Journal of Affective Disorders 232 (2018) 9–16

Contents lists available at ScienceDirect

Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad

Research paper

The Koukopoulos Mixed Depression Rating Scale (KMDRS): An International T

Mood Network (IMN) validation study of a new mixed mood rating scale
⁎
Gabriele Sania,b,f,h, , Paul A. Vöhringerb,c,d,h, Sergio A. Barroilhetb,e,h, Alexia E. Koukopoulosf,g,
S. Nassir Ghaemih,i,j
a
NeSMOS Department, Sapienza University of Rome, Italy
b
Department of Psychiatry, Tufts Medical Center, Boston, MA, USA
c
Psychiatry Department, Hospital Clinico University of Chile, Chile
d
Millenium Institute for Depression and Personality Research, Ministry of Economy, Macul, Santiago, Chile
e
University of Chile, Chile
f
Centro Lucio Bini, Rome, Italy
g
Umberto I Hospital, Psychiatry Department, Rome, Italy
h
Tufts University School of Medicine, USA
i
Department of Psychiatry, Harvard Medical School, USA
j
Translational Medicine-Neuroscience, Novartis Institutes of Biomedical Research, Cambridge, MA, USA

A R T I C L E I N F O A B S T R A C T

Keywords: Background: It has been proposed that the broad major depressive disorder (MDD) construct is heterogenous.
Irritability Koukopoulos has provided diagnostic criteria for an important subtype within that construct, "mixed depression"
Mixed depression (MxD), which encompasses clinical pictures characterized by marked psychomotor or inner excitation and rage/
Psychomotor agitation anger, along with severe depression. This study provides psychometric validation for the first rating scale spe-
Rating scale
cifically designed to assess MxD symptoms cross-sectionally, the Koukopoulos Mixed Depression Rating Scale
(KMDRS).
Methods: 350 patients from the international mood network (IMN) completed three rating scales: the KMDRS,
Montgomery-Asberg Depression Rating Scale (MADRS) and Young Mania Rating Scale (YMRS). KMDRS’ psy-
chometric properties assessed included Cronbach's alpha, inter-rater reliability, factor analysis, predictive va-
lidity, and Receiver Operator Curve analysis.
Results: Internal consistency (Cronbach's alpha = 0.76; 95% CI 0.57, 0.94) and interrater reliability (kappa =
0.73) were adequate. Confirmatory factor analysis identified 2 components: anger and psychomotor excitation
(80% of total variance). Good predictive validity was seen (C-statistic = 0.82 95% CI 0.68, 0.93). Severity cut-
off scores identified were as follows: none (0–4), possible (5–9), mild (10–15), moderate (16–20) and severe
(> 21) MxD.
Limitations: Non DSM-based diagnosis of MxD may pose some difficulties in the initial use and interpretation of
the scoring of the scale. Moreover, the cross-sectional nature of the evaluation does not verify the long-term
stability of the scale.
Conclusions: KMDRS was a reliable and valid instrument to assess MxD symptoms.

1. Introduction excitation can be reflected in physical agitation, but also in marked

mood lability and rage or inner tension. This marked anger and lability/
It has been proposed that the construct of depression – even the reactivity differentiates these depressed patients notably from classic
relatively well-defined major depressive episode – is heterogenous, and melancholic states, where usually psychomotor retardation, anhedonia
therefore invalid as a single diagnosis (Ghaemi et al., 2012). An im- and anergia are the key components of the clinical picture (Parker et al.,
portant subtype within that construct has been called "mixed depres- 2013). Such mixed depressive states may reflect the nature of
sion" (MxD), by which is meant marked psychomotor or inner excita- depression, which cannot be separated into purely unipolar or
tion along with severe depression (Kraepelin, 1899). This psychomotor purely bipolar types (Akiskal et al., 2005), and are associated with

⁎
Correspondence to: NeSMOS Department, Sapienza University, Via di Grottarossa 1035-1039, Rome, Italy.
E-mail address: gabriele.sani@uniroma1.it (G. Sani).

https://doi.org/10.1016/j.jad.2018.01.025
Received 20 November 2017; Received in revised form 14 January 2018; Accepted 29 January 2018
Available online 16 February 2018
0165-0327/ © 2018 Elsevier B.V. All rights reserved.
G. Sani et al. Journal of Affective Disorders 232 (2018) 9–16

severe course of illness, psychotic symptoms (Perugi et al., 2013), many depression.
hospitalizations (Pacchiarotti et al., 2011) high risk of suicide (Sani
et al., 2011), and non-response to, or worsening with, antidepressants 2. Methods
(Sani et al., 2014a).
Mixed depression has not been accepted as central to mood condi- This study utilized a database of 350 outpatients, 192 with a diag-
tions in the DSM system (Koukopoulos et al., 2013; Koukopoulos and nosis of Major Depressive Disorder (MDD) and 158 with a diagnosis of
Sani, 2014). In DSM-5, a construct for MDD with mixed features was Bipolar Disorder (BD) according to DSM-IV criteria, presenting a DSM-
included, in which depression was associated only with classic manic IV major depressive episode enrolled at the IMN network. Ninety five
symptoms, such as high libido or overactivity or flight of ideas, ex- subjects (83 MDD, 12 BD, 57 Caucasian, 12 Hispanic, 26 African-
cluding all symptoms, such as irritability, that can be present both in American patients) from Boston (USA), 153 (23 MDD, 130 BD, all
mania and in depression. This construct has been criticized as being Caucasian) from Rome (Italy), and 102 (92 MDD, 10 BD, 20 Caucasian,
rare in the clinical practice and not necessary to diagnose MxD, as 82 Hispanic) from Santiago (Chile) were included. All sites obtained
defined by Koukopoulos and others (Koukopoulos et al., 2007). The IRB approval from their local academic institutions and patients signed
traditional concept of mania in psychopathology, as used by Kraepelin an informed consent before enrollment. Patients were recruited be-
and others for centuries, is not the same as the DSM-defined manic tween 2012 and 2016.
criteria. “Mania” traditionally meant being sped up in thought, affect, Mixed depression was defined according to the definition proposed
and behavior. Depression, which was called “melancholia,” meant by Koukopoulos (2007). Clinical features of the sample were described
being slowed down in thought, affective, and behavior. If one has some as percentages for categorical/binary variables, and means with SD for
thought, affect, and behavior that is sped up, and some that is slowed continuous ones. Initially, as a face validity process three experts in
down, that is what Kraepelin and others meant by “mixed states.” The mood disorders research (SNG, GS, PV), examined each item's content
DSM definition of melancholia is not the same, since it includes agita- for the KMDRS, MADRS and YMRS. Then, as a criterion validity process
tion, which has to be excluded in the pre-DSM usage of melancholia. using subject matter knowledge, these experts decided which KMDRS
The DSM definition of mania is not the same, since it is more narrow items were correlated with MADRS or YMRS items and therefore
and requires specific types of sped-up experiences, such as flight of pairwise spearman correlations were obtained from those paired items
ideas, and increased goal-directed activities. Mood states, whether eu- of MADRS and YMRS with KMDRS. Additionally, to obtain an accurate
phoric or irritable or sad, are epiphenomenal or optional in the classic content validity, a confirmatory factorial analysis was done, using a
19th century and later definitions of mania and melancholia (one can scree plot to confirm the number of factors proposed a priori.
have depression without sadness, or mania without happiness), but Crombach's alpha was calculated for internal reliability. Interrater re-
they are central to DSM-defined mania or depression. The traditional liability between IMN clinicians was performed on-line and was as-
pre-DSM psychopathological construct of mania as reflecting a core of sessed by kappa values. Predictive validity was assessed by logistic
psychomotor excitation has been confirmed by post-DSM psycho- regression models using the scores of the KMDRS. ROC curve was ob-
pathology research as well (Scott et al., 2017), even though the DSM tained to assess its predictive capacity (C-statistic). In order to assess
definitions have remained largely unchanged despite this research. how the scale differentiated clinical diagnosis (mixed vs non-mixed) in
Based on decades of psychopathology research, it can be asserted that depressed patients, the aim was to capture with the KMDRS the “mixed
DSM criteria are too narrow for mania and too broad for depression. depression” concept, as originally propounded by Koukopoulos and
Such DSM criteria fail to capture the essence of mixed states because recently validated (Sani et al., 2014b). This concept entails a full clin-
the DSM approach seeks to split mood states into two parts, depression ical depressive episode, along with manic-like symptoms of psycho-
and mania, in a way that may be invalid empirically. The DSM ap- motor excitation. In accordance with the diagnostic criteria, a con-
proach does not take psychomotor excitation as central to mania, nor comitant full manic episode is not required and the manic symptoms do
psychomotor slowing as central to depression, with mixed states re- not meet full DSM manic episode criteria in most cases. To oper-
flecting a combination. If these clinical constructs are correct, then the ationalize this concept, the sample was defined as meeting usual
DSM approach is unable to capture mixed states defined in this manner. MADRS cutoff scores of 20 or greater for a clinical depressive episode,
Prior work has validated Koukopoulos specific diagnostic criteria for and as being below usual YMRS scores of 20 or greater for a manic
MxD (Sani et al., 2014b) (Table 1). In these patients, depressed/anxious episode. Further, since some manic symptoms were expected to be
mood, inner unrest and inner/psychic agitation dominate the clinical present, YMRS scores would be above the usual cut-off for full remis-
picture. sion of 4 points. Hence, the operationalized assessment of mixed de-
Based on these clinically validated diagnostic criteria, Koukopoulos pression was MADRS score of > 20 and YMRS scores of 4–19. To obtain
and collaborators developed a rating scale specifically designed to en- clinically meaningful cut-off points for the KMDRS, we compared the
able clinicians and research investigators to assess the presence and sample distribution for the total scores of MADRS with the sample
severity of MxD. The “Koukopoulos Mixed Depression Rating Scale” distribution for the total scores of KMDRS. Testing for normality in both
(KMDRS), published here for the first time (see Appendix), was devel- distribution was done before comparing then, using histograms, Q-Q
oped for Koukopoulos’ construct of MxD. This report validates the plots and Shapiro-Wilkins test. Cut-off points were considered as fol-
KMDRS in a large international sample of patients with clinical lows: Below 50th percentiles scores were considered no depression,
between 50th and 63th percentiles scores mild depression, between
63th percentiles and 75th percentiles moderate, and scores above the
Table 1
Koukopoulos criteria for mixed depression.
75th percentiles were deemed severe depression cases. Parameters were
reported as effects sizes when possible, along their 95% confidence
Positive if: Major depressive episode + at least 3 of 8 items intervals. All statistics were done using STATA 12. All plots are avail-
able upon request.
Absence of retardation
Talkativeness
Psychic agitation or inner tension 3. Results
Description of suffering or spells of weeping
Racing or crowded thoughts Clinical and demographic characteristics of the sample are provided
Irritability or unproved rage
on Table 2. Some similarities and differences can be noted between the
Mood lability or marked reactivity
Early insomnia bipolar and MDD subgroups in the IMN sample. The MDD group, as
expected, had more overall depressive symptomatology (higher MADRS

10
G. Sani et al. Journal of Affective Disorders 232 (2018) 9–16

Table 2
Clinical and demographic characteristic of the sample (n = 350).

Total sample K-MxD Non-MxD RRa 95% CI

N = 350 N = 169 N = 181
Variables % % %

Diagnosis
MDD 55 50 60 0.83 0.67 − 1.03
Bipolar 45 50 40 1.20 0.93 − 1.50
Sex
Male 50 51 48 1.07 0.86 − 1.33
Female 50 49 52 0.96 0.76 − 1.19
Ethnicity
Caucasian 66 61 70 0.81 0.65 − 1.01
Hispanic 27 28 25 1.55 1.17 − 2.05
African-Americans 7 11 5 1.28 0.82 − 2
Employed
Yes 60 60 59 1.31 1.04 − 1.66
No 40 40 41 0.97 0.77 − 1.21
Civil Status
Single 22 25 20 1.13 0.88 − 1.44
Married/Cohabitation 36 50 55 0.89 0.71 − 1.1
Divorced 13 18 15 1.15 0.88 − 1.51
Widow 12 7 10 0.82 0.53 − 1.27
Current Substance Abuse
Yes 12 14 10 1.15 0.85 − 1.55
No 88 86 90 0.86 0.64 − 1.16
Past Substance Abuse
Yes 11 13 10 1.15 0.85 − 1.57
No 89 87 90 0.86 0.63 − 1.16
Past/current suicidality
Yes 42 46 38 1.16 0.94 − 1.45
No 58 54 62 0.85 0.68 − 1.06
Koukopoulos criteria for MxD
1. Absence of retardation 60 87 36 2.38 1.95 − 2.91
2. Talkativeness 23 45 3 13.6 6.06 − 30.32
3. Psychic agitation or 47 74 21 3.46 2.58 − 4.63
inner tension
4. Description of suffering or 44 66 23 2.85 2.14 − 3.8
spells of weeping
5. Racing or crowded thoughts 11 20 2 7.06 2.82 − 17.68
6. Irritability or unproved rage 28 47 11 4.23 2.71 − 6.59
7. Mood lability or marked reactivity 35 64 15 (8) 7.71 4.68 − 12.68
8. Early insomnia 37 60 16 3.60 2.54 − 5.11
Mean (SD) Mean (SD) Mean (SD) MD 95% CI
Age (years) 41.7 (20.6) 41.6 (9.5) 41.7 (21.6) 0.1 −3.45 − 3.65
Scores on clinical measures
at assessment
CGI-lifetime severity 4.4 (1) 4.5 (0.9) 4.3 (1.1) −0.2 −0.41 − 0.01
CGI-current symptoms 4 (1) 4.3 (1) 3.8 (0.9) −0.5 −0.69 to −0.3
GAF 55.9 (12) 52.2 (12.2) 59.3 (10.8) 7.1 − 9.51

CGI = clinical global impression; GAF = global assessment functioning; MDD = major depressive disorder; RR = Relative Risk; CI = confidence interval.
a
RR compares MxD and non-MxD groups.

scores), but both groups had similar mixed depressive symptomatology between the three scales based on the symptoms or signs described. It
(similar KMDRS scores). Past suicidality was more common in the bi- will be noted that the KMDRS seemed to overlap slightly more with the
polar subgroup. YMRS (8/11 items) than with the MADRS (6/10 items). One item in the
As shown in Table 3, confirmatory factor analysis identified two KMDRS (muscular tension) was not captured in either of the two other
underlying components that explain 80% of the variance. Additionally, scales in terms of content description.
Table 4 shows that an “Anger/tension/impulsivity” factor comprised Pairwise correlation of the above items found effect sizes and sta-
KMDRS items number 6, 7, 8, 9, 10, 11 and 12, and a “Psychomotor tistical significance. Of the original 20 similar items identified by con-
excitation” factor, comprised items number 1, 2, 3, 4, 5, 13 and 14. tent analysis, 4 items did not have statistically significant correlations
Initial content analysis identified those items which overlapped in pairwise analysis, leaving 16 items as correlated. Selection of the
largest correlation per item that was statistically significant led to a
model of overlap between the three scales as shown on Table 5. Overall,
Table 3 significant correlations in the final model were shown with 4 MADRS
Confirmatory factorial analysis of KMDRS.
items (1,2,3,10) and 7 YMRS items (2−9)
KMDRS Eigenvalue Difference Proportion Cumulative MADRS item 3, “inner tension”, best correlated twice with two
different KMDRS items (4 and 9). In the final model, we reserved that
Factor 1 2.42726 1.04855 0.5061 2.42726 MADRS item 3 for KMDRS item 9, which has the same title and content
Anger/tension/
of “inner tension.” The other KMDRS item 4 (emotional lability) was fit
impulsivity
Factor 2 1.37871 0.42610 0.2975 0.8036 with the next best MADRS item correlation (item 2, reported sadness).
Psychomotor excitation Two KMDRS items (2, vivacious facial expression, and 10, muscular

11
G. Sani et al. Journal of Affective Disorders 232 (2018) 9–16

Table 4
Content analysis and pairwise correlations of the three rating scales.

KMDRS item KMDRS description Similar MADRS item MADRS description Similar YMRS item YMRS description Pairwise correlation (r)

1 Expression of suffering 1,2,3, 8 Apparent sadness None NA MADRS-1 (0.40)*

Reported sadness MADRS-2 (0.42)*
Inner tension, MADRS-3 (0.27)*
Inability to feel MADRS-8 (0.30)*
2 Vivacious facial expression none NA 2 Increased motor activity/energy YMRS-2 (0.06)
3 Speech none NA 6 Speech YMRS-6 (0.57)*
4 Emotional lability 1, 2, 3 Apparent sadness 1, 9 Elated mood, MADRS-1 (0.32),*
Reported sadness Disruptive/aggressive behavior MADRS-2 (0.30),*
Inner tension MADRS-3 (0.46),*
YMRS-1 (−0.006)
YMRS-9 (0.20)*
5 Psychomotor activity 3, 7 Inner tension 2, 9 Increased motor activity/energy MADRS-3 (0.17)*,
MADRS7 (−0.008)
Lassitude Disruptive/aggressive behavior YMRS 2 (0.28)*
YMRS-9 (0.09)
6 Subjective irritability None NA 5 Irritability YMRS-5 (0.12)*
7 Overt anger None NA 9 Disruptive/aggressive behavior YMRS-9 (0.14)*
8 Accelerated thinking None NA 7 Language/thought disorder YMRS-7 (0.17)*
9 Inner tension 3 Inner tension None NA MADRS-3 (0.22)*
10 Muscular tension None NA None NA NA
11 Insomnia 4 Reduced sleep 4 Sleep MADRS-4 (0.30)*
YMRS-4 (0.30)*
12 Suicidal impulsivity 10 Suicidal thoughts None NA MADRS −10 (0.18)*
13 Sexuality None NA 3 Sexual interest YMRS-3 (0.4)*
14 Psychotic symptoms None NA 8 Thought content YMRS-8 (0.62)*

tension) did not have any similar or correlating MADRS/YMRS items. In Table 6
the final model, those two item scores from the KMDRS scale were Discriminant validation of KMDRS in Koukopoulos' criteria mixed depression.
imputed into the model of corresponding MADRS/YMRS items.
Koukopoulos' criteria Non-mixed Total
To show that the selected MADRS/YMRS items capture the corre- Mixed depression depression
sponding KMDRS items, Spearman correlation was performed on the (n = 169) (n = 181) (n = 350)
total score of the combined selected MADRS/YMRS items versus the
*
total KMDRS score (Table 6). The model shown had a strong correlation KMDRS 15.0 ± 5.4 7.9 ± 4.1 11.5 ± 6.0
MADRS 23.0 ± 8.7 22.0 ± 10.1 22.5 ± 9.4
of 0.575 (P < 0.0001).
YMRS 4.9 ± 4.5** 3.3 ± 3.3 4.1 ± 4.0
Predictive validity testing drawn that 43.7% of the sample had Modified MADRS/ 15.9 ± 4.7*** 12.1 ± 4.9 13.9 ± 5.2
YMRS scores of 4–19 (manic symptoms below the threshold of usual full YMRS model
manic episodes). 58.0% had MADRS score of > 20. Combining both Factor 1 7.86 ± 4.1 3.22 ± 2.22 5.71 ± 4.08
Anger/tension/
groups, 21.1% (n = 74) of subjects were operationalized as above to
impulsivity
approximate Koukopoulos’ concept of mixed depression. Factor 2 Psychomotor 8.6 ± 4.81 5.49 ± 3.50 7.15 ± 4.52
As shown in Table 7, logistic regression modeling of the above op- excitement
erationalized sample of mixed depression (n = 74) with the KMDRS as
the primary predictor was highly significant (OR = 1.10 ± 0.03 (SE), MADRS score differences were not statistically significant ((t-0.99, p = 0.30).
* Pairwise comparison, t = 13.7, p < 0.0001, mixed versus non-mixed sample.
[95% confidence intervals 1.04,1.16], p < 0.00001). The ROC curve
** Pairwise comparison, t = 3.8, p < 0.002, mixed versus non-mixed sample.
drawn for this model obtained a C-statistic of 0.82 (95% CI 0.68, 0.93).
*** Pairwise comparison, t = 7.4, p < 0.0001, mixed versus non-mixed sample.
To test reliability, internal consistency was measured, calculating
Cronbach's alpha testing, which was good (alpha = 0.76, 95% CI 0.57,
0.94).
Discriminant validation was assessed by examining KMDRS scores

Table 5
Final model correlating KMDRS to corresponding MADRS/YMRS item.

KMDRS item KMDRS description Best correlated MADRS or YMRS item MADRS or YMRS description Pairwise correlation

1 Expression of suffering MADRS−2 Reported sadness 0.42*

2 Vivacious facial expression None NA 0.06
3 Speech YMRS−6 Speech 0.57*
4 Emotional lability MADRS−1 Apparent sadness 0.32*
5 Psychomotor activity YMRS−2 Increased motor activity/energy 0.28*
6 Subjective irritability YMRS−5 Irritability 0.12*
7 Overt anger YMRS−9 Disruptive/aggressive behavior 0.14*
8 Accelerated thinking YMRS−7 Language/thought disorder 0.17*
9 Inner tension MADRS−3 Inner tension 0.22*
10 Muscular tension None NA NA
11 Insomnia YMRS−4 Reduced sleep .30*
12 Suicidal impulsivity MADRS−10 Suicidal thoughts 0.18*
13 Sexuality YMRS−3 Sexual interest 0.40*
14 Psychotic symptoms YMRS−8 Thought content 0.62*

12
G. Sani et al. Journal of Affective Disorders 232 (2018) 9–16

Table 7 more suicide attempts in such subjects. Instead, benefit has been shown
Logistic regression model for Predictive validity of KMDRS. with dopamine blocking agents (Centorrino et al., 2005), including the
first randomized trial of any agent for mixed depression (Patkar et al.,
Variable OR CI95% p-value
2012), or electroconvulsive therapy (Perugi et al., 2017). Other studies
KMDRS total score 1.097928 1.041943 − 1.156921 0.000 have identified as well the importance of paying attention to the con-
cept of mixed depressive states. Koukopoulos' definition of MxD was
validated in a prior study, with association of MxD with more severe
versus the other scales in patients in the sample who met the definition course of illness, antidepressant nonresponse, and suicidality (Sani et,
of mixed depression as proposed by Koukopoulos' criteria (see 2011; Sani et al., 2014a). The largest study of mixed depression was the
methods). 48.2% of the sample (n = 169) met that mixed depression BRIDGE study (Angst et al., 2011), with over 5000 patients with clinical
definition. depression; in that study, mixed depression, defined as depression plus
Interrater reliability was measured at the two main sites of the IMN three or more excitatory symptoms, was present in almost half of the
sample (Boston and Santiago) was determined to be adequate (kappa sample. Diagnostic validation of the mixed depressive sample was
= 0.73 95% CI 0.64, 0.85). suggested by elevated family history rates of bipolar illness as well
As described in the methods section, based on the similarity of the increased risk of antidepressant-induced mania. A similar set of criteria
normal distributions of the MADRS and KMDRS scales, the following together with one external validator such as family history and specific
cut-off scores were obtained for no, possible, mild, moderate and severe illness course identified about 30% of depressive patients as having
mixed depression, respectively: 0–4, 5–9, 10–15, 16–20 and above 21. mixed states (Perugi et al., 1997). MxD appears to occur commonly in
persons with affective temperaments of cyclothymia or hyperthymia
4. Discussion (Akiskal, 1992; Akiskal et al., 1998; Koukopoulos et al., 2005; Röttig
et al., 2007; Pacchiarotti et al., 2013) Thus, it is crucial to detect,
This analysis is the first study of a new rating scale for mixed de- correctly diagnose and adequately treat these patients.
pression, the Koukopoulos Mixed Depression Rating Scale (KMDRS). It
found that the KMDRS was a reliable and valid rating scale for the 4.1. Limitations
assessment of the mixed depression (MxD) construct. Factor analysis
identified two components that capture the main symptoms of MxD: a Before presenting our conclusion, we must acknowledge some issues
factor for anger/tension/impulsivity and a factor for psychomotor ex- that may limit our findings. Firstly, the non DSM-based diagnosis of
citation. Good predictive validity was shown for discrimination of pa- MxD may pose some difficulties in the initial use and interpretation of
tients meeting MxD diagnostic criteria. Cut-off scores were identified the scoring of the scale. However, as noted before, DSM-5 diagnostic
for mild, moderate and marked severity of MxD symptoms. criteria for depression with mixed features have, in our view, limited
There are no prior studies of this scale, since it is newly developed clinical utility, therefore, new tools based on the same criteria would
and published in this paper. No other rating scales exist specifically for incur in those same limitations. Furthermore, the cross-sectional as-
the MxD concept, as defined by Koukopoulos. Only one other rating sessment does not allow the evaluation of the long-term stability of the
scale has been developed specifically to assess mixed states of manic scale. A longitudinal study that includes the use of the KMDRS during
and depressive symptoms (Cavanagh et al., 2009). That scale limited different phases of the mood disorder is needed to confirm the present
itself to DSM-defined manic criteria, and identified two components of observation.
mixed states in factor analysis: one factor for psychomotor excitation,
and another for neurovegetative symptoms and pain. The KMDRS dif- 5. Conclusion
fers from this prior scale by not limiting itself to DSM-defined manic
criteria. As discussed in the introduction, the excitatory process un- This study provides the first psychometric validation of a new scale
derlying MxD is not the same as DSM-defined manic symptoms, but developed for Koukopoulos’ mixed depression construct. The KMDRS
reflects the traditional core of mania defined as psychomotor excitation, was found to be a reliable and valid instrument to assess MxD symp-
as supported by recent psychopathology research as well (Scott et al., toms. Two main factors identified were anger/tension/impulsivity and
2017). For this reason, we suggest the term excitatory symptoms instead psychomotor excitation. Severity cut-off points were obtained. The
of manic symptoms to describe the manic aspect of MxD. In our view, KMDRS can be used in future clinical trials of mixed depression.
this term is closer to the neurophysiological mechanism underlying
MxD and has strong clinical and psychopathological validity. Further, Acknowledgements
unlike the other scale, the KMDRS captures the psychopathology of
marked irritability/rage, which is central to Koukopoulos' construct of We express our gratitude to the patients who contributed their time
MxD. and effort to participate in this study
The only other relevant study used the Hypomania Checklist to
assess manic symptoms during an acute clinical depressive episode Funding
(Angst et al., 2005). The current study differs from this prior report in
that the KMDRS was developed specifically to assess MxD, unlike the This study was supported by the Fund for Innovation and
HCL which was developed to assess hypomanic episodes. Further, the Competitiveness (FIC) of the Chilean Ministry of Economy,
HCL is a screening tool, not a rating scale of symptom severity. Also, the Development and Tourism, through the Millennium Scientific Initiative,
HCL is self-rated, not clinician-rated, which introduces some risk of Grant Number IS130005.
reporting bias. Again, the HCL is limited to DSM-defined manic symp-
toms, and does not capture marked anger/rage and psychomotor ex- Appendix. The Koukopoulos Mixed Depression Rating Scale
citation more broadly, unlike the KMDRS.
The clinical relevance of the development of the KMDRS is that it Koukopoulos Mixed Depression Rating Scale ©
may help in the conduct of clinical trials of MxD and further in assessing
by
such symptomatology in clinical practice. The identification of MxD
patients may help improve clinical care of this subgroup of depressed Athanasios KOUKOPOULOS, Gabriele SANI, Leonardo TONDO,
subjects. As noted in the introduction, it has been found that anti- Daniela REGINALDI, Paolo DECINA, Alexia Emilia KOUKOPOULOS,
depressants appear to worsen MxD, being associated with 2.5 times Giovanni Walter MANFREDI, Paolo GIRARDI

13
G. Sani et al. Journal of Affective Disorders 232 (2018) 9–16

INSTRUCTIONS up by few (often only one) depressive preoccupation.

This scale was developed in order to enable clinicians and research 9) INNER TENSION
investigators to collect data assessing the presence and severity of Consider spontaneous reports before direct questioning. This
symptoms of excitatory or mixed nature in individuals diagnosed as symptom is of cardinal importance. When present even to a mod-
suffering from a Major Depressive Episode (MDE) according to the erate degree, tension is reported spontaneously as being very dis-
DSM-IV criteria. Case records and previous assessments supporting the tressing. When severe, tension may be described as a painful and
diagnosis of MDE should be reviewed before interviewing the patient. tormenting feeling. Patients may complain of blockade of thoughts,
For Agitated Depression with clear motor agitation the RDC 1978 cri- sensations and emotions. There is no need for autonomic or motor
teria are considered sufficient by the authors, but an item on motor accompaniments for this symptom to be rated present. If motor
agitation is included in this scale. agitation actually occurred, rate it separately.
As the more typical depressive and anxiety symptoms present 10) MUSCULAR TENSION
during an episode of depression are not covered by the scale, this in- It is an unpleasant sensation of muscular tension without the ability
strument is best used together with other scales assessing these symp- of relaxing and the tension is unrelated to any specific voluntary
toms like the Hamilton Rating Scale for Depression, the Hamilton muscular effort. The patient may complain of muscular soreness.
Anxiety Scale and the Montgomery-Asberg Depression Rating Scale. 11) INITIAL AND MIDDLE INSOMNIA
The evaluation of manic symptoms could be investigated by the Young The patients have difficulties in falling asleep. It is often associated
Mania Rating Scale. with racing or crowded thoughts. As far as middle insomnia is
Unless otherwise specified, the time period under study for each concerned, waking up for a few minutes and falling back to sleep
item is the week prior to the interview. should not be taken into account. Only waking up with agitation
Typically, ratings will be made on the basis of observations made and/or difficulty falling back to sleep should be rated.
during interviews, of self-reports and/or information reported by reli- 12) SUICIDAL IMPULSIVENESS
able informants. The first five items are to be rated primarily on the This item will consider mainly the impulsive onset of suicidal
basis of observations made during the diagnostic interviews. ideation and/or the impulsive characteristics of suicidal attempts.
(NB: Patients tend to hide aggressive and hypersexual behavior, in Suicidal thoughts not impulsive must be rated zero.
their presence the relatives also do not mention them. It is therefore 13) SEXUALITY
useful to have a talk with the relatives separately). This item will consider the activity level of sex drive, whether or
not consummated. The patient rarely reports on sexual activity
1) EXPRESSION OF SUFFERING while depressed. Thus the most useful source of information will be
Rate both the extent of internal feelings and their expression, i.e. the interview with the patient's partner.
experience and display. Also determine signs of suffering from 14) PSYCHOTIC SYMPTOMPS
behavior during interview, particularly dramatic speech and ges- This item will consider the presence of thought or perception al-
turing. This item need not be rated on behavioral observation terations. Whenever present, congruous or incongruous delusions
alone, reports are also considered. are scored equally.
2) VIVACIOUS FACIAL EXPRESSION
Rate on the basis of behavior during interview. KOUKOPOULOS MIXED DEPRESSION RATING SCALE©
3) AMOUNT OF SPEECH Date (dd/mm/yyyy): _ _/_ _/ _ _ _ _
Rate on the basis of the quantity of verbal productions, regardless Patient's Name: ________________________
of content. This item need not to be rated on behavioral observa- Date of Birth (dd/mm/yyyy): _ _/_ _/ _ _ _ _
tion alone, reports are also considered. Gender: M F
4) EMOTIONAL LABILITY Rater: _________________________
This item concerns the changing from anxiety to sadness, to des- The time period under study for each item is the week preceding the
pair, to anger, to normal mood, to elation. The item should be rated interview. The first five items should be rated primarily, but not only,
on observations during the interview. Reliable and careful reports on the basis of observations made during the diagnostic interviews.
may also be considered.
5) MOTOR ACTIVITY/RESTLESSNESS 1) EXPRESSION OF SUFFERING
Rate on the basis of observed motor movements. If reports are 0 = laconic expressions of depressive suffering
considered, as they should, make certain that the motor hyper- 1 = animated and prolonged descriptions of suffering
activity actually occurred and was not merely a subjective feeling. 2 = dramatic utterances of suffering and despair
6) SUBJECTIVE FEELINGS OF IRRITABILITY AND UNPROVOKED 3 = outbursts of complaining and spells of crying
ANGER 2) VIVACIOUS FACIAL EXPRESSION
This symptom is of cardinal importance. This item should rate only 0 = reduced facial expression
feelings of irritation or anger which may be recognized by patients 1 = facial expression clearly manifesting emotions
as being unprovoked. Overt expressions of these feelings are rated 2 = vivacious expression of emotions
in the next item. 3 = dramatic expression of emotions
7) OVERT EXPRESSION OF IRRITABILITY AND ANGER 3) AMOUNT OF SPEECH
Rate overt expressions of irritability, annoyance and anger, in- 0 = retarded speech
cluding being argumentative, shouting, losing temper, as well as 1 = normal speech
throwing things and being assaultive. Do not include mere sub- 2 = talkative, conversation not compromised
jective feelings no matter how intense. Patients rarely report ag- 3 = clearly logorrhoeic; the conversation is compromised
gressive behavior. Reports from family members should be con- 4) EMOTIONAL LABILITY
sidered. 0 = absent
8) RACING OR CROWDED THOUGHTS 1 = shifts of mood limited to a depressive-dysphoric range, i.e.
This symptom is of cardinal importance. Note spontaneous reports changes from sadness to anger
before direct questioning. Rate on the basis of subjective experi- 2 = shifts are still within the depressive-dysphoric range, but
ence that thoughts/memories are more than usual, and/or thinking emotions are more intense, i.e. despair, rage
is accelerated. Depressive ruminations are different: they are made 3 = shifts from depressed mood to elation

14
G. Sani et al. Journal of Affective Disorders 232 (2018) 9–16

5) PSYCHOMOTOR ACTIVITY 0 = absent

Make certain that motor hyperactivity/restlessness is visible (or 1 = complaints of difficulty falling asleep of no more than 1/2 h
actually occurred) and is not merely a subjective feeling. (Have you 2 = complaints of difficulty falling asleep of more than one hour
been so fidgety and restless that you couldn’t sit still? Do you have and/or two or more awakenings during the night
to keep pacing up and down?) 3 = practically sleepless because of difficulty falling asleep and of
0 = psychomotor retardation several awakenings during the night
2 = normal rate of psychomotricity 12) SUICIDAL IMPULSIVENESS -
4 = evident restlessness but able to remain seated Have you had thoughts about killing or hurting yourself? (Were
6 = unable to sit still or paces about a great deal these thoughts impulsive, coming suddenly? Often in moments of
6) SUBJECTIVE FEELINGS OF IRRITABILITY AND UNPROVOKED anger?) (Have you actually done anything?
ANGER 0 = absent
How annoyed, angry, or resentful have you felt - whether you 2 = suicidal thoughts possibly impulsive
showed it or not? 4 = impulsive suicidal thoughts definitely present
(How strongly did you feel this way? How much of the time did you 6 = impulsive suicidal attempt(s) (like trying to jump out of a car)
feel this way?) 13) SEXUALITY
0 = absent When a person gets depressed, (s)he has less sexual desire than
1 = complaints of irritability for minor reasons usual. Have you noticed occasionally an increased sexual desire
2 = feels quite angry without any reason and/or activity?
3 = feels like breaking things (or feels acting against oneself or 0 = reduced sexual activity and/or desire
others) 1 = sexual activity normal
7) OVERT EXPRESSION OF IRRITABILITY AND ANGER 2 = occasional hypersexuality
How did you show your (anger, annoyance, irritability)? (Did you 3 = frequent hypersexuality
get into arguments?) (Did you lose your temper, throw or break 14) PSYCHOTIC SYMPTOMS
things?) Have you ever thought that others were doing something against
(What about hitting anybody?) you? Have you ever thought to be in the middle of others’ interest?
0 = absent Have you ever heard noises or voices?
2 = quick to express annoyance, impatience, occasional verbal 0 = absent
aggression 1 = suspiciousness
4 = occasional violence against things 2 = ideas of reference
6 = occasional violence against oneself or others 3 = delusions and/or hallucinations
8) CROWDED AND/OR ACCELERATED THINKING Copyright 2015, Athanasios Koukopoulos and S. Nassir Ghaemi.
Note spontaneous reports before direct questioning.
Have you had more thoughts than usual or more than you can References
handle? (Did it interfere with your falling asleep?) (Have your
thoughts raced through your mind?) Akiskal, H.S., 1992. The distinctive mixed states of bipolar I, II, and III. Clin.
0 = absent Neuropharmacol. 15 (Suppl. 1), 632A–633A.
Akiskal, H.S., Hantouche, E.G., Bourgeois, M.L., Azorin, J.M., Sechter, D., Allilaire, J.F.,
1 = depressive or painful ruminations Lancrenon, S., Fraud, J.P., Châtenet-Duchêne, L., 1998. Gender, temperament, and
2 = crowded thoughts or memories, some of which may not be the clinical picture in dysphoric mixed mania: findings from a French national study
painful. (EPIMAN). J. Affect. Disord. 50, 175–186.
Akiskal, H.S., Benazzi, F., Perugi, G., Rihmer, Z., 2005. Agitated "unipolar" depression re-
3 = persistent racing thoughts or musical tunes in the head. conceptualized as a depressive mixed state: implications for the antidepressant-sui-
9) INNER TENSION cide controversy. J. Affect. Disord. 85, 245–258.
Note spontaneous report before questioning. This is a particular Angst, J., Adolfsson, R., Benazzi, F., Gamma, A., Hantouche, E., Meyer, T.D., Skeppar, P.,
Vieta, E., Scott, J., 2005. The HCL-32: towards a self-assessment tool for hypomanic
and distressing sensation of nervous tension. This sensation may be symptoms in outpatients. J. Affect. Disord. 88, 217–233.
accompanied by a feeling of being blocked in your mental activ- Angst, J., Azorin, J.M., Bowden, C.L., Perugi, G., Vieta, E., Gamma, A., Young, A.H.,
ities. (Do you feel on edge, or keyed-up or stressed?) (Does this BRIDGE Study Group, 2011. Prevalence and characteristics of undiagnosed bipolar
disorders in patients with a major depressive episode: the BRIDGE study. Arch. Gen.
tension prevent you from doing things, from thinking?)
Psychiatry 68, 791–798.
0 = absent Cavanagh, J., Schwannauer, M., Power, M., Goodwin, G.M., 2009. A novel scale for
1 = this sensation is reported only on questioning measuring mixed states in bipolar disorder. Clin. Psychol. Psychother. 16, 497–509.
2 = this sensation is spontaneously reported as a distressing part of Centorrino, F., Fogarty, K.V., Cimbolli, P., Salvatore, P., Thompson, T.A., Sani, G.,
Cincotta, S.L., Baldessarini, R.J., 2005. Aripiprazole: initial clinical experience with
the condition 142 hospitalized psychiatric patients. J. Psychiatr. Pract. 11, 241–247.
3 = the patient reports being tormented by a sensation of tension Ghaemi, S.N., Vöhringer, P.A., Vergne, D.E., 2012. The varieties of depressive experience:
and inner agitation. Diastolic blood pressure may be > 90 mmHg. diagnosing mood disorders. Psychiatr. Clin. N. Am. 35, 73–86.
Koukopoulos, A., Albert, M.J., Sani, G., Koukopoulos, A.E., Girardi, P., 2005. Mixed de-
10) MUSCULAR TENSION pressive states: nosologic and therapeutic issues. Int. Rev. Psychiatry 17, 21–37.
Have you had difficulty relaxing your muscles since you have been Koukopoulos, A., Sani, G., Koukopoulos, A.E., Manfredi, G., Pacchiarotti, I., Girardi, P.,
depressed? (Do your muscles feel tense?) (Where did you feel the 2007. Melancholia agitata and mixed depression. Acta Psychiatr. Scand. 115 (Suppl.
433), 50–57.
tension?) (Were you able to relax?) Koukopoulos, A., Sani, G., Ghaemi, S.N., 2013. Mixed features of depression: why DSM-5
0 = absent is wrong (and so was DSM-IV). Br. J. Psychiatry 203, 3–5.
1 = muscular tension is reported only on questioning Koukopoulos, A., Sani, G., 2014. DSM-5 criteria for depression with mixed features: a
farewell to mixed depression. Acta Psychiatr. Scand. 129, 4–16.
2 = slight recurrent muscular tension with some ability to relax
Kraepelin, E., 1899. Psychiatrie, sixth ed. J.A. Barth, Leipzig.
3 = definite experience of muscular tension without the ability to Pacchiarotti, I., Mazzarini, L., Kotzalidis, G.D., Valentí, M., Nivoli, A.M., Sani, G., Torrent,
relax, or visible tension (possibly tremors). The patient may com- C., Murru, A., Sanchez-Moreno, J., Patrizi, B., Girardi, P., Vieta, E., Colom, F., 2011.
Mania and depression. Mixed, not stirred. J. Affect. Disord. 133, 105–113.
plain of muscle soreness, especially in the neck and back.
Pacchiarotti, I., Nivoli, A.M., Mazzarini, L., Kotzalidis, G.D., Sani, G., Koukopoulos, A.,
11) INITIAL AND MIDDLE INSOMNIA Scott, J., Strejilevich, S., Sánchez-Moreno, J., Murru, A., Valentí, M., Girardi, P.,
Have you had difficulty falling asleep? (How long does it take to Vieta, E., Colom, F., 2013. The symptom structure of bipolar acute episodes: in search
fall asleep once you go to bed? Do you wake up in the middle of the for the mixing link. J. Affect. Disord. 149, 56–66.
Parker, G., McCraw, S., Blanch, B., Hadzi-Pavlovic, D., Synnott, H., Rees, A.M., 2013.
night? Do you fall asleep again once you wake up?

15
G. Sani et al. Journal of Affective Disorders 232 (2018) 9–16

Discriminating melancholic and non-melancholic depression by prototypic clinical Röttig, D., Röttig, S., Brieger, P., Marneros, A., 2007. Temperament and personality in
features. J. Affect. Disord. 144, 199–207. bipolar I patients with and without mixed episodes. J. Affect. Disord. 104, 97–102.
Patkar, A., Gilmer, W., Pae, C.U., Vöhringer, P.A., Ziffra, M., Pirok, E., Mulligan, M., Sani, G., Tondo, L., Koukopoulos, A., Reginaldi, D., Kotzalidis, G.D., Koukopoulos, A.E.,
Filkowski, M.M., Whitham, E.A., Holtzman, N.S., Thommi, S.B., Logvinenko, T., Manfredi, G., Mazzarini, L., Pacchiarotti, I., Simonetti, A., Ambrosi, E., Angeletti, G.,
Loebel, A., Masand, P., Ghaemi, S.N., 2012. A 6 week randomized double-blind Girardi, P., Tatarelli, R., 2011. Suicide in a large population of former psychiatric
placebo-controlled trial of ziprasidone for the acute depressive mixed state. PLoS One inpatients. Psychiatry Clin. Neurosci. 65, 286–295.
7, e34757. Sani, G., Napoletano, F., Vöhringer, P.A., Sullivan, M., Simonetti, A., Koukopoulos, A.,
Perugi, G., Akiskal, H.S., Micheli, C., Musetti, L., Paiano, A., Quilici, C., Rossi, L., Cassano, Danese, E., Girardi, P., Ghaemi, N., 2014a. Mixed depression: clinical features and
G.B., 1997. Clinical subtypes of bipolar mixed states: validating a broader European predictors of its onset associated with antidepressant use. Psychother. Psychosom. 83,
definition in 143 cases. J. Affect. Disord. 43, 169–180. 213–221.
Perugi, G., Medda, P., Reis, J., Rizzato, S., Giorgi Mariani, M., Mauri, M., 2013. Clinical Sani, G., Vöhringer, P.A., Napoletano, F., Holtzman, N.S., Dalley, S., Girardi, P., Ghaemi,
subtypes of severe bipolar mixed states. J. Affect. Disord. 151, 1076–1082. S.N., Koukopoulos, A., 2014b. Koukopoulos‫ ׳‬diagnostic criteria for mixed depression:
Perugi, G., Medda, P., Toni, C., Mariani, M.G., Socci, C., Mauri, M., 2017. The role of a validation study. J. Affect. Disord. 164, 14–18.
electroconvulsive therapy (ECT) in bipolar disorder: effectiveness in 522 patients Scott, J., Murray, G., Henry, C., Morken, G., Scott, E., Angst, J., Merikangas, K.R., Hickie,
with bipolar depression, mixed-state, mania and catatonic features. Curr. I.B., 2017. Activation in bipolar disorders: a systematic review. JAMA Psychiatry 74,
Neuropharmacol. 15, 359–371. 189–196.

16