Copyright ª Blackwell Munksgaard 2007

Bipolar Disorders 2007: 9: 1–9 BIPOLAR DISORDERS

1
2 Original Article
3
4
5
6 The Bipolar Depression Rating Scale (BDRS):
7
8 its development, validation and utility
9
10
11
12 Berk M, Malhi GS, Cahill C, Carman AC, Hadzi-Pavlovic D, Hawkins Michael Berka,b,c, Gin S Malhid,e,f,
MT, Tohen M, Mitchell PB. The Bipolar Depression Rating Scale Catherine Cahillf,g, A Catherine
13 (BDRS): its development, validation and utility. Carmana,b, Dusan Hadzi-Pavlovicf,g,
14 Bipolar Disord 2007: xx: xxx–xxx. ª Blackwell Munksgaard, 2007 Mary T Hawkinsa,b, Mauricio
15 Tohenc,h and Philip B Mitchellf,g
16 Objectives: Unipolar and bipolar depression differ neurobiologically a
and in clinical presentation. Existing depression rating instruments, used Barwon Health and the Geelong Clinic, Geelong,
17 b
in bipolar depression, fail to capture the necessary phenomenonlogical 1 Department of, University of Melbourne,
18 2 Melbourne , cOrygen Research Centre, Melbourne,
19 nuances, as they are based on and skewed towards the characteristics of d
unipolar depression. Both clinically and in research there is a growing 43 Victoria, CADE Clinic, Royal North Shore Hospital,
20 need for a new observer-rated scale that is specifically designed to assess
e
Neuroscience Research Group, Psychological
21 bipolar depression. Medicine, Northern Clinical School, University of
22 5 Sydney, Sydney, fMood Disorders Unit, Black Dog
23 Methods: An instrument reflecting the characteristics of bipolar 6 Institute, Prince of Wales Hospital, Sydney,
depression was drafted by the authors, and administered to 122 7 gDepartment of , School of Psychiatry, University of
24
participants aged 18–65 (44 males and 78 females) with a diagnosis of New South Wales, Sydney, NSW, Australia, hEli Lilly
25 & Co., Indianapolis, IN, USA
26 DSM-IV bipolar disorder, who were currently experiencing symptoms of
depression. The Bipolar Depression Rating Scale (BDRS) was
27 administered together with the Hamilton Depression Rating Scale
28 8,9 (HAM-D), Montgomery Asberg Depression Rating Scale (MADRS) and
29 Key words: assessment – bipolar depression –
Young Mania Rating Scale (YMRS).
bipolar disorder – mixed states – rating scale –
30
validation
31 Results: The BDRS has strong internal consistency (CronbachÕs
32 alpha = 0.917), and robust correlation coefficients with the MADRS
Received 23 June 2005, revised and accepted for
33 (r = 0.906) and HAM-D (r = 0.744), and the mixed subscale correlated
publication 12 May 2006
with the YMRS (r = 0.757). Exploratory factor analysis showed a
34 3-factor solution gave the best account of the data. These factors Corresponding author: Professor Michael Berk,
35 corresponded to depression (somatic), depression (psychological) and Barwon Health and Geelong Clinic, Swanston
36 mixed symptom clusters. Centre, PO Box 281, Geelong, Victoria 3220,
37 Australia. Fax: + 61 3 5226 7436;
38 Conclusions: This study provides evidence for the validity of the BDRS e-mail: mikebe@barwonhealth.org.au
39 for the measurement of depression in bipolar depression. These results
40 suggest good internal validity, provisional evidence of interrater This project was funded by an unrestricted
reliability and strong correlations with other depression rating scales. research grant provided by Eli Lilly & Co.
41
42
43
44 Depression rating scales serve an important func- also more likely to be associated with psychotic
45 10 tion in both research and clinical practice. Conse- features, substance misuse and a positive family
46 quently, it is essential that they are accurate and history of BD (2, 3). Studies that have compared
47 the information gathered is clinically salient. Cur- the 2 subtypes of depression have shown that
48 rently available depression rating scales focus on bipolar depression episodes are shorter and more
49 the phenomenology of unipolar major depression, likely to recur (4), and that the risk of suicide is
50 and have been developed and validated in that greater (5). Atypical depressive features such as
51 population. Bipolar depression has a distinct hypersomnia, hyperphagia, fatigue and rejection
52 clinical profile differentiating it from unipolar sensitivity occur with greater frequency in BD, and
53 major depression. For instance, bipolar disorder weight loss is less common. Somatic symptoms are
54 (BD) occurs equally in men and women and has a more common in unipolar depression (5–7).
55 younger age of onset (1), with episodes of illness Indeed, Mitchell et al. identified a number of key
56 beginning and ending more abruptly. Bipolarity is distinguishing features that they describe as a
1
B D I
Journal Name
5
Manuscript No.
3 6
B Dispatch: 19.6.07
Author Received:
Journal: BDI CE: Balaji
No. of pages: 9 PE: Sri
 Berk et al.

1 bipolar ÔsignatureÕ in a bipolar I depressed sample DSM-IV criteria, currently experiencing symptoms
2 (2, 8). These include atypical features and psycho- of depression (but not necessarily fulfilling criteria
3 motor retardation along with symptoms of melan- for a major depressive episode) were included in
4 cholia, namely, worthlessness, anhedonia and the study. The study cohort sampled a broad range
5 unvarying mood, with fewer anxiety symptoms. of illness severity, with both inpatient and outpa-
6 Benazzi (9) has reported that ÔatypicalÕ symptoms tient populations from public and private sector
7 were more common in bipolar II depressed subjects clinics. Participants were recruited through the
8 but there were no differences in rates of psycho- Swanston Centre Psychiatric Unit, Barwon Health
9 motor retardation. However, some doubts have and the Geelong Clinic, Healthscope in Geelong,
10 been raised as to whether depression severity and Australia (n = 66) and through the Mood Disor-
11 melancholia can be used to distinguish bipolar II ders Unit, Bipolar Clinic, Black Dog Institute,
12 disorder from unipolar depression (10–12). Bipolar Prince of Wales Hospital, Sydney, Australia
13 depression is often further complicated by mixed (n = 56). Patients with comorbid psychiatric dis-
14 states, which occur in as many as 49.5% of bipolar order (n = 22), were included if their diagnosis
15 II patients (13, 14). Other features that have been included BD as the primary illness. Participants
16 reported to differ between unipolar and bipolar (n = 34) with medical comorbidity were included,
17 depression have been reviewed (15). These include but participants with severe cognitive impairment
18 a more recurrent pattern of illness, postnatal were excluded from the study.
19 episodes and more abrupt onset and offset of
20 episodes in bipolar depression. Irritability, lability
Procedure
21 and mixed symptoms are also more common in
22 BD, as are psychotic symptoms and premorbid All participants were interviewed by a psychiatrist
23 hyperthymic or cyclothymic temperaments. In 15 (MB or GSM) and 1 other trained researcher
24 contrast, tearfulness, anxiety and initial insomnia 16 (ACC, MTH or CC), using the BDRS, 21-item
25 are more common in unipolar depression, and HAM-D, MADRS and Young Mania Rating
26 episodes of illness may be longer in unipolar Scale (YMRS) (21). The order of the application
27 depression. General depression rating scales, such 17 of the rating scales was randomized in order to
28 as the Inventory of Depression Symptomatology limit fatigue and regression to the mean effects.
29 (IDS) (16), the Hamilton Depression Rating Scale Interrater reliability of the BDRS was assessed by
30 11 (HAM-D) (17) and the Montgomery Asberg both independent and concurrent rating. Data
31 12 Depression Rating Scale (MADRS) (18), have were de-identified prior to statistical analysis,
32 been used in BD, although they fail to capture key utilizing SPSS V13.0 for Windows.
33 elements of bipolar depression, including mixed
34 symptoms (15, 19, 20).
Scale development
35
36 An original 36-item scale was developed. Items
Objectives
37 were based on literature review, clinical experience
38 The Bipolar Depression Rating Scale (BDRS) was and internal consensus in a consultative iterative
39 developed to be sensitive to both typical and process. This was revised to a 24-item scale based
40 atypical symptoms of depression of varying sever- on published literature, research into phenomeno-
41 ity and to assess mixed symptomatology, thus logical differences between unipolar and bipolar
42 reflecting the characteristic features of the depres- depression (2, 8, 15), and internal methodological
43 sive phase of BD. In developing the semi-struc- and statistical review, which was the subject of this
44 tured, observer-rated BDRS, items were derived by study. The process of scale development and
45 13;14 the authors (MB, PBM and GSM) from phenom- validation was guided by an a priori statistical
46 enological research and literature reviews into the 18 plan, developed by a statistician, (DH-P) who has
47 signature of bipolar depression and differences extensive experience in scale development. In the
48 between unipolar and bipolar depression, (2, 7, 8, original 36-item version, duplicate wordings and
49 15) and their clinical experience. operational variants of many items were trialled
50 and discarded. Further items deleted from the
51 original 36-item version included response latency,
52 Methods insight, diminished libido, psychological retarda-
53 tion and emotional responsiveness. Several items
Participants
54 that were split into 2 questions in the original scale
55 Participants (n = 122, 44 male and 78 female,) were subsumed into 1 item (specifically the depres-
56 aged 18–65 years, with a diagnosis of BD, meeting sion and anxiety items). With both the sleep and
2
 Bipolar Depression Rating Scale (BDRS)

1 appetite items, increase and decrease were confla- Factor extraction was done using unweighted least
2 ted into a single item in order to capture both squares (more robust with non-normal 4-point
3 typical and atypical symptoms and give them equal data such as ours) and followed with oblique
4 weight. The 24-item scale was factor analysed, and (Promax) factor rotation as dimensions were
5 after exclusion of the poorly performing items, the expected to be correlated.
6 final 20-item version (see Appendix) was re-ana-
7 lysed. This did not significantly alter the loading of Reliability analyses. Internal consistency reliability
8 the items. Items that performed relatively poorly in was assessed using CronbachÕs alpha, which gives
9 the study of the 24-item version [e.g., had poor an estimate of the internal consistency of the scale.
10 interrater reliability, low factor analysis loading Interrater reliability at the item level was measured
11 (did not load at least 0.3) or correlated poorly with using CohenÕs kappa statistic, which measures the
12 other items] were successively eliminated. All 4 agreement between 2 raters as between – 1 and 1
13 items excluded in the final 20-item version (motor (perfect agreement). CohenÕs kappa can be mis-
14 retardation, functional impairment, increased leading when there is only a small number of
15 interpersonal sensitivity and slowed time percep- ratings and the items have more than 2 levels, so
16 tion) were from the depression subscale of the that it is common to recode items as dichotomous
17 BDRS. (absent ⁄ present). Interrater reliability at the scale
18 The scale was designed to be administrated by level was assessed using intraclass correlations for
19 psychiatrists and other trained observers. Symp- agreement.
20 tom severity was operationalized to be consistent
21 throughout the scale (0 = no symptoms present, Associations between measures. These were all
22 1 = mild, 2 = moderate, 3 = severe). Items were assessed using Pearson correlations (except as
23 developed to take account of both objective above).
24 assessment and self-report by the patient.
25 A detailed semi-structured rating manual was
26 developed in concert with the scale development Results
27 process, with operationalized anchor points
Demographic data
28 defined by response to key questions. There are
29 between 4 and 7 key questions per scale item. For The demographic data and the breakdown of
30 each item, instructions are given regarding rating medication usage in the sample are shown in
31 of observed and reported features. The manual Table 1. Almost half of the sample (46.7%) had
32 includes examples to guide raters and improve experienced more than 20 episodes of clinical
33 standardization and reliability of rating. Detailed depression across their lifespan and 73.6% had at
34 instructions for dealing with confounds, including least 5 episodes of depression. People reported less
35 rapid cycling, discrepancies between observed and mania overall, with 26.5% (n = 30) reporting 2–4
36 reported data, effects of medications, timing, items lifetime episodes of mania and 30% (n = 37)
37 that fall between anchor points, and medical indicating more than 20 episodes. The mean
38 influences that affect scoring were included in the duration of illness was 13.49 years [standard devi-
39 rating manual. Scoring was further standardized ation (SD) = 11.59] with current phase character-
40 with instructions as to how objective clinical istically lasting 1–3 months (47.3%). With regard
41 judgement should temper qualitative responses, in to symptom severity, the mean score on the HAM-
42 situations where subjective and objective responses D was 21.5, while the mean scores on the MADRS
43 are significantly discrepant. The use of the specific and YMRS were 24.0 and 4.2, respectively. As
44 timeframe of Ônow and the last couple of daysÕ was expected, higher scores were seen in inpatients.
45 chosen to maximize the utility of assessment of the The 20-item scale was factor analysed with a
46 current mood state and to be sensitive to the rapid view to eliciting the optimal number of factors to
47 and ultra-rapid cycling commonly seen in the describe the scale. An unweighted least squares
48 disorder. Both the BDRS itself and the rating factor analysis was followed by oblique (Promax)
49 manual are freely available at http://www.barwon- rotations of 2–5 factors. Prior to rotation the
50 health.org.au/bdrs/. eigenvalues for the first 5 factors were 8.34, 2.68,
51 1.13, 0.95 and 0.83, with corresponding percent-
52 ages of variance accounted for of 41.7%, 13.4%,
Statistical analyses
53 5.7%, 4.8% and 4.1%. Therefore, the 3-factor
54 Factor analysis. This was used to identify the rotation was judged to provide the most useful
55 dimensions underlying the items and so construct account of the data. Table 2 shows the factor
56 scales, other than the total, which might be of use. loadings and the correlations of each individual
3
 Berk et al.

1 Table 1. Clinical characteristics of participants (n = 122) toms that load together are shown in bold. The
2 Number of %
correlations of each item with the BDRS total
3 Clinical descriptor participants score were significant at the 0.05 level with the
4 exception of item 18 (increased motor drive). This
5 Marital status Married 52 43.3 lower correlation followed an expected trend for all
Never married 46 38.3
6 Divorced ⁄ widowed ⁄ separated 22 18.3
the mixed subscale symptoms to correlate less
7 Employment Employed ⁄ student 61 50.4 strongly with the total BDRS score. Pearson
8 status Unemployed 42 34.7 correlations between the 3 factors (somatic, psy-
9 Retired 5 4.1 chological and mixed) and the BDRS, HAM-D,
10 Home duties 13 10.7 MADRS and YMRS totals are shown in Table 3.
Education Primary 18 15.5
11 Secondary 41 35.3
A factor correlation matrix, showing the correla-
12 Tertiary 57 49.1 tions of the factors with each other, is shown in
13 Diagnosis Bipolar I 79 64.8 Table 4.
14 Bipolar II 28 23.0 The internal consistency reliability of the scale
15 Bipolar NOS 11 9.0 was assessed by calculation of CronbachÕs alpha.
Cyclothymia 1 0.8
16 Bipolar due to GMC 3 2.5
This was calculated to be 0.92. Correlation coeffi-
17 Medication Lithium 49 40.2 cients of the BDRS with the HAM-D indicated
18 Anticonvulsant 64 52.5 strong positive correlation between these 2 scales
19 Antidepressant 64 52.5 with r = 0.74. The correlation coefficient between
20 Antipsychotic 35 28.7 the BDRS and the MADRS was even stronger, with
Benzodiazepine 5 4.1
21 Non-psychiatric 38 31.1
a value of 0.91. The correlations of the depression
22 No psychotropic agent 10 8.2 subscale (items 1–15) of the BDRS with the MAD-
23 RS and HAM-D were stronger (0.96 and 0.84,
24 Where percentage scores do not total 100, this is due to missing respectively). The total BDRS scores, as expected,
25 variables. Total percentages not equal to 100 are due to rounding. correlated weakly with the YMRS (0.19), while the
NOS = not otherwise specified; GMC = general medical condi-
26 tion.
mixed subscale items (items 16–20) were more
27 strongly correlated with the YMRS (r = 0.76).
28 item with the total BDRS score. The 3 factors were To assess the performance of individual items,
29 labelled as depression (psychological), depression interrater reliability was analysed for a subsample
30 (somatic) and mixed. In each column the symp- of patients (n = 38). Each of these participants
31
32 Table 2. Item-total correlations, factor loadings from a 3-factor oblique rotation and kappa co-efficient for the 20-item Bipolar Depression Rating Scale (BDRS)
33
34 Factor loadingsa
35 Psychological depression Somatic depression Mixed Item-total correlationsb Kappac
36
37 1 Depression 0.419 0.488 )0.070 0.777 0.387
2 Sleep disturbance 0.006 0.587 0.174 0.551 0.408
38
3 Appetite disturbance )0.148 0.838 0.150 0.611 0.638
39 4 Social impairment 0.576 0.203 )0.115 0.665 0.612
40 5 Activity ⁄ energy reduction 0.316 0.566 )0.246 0.702 0.630
41 6 Reduced motivation 0.287 0.593 )0.130 0.728 0.775
42 7 Reduced concentration 0.097 0.604 0.154 0.648 0.541
8 Anxiety 0.451 0.189 0.298 0.635 0.234
43
9 Anhedonia 0.511 0.373 )0.112 0.751 0.678
44 10 Flattened affect 0.400 0.310 )0.281 0.552 0.814
45 11 Worthlessness 0.825 )0.035 0.002 0.704 0.599
46 12 Helplessness 0.808 0.025 0.057 0.759 0.650
47 13 Suicidal ideation 0.621 0.153 0.097 0.715 0.780
14 Guilt 0.900 0.096 0.702 0.435
48 )0.137
15 Psychotic symptoms 0.352 0.209 0.357 0.576 0.771
49 16 Irritability 0.389 )0.125 0.349 0.321 0.439
50 17 Lability 0.214 0.062 0.603 0.376 0.460
51 18 Increased motor drive )0.188 0.055 0.576 0.001 0.680
52 19 Increased speech )0.247 0.272 0.684 0.148 0.212
20 Agitation 0.180 0.595 0.153 0.203
53 )0.153

54 34 aLoadings in bold represent the highest loading of each symptom onto 1 of the 3 factors.
55 b
Correlations are based on total minus the item.
c
56 Calculated for a subset of 38 participants.

4
 Bipolar Depression Rating Scale (BDRS)

1 Table 3. PearsonÕs correlation between scales based on the 3 factors the 2 depression factors show high interrater
2 (mixed, somatic and psychological) and the BDRS, HAM-D, MADRS and reliability, while the mixed factor shows only
YMRS totals
3 moderate reliability.
4 Somatic Psychological BDRS The pattern of item endorsement can be seen in
5 Factors Mixed depression depression totala Table 5. These show the highest item endorsement
6 BDRS total 0.494 0.904 0.947 –
was for memory and concentration difficulties,
7 BDRS total 0.312 0.760 0.768 – followed by depressed mood, motivation and
8 (minus scale) anxiety.
9 MADRS 0.296 0.860 0.881 0.906
10 HAM-D 0.340 0.666 0.717 0.744
YMRS 0.766 0.029 0.030 0.188 Discussion
11
12 All correlations were significant at the 0.01 level (1-tailed). The BDRS is the first instrument specifically
13 a
Scales formed by summing items with high loadings as given in designed to detect and measure the specific signa-
14 Table 2. ture of bipolar depression. The results of this study
15 BDRS = Bipolar Depression Rating Scale; HAM-D = Hamilton confirm that the BDRS is a valid scale for the
Depression Rating Scale; MADRS = Montgomery Asberg
16 Depression Rating Scale; YMRS = Young Mania Rating Scale.
measurement of depression in BD, with good
17 internal validity and provisional data on interrater
18 reliability. An instructive ratersÕ manual was devel-
19 Table 4. Factor correlation matrix showing correlations of the 3 factors oped to accompany the scale to enhance reliability.
(psychological depression, somatic depression and mixed) with each other
20 The internal consistency of the scale was shown to
21 Psychological Somatic Mixed be strong.
22 Factor depression depression Robust correlation coefficients of the BDRS
23 with the MADRS and HAM-D indicate that the
Psychological depression 1.000 0.750 0.026
24 Somatic depression 0.750 1.000 0.048 scale assesses depressive symptomatology well,
25 Mixed 0.026 0.048 1.000 and, as expected, it correlated weakly with the
26 YMRS. The correlation of the mixed subscale with
27 the YMRS confirms the utility of the scale in
28 was rated once by a psychiatrist and once by a accurately assessing mixed symptomatology. The
29 research psychologist. As numbers were too small
30 to reliably calculate kappa coefficients for items Table 5. Pattern of mean item endorsement on the Bipolar Depression
31 with 4 response levels, data were recoded into Rating Scale by 122 participants with bipolar depressed symptoms
32 absent versus present for comparison. The kappa Percentage Percentage
33 coefficients were rated according to criteria speci- of sample of sample
34 fied by Landis and Koch (22) and no ratings fell with with
35 below 0.2 (Table 2). Standard minimum maximum
Mean deviation score score
36 There was good agreement for the scale scores,
37 with 58–66% of ratings within 1 point of each Impaired memory 1.71 0.88 11.5 16.4
38 other. Intraclass correlations were calculated for and concentration
39 the total score and the 3 scale scores to measure Depressed mood 1.65 0.90 10.7 18.0
Reduced motivation 1.60 1.06 19.7 23.8
40 interrater reliability. The intraclass correlation for
Anxiety 1.54 0.90 14.8 13.1
41 the total scale was 0.88 [95% confidence interval Reduced social 1.50 1.01 22.1 16.4
42 (CI) 0.79–0.93], for depression (psychological) engagement
43 19 0.86 (95% CI 0.75–0.92), for depression (somatic) Reduced energy 1.48 0.86 16.4 7.4
44 20 0.85 (95% CI 0.73–0.92) and for the mixed factor and activity
21 0.42 (95% CI 0.13–0.65). Some of these correla- Sleep disturbance 1.48 1.04 19.7 21.3
45
Anhedonia 1.43 0.95 21.3 11.3
46 tions were affected by outliers: for the mixed Worthlessness 1.39 1.09 27.9 19.7
47 factor, excluding 1 patient for whom there was a Hopelessness and 1.38 1.11 31.1 18.0
48 difference of 9 points, whereas all other patients helplessness
49 showed at most a difference of 5 points, gave an Affective flattening 1.33 0.98 24.6 12.3
Guilt 1.25 0.96 27.0 9.0
50 22;23 intraclass correlation of 0.58 (95% CI 0.31–0.77);
Appetite disturbance 1.16 1.04 36.9 10.7
51 for the psychological factor, removing 3 outliers Irritability 0.93 0.84 34.4 4.1
52 (differences of 8, 8 and 12) gave an intraclass Suicidal ideation 0.90 1.09 50.0 14.8
53 24;25 correlation of 0.94 (95% CI 0.89–0.97). The Lability 0.89 0.91 41.8 5.7
54 removal of 3 outliers also changed the correlation Psychotic symptoms 0.45 0.79 72.1 4.6
26 for the total score to 0.94 (95% CI 0.89–0.97). Increased speech 0.30 0.69 81.1 2.5
55
Agitation 0.28 0.63 79.5 2.5
56 However they are analysed, the total score and
5
 Berk et al.

1 greater correlation of the depressive items of the were virtually absent in other participants involved
2 BDRS with the HAM-D and MADRS when with the study. This is compatible with data that
3 compared with the BDRS total correlations con- suggest mixed states are present in a significant
4 firms the use of the latter scales in characterizing minority of individuals with bipolar depression (13,
5 purely depressive symptoms, but suggests they are 14). It is predictable that mixed symptoms would
6 less accurate in measuring the full symptomatic occur with lower frequency than core depressive
7 picture of bipolar depression. items in a depressed sample. Nevertheless, the
8 Factor analysis of the scale showed a 3-factor ability to rate and draw attention to mixed features
9 solution gave the best account of the data. This within bipolar depression is an area of great
10 was in line with expectations as the scale was clinical significance. It is likely that this may be a
11 27 already divided into 2 clear subscales: the marker signifying the need for a divergent thera-
12 depressed subscale and the mixed subscale. Like peutic approach. Antidepressants may worsen
13 many depression scales, such as the Beck Depres- symptoms in depressive mixed states, (28, 29),
14 sion Inventory (BDI) and BDI-II, the Depression which may require preferential treatment with
15 Anxiety Stress Scale (DASS) (23–25) and HAM-D, mood stabilizers (30, 31) or atypical agents (32).
16 the depression subscale of the BDRS loaded into Strengths of the sample include the fact that it
17 psychological depression (anhedonia, worthless- spanned inpatients and outpatients, public and
18 ness, helplessness, guilt) and somatic depression private sectors, examined individuals with a range
19 (reduced concentration and activity and disrupted of severities and involved clinicians and patients at
20 sleep and appetite). The third factor of the scale 2 study sites. Limitations and future directions
21 encompassed the mixed symptoms. As would be arising from the trial include the fact that it did not
22 expected, both depression factors correlated encompass diverse cultures and language groups. It
23 strongly with each other and neither correlated is planned to further translate and validate the
24 strongly with the mixed factor. The slightly weaker scale in non-English speaking samples. Confirmat-
25 loading of symptoms on somatic depression sup- ory validation of the scale would be valuable. The
26 ports the hypothesis that depression is best meas- sample size of 122 was relatively small for factor
27 ured by psychological symptoms, but also may analysis of a 24-item scale, and the cohort for
28 indicate the greater likelihood of such symptoms 29 which interrater reliability ratings were available
29 28 being reported when the subject is depressed, even was small, and therefore data on interrater relia-
30 if they are not diagnostically distinctive (23). As bility should be seen as preliminary. Further
31 predicted, all the mixed symptoms loaded together, evidence is needed to support interrater reliability
32 with the exception of irritability. Irritability is a and validity. It would be useful to study the
33 key item in many scales that assess depressive trajectory of depression using this instrument to
34 symptomatology, and is common in bipolar detect patterns of change over time, and such a
35 depression (26, 27). Its loading on both the confirmatory trial is underway. A small prelimin-
36 depression and mixed factors reflects its commo- ary study evaluating the scaleÕs utility in assessing
37 nality across the phases of BD. Psychotic symp- bipolar compared to unipolar depression has been
38 toms similarly loaded satisfactorily on both undertaken. There was higher illness severity in the
39 psychological depression and mixed factors but unipolar cohort in this study. Differences in phe-
40 this item was included in the mixed factor for nomenology between the groups included more
41 analysis, reflecting its greater association with lability, hypersomnia and increased motor drive in
42 mixed rather than depressive states (2). the bipolar cohort, and more agitation in the
43 Patterns of mean item endorsement give an unipolar group (33). Lastly, examinations of the
44 indication of how the symptoms perform in terms scale in subtypes of BD would be useful.
45 of salience for this sample. The highest mean The BDRS is the first clinician-administered
46 endorsement fell between the mild and moderate depression rating scale tailored to the clinical
47 anchor points. This highlights the fact that the profile of bipolar depression. It includes items for
48 sample population under investigation were dis- rating mixed features, as well as being sensitive to
49 playing a range of severity of symptoms of many phenomenological elements found com-
50 depression and is consistent with the HAM-D monly in bipolar depression, such as hypersomnia
51 and MADRS scores. The most highly endorsed and hyperphagia, which are not picked up by
52 items were impaired memory and concentration, conventional depression measures (15). It is a
53 depressed mood and reduced motivation. The least potentially useful scale for the measurement of
54 endorsed items were agitation, increased motor depression in BD. This study provides supportive
55 drive and increased speech, reflecting the fact that evidence of the validity of the BDRS, with good
56 these items rated highly in a few participants yet internal validity, provisional interrater reliability,
6
 Bipolar Depression Rating Scale (BDRS)

1 and strong correlations with other depression 3 Severe (marked increase in food intake or
2 rating scales. cravings)
3 4 Reduced Social Engagement (Reports reduced
4 social and interpersonal engagement or interactions)
Appendix
5 0 Nil (normal)
6 Bipolar Depression Rating Scale 1 Mild (slight reduction in social engagement
7 Items in the scale are used to rate the severity of with no impairment in social or interpersonal
8 depression or mixed symptoms in patients now and function)
9 during the past couple of days based on a clinical 2 Moderate (clear reduction in social engage-
10 interview. Higher scores indicate more severe ment with some functional sequelae, e.g., avoids
11 symptoms. some social engagements or conversations)
12 1 Depressed Mood (Self-reported and/or observed 3 Severe (marked reduction in social interaction
13 depression as evidenced by gloom, sadness, pessim- or avoidance of almost all forms of social contact,
14 ism, hopelessness, and helplessness) e.g., refuses to answer the phone or see friends or
15 0 Nil family)
16 1 Mild (brief or transient periods of depression, 5 Reduced Energy and Activity (Reduced energy,
17 or mildly depressed mood) drive and goal-directed behaviour)
18 2 Moderate (depressed mood is clearly but not 0 Nil
19 consistently present and other emotions are 1 Mild (able to engage in usual activities but with
20 expressed, or depression is of moderate intensity) increased effort)
21 3 Severe (pervasive or continuous depressed 2 Moderate (significant reduction in energy
22 mood of marked intensity) leading to reduction of some role-specific activities)
23 2 Sleep disturbance: score either A or B (Change 3 Severe (leaden paralysis or cessation of almost
24 in total amount of sleep over a 24-hour cycle, rated all role-specific activities, e.g., spends excessive
25 independently of the effect of external factors) time in bed, avoids answering the phone, poor
26 A Insomnia (reduction in total sleep time) personal hygiene)
27 0 Nil 6 Reduced Motivation (Reports of subjective
28 1 Mild (up to 2 hours) reduction in drive, motivation, and consequent goal-
29 2 Moderate (2–4 hours) directed activity)
30 3 Severe (more than 4 hours) 0 Nil (normal motivation)
31 OR 1 Mild (slight reduction in motivation with no
32 B Hypersomnia (increase in total sleep time, reduction in function)
33 inclusive of daytime sleep) 2 Moderate (reduced motivation or drive with
34 0 Nil significantly reduced volitional activity or requires
35 1 Mild (less than 2 hours, or normal amount but substantial effort to maintain usual level of func-
36 non-restorative) tion)
37 2 Moderate (greater than 2 hours) 3 Severe (reduced motivation or drive such that
38 3 Severe (greater than 4 hours) goal-directed behaviour or function is markedly
39 3 Appetite Disturbance: score either A or B reduced)
40 (Change in appetite and food consumption, rated 7 Impaired Concentration and Memory (Subject-
41 independently of the effect of external factors) ive reports of reduced attention, concentration or
42 A Loss of Appetite memory, and consequent functional impairment)
43 0 Nil 0 Nil
44 1 Mild (no change in food intake, but has 1 Mild (slight impairment of attention, concen-
45 to push self to eat or reports that food has lost tration, or memory with no functional impairment)
46 taste) 2 Moderate (significant impairment of attention,
47 2 Moderate (some decrease in food intake) concentration, or forgetfulness with some func-
48 3 Severe (marked decrease in food intake, hardly tional impairment)
49 eating) 3 Severe (marked impairment of concentration
50 OR or memory with substantial functional impairment,
51 B Increase in Appetite e.g., unable to read or watch TV)
52 0 Nil 8 Anxiety (Subjective reports of worry, tension,
53 1 Mild (no change in food intake, but increased and/or somatic anxiety symptoms e.g., tremor,
54 hunger) palpitations, dizziness, light-headedness, pins and
55 2 Moderate (some increase in food intake, e.g., needles, sweating, dyspnoea, butterflies in the stom-
56 comfort eating) ach, or diarrhoea)
7
 Berk et al.

1 0 Nil 2 Moderate (thoughts of dying or death, but

2 1 Mild (transient worry or tension about minor with no active suicide thoughts or plans)
3 matters) 3 Severe (thoughts or plans of suicide)
4 2 Moderate (significant anxiety, tension 14 Guilt (Subjective sense of self-blame, failure,
5 or worry, or some accompanying somatic features) or remorse for real or imagined past errors)
6 3 Severe (marked continuous anxiety, tension or 0 Nil
7 worry that interferes with normal activity, or panic 1 Mild (slight decrease in self-esteem or
8 attacks) increased self-criticism)
9 9 Anhedonia (Subjectively reduced ability to 2 Moderate (significant thoughts of failure, self-
10 experience pleasure in usual activities) criticism, inability to cope, or ruminations regard-
11 0 Nil ing past failures and the effect on others; able to
12 1 Mild (slight reduction in pleasure from usually recognize as excessive)
13 pleasurable activities) 3 Severe (marked, pervasive, or persistent guilt,
14 2 Moderate (significant reduction in pleasure e.g., feelings of deserving punishment, or does not
15 from usually pleasurable activities; some pleasure clearly recognize as excessive)
16 from isolated activities retained) 15 Psychotic Symptoms (Presence of overvalued
17 3 Severe (complete inability to experience plea- ideas, delusions, or hallucinations)
18 sure) 0 Nil (absent)
19 10 Affective Flattening (Subjective sense of 1 Mild (mild overvalued ideas, e.g., self-criticism
20 reduced intensity or range of feelings or emotions) or pessimism without clear effect on behaviour)
21 0 Nil 2 Moderate (significant overvalued ideas with
22 1 Mild (slight constriction of range of affect, clear effect on behaviour, e.g., strong guilt feelings,
23 or transient reduction in range or intensity of clear thoughts that others would be better off
24 feelings) without them)
25 2 Moderate (significant constriction of range or 3 Severe (clear psychotic symptoms, e.g., delu-
26 intensity of feelings with preservation of some sions or hallucinations)
27 emotions, e.g., unable to cry) 16 Irritability (Reports uncharacteristic subjective
28 3 Severe (marked and pervasive constriction of irritability, short fuse, easily angered, manifested by
29 range of affect or inability to experience usual verbal or physical outbursts)
30 emotions) 0 Nil
31 11 Worthlessness (Subjective sense, or thoughts, 1 Mild (slight subjective irritability; may not be
32 of decreased self-value or self-worth) overtly present)
33 0 Nil 2 Moderate (verbal snappiness and irritability
34 1 Mild (slight decrease in sense of self-worth) that is clearly observable in the interview)
35 2 Moderate (some thoughts of worthlessness and 3 Severe (reports of physical outbursts, e.g.,
36 decreased self-worth) throwing ⁄ breaking objects, or markedly abusive
37 3 Severe (marked, pervasive, or persistent feel- verbal outbursts)
38 ings of worthlessness, e.g., feels others better off 17 Lability (Observed mood lability or reported
39 without them, unable to appreciate positive attrib- mood swings)
40 utes) 0 Nil
41 12 Helplessness and Hopelessness (Subjective 1 Mild (subjective reports of mild increase in
42 sense of pessimism or gloom regarding the future, mood lability)
43 inability to cope, or sense of loss of control) 2 Moderate (mood lability clearly observable,
44 0 Nil moderate in intensity)
45 1 Mild (occasional and mild feelings of not being 3 Severe (marked and dominant mood lability,
46 able to cope as usual, or pessimism) frequent or dramatic swings in mood)
47 2 Moderate (often feels unable to cope, or 18 Increased Motor Drive (Subjective reports and
48 significant feelings of helplessness or hopelessness objective evidence of increased motor drive and
49 which lift at times) motor activity)
50 3 Severe (marked and persistent feelings of 0 Nil (normal motor drive)
51 pessimism, helplessness, or hopelessness) 1 Mild (slight increase in drive, not observable in
52 13 Suicidal Ideation (Thoughts or feelings that interview)
53 life is not worthwhile; thoughts of death or suicide) 2 Moderate (clear and observable increase in
54 0 Nil energy and drive)
55 1 Mild (thoughts that life is not worthwhile or is 3 Severe (marked or continuous increase in
56 meaningless) drive)
8
 Bipolar Depression Rating Scale (BDRS)

1 19 Increased Speech (Observed increase in either temperamental and psychopathologic characteristics in 108
2 the rate or quantity of speech, or observed flight of female inpatients. Eur Arch Psychiatry Clin Neurosci 1991;
240: 234–239.
3 ideas) 15. Berk M, Malhi GS, Mitchell PB et al. Scale matters: the
4 0 Nil need for a Bipolar Depression Rating Scale (BDRS). Acta
5 1 Mild (slight increase in the rate or quantity of Psychiatr Scand 2004; 110(Suppl. 422): 39–45.
6 speech) 16. Trivedi MH, Rush AJ, Ibrahim HM et al. The Inventory
7 2 Moderate (racing thoughts, significantly more of Depressive Symptomatology, Clinician Rating (IDS-C)
and Self-Report (IDS-SR), and the Quick Inventory of
8 talkative, clearly distractible, or some circumstan- Depressive Symptomatology, Clinician Rating (QIDS-C)
9 tiality; does not impede interview) and Self-Report (QIDS-SR) in public sector patients with
10 3 Severe (flight of ideas; interferes with inter- mood disorders: a psychometric evaluation. Psychol Med
11 view) 2004; 34: 73–82.
12 20 Agitation (Observed restlessness or agitation) 17. Hamilton M. A rating scale for depression. J Neurol
Neurosurg Psychiatry 1960; 23: 56–73.
13 0 Nil (normal) 18. Montgomery SA, Asberg M. A new depression scale
14 1 Mild (slight restlessness) designed to be sensitive to change. Br J Psychiatry 1979;
15 2 Moderate (clear increase in level of agitation) 134: 382–389.
16 3 Severe (marked agitation, e.g., near continuous 19. Seretti A, Olgiati P. Profiles of ÔmanicÕ symptoms in bipolar
17 pacing or wringing hands) I, bipolar II and major depressive disorders. J Affect
Disord 2005; 84: 159–166.
18 20. Hantouche EG, Akiskal HS. Bipolar II versus unipolar
19 depression: psychopathologic differentiation by dimen-
References
20 sional measures. J Affect Disord 2005; 84: 127–132.
21 1. Weissman MM, Bland RC, Canino GJ et al. Cross- 21. Young RC, Biggs JT, Ziegler VE, Meyer DA. A rating
22 national epidemiology of major depression and bipolar scale for mania: reliability, validity and sensitivity. Br J
disorder. JAMA 1996; 276: 293–299. Psychiatry 1978; 133: 429–435.
23 2. Mitchell PB, Wilhelm K, Parker G, Austin MP, Rutgers P, 22. Landis JR, Koch GG. The measurement of observer
24 Malhi GS. The clinical features of bipolar depression: a agreement for categorical data. Biometrics 1977; 33: 159–
25 comparison with major depressive disorder controls. J Clin 174.
26 Psychiatry 2001; 62: 212–216. 23. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J.
27 3. Tohen M, Greenfield SF, Weiss RD, Zarate CA, Vagge An inventory for measuring depression. Arch Gen Psychi-
LM. The effect of comorbid substance use disorders on the atry 1961; 4: 561–571.
28 course of bipolar disorder: a review. Harv Rev Psychiatry 24. Beck AT, Steer RA, Garbin MG. Psychometric Properties
29 1998; 6: 133–141. of the Beck Depression Inventory Manual, 2nd edn. San
30 4. Manning JS, Connor PD, Sahai A. The bipolar spectrum. 30 Antonio, TX: Psychological Corporation, 1988: ?????–????.
31 A review of current concepts and implications for the 25. Lovibond SH, Lovibond PF. Manual for the Depression
32 management of depression in primary care. Arch Fam Anxiety Stress Scales (DASS). In: ????? ????? ed. Psychol-
Med 1998; 7: 63–72. ogy Foundation Monograph. ??????, NSW: University of
33 5. Angst J, Sellaro R. Historical perspectives and natural 31,32,33 New South Wales, 1993: ????–????.
34 history of bipolar disorder. Biol Psychiatry 2000; 48: 445– 26. Deckersbach T, Perlis RH, Frankle WG et al. Presence of
35 457. irritability during depressive episodes in bipolar disorder.
36 6. Judd LL, Akiskal HS. The prevalence and disability CNS Spectr 2004; 9: 227–231.
37 of bipolar spectrum disorder in the US population: 27. Cassano GB, Rucci P, Frank E et al. The mood spectrum
re-analysis of the ECA database taking into account sub- in unipolar and bipolar disorder: arguments for a unitary
38 threshold cases. J Affect Disord 2003; 73: 123–131. approach. Am J Psychiatry 2004; 161: 1264–1269.
39 7. Cuellar AK, Johnson SL, Winters R. Distinctions between 28. Altshuler LL, Post RM, Leverich GS, Mikalauskas K,
40 bipolar and unipolar depression. Clin Psychol Rev 2005; Rosoff A, Ackerman L. Antidepressant-induced mania and
41 25: 307–339. cycle acceleration: a controversy revisited. Am J Psychiatry
42 8. Mitchell PB, Malhi GS. Bipolar depression: phenomeno- 1995; 152: 1130–1138.
logical overview and clinical characteristics. Bipolar Dis- 29. Post RM, Rubinow DR, Uhde TW et al. Dysphoric
43 ord 2004; 6: 530–539. mania. Clinical and biological correlates. Arch Gen Psy-
44 9. Benazzi F. Psychomotor changes in melancholic and chiatry 1989; 46: 353–358.
45 atypical depression: unipolar and bipolar II subtypes. 30. Berk M, Dodd S, Malhi GS. ÔBipolar missed statesÕ: the
46 Psychiatry Res 2002; 112: 211–220. diagnosis and clinical salience of bipolar mixed states. Aust
47 10. Kupfer DJ, Carpenter LL, Frank E. Is bipolar disorder II a N Z J Psychiatry 2005; 39: 215–221.
unique disorder? Compr Psychiatry 1998; 29: 228–236. 31. Swann AC, Bowden CL, Morris D et al. Depression
48 11. Berk M, Dodd S. Bipolar II disorder: a review. Bipolar during mania. Treatment response to lithium or divalp-
49 Disord 2005; 7: 11–21. roex. Arch Gen Psychiatry 1997; 54: 37–42.
50 12. Yatham LN. Diagnosis and management of patients with 32. Baker RW, Brown E, Akiskal HS et al. Efficacy of
51 bipolar II disorder. J Clin Psychiatry 2005; 66(Suppl. 1): olanzapine combined with valproate or lithium in the
52 13–17. treatment of dysphoric mania. Br J Psychiatry 2004; 185:
13. Benazzi F. Depressive mixed state: a feature of the natural 472–478.
53 course of bipolar II (and major depressive) disorder? 33. Cahill CM, Malhi GS, Berk M, Hadsi-Pavlovic D,
54 Psychopathology 2004; 37: 207–212. Mitchell PB, Lovibond PF. Exploration of symptomatic
55 14. DellÕOsso L, Placidi GF, Nassi R, Freer P, Cassano GB, differences between bipolar and unipolar depression. Aust
56 Akiskal HS. The manic-depressive mixed state: familial, N Z J Psychiatry 2005; 39(Suppl. 2): 48.

9
 Author Query Form
Journal: BDI

Article: 536
Dear Author,
During the copy-editing of your paper, the following queries arose. Please respond to these by marking up your
proofs with the necessary changes/additions. Please write your answers on the query sheet if there is insufficient
space on the page proofs. Please write clearly and follow the conventions shown on the attached corrections
sheet. If returning the proof by fax do not write too close to the paper’s edge. Please remember that illegible
mark-ups may delay publication.

Many thanks for your assistance.

Query Query Remarks

reference
Q1 Au: insert name of department
Q2 Au: Melbourne has been inserted
Q3 Au: give CADE in full
Q4 Au: insert town
Q5 Au: Sydney has been inserted
Q6 Au: Prince of Wales Hospital, Sydney has been inserted
Q7 Au: insert name of department
Q8 Au: HAMD has been changed to HAM-D throughout because it is usually
abbreviated this way
Q9 Au: Depression has been inserted
Q10 Au: both in has been changed to in both
Q11 Au: HAMD has been changed to HAM-D throughout because it is usually
abbreviated this way
Q12 Au: Depression has been inserted
Q13 Au: PM has been changed to PBM to match authorÕs initials in author list
Q14 Au: GM has been changed to GSM to match authorÕs initials in author list
Q15 Au: GM has been changed to GSM to match authorÕs initials in author list
Q16 Au: MH has been changed to MTH to match the author list
Q17 Au: -ise spellings have been changed to -ize spellings throughout for
consistency (previously both were used)
Q18 Au: DHP has been changed to DH-P to match the author list
Q19 Au: 95% CI has been inserted
Q20 Au: 95% CI has been inserted
Q21 Au: 95% CI has been inserted
Q22 Au: 95% CI has been inserted
Q23 Au: 0.0.77 has been changed to 0.77
Q24 Au: 95% CI has been inserted
Q25 Au: 0.0.97 has been changed to 0.97
Q26 Au: 95% CI has been inserted
Q27 Au: divides already has been changed to was already divided
Q28 Au: the subject is has been inserted
Q29 Au: the cohort who had interrater reliability ratings has been changed to the
cohort for which interrater reliability ratings were available
Q30 Au: insert page range
Q31 Au: insert name(s) and initials of editor(s); if there is more than 1 editor,
change ed to eds
Q32 Au: insert town of publisher
Q33 Au: insert page range
Q34 Au: please explain why some loadings are in italics
 MARKED PROOF
Please correct and return this set
Please use the proof correction marks shown below for all alterations and corrections. If you
wish to return your proof by fax you should ensure that all amendments are written clearly
in dark ink and are made well within the page margins.

Instruction to printer Textual mark Marginal mark

Leave unchanged under matter to remain
Insert in text the matter New matter followed by
indicated in the margin or
Delete through single character, rule or underline
or or
through all characters to be deleted
Substitute character or
through letter or new character or
substitute part of one or
more word(s) through characters new characters
Change to italics under matter to be changed
Change to capitals under matter to be changed
Change to small capitals under matter to be changed
Change to bold type under matter to be changed
Change to bold italic under matter to be changed
Change to lower case Encircle matter to be changed
Change italic to upright type (As above)
Change bold to non-bold type (As above)
or
Insert ‘superior’ character through character or
under character
where required
e.g. or
Insert ‘inferior’ character (As above) over character
e.g.
Insert full stop (As above)
Insert comma (As above)
or and/or
Insert single quotation marks (As above)
or

or and/or
Insert double quotation marks (As above)
or
Insert hyphen (As above)
Start new paragraph
No new paragraph
Transpose
Close up linking characters

Insert or substitute space through character or

between characters or words where required

Reduce space between between characters or

characters or words words affected

View publication stats