M., Haas, G.L., Keshavan, M.S., 2006b. Insight and frontal cortical function in schizophrenia: a review. Schizophr. Res. 86, 54–70. Stoll, A.L., Renshaw, P.F., Yurgelun-Todd, D.A., Cohen, B.M., 2000. Neuroimaging in bipolar disorder: what have we learned? Biol. Psychiatry 15, 505–517. Strakowski, S.M., DelBello, M.P., Adler, C.M., 2005. The functional neuroanatomy of bipolar disorder: a review of neuroimaging findings. Mol. Psychiatry 10, 105–116.

63

Stuss, D.T., Benson, D.F., 1986. The Frontal Lobes. Raven Press, New York. Takai, A., Uematsu, M., Ueki, H., Sone, K., 1992. Insight and its related factors in chronic schizophrenic patients: a preliminary study. Eur. Psychiatry 6, 159–170. Tutus, A., Simsek, A., Sofuoglu, S., Nardali, M., Kugu, N., Karaaslan, F., Gonul, A.S., 1998. Changes in regional cerebral blood flow demonstrated by single photon emission computed tomography in depressive disorders: comparison of unipolar and bipolar subtypes. Psychiatry Res. 28, 169–177. Varga, M., Magnusson, A., Flekkoy, K., Ronneberg, U., Opjordsmoen, S., 2006. Insight, symptoms and neurocognition in bipolar I patients. J. Affect. Disord. 91, 1–9.