Trajectories of depressive symptoms in the acute phase of psychosis: Implications for treatment

Accepted Manuscript Trajectories of depressive symptoms in the acute phase of psychosis: Implications for treatment E. Kjelby, R. Gjestad, I. Sinkeviciute, R.A. Kroken, E.-M. Løberg, H.A. Jørgensen, E. Johnsen PII:

S0022-3956(18)30043-8

DOI:

10.1016/j.jpsychires.2018.06.003

Reference:

PIAT 3394

To appear in:

Journal of Psychiatric Research

Received Date: 16 January 2018 Revised Date:

31 May 2018

Accepted Date: 1 June 2018

Please cite this article as: Kjelby E, Gjestad R, Sinkeviciute I, Kroken RA, Løberg E-M, Jørgensen HA, Johnsen E, Trajectories of depressive symptoms in the acute phase of psychosis: Implications for treatment, Journal of Psychiatric Research (2018), doi: 10.1016/j.jpsychires.2018.06.003. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Trajectories of depressive symptoms in the acute phase of psychosis: implications for treatment Kjelby E1§, Gjestad R1,2, Sinkeviciute I1,2, Kroken RA1,3,4, Løberg E-M1,4,5,6, Jørgensen HA3,

RI PT

Johnsen E1,3,4

1

Division of Psychiatry, Haukeland University Hospital, Bergen, Norway

2

Centre for Research and Education in Forensic Psychiatry, Haukeland University Hospital,

Department of Clinical Medicine, Section of Psychiatry, Faculty of Medicine and Dentistry,

University of Bergen, Norway.

M AN U

3

SC

Bergen, Norway

NORMENT Centre of Excellence, University of Oslo, Norway

5

Department of Addiction Medicine, Haukeland University Hospital, Bergen, Norway

6

Department of Clinical Psychology, University of Bergen, Norway

TE D

4

§

Email addresses:

EP

Corresponding author. Address: Division of Psychiatry, Haukeland University Hospital, PB 1400, 5021 Bergen, Norway. Tel: +47 55 95 86 65

AC C

EK: [email protected] RG: [email protected] IS: [email protected] RAK: [email protected] EML: [email protected] HAJ: [email protected] EJ: [email protected]

1

ACCEPTED MANUSCRIPT Abstract Depression is common in schizophrenia and associated with negative outcomes. Previous studies have identified heterogeneity in treatment response in schizophrenia. We aimed to investigate different trajectories of depression in patients suffering from psychosis and

RI PT

predictors of change in depressive symptoms during antipsychotic treatment. Two hundred and twenty-six patients >18 years acutely admitted due to psychosis were consecutively

included and the follow-up was 27 weeks. The Calgary Depression Scale for Schizophrenia

SC

(CDSS) sum score was the primary outcome. Latent growth curve (LGCM) and Growth Mixture Models (GMM) were conducted. Predictors were the Positive sum score of the

M AN U

Positive and Negative Syndrome Scale for Schizophrenia (PANSS), Schizophrenia spectrum/non-spectrum psychoses, gender and being antipsychotic naive at inclusion. We found support for three depression-trajectories, including a high- (14.7%), a low depressionlevel (69.6%) class and a third depressed class quickly decreasing to a low level (15.7%).

TE D

Change in CDSS was associated with change in PANSS positive score in all time intervals (4 weeks: b = 0.18, p<0.001, 3 months: 0.21, p<0.023, 6 months: 0.43, p<0.001) and with a diagnosis within schizophrenia spectrum but not with antipsychotic naivety or gender. The

EP

schizophrenia-spectrum patients had less depressive symptoms at inclusion (-2.63, p<0.001).

AC C

In conclusion, an early responding and a treatment refractory group were identified. The treatment-refractory patients are candidates for enhanced anti-depressive treatment, for which current evidence is limited. The post-psychotic depression group was characterized by depressive symptoms in the acute phase as well. We could not identify differentiating characteristics of the depression trajectories.

2

ACCEPTED MANUSCRIPT 1. Introduction Depression in psychotic disorders is a major cause of impaired functioning (Conley et al., 2007). The modal rate in schizophrenia is about 25% (Siris, 2000). In schizophrenia depression is associated with a poorer quality of life (Tollefson and Andersen, 1999),

RI PT

increased rates of relapse (Sands and Harrow, 1999), longer duration of hospitalizations

(Hausmann and Fleischhacker, 2002) and suicide (Hawton et al., 2005; Palmer et al., 2005). The phenomenology of depression in schizophrenia is heterogeneous and different subtypes

SC

have been described: depressive symptoms as part of the prodromal phase (Birchwood et al., 1989; Yung and McGorry, 1996), depressive symptoms during and preceding a psychotic

M AN U

episode (Lancon et al., 2001; Sax et al., 1996) and a post-psychotic depression subtype. Two patterns of the post-psychotic subtype have been described (Kohler and Lallart, 2002): One where depressive symptoms emerge after a psychotic episode (Birchwood et al., 2005; McGlashan and Carpenter, 1976; Mulholland and Cooper, 2000; Upthegrove et al., 2014) and

TE D

a second pattern where depressive symptoms are revealed after being overshadowed by positive symptoms in the acute phase (Knights and Hirsch, 1981). Patients with similar depression profiles may differ in their response to treatment. Thus, there is a need to

EP

investigate trajectories.

AC C

The evidence concerning treatment of depression in schizophrenia is limited. A dysphoric effect of antipsychotics has been suspected (Bressan et al., 2002), but studies find little evidence for a depression-inducing effect (Birchwood et al., 2000; Hirsch et al., 1989). In fact, anti-depressive properties have been indicated for several second generation antipsychotics (SGAs) (Kaneriya et al., 2016; Leucht et al., 2009; Suppes et al., 2016; Tohen et al., 2012). In previous results from the present dataset there was a decline in depressive symptoms in all 4 antipsychotic trial arms (olanzapine, quetiapine, risperidone and

3

ACCEPTED MANUSCRIPT ziprasidone), but no differences in anti-depressive effectiveness among the groups (Johnsen et al., 2010; Kjelby et al., 2011). Meta-analyses do not find robust and stable proof of efficacy for antidepressants in depression in schizophrenia (Buoli et al., 2016; Hasan et al., 2015). There is some evidence that lithium

RI PT

may be effective for a subgroup of patients (Hasan et al., 2015). Cognitive behavioral therapy (CBT) has demonstrated efficacy for affective symptoms in schizophrenia in some trials (Hazell et al., 2016; Jones et al., 2012; Sensky et al., 2000). However, guidelines for the

SC

treatment of depression in psychotic disorders remain unclear due to unresolved issues related to among other, the heterogeneity of depression in psychosis. Thus, there is a need to

M AN U

disentangle this heterogeneity to be able to target the symptoms more directly. Predictors of depression and of the course of depression in schizophrenia are insufficiently known, even less so for depressive symptoms in other psychotic disorders, which may be difficult to distinguish from schizophrenia at acute admissions and in the early phase

TE D

(Addington et al., 2006; Bromet et al., 2011). A more thorough knowledge of predictors would help understand depressive symptom change processes, thus facilitating targeted treatment approaches. Gender has been investigated as a predictor of depression prevalence

EP

with mixed findings (Addington et al., 1996; Caton et al., 2014; Goldstein and Link, 1988; Shtasel et al., 1992). Gender as a predictor of course of depression has not been examined.

AC C

Reduction of positive psychotic symptoms has been found to correlate with improvement of depression, however with less valid depression rating instruments (Norman and Malla, 1994; Tapp et al., 2001). Impact of schizophrenia spectrum vs. non-spectrum psychoses and antipsychotic-naivety on anti-depressive effectiveness has not been investigated before (Zhu et al., 2017). Antipsychotic-naivety is associated with an increased response of positive psychotic symptoms to antipsychotics and would thus be important to investigate for a similar impact on anti-depressive effects (Jager et al., 2007; Zhu et al., 2017). Schizophrenia

4

ACCEPTED MANUSCRIPT spectrum diagnoses are generally investigated isolated from other psychotic disorders. Investigating schizophrenia spectrum as a separate predictor of depression would determine if this subgrouping predicts different level and change in depressive symptoms. Trajectory studies have been conducted in schizophrenia to investigate course of illness

RI PT

(Buchy et al., 2010; Czobor and Volavka, 1996; Kaplan et al., 2016; Levine et al., 2011;

Madsen et al., 2016; Marengo et al., 2000) and antipsychotic treatment response (Levine and Leucht, 2010; Levine et al., 2012; Marques et al., 2011; Stauffer et al., 2011). Trajectories of

SC

response, rather than simple responder/non-responder models, take better into account the heterogeneity of the extensive information in trials, thus providing a better understanding of

M AN U

treatment-response for subgroups of patients (Hunter et al., 2010; Kapur et al., 2009). No studies have investigated trajectories of depression in antipsychotic trials as extensively as the present one.

We aimed to investigate heterogeneity in treatment response of depressive symptoms in an

TE D

antipsychotic trial in patients suffering from acute phase psychosis by uncovering depressive trajectories. Furthermore, we aimed to investigate possible predictors of depressive symptomcourse: the Positive sum score of the Positive and Negative Syndrome Scale for

EP

Schizophrenia (PANSS-P), schizophrenia spectrum, gender and being antipsychotic naïve, thus aiming to determine characteristics of patients with and without improvement in

AC C

depressive symptoms. This contribution to knowledge of the phenomenology of depressioncourse in acutely psychotic patients may subsequently pave the way to a more focused approach to treatment of depression in patients with acute psychosis.

2. Methods 2.1 Study design

5

ACCEPTED MANUSCRIPT The Bergen Psychosis Project (BPP) is a prospective, pragmatic antipsychotic drug trial aiming at including all acutely admitted patients with psychosis who are eligible for oral antipsychotic drug treatment. Patients were consecutively recruited from the Division of Psychiatry at Haukeland University Hospital in Bergen, Norway, with a catchment population

RI PT

of about 400,000. The methods of the BPP were described in more detail in a previous

publication (Johnsen et al., 2010). The BPP was approved by the Regional Committee for Medical and Health Research Ethics West-Norway (REC) and the Norwegian Social Science

SC

Data Services. Funding of the project was initiated by the Research Council of Norway, followed by the Western Norway Regional Health Authority and Haukeland University

the pharmaceutical industry.

2.2 Sample

M AN U

Hospital, Division of Psychiatry. The BPP did not receive any financial or other support from

TE D

The study gained ethical approval to include participants before consent was provided, thus entailing a clinically relevant representation in the study (Johnsen et al., 2010). Informed written consent for the follow-up part of the project was obtained at visit 2 (at discharge or at

EP

6 weeks from baseline at the latest).

Patients > 18 years were eligible for the study if they were admitted to the psychiatric acute

AC C

ward for symptoms of psychosis as determined by a score of ≥ 4 on one or more of the items Delusions, Hallucinations, Grandiosity, Suspiciousness/Persecution or Unusual Thought Content on the Positive and Negative Syndrome Scale (PANSS) (Kay et al., 1987) and were candidates for oral antipsychotic drug therapy with risperidone, olanzapine, quetiapine or ziprasidone. Participants met the International Classification of Diseases (ICD-10) diagnostic criteria (World Health Organization, 2007) for schizophrenia (F20), schizotypal disorder (F21), persistent delusional disorders (F22), acute and transient psychotic disorders (F23), schizoaffective disorder (F25), other non-organic psychotic disorders (F28), unspecified non6

ACCEPTED MANUSCRIPT organic psychosis (F29), drug-induced psychosis (F1x.5) and major depressive disorder with psychotic features (F31.5, F32.3, F33.3) (Figure 1). The diagnoses were obtained from the medical record at discharge. Patients were excluded from the study if they were unable to use oral antipsychotics, were

RI PT

unable to cooperate reliably during investigations, did not understand spoken Norwegian language, were candidates for electroconvulsive therapy, were suffering from an organic brain disorder or were medicated with clozapine on admission. Patients with drug-induced

SC

psychoses were included only when the condition did not resolve within a few days and when

M AN U

antipsychotic drug therapy was indicated.

2.3 Measures

The patients were assessed at baseline (T1), at discharge or at 6 weeks (mean 4.1 weeks) if not discharged earlier (T2) and at follow-up visits after 3 (T3) and 6 months (T4). Before

TE D

inclusion, eligible patients were interviewed by the investigators using the SCI-PANSS (Opler et al., 1999). Intra-class correlation coefficients (ICC) were calculated based on inter-rater assessments and showed high inter-rater reliability (0.92). Furthermore, the patients

EP

underwent assessments using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) (Gold et al., 1999; Randolph, 1998) and the Clinical

AC C

Drug and Alcohol Use Scales (CDUS/CAUS) (Drake et al., 1996). The CDUS and CAUS are scored by the clinician/rater from 1 to 5 where 1 is abstinence and 2 is use without impairment. This score was dichotomized as 1 and 2 defining abstinence or use without impairment and 3-5 defining drug/alcohol abuse or -dependence. The Clinical Global Impression-Severity of Illness scale (CGI-S) (Guy, 1976) and Global Assessment of Functioning-Split Version, Functions scale (GAF-F) (American Psychiatric Association, 2000) were conducted to assess the general symptom and functioning level.

7

ACCEPTED MANUSCRIPT Symptoms of depression were assessed by means of the Calgary Depression Scale for Schizophrenia (CDSS) (Addington et al., 1990). The CDSS consists of 9 items, each giving a score of 0 to 3 points and has been specifically developed to assess the level of core depressive symptoms in schizophrenia, distinguishing depressive symptoms from negative

RI PT

symptoms and extra-pyramidal side-effects (Addington et al., 1994; Lako et al., 2012;

Schennach et al., 2012). Addington et al. have previously shown that a CDSS sum score > 6 has a specificity of 82% and a sensitivity of 85% for predicting a major depressive episode

SC

(Addington and Addington, 2015).

antipsychotic medication before.

M AN U

Participants were determined antipsychotic-naive if they had not been treated with

Schizophrenia spectrum diagnoses were defined as: schizophrenia (F20), schizotypal disorder (F21), delusional disorder (F22), acute polymorphic psychotic disorder with symptoms of schizophrenia (F23.1), acute schizophrenia-like psychotic disorder (F23.2), other acute

TE D

predominantly delusional psychotic disorders (F23.3) and schizoaffective disorder (F25) (World Health Organization, 2007).

The PANSS Excitement Component (PANSS-EC) was applied as a measure of agitation. The

EP

PANSS-EC consists of the PANSS items P4 Hyperactivity, P7 Hostility, G4 Tension, G8 Uncooperativeness and G14 Poor Impulse Control. The PANSS-EC has been validated as an

AC C

assessment of agitation and aggression (Montoya et al., 2011).

2.4 Statistical procedures 2.4.1 Analysis strategy

Initially, a latent growth curve model (LGCM) with CDSS sum score as outcome was established, as LGCM applies well to analyses of repeated measures as a function of time (Kline, 2011). This method is modeling not only the mean level and change over time, but 8

ACCEPTED MANUSCRIPT also the individual levels and changes. Next, predictors were added to the model. Then, a growth mixture model (GMM) identified unobserved sub-populations with different depressive symptom-trajectories. Finally, differences in clinical characteristics between

RI PT

trajectory-groups were investigated.

2.4.2 Statistical methods

SPSS 23 was used for descriptive statistics and cross-tabulation with χ2 and ANOVA-analyses

SC

(IBM, 2015). Mplus 7.4 was used for analyzing Latent Growth Curve (LGC) and Growth

M AN U

Mixture Models (Muthén and Muthén, 2015). The estimator was set to Maximum Likelihood with Robust standard errors (MLR) in order to handle non-normality in data (Kline, 2011). Full information maximization likelihood method uses all available data under the “Missing at Random” assumption (Kline, 2011; Muthén and Muthén, 2014) and minimize the effect of missing data (Bollen and Curran, 2006). Analyses did not show significant differences

TE D

between study drop outs and –completers at baseline, supporting a “Missing Completely at Random” assumption. However, this does not rule out the possibility that missingness is non-

AC C

EP

ignorable (missing not at random) (Schafer and Graham, 2002).

2.4.2.1 LGCM-analyses

Unconditional piecewise growth models were estimated since a linear trajectory or a quadratic model did not fit data well (Bollen and Curran, 2006), with one slope factor describing the change from T1-T2 and a second factor the T2-T4 change. The level and change had to be analyzed in three pieces (a contrast difference model) when adding the predictors (Newsom, 2015), since different patterns of trajectories were evident for the schizophrenia spectrum and

9

ACCEPTED MANUSCRIPT the non-schizophrenia spectrum groups. An identical model was specified with the same number of change factors for the PANSS positive scale over the same period. The entered predictors were level and change in PANSS positive, diagnosis (schizophrenia spectrum vs. other psychosis diagnoses), antipsychotic-naivety and gender. In order to

RI PT

investigate possible moderator effects, interaction terms were added to the model. However, interaction effects not being found to be statistically significant were removed from the model

SC

and the model re-estimated (backward stepwise).

M AN U

2.4.2.2 Growth mixture model analyses

Based on the established unconditional LGC model including the intercept and two slope factors, GMM models were analyzed (Wang and Wang, 2012). In addition, the default setting in Mplus with equal variance and covariance estimates over classes was used. This strategy

TE D

kept the number of estimated parameters as low as possible. Models with increasing number of classes were estimated in order to decide the number of classes representing the data. This evaluation was based on an overall evaluation including the entropy index and model fit

EP

indices: Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC) and Sample-size Adjusted BIC (SABIC), with lower fit and higher entropy values indicating

AC C

better model fit (Kline, 2011; Muthén and Muthén, 2015).

10

ACCEPTED MANUSCRIPT Statistically significant improvement of adding classes was tested by Vuong-Lo-MendellRubin Likelihood Ratio Test (VLMR) and the parametric bootstrapped likelihood ratio test (BLRT) (Muthén and Muthén, 2015). BIC and bootstrapped LRT have been found to perform best of these indices across several simulated models (Nylund et al., 2007), however, results

RI PT

also show their performance to be sensitive for sample size. A later simulation study showed SABIC to be the preferable index (Kim, 2014). The estimated sample size and relative class frequencies were also used to determine the number of classes. A sample size of 25 classified

M AN U

SC

subjects (5%) has been proposed (Wickrama et al., 2016).

2.4.2.3 Trajectory group-comparisons

After classification, the relationships between estimated classes and relevant clinical variables were explored with the auxiliary distal outcome option in Mplus (a 3-step-procedure) as the

TE D

entropy index was beyond the preferred threshold of .80 (Asparouhov and Muthén, 2014; Kim, 2014; Li and Hser, 2011; Peugh and Fan, 2015; Wang and Wang, 2012; Wickrama et al., 2016). The auxiliary variables were entered as distal outcomes in order to explore

AC C

EP

bivariate relationships (Wickrama et al., 2016).

2.4.2.4 Sensitivity analysis A sensitivity analysis was conducted, due to a majority of patients with uni- or bipolar depression clustering in the early depressive response group. Participants diagnosed with primary affective psychoses were excluded from the analysis and new LGCM- and GMManalyses were conducted.

11

ACCEPTED MANUSCRIPT

3. Results

The study enrolment is displayed in Figure 2. One third of the 226 patients were female

RI PT

(32.7%), the mean age at inclusion was 34.1 (SD 13.5), 44.2% were antipsychotic-naive and 54.9% had a diagnosis within schizophrenia spectrum disorders (Figure 1). The descriptive

M AN U

3.1 Depressive symptom level and change over time

SC

statistics for the outcome variable for the four visits are presented in Table 1.

The model baseline CDSS sum score was 6.51, the reduction from T1 to T2 was -0.65 per week (p<0.001) and the small reduction over the last three visits (T2-T4: -0.02) was not statistically significant (p = 0.254). The unconditional piecewise level and change model showed close fit with data (χ2 = 3.05, df = 2, p = 0.22, CFI = 0.98, TLI = 0.93, RMSEA =

TE D

0.048, RMSEA confidence interval (CI) = 0.000-0.149, RMSEA close fit = 0.39).

EP

3.2 Predictors of level and changes in CDSS

In summary the following were the results for predictors of depression level and change,

AC C

displayed as a plot based on the LGCM-model with predictors in Figure 3: Reduction in positive psychotic symptoms was associated with reduction in depression. The schizophreniaspectrum subgroup had less depressive symptoms at inclusion and was associated with change in depression. Gender and antipsychotic-naivety was not associated with course of depression as a main effect. In greater detail the following was found: Gender did not contribute in the model when the other variables were accounted for, either as a main effect or in interaction terms and was therefore removed from the model. Reduction in the CDSS score was associated with 12

ACCEPTED MANUSCRIPT decrease in the PANSS positive subscale score in all time intervals (b: unstandardized regression weights) = 0.18, p<0.001; b=0.21, p<0.023; b=0.43, p<0.001). Baseline level of the CDSS score, however, was not predicted by baseline level of the PANSS positive subscale score.

RI PT

Lower baseline CDSS score was associated with being diagnosed within the schizophreniaspectrum compared to other psychoses (b=-2.63, p<0.001), but not with being antipsychoticnaive (b=-1.23, p=0.077). Change in the CDSS score from baseline to T2 was associated with

SC

a smaller reduction in the schizophrenia group (b=0.64, p=0.021), but not with being

antipsychotic-naive (b=0.53, p=0.082). However the interaction-effect between diagnosis-

M AN U

group and antipsychotic-naivety was statistically significant (b=-0.91, p=.004). Change in the CDSS score in the interval T2-T3 was associated with reduction in the schizophreniaspectrum group (b=-0.25, p<0.003), but not with being medication-naïve (b=-0.211, p=0.061). The interaction-effect was, however, statistically significant (b=0.50, p<.001). Neither

TE D

diagnosis-group nor being antipsychotic-naive predicted the CDSS score-change from T3-T4 (b=0.095, p=0.087; b=-0.021, p=0.689). The end model fitted data well (χ2 = 16.82, df = 19, p

AC C

fit = 0.94).

EP

= 0.60, CFI = 1.00, TLI = 1.03, RMSEA = 0.000, RMSEA c.i. = 0.000-0.051, RMSEA close

3.3 Trajectories of depressive symptoms: mixture analysis In summary, a three class model was supported with one high depression level class, one depressed but improving quickly class and one class with a low level of depressive symptoms (Figure 4, Supplementary Figure 1). In greater detail, the results for the Growth mixture model (GMM) analyses were the following: Fit information and Likelihood Ratio Test significance difference tests showed up to five possible classes (Table 2). However, some iteration problems were experienced for the 13

ACCEPTED MANUSCRIPT five class model, and the starting values had to be increased considerably in order to replicate the best log likelihood. AIC indicated five classes, BIC three classes, SABIC five classes, VLMR four classes, BLRT two classes and the entropy four or five classes. The table of estimated average class probabilities for most likely latent class membership (Supplementary

RI PT

Table 1) shows class 1 to be best predicted in the 2-class model (.95), while 87% of subjects estimated to be in class 2 was estimated to be in this class. In the 3-class model, class 3 was most strongly predicted (.95) and class one least accurately predicted (.73). Overall, the

SC

entropy results showed almost equal predictions in all models (.78-.82). The highest number of subjects in the smallest class was found for the three class model. Based on fit information,

M AN U

class-sizes, model results and an overall evaluation we concluded that the three class model was the best fitting model, with the lowest BIC value and best entropy result (Ram and Grimm, 2009).

The three class model showed one class (15.7%) starting at a very high level and then

EP

(14.7% and 69.6%).

TE D

decreasing to a low-level at T2. The other two classes represented a high and a low-level class

3.4 Trajectory-group-comparisons

AC C

Comparisons of the clinical characteristics in the three identified depressive symptomtrajectories are displayed in Table 3. Participants in the low depressive symptoms-group were significantly more often diagnosed within the schizophrenia-spectrum, displayed more disorganized symptoms, were less cooperative and characterized by lower levels of insight, lower general psychopathology-score, less hopelessness, negative symptoms and suicidal ideation. Alcohol misuse or dependence at inclusion was more frequent in the persistently depressed group than in both the low depression and the early response group. Symptoms of agitation (PANSS Excitement component) were lower in the persistently depressed group 14

ACCEPTED MANUSCRIPT than in the two other groups. There were no further statistically significant differences between the high depression level- and the early response-group. While both the low leveland the early response groups improved considerably regarding positive and negative symptoms, the high depression level group displayed substantially unchanged positive and

RI PT

negative symptoms (data not shown).

3.5 Antidepressant prescription

SC

The distribution of antidepressant prescription is presented in Table 4. The prescription rate

M AN U

was and remained high for the high level depression group, was low in the low-level group and with a rate in-between and reducing in the early response group.

3.6 Sensitivity analysis

TE D

The results after excluding primary affective psychoses were almost identical with the original GMM identifying essentially the same 3 trajectories. The distribution of

EP

schizophrenia spectrum-diagnoses between the trajectories was no longer statistically

AC C

significantly different.

4. Discussion

We identified three different trajectories of depressive symptoms in the acute phase of psychosis: one high level, one early response and one low depressive level class. The high level class remained depressed despite a high rate of antidepressant prescription. The reduction in depressive symptoms was greater in patients who had larger reduction in PANSS positive subscale score. The schizophrenia-spectrum patients had less depressive symptoms at 15

ACCEPTED MANUSCRIPT inclusion. The low depression-level group differed mostly from the other trajectory groups. Patients with post-psychotic depressive symptoms were also depressed in the acute phase.

RI PT

4.1 Depressive symptom change over time In the total group the depressive symptom level on average decreased, particularly early in the follow-up. This corresponds with earlier research, where depressive symptoms are

pronounced at the beginning of a psychotic episode, but decrease parallel to the reduction in

SC

positive psychotic symptoms. However, this phenomenon is previously mainly described in

M AN U

first-episode patients (Hafner et al., 2005; Koreen et al., 1993). Our data suggest that the overall pattern for patients with one or more episodes corresponds with first-episode findings.

4.2 Predictors of reduction in CDSS

TE D

Several clinical characteristics were associated with more pronounced improvements in depressive symptoms. Patients with the greatest improvements in the PANSS Positive subscale score reduced the CDSS sum score most. The relationship between depressive

EP

symptoms and positive symptoms has been described in a few earlier works (Dollfus et al., 1993; Marengo et al., 2000), however with wider time-intervals, repeated cross-sectional

AC C

comparisons, in more long-term data and with fewer participants. Statistically significant associations between antipsychotic-naivety and change in the CDSS were not found. This is in contrast to findings that antipsychotic-naive patients have greater reductions of positive symptoms (Jager et al., 2007; Zhu et al., 2017). Our results do not indicate that this is true for depressive symptoms. The schizophrenia-spectrum subgroup had less depressive symptoms at inclusion, with less improvement than non-schizophrenia patients the first 4 weeks and more improvement in the second time-period. Implications of these varying time-effects are unclear, indeed the concept 16

ACCEPTED MANUSCRIPT of schizophrenia spectrum is widely discussed. The definitions of schizophrenia spectrum disorders vary (American Psychiatric Association, 2013; Jablensky, 2010), thus the present

RI PT

definition may not yield results directly comparable to other schizophrenia spectrum findings.

4.3 Trajectories of depressive symptoms and antidepressant treatment

The three-class model was selected based on an overall evaluation of models. The high-level

SC

patients remained severely depressed despite treatment and a time lapse of 6 months since the beginning of the psychotic episode. We found that a majority of depressed patients were

M AN U

prescribed antidepressants, nevertheless remained highly depressed. This is in line with former reviews and treatment guidelines demonstrating a low efficacy of antidepressants in depression in schizophrenia (Hasan et al., 2015; Whitehead et al., 2002). Furthermore, the persistently depressed participants were characterized by substantially unchanged positive and

TE D

negative symptoms, thus constituting a cross-dimensional treatment-resistant group. Identifying early determinants for this treatment-refractory group would be important. Disappointingly, few identifiers that could separate refractory participants from the early

EP

response group were discovered. However, alcohol misuse or dependence was more frequent at inclusion in the treatment refractory group. This finding must however be interpreted with

AC C

caution due to few participants with reported alcohol abuse in the smallest trajectories. The lack of early clinical identifiers for persistent depressive symptoms underlines the importance of close symptom-surveillance after initiating treatment and avoiding delayed change of treatment in the case of treatment-resistance. Earlier focused treatment might also reduce suicide-risk, which depression is strongly associated with (Cassidy et al., 2017). Although non-significant, there were fewer antipsychotic-naive patients in the high-level group, which might be a possible indicator of chronicity in this group, perhaps leading to a less positive outcome. Candidacy for electroconvulsive treatment was an exclusion criterion, 17

ACCEPTED MANUSCRIPT but may be considered a possible treatment option for such treatment-refractive individuals with schizophrenia (National Institute for Health and Care Excellence, 2003; Pompili et al., 2013; Tharyan and Adams, 2005). Although clozapine treated patients were excluded, clozapine can also be of benefit in severely depressed patients with schizophrenia (Leucht et

RI PT

al., 2009) and might be a treatment option for treatment-resistant depression in persons with schizophrenia.

The early-response group seems to consist of highly treatment-responsive patients. As a

SC

parallel, the rapid response of psychotic symptoms to antipsychotic medication treatment has been described in some papers (Agid et al., 2003; Leucht et al., 2005). This group seems to be

M AN U

characterized by a simultaneous reduction of depression symptoms as the psychosis resolves. Our data support the World Federation of Societies of Biological Psychiatry (WFSBP) – guideline, which recommends that adding antidepressants, if considered, should be limited to depressive symptoms that do not improve with decrease in psychotic symptoms (Hasan et al.,

TE D

2015).

The low depression level-class, which was by far the largest, is perhaps the least clinically relevant group in this setting, as it identifies patients with few depressive symptoms. For this

EP

class, treatment focus should be directed at other challenges of the illness. The high-level group with persistent depression may be considered a “revealed depression”

AC C

subtype of post-psychotic depression, a term originally coined by Knights et al (Knights and Hirsch, 1981). We could not identify a group with emerging depressive symptoms after initiated antipsychotic treatment. If there was an emerging post-psychotic subgroup in the study, defined as being non-depressed in the acute phase, the inability to identify it could be explained by the high attrition rate. There is a possibility that more depressed patients dropped out as in antidepressant trials (Warden et al., 2009; Weissman et al., 2012). Whether depressive symptoms predict attrition in antipsychotic trials is not well known (Kemmler et

18

ACCEPTED MANUSCRIPT al., 2005). Nevertheless, attrition-analyses in this trial have not found differences at inclusion between the patients who dropped out and the patients who completed the trial (data not shown) (Johnsen et al., 2010). Compared with known trajectory findings for primary affective disorders some similar results

RI PT

emerge, although in widely differing cohorts, some population-based and some limited to samples of clinical depression; several of these also reveal treatment-resistant trajectories, a similar number of trajectories and most patients subgrouping under the low symptom level

SC

trajectories. In contrast to our predictor findings, women were overrepresented in trajectories with greater symptom burden. Severity of depressive episode predicted a more chronic

M AN U

course. Some of the other predictors could not be directly compared (Moos and Cronkite, 1999; Musliner et al., 2016a; Musliner et al., 2016b).

TE D

4.4 Sensitivity analysis

The decision to include primary affective psychoses might affect the trajectories identified by the GMM. However, sensitivity analyses conducted without the primary affective psychoses

EP

revealed essentially unaltered trajectories. The early depression response-class was reduced most substantially in size (from 34 to 23). We decided not to exclude drug-related psychoses

AC C

in the sensitivity analyses, due to the fact that these participants were included only if the psychosis did not resolve within a few days of admission and since drug-related psychoses may be difficult to separate from primary psychotic disorders and exhibit a high transition rate to schizophrenia (Medhus et al., 2015).

19

ACCEPTED MANUSCRIPT 4.5 Strengths and limitations The main strengths of this study are the acute clinical setting and clinically relevant representation of consecutively recruited participants, making the results more generalizable to everyday clinical circumstances.

RI PT

Some limitations apply: The attrition rate was high, which makes the models more uncertain. The attrition rate was however, comparable to for instance the CATIE trial, which had a similar study design (Lieberman et al., 2005). The trial was not primarily designed for

SC

investigating depressive symptoms in psychosis. The diagnoses were based on medical

records which might be biased and the validity of the diagnosis was not confirmed by a

M AN U

structured diagnostic interview. Although the proportion of drug addiction was comparable to other naturalistic schizophrenia trials, the mean overall level of alcohol-addiction was low, limiting representability to cohorts with more alcohol-abuse (Jones et al., 2006; Lieberman et al., 2005). More complex level and change models could have been tested, e.g. the latent

TE D

contrast score model in combination with GMM or latent class growth analysis (LCGA). However, some of these would require larger sample size in order to give robust estimates.

EP

Larger samples could also result in higher number of latent classes.

AC C

4.6 Clinical and research implications Due to the inconclusive literature on treatment of depression in schizophrenia, more trials with different arms for treatment of persistent depressive symptoms after a psychotic episode, are needed. Trial arms may include both antidepressants, CBT, switching to clozapine and ECT. Latent growth models and Mixture models may to our experience serve as methods of choice to discover clinically important predictors and patterns of symptom developments for groups of patients. 20

ACCEPTED MANUSCRIPT

5. Conclusions Of clinical importance, an early responding and a treatment refractory group were identified. Positive psychotic symptoms and a diagnosis within schizophrenia spectrum predicted the

RI PT

course of depressive symptoms. The main message from the present paper is to carefully identify the psychotic patients that do not improve from their debilitating depressive

symptoms during an acute phase of psychosis. The highly depressed group remained

SC

depressed despite a high rate of antidepressant treatment. These patients are candidates for an enhanced treatment-plan, for which current evidence is limited. Observation seems to be

M AN U

sufficient for the quickly improving-group. We could not identify differentiating

Trial Registration

TE D

characteristics of the different depression trajectories.

ClinicalTrials.gov ID; NCT00932529; URL: http://www.clinicaltrials.gov/

EP

Role of funding source

The Research Council of Norway initiated funding, followed by Haukeland University

AC C

Hospital, Division of Psychiatry. The supporters had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review or approval of the manuscript.

Contributors EK drafted the manuscript and participated in the data analyses. RG helped draft the manuscript, provided statistical analyses and made substantial contributions to the analysis and interpretation of the data. IS helped draft the manuscript and participated in the 21

ACCEPTED MANUSCRIPT interpretation of the data. RAK helped draft the manuscript and participated in the data collection. EML participated in designing the study and helped draft the manuscript. HAJ participated in designing the study, helped draft the manuscript and participated in data

manuscript. All authors read and approved the final manuscript.

Acknowledgements

RI PT

collection. EJ participated in designing the study, collected the data and helped draft the

SC

The authors thank research nurses Ingvild Helle and Marianne Langeland at the Research Department, Division of Psychiatry, Haukeland University Hospital for their contributions.

M AN U

We also wish to thank the Division of Psychiatry, Haukeland University Hospital for financial support, and the clinical departments for their enthusiasm and cooperation.

TE D

References

AC C

EP

Addington, D., Addington, J., 2015. About the Calgary Depression Scale for Schizophrenia. http://www.ucalgary.ca/cdss/node/9. (Accessed 2017-05-02. Addington, D., Addington, J., Maticka-Tyndale, E., 1994. Specificity of the Calgary Depression Scale for schizophrenics. Schizophr Res 11(3), 239-244. Addington, D., Addington, J., Patten, S., 1996. Gender and affect in schizophrenia. Can J Psychiatry 41(5), 265-268. Addington, D., Addington, J., Schissel, B., 1990. A depression rating scale for schizophrenics. Schizophr Res 3(4), 247-251. Addington, J., Chaves, A., Addington, D., 2006. Diagnostic stability over one year in firstepisode psychosis. Schizophr Res 86(1-3), 71-75. Agid, O., Kapur, S., Arenovich, T., Zipursky, R.B., 2003. Delayed-onset hypothesis of antipsychotic action: a hypothesis tested and rejected. Arch Gen Psychiatry 60(12), 12281235. American Psychiatric Association, 2000. Diagnostic and Statistical Manual of Mental Disorders, 4th Edition. American Psychiatric Publishing, Arlington, VA. American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders, 5th edition, 5th ed. American Psychiatric Publishing,, Arlington, VA. Asparouhov, T., Muthén, B., 2014. Auxiliary Variables in Mixture Modeling: Three-Step Approaches Using Mplus. Structural Equation Modeling: A Multidisciplinary Journal 21(3), 329-341. Birchwood, M., Iqbal, Z., Chadwick, P., Trower, P., 2000. Cognitive approach to depression and suicidal thinking in psychosis. 1. Ontogeny of post-psychotic depression. Br J Psychiatry 177, 516-521. 22

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Birchwood, M., Iqbal, Z., Upthegrove, R., 2005. Psychological pathways to depression in schizophrenia: studies in acute psychosis, post psychotic depression and auditory hallucinations. Eur Arch Psychiatry Clin Neurosci 255(3), 202-212. Birchwood, M., Smith, J., Macmillan, F., Hogg, B., Prasad, R., Harvey, C., Bering, S., 1989. Predicting relapse in schizophrenia: the development and implementation of an early signs monitoring system using patients and families as observers, a preliminary investigation. Psychol Med 19(3), 649-656. Bollen, K.A., Curran, P.J., 2006. Latent Curve Models: A Structural Equation Perspective. Wiley-Interscience, Hoboken, N.J. Bressan, R.A., Costa, D.C., Jones, H.M., Ell, P.J., Pilowsky, L.S., 2002. Typical antipsychotic drugs -- D(2) receptor occupancy and depressive symptoms in schizophrenia. Schizophr Res 56(1-2), 31-36. Bromet, E.J., Kotov, R., Fochtmann, L.J., Carlson, G.A., Tanenberg-Karant, M., Ruggero, C., Chang, S.W., 2011. Diagnostic shifts during the decade following first admission for psychosis. Am J Psychiatry 168(11), 1186-1194. Buchy, L., Bodnar, M., Malla, A., Joober, R., Lepage, M., 2010. A 12-month outcome study of insight and symptom change in first-episode psychosis. Early Interv Psychiatry 4(1), 79-88. Buoli, M., Serati, M., Ciappolino, V., Altamura, A.C., 2016. May selective serotonin reuptake inhibitors (SSRIs) provide some benefit for the treatment of schizophrenia? Expert Opin Pharmacother 17(10), 1375-1385. Cassidy, R.M., Yang, F., Kapczinski, F., Passos, I.C., 2017. Risk Factors for Suicidality in Patients With Schizophrenia: A Systematic Review, Meta-analysis, and Meta-regression of 96 Studies. Schizophr Bull. Caton, C.L., Xie, H., Drake, R.E., McHugo, G., 2014. Gender differences in psychotic disorders with concurrent substance use. Journal of dual diagnosis 10(4), 177-186. Conley, R.R., Ascher-Svanum, H., Zhu, B., Faries, D.E., Kinon, B.J., 2007. The burden of depressive symptoms in the long-term treatment of patients with schizophrenia. Schizophr Res 90(1-3), 186-197. Czobor, P., Volavka, J., 1996. Positive and negative symptoms: is their change related? Schizophr Bull 22(4), 577-590. Dollfus, S., Petit, M., Menard, J.F., 1993. Relationship between depressive and positive symptoms in schizophrenia. J Affect Disord 28(1), 61-69. Drake, R.E., Rosenberg, S.D., Mueser, K.T., 1996. Assessing substance use disorder in persons with severe mental illness. New Dir Ment Health Serv(70), 3-17. Gold, J.M., Queern, C., Iannone, V.N., Buchanan, R.W., 1999. Repeatable battery for the assessment of neuropsychological status as a screening test in schizophrenia I: sensitivity, reliability, and validity. Am J Psychiatry 156(12), 1944-1950. Goldstein, J.M., Link, B.G., 1988. Gender and the expression of schizophrenia. Journal of Psychiatric Research 22(2), 141-155. Guy, W., 1976. ECDEU Assessment Manual for Psychopharmacology. Department of Health, Education, and Welfare, Rockville, MD, USA. Hafner, H., Maurer, K., Trendler, G., an der Heiden, W., Schmidt, M., 2005. The early course of schizophrenia and depression*. Eur Arch Psychiatry Clin Neurosci 255(3), 167-173. Hasan, A., Falkai, P., Wobrock, T., Lieberman, J., Glenthøj, B., Gattaz, W.F., Thibaut, F., Möller, H.-J., 2015. World Federation of Societies of Biological Psychiatry (WFSBP) Guidelines for Biological Treatment of Schizophrenia Part 3: Update 2015 Management of special circumstances: Depression, Suicidality, substance use disorders and pregnancy and lactation. The World Journal of Biological Psychiatry 16(3), 142-170.

23

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Hausmann, A., Fleischhacker, W.W., 2002. Differential diagnosis of depressed mood in patients with schizophrenia: a diagnostic algorithm based on a review. Acta Psychiatr Scand 106(2), 83-96. Hawton, K., Sutton, L., Haw, C., Sinclair, J., Deeks, J.J., 2005. Schizophrenia and suicide: systematic review of risk factors. Br J Psychiatry 187, 9-20. Hazell, C.M., Hayward, M., Cavanagh, K., Strauss, C., 2016. A systematic review and metaanalysis of low intensity CBT for psychosis. Clinical psychology review 45, 183-192. Hirsch, S.R., Jolley, A.G., Barnes, T.R., Liddle, P.F., Curson, D.A., Patel, A., York, A., Bercu, S., Patel, M., 1989. Dysphoric and depressive symptoms in chronic schizophrenia. Schizophr Res 2(3), 259-264. Hunter, A.M., Muthén, B.O., Cook, I.A., Leuchter, A.F., 2010. Antidepressant response trajectories and quantitative electroencephalography (QEEG) biomarkers in major depressive disorder. Journal of Psychiatric Research 44(2), 90-98. IBM, 2015. IBM SPSS Statistics for Windows, 23.0 ed. IBM Corp., Armonk, NY. Jablensky, A., 2010. The diagnostic concept of schizophrenia: its history, evolution, and future prospects. Dialogues in Clinical Neuroscience 12(3), 271-287. Jager, M., Riedel, M., Messer, T., Laux, G., Pfeiffer, H., Naber, D., Schmidt, L.G., Gaebel, W., Huff, W., Heuser, I., Kuhn, K.U., Lemke, M.R., Ruther, E., Buchkremer, G., Gastpar, M., Bottlender, R., Strauss, A., Moller, H.J., 2007. Psychopathological characteristics and treatment response of first episode compared with multiple episode schizophrenic disorders. Eur Arch Psychiatry Clin Neurosci 257(1), 47-53. Johnsen, E., Kroken, R.A., Wentzel-Larsen, T., Jorgensen, H.A., 2010. Effectiveness of second-generation antipsychotics: a naturalistic, randomized comparison of olanzapine, quetiapine, risperidone, and ziprasidone. BMC Psychiatry 10, 26. Jones, C., Hacker, D., Cormac, I., Meaden, A., Irving, C.B., 2012. Cognitive behaviour therapy versus other psychosocial treatments for schizophrenia. Cochrane Database Syst Rev 4. Jones, P.B., Barnes, T.R., Davies, L., Dunn, G., Lloyd, H., Hayhurst, K.P., Murray, R.M., Markwick, A., Lewis, S.W., 2006. Randomized controlled trial of the effect on Quality of Life of second- vs first-generation antipsychotic drugs in schizophrenia: Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS 1). Arch Gen Psychiatry 63(10), 1079-1087. Kaneriya, S.H., Robbins-Welty, G.A., Smagula, S.F., Karp, J.F., Butters, M.A., Lenze, E.J., Mulsant, B.H., Blumberger, D., Anderson, S.J., Dew, M.A., Lotrich, F., Aizenstein, H.J., Diniz, B.S., Reynolds, C.F., 3rd, 2016. Predictors and Moderators of Remission With Aripiprazole Augmentation in Treatment-Resistant Late-Life Depression: An Analysis of the IRL-GRey Randomized Clinical Trial. JAMA psychiatry 73(4), 329-336. Kaplan, K.J., Harrow, M., Clews, K., 2016. The Twenty-Year Trajectory of Suicidal Activity Among Post-Hospital Psychiatric Men and Women with Mood Disorders and Schizophrenia. Arch Suicide Res. Kapur, S., Arenovich, T., Agid, O., Sajeev, G., Muthen, B., Chen, L.C., Kinon, B., 2009. Trajectories of antipsychotic response: moving beyond the responder/non-responder dichotomy. Schizophr Bull 35 (suppl 1), 366. Kay, S.R., Fiszbein, A., Opler, L.A., 1987. The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull 13(2), 261-276. Kemmler, G., Hummer, M., Widschwendter, C., Fleischhacker, W.W., 2005. Dropout rates in placebo-controlled and active-control clinical trials of antipsychotic drugs: a meta-analysis. Arch Gen Psychiatry 62(12), 1305-1312. Kim, S.-Y., 2014. Determining the number of latent classes in single-and multiphase growth mixture models. Structural equation modeling: a multidisciplinary journal 21(2), 263-279. 24

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Kjelby, E., Jorgensen, H.A., Kroken, R.A., Loberg, E.M., Johnsen, E., 2011. Anti-depressive effectiveness of olanzapine, quetiapine, risperidone and ziprasidone: a pragmatic, randomized trial. BMC Psychiatry 11, 145. Kline, R.B., 2011. Principles and practice of structural equation modeling, Third ed. The Guilford Press, New York Knights, A., Hirsch, S.R., 1981. "Revealed" Depression and drug treatment for schizophrenia. Arch Gen Psychiatry 38(7), 806-811. Kohler, C.G., Lallart, E.A., 2002. Postpsychotic depression in schizophrenia patients. Curr Psychiatry Rep 4(4), 273-278. Koreen, A.R., Siris, S.G., Chakos, M., Alvir, J., Mayerhoff, D., Lieberman, J., 1993. Depression in first-episode schizophrenia. Am J Psychiatry 150(11), 1643-1648. Lako, I.M., Bruggeman, R., Knegtering, H., Wiersma, D., Schoevers, R.A., Slooff, C.J., Taxis, K., 2012. A systematic review of instruments to measure depressive symptoms in patients with schizophrenia. J Affect Disord 140(1), 38-47. Lancon, C., Auquier, P., Reine, G., Bernard, D., Addington, D., 2001. Relationships between depression and psychotic symptoms of schizophrenia during an acute episode and stable period. Schizophr Res 47(2-3), 135-140. Leucht, S., Busch, R., Hamann, J., Kissling, W., Kane, J.M., 2005. Early-onset hypothesis of antipsychotic drug action: a hypothesis tested, confirmed and extended. Biol Psychiatry 57(12), 1543-1549. Leucht, S., Corves, C., Arbter, D., Engel, R.R., Li, C., Davis, J.M., 2009. Second-generation versus first-generation antipsychotic drugs for schizophrenia: a meta-analysis. The Lancet 373(9657), 31-41. Levine, S.Z., Leucht, S., 2010. Elaboration on the early-onset hypothesis of antipsychotic drug action: treatment response trajectories. Biol Psychiatry 68(1), 86-92. Levine, S.Z., Lurie, I., Kohn, R., Levav, I., 2011. Trajectories of the course of schizophrenia: from progressive deterioration to amelioration over three decades. Schizophr Res 126(1-3), 184-191. Levine, S.Z., Rabinowitz, J., Faries, D., Lawson, A.H., Ascher-Svanum, H., 2012. Treatment response trajectories and antipsychotic medications: examination of up to 18 months of treatment in the CATIE chronic schizophrenia trial. Schizophr Res 137(1-3), 141-146. Li, L., Hser, Y.-I., 2011. On inclusion of covariates for class enumeration of growth mixture models. Multivariate behavioral research 46(2), 266-302. Lieberman, J.A., Stroup, T.S., McEvoy, J.P., Swartz, M.S., Rosenheck, R.A., Perkins, D.O., Keefe, R.S., Davis, S.M., Davis, C.E., Lebowitz, B.D., Severe, J., Hsiao, J.K., 2005. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med 353(12), 1209-1223. Madsen, T., Karstoft, K.-I., Secher, R.G., Austin, S.F., Nordentoft, M., 2016. Trajectories of suicidal ideation in patients with first-episode psychosis: secondary analysis of data from the OPUS trial. The Lancet Psychiatry 3(5), 443-450. Marengo, J., Harrow, M., Herbener, E.S., Sands, J., 2000. A prospective longitudinal 10-year study of schizophrenia's three major factors and depression. Psychiatry Res 97(1), 61-77. Marques, T.R., Arenovich, T., Agid, O., Sajeev, G., Muthen, B., Chen, L., Kinon, B.J., Kapur, S., 2011. The different trajectories of antipsychotic response: antipsychotics versus placebo. Psychol Med 41(7), 1481-1488. McGlashan, T.H., Carpenter, W.T., Jr., 1976. Postpsychotic depression in schizophrenia. Arch Gen Psychiatry 33(2), 231-239. Medhus, S., Rognli, E.B., Gossop, M., Holm, B., Morland, J., Bramness, J.G., 2015. Amphetamine-induced psychosis: Transition to schizophrenia and mortality in a small prospective sample. Am J Addict 24(7), 586-589. 25

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Montoya, A., Valladares, A., Lizan, L., San, L., Escobar, R., Paz, S., 2011. Validation of the Excited Component of the Positive and Negative Syndrome Scale (PANSS-EC) in a naturalistic sample of 278 patients with acute psychosis and agitation in a psychiatric emergency room. Health Qual Life Outcomes 9, 18. Moos, R.H., Cronkite, R.C., 1999. Symptom-based predictors of a 10-year chronic course of treated depression. J Nerv Ment Dis 187(6), 360-368. Mulholland, C., Cooper, S., 2000. The symptom of depression in schizophrenia and its management. Adv in Psych Treatm 6(3), 169-177. Musliner, K.L., Munk-Olsen, T., Eaton, W.W., Zandi, P.P., 2016a. Heterogeneity in longterm trajectories of depressive symptoms: Patterns, predictors and outcomes. J Affect Disord 192, 199-211. Musliner, K.L., Munk-Olsen, T., Laursen, T.M., Eaton, W.W., Zandi, P.P., Mortensen, P.B., 2016b. Heterogeneity in 10-Year Course Trajectories of Moderate to Severe Major Depressive Disorder: A Danish National Register-Based Study. JAMA psychiatry 73(4), 346353. Muthén, L.K., Muthén, B.O., 2014. Mplus 7.3. Los Angeles, CA. Muthén, L.K., Muthén, B.O., 2015. Mplus User's Guide. Seventh Edition. (Accessed 02.20.2015 2015). National Institute for Health and Care Excellence, 2003. Guidance on the use of electroconvulsive therapy. www.nice.org.uk/guidance/ta59. (Accessed 2017-05-23 2017). Newsom, J.T., 2015. Longitudinal Structural Equation Modeling: A Comprehensive Introduction (Multivariate Applications Series), 1st Edition ed. Taylor & Francis, New York. Norman, R.M.G., Malla, A.K., 1994. Correlations over time between dysphoric mood and symptomatology in schizophrenia. Comprehensive Psychiatry 35(1), 34-38. Nylund, K.L., Asparouhov, T., Muthén, B.O., 2007. Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural equation modeling 14(4), 535-569. Opler, L.A., Kay, S.R., Lindenmayer, J.P., Fiszbein, A., 1999. Structured clinical interview: The positive and negative syndrome scale (SCI-PANSS). Multi-Health Systems, North Tonawanda, NY. Palmer, B.A., Pankratz, V.S., Bostwick, J.M., 2005. The lifetime risk of suicide in schizophrenia: a reexamination. Arch Gen Psychiatry 62(3), 247-253. Peugh, J., Fan, X., 2015. Enumeration index performance in generalized growth mixture models: A Monte Carlo test of Muthén’s (2003) hypothesis. Structural Equation Modeling: A Multidisciplinary Journal 22(1), 115-131. Pompili, M., Lester, D., Dominici, G., Longo, L., Marconi, G., Forte, A., Serafini, G., Amore, M., Girardi, P., 2013. Indications for electroconvulsive treatment in schizophrenia: a systematic review. Schizophr Res 146(1-3), 1-9. Ram, N., Grimm, K.J., 2009. Growth Mixture Modeling: A Method for Identifying Differences in Longitudinal Change Among Unobserved Groups. International journal of behavioral development 33(6), 565-576. Randolph, C., 1998. RBANS Repeatable Battery for the Assessment of Neuropsychological Status. Manual. . The Psychological Corporation, San Antonio. Sands, J.R., Harrow, M., 1999. Depression during the longitudinal course of schizophrenia. Schizophr Bull 25(1), 157-171. Sax, K.W., Strakowski, S.M., Keck, P.E., Jr., Upadhyaya, V.H., West, S.A., McElroy, S.L., 1996. Relationships among negative, positive, and depressive symptoms in schizophrenia and psychotic depression. Br J Psychiatry 168(1), 68-71. Schafer, J.L., Graham, J.W., 2002. Missing data: our view of the state of the art. Psychological methods 7(2), 147-177. 26

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Schennach, R., Obermeier, M., Seemuller, F., Jager, M., Schmauss, M., Laux, G., Pfeiffer, H., Naber, D., Schmidt, L.G., Gaebel, W., Klosterkotter, J., Heuser, I., Maier, W., Lemke, M.R., Ruther, E., Klingberg, S., Gastpar, M., Riedel, M., Moller, H.J., 2012. Evaluating depressive symptoms in schizophrenia: a psychometric comparison of the Calgary Depression Scale for Schizophrenia and the Hamilton Depression Rating Scale. Psychopathology 45(5), 276-285. Sensky, T., Turkington, D., Kingdon, D., Scott, J.L., Scott, J., Siddle, R., O'Carroll, M., Barnes, T.R., 2000. A randomized controlled trial of cognitive-behavioral therapy for persistent symptoms in schizophrenia resistant to medication. Arch Gen Psychiatry 57(2), 165-172. Shtasel, D.L., Gur, R.E., Gallacher, F., Heimberg, C., Gur, R.C., 1992. Gender differences in the clinical expression of schizophrenia. Schizophr Res 7(3), 225-231. Siris, S.G., 2000. Depression in schizophrenia: perspective in the era of "Atypical" antipsychotic agents. Am J Psychiatry 157(9), 1379-1389. Stauffer, V., Case, M., Kollack-Walker, S., Ascher-Svanum, H., Ball, T., Kapur, S., Kinon, B.J., 2011. Trajectories of response to treatment with atypical antipsychotic medication in patients with schizophrenia pooled from 6 double-blind, randomized clinical trials. Schizophr Res 130(1-3), 11-19. Suppes, T., Silva, R., Cucchiaro, J., Mao, Y., Targum, S., Streicher, C., Pikalov, A., Loebel, A., 2016. Lurasidone for the Treatment of Major Depressive Disorder With Mixed Features: A Randomized, Double-Blind, Placebo-Controlled Study. Am J Psychiatry 173(4), 400-407. Tapp, A., Kilzieh, N., Wood, A.E., Raskind, M., Tandon, R., 2001. Depression in patients with schizophrenia during an acute psychotic episode. Compr Psychiatry 42(4), 314-318. Tharyan, P., Adams, C.E., 2005. Electroconvulsive therapy for schizophrenia. Cochrane Database Syst Rev(2), Cd000076. Tohen, M., McDonnell, D.P., Case, M., Kanba, S., Ha, K., Fang, Y.R., Katagiri, H., Gomez, J.C., 2012. Randomised, double-blind, placebo-controlled study of olanzapine in patients with bipolar I depression. Br J Psychiatry 201(5), 376-382. Tollefson, G.D., Andersen, S.W., 1999. Should we consider mood disturbance in schizophrenia as an important determinant of quality of life? J Clin Psychiatry 60 Suppl 5, 2329; discussion 30. Upthegrove, R., Ross, K., Brunet, K., McCollum, R., Jones, L., 2014. Depression in first episode psychosis: the role of subordination and shame. Psychiatry Res 217(3), 177-184. Wang, J., Wang, X., 2012. Structural Equation Modeling: Applications Using Mplus. John Wiley & Sons Ltd., West Sussex, UK. Warden, D., Rush, A.J., Wisniewski, S.R., Lesser, I.M., Kornstein, S.G., Balasubramani, G.K., Thase, M.E., Preskorn, S.H., Nierenberg, A.A., Young, E.A., Shores-Wilson, K., Trivedi, M.H., 2009. What predicts attrition in second step medication treatments for depression?: a STAR*D Report. Int J Neuropsychopharmacol 12(4), 459-473. Weissman, J., Flint, A., Meyers, B., Ghosh, S., Mulsant, B., Rothschild, A., Whyte, E., Group, S.-P.S., 2012. Factors associated with non-completion in a double-blind randomized controlled trial of olanzapine plus sertraline versus olanzapine plus placebo for psychotic depression. Psychiatry Res 197(3), 221-226. Whitehead, C., Moss, S., Cardno, A., Lewis, G., 2002. Antidepressants for people with both schizophrenia and depression. Cochrane Database Syst Rev(2), CD002305. Wickrama, K.K., Lee, T.K., O’Neal, C.W., Lorenz, F.O., 2016. Higher-order growth curves and mixture modeling with Mplus: a practical guide. Routledge. World Health Organization, 2007. International Statistical Classification of Diseases and Related Health Problems, 10th Revision. http://apps.who.int/classifications/icd10/browse/2010/en. (Accessed 2018-01-25.

27

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Yung, A.R., McGorry, P.D., 1996. The prodromal phase of first-episode psychosis: past and current conceptualizations. Schizophr Bull 22(2), 353-370. Zhu, Y., Li, C., Huhn, M., Rothe, P., Krause, M., Bighelli, I., Schneider-Thoma, J., Leucht, S., 2017. How well do patients with a first episode of schizophrenia respond to antipsychotics: A systematic review and meta-analysis. Eur Neuropsychopharmacol.

28

ACCEPTED MANUSCRIPT Table 1: Descriptive statistics for the Calgary Depression Scale for

T1 (baseline) T2 ( ̅ = 4,5 weeks) T3 (3 months) T4 (6 months)

N 225 109 52 39

Mean 6.51 3.90 4.08 3.67

SD Skewness 5.29 0.80 4.05 1.51 3.81 0.95 4.10 1.40

Captions:

AC C

EP

TE D

M AN U

SC

SD (standard deviation)

RI PT

Schizophrenia Sum score

ACCEPTED MANUSCRIPT Table 2: Statistics identifying number of trajectories (Growth Mixture Model)

N in smallest class (%) 32 (14.2) 33 (14.6) 3 (1.3) 6 (2.7)

RI PT

Fit indices from growth mixture models (GMM) for 1-5 classes. Classes AIC BIC SABIC VLMR BLRT Entropy p-value p-value 1 2432.39 2470.02 2435.16 2 2368.95 2420.26 2372.72 .000 .006 .78 3 2340.24 2405.23 2345.02 .000 .158 .79 4 2328.48 2407.15 2334.25 .030 .168 .82 5 2322.40 2414.75 2329.18 .140 .079 .82

AC C

EP

TE D

M AN U

SC

AIC: Akaikes Information Criterion BIC: Bayes Information Criterion SABIC: Sample-size Adjusted BIC Bootstrapped Likelihood Ratio Test significance difference tests for k versus k-1 classes: VLMR: Vuong-Lo-Mendell-Rubin Likelihood Ratio and BLRT: Parametric bootstrapped likelihood ratio test N in smallest class (%) is the number of subjects in the smallest group based on the most likely latent class membership.

ACCEPTED MANUSCRIPT Table 3: Demographics and clinical characteristics of trajectory-groups (3-stepanalysis)

CDUS Abstinence or use without impairment Misuse or dependence

42.1 57.9

28.8 71.2

74.3

82.0

25.7

18.0

75.4

69.2 30.8 46.9 63.5 36.5

83.9

0.91 (0.64)

4.09 (0.13)

14.6 (0.001)

0.84 (0.66)

16.1

26.1 (<0.001)

100

90.9

0

9.1

Mean (SD) 33.1 (12.9) 38.7 (8.4)

Mean (SD) 35.0 (13.8) 38.1 (8.8)

18.6 (4.1)

19.1 (4.2)

20.3 (4.5)

5.43 (0.07)

22.4 (7.5)

20.1 (7.4)

19.0 (7.2)

6.03 (0.049)

37.9 (7.1)

37.8 (7.2)

33.0 (6.1)

24.6 (<0.001)

78.9 (13.6) 1.8 (1.2) 4.3 (1.0) 3.1 (1.2) 1.1 (0.4)

76.9 (13.5) 1.8 (1.2) 4.6 (0.9) 3.3 (1.1) 1.0 (0.1)

72.3 (12.9) 2.6 (1.6) 3.3 (1.3) 1.8 (1.1) 1.4 (0.8)

8.72 (0.013) 21.4 (<0.001) 58.9 (<0.001) 72.8 (<0.001) 38.3 (<0.001)

2.9 (1.3)

2.7 (1.3)

4.1 (1.3)

47.9 (<0.001)

3.3 (1.7)

3.2 (1.8)

2.3 (1.5)

15.6 (<0.001)

1.5 (0.8) 1.5 (0.7) 28.9 (6.4) 5.2 (0.6) 6.9 (2.1)

1.8 (0.8) 1.6 (0.9) 29.1 (6.2) 5.4 (0.6) 8.2 (2.9)

0.3 (0.5) 0.1 (0.3) 31.3 (5.7) 5.1 (0.6) 8.6 (3.1)

173.8 (<0.001) 210.7 (<0.001) 7.07 (0.029) 5.31 (0.07) 15.1 (0.001)

24.6

Mean (SD) 30.7 (11.6) 37.9 (9.6)

AC C

Captions:

2

Χ (p-value)

RI PT

59.2 40.8 49.5

EP

Age Neurocognitive sum score (RBANS t-scores) PANSS Positive subscale sum PANSS Negative subscale sum PANSS General psychopathology sum PANSS total PANSS P2 Disorganization PANSS G2 Anxiety PANSS G3 Guilt feelings PANSS G8 Uncooperativeness PANSS G12 Lack of judgment and insight PANSS G16 Active social avoidance CDSS item 2 Hopelessness CDSS item 8 Suicide GAF-F-score CGI-S PANSS Excitement component

66.8 33.2 26.3

Group 3 -low depression level (%)

TE D

CAUS Abstinence or use without impairment Misuse or dependence

Group 2 -early depression response (%)

SC

Gender Male Female Antipsychotic-naive a Diagnosis ICD-10) Schizophrenia spectrum Other psychoses

Group 1 - high depression level (%)

M AN U

Variable

2

Χ (p-value) 3.65 (0.16) 0.16 (0.92)

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ICD-10 (International Classification of Diseases, version 10), CDSS (Calgary Depression Scale for Schizophrenia), CDUS (Clinical Drug Use Scale), CAUS (Clinical Alcohol Use Scale), SD (Standard Deviation), PANSS (the Positive and Negative Syndrome Scale for Schizophrenia), RBANS (Repeatable Battery for the Assessment of Neuropsychological Status), GAF-F (Global Assessment of Functioning), CGI-S (Clinical Global Impression – Severity Scale), PANSS Excitement component (the sum of PANSS items P4 Hyperactivity, P7 Hostility, G4 Tension, G8 Uncooperativeness and G14 Poor Impulse Control).

ACCEPTED MANUSCRIPT Table 4: Antidepressant prescription 4 weeks 17% 22% 65%

6 months 19% 20% 71%

AC C

EP

TE D

M AN U

SC

RI PT

Low-level group Early response group High level group

3 months 17% 38% 88%

ACCEPTED MANUSCRIPT

RI PT

Figure 1: Distribution of diagnoses

Substance abuse-related psychosis (12.8%)

SC

Schizophrenia and schizophreniform psychosis (26.5%)

M AN U

Schizoaffective disorder (2.7%)

AC C

EP

TE D

Schizotypal disorder (2.2%) Delusional disorder and acute delusional disorder (23.5%) Other primary psychotic disorders (13.3 %) Uni- or bipolar depression with psychotic symptoms (10.2%) Other (4.9%) Missing (4%)

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Figure 2: Flow of patients through the study

ACCEPTED MANUSCRIPT

Assessed for eligibility

RI PT

(100%)1

Excluded

Not meeting inclusion criteria (1.2%)1 Unable to assess (uncoop, organic braindis.) (46.5%)1 Randomization not acceptable (6.8%)1 Administrative causes (15.0%)1

Randomized

M AN U

SC

(N=226) (30.5%)

Risperidone

Olanzapine

Quetiapine

Ziprasidone

Allocated to drug (N=57) Received allocated drug (N=44) Chose another drug (N=12) Unknown choice of drug (N=1) Did not take any drug doses of received allocated drug (N=5)

Allocated to drug (N=54) Received allocated drug (N=45) Chose another drug (N=9) Did not take any drug doses of received allocated drug (N=5)

Allocated to drug (N=52) Received allocated drug (44) Chose another drug (N=8) Did not take any drug doses of received allocated drug (N=3)

Allocated to drug (N=63) Received allocated drug (=52) Chose another drug (N=11) Did not take any drug doses of received allocated drug (N=4)

TE D

Lost to follow-up2 Uncoop (N=13) Polypharmacy (N=4) Discharge (N=8) Depot (N=0) Other (N=1) Total (N=26)

AC C

EP

Lost to follow-up2 Uncoop (N=7) Polypharmacy (N=6) Discharge (N=15) Depot (N=1) Other (N=1) Total (N=30)

Lost to follow-up2 Uncoop (N=6) Polypharmacy (N=1) Discharge (N=17) Depot (N=0) Other (N=1) Total (N=25)

Follow-up

Follow-up

Follow-up

Follow-up

Discharge/ 6 weeks (N=30)3 3 months (N=15) 6 months (N=11)

Discharge/ 6 weeks (N=23)3 3 months (N=14) 6 months (N=11)

Discharge/ 6 weeks (N=27) 3 months (N=11) 6 months (N=8)

Discharge/ 6 weeks (N=29) 3 months (N=12) 6 months (N=9)

Lost to follow-up2 Uncoop (N=7) Polypharmacy (N=6) Discharge (N=12) Depot (N=1) Other (N=8) Total (N=34)

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Flow of patients through the study. Not meeting inclusion criteria = score below 4 on all the items: delusions, hallucinatory behaviour, grandiosity, suspiciousness/persecution or unusual thought content in the Positive and Negative Syndrome Scale (PANSS); Uncoop. = the patient was not able or willing to cooperate with testing and assessments; Organic braindis. = Organic brain disorder, principally dementia; Randomization not acceptable = patient or treating clinician not willing to change existing antipsychotic medication; Administrative causes = principally patient discharged before assessments could be made. 1 Enrolment started March 2003 until 2008, week 26. Full details on enrolment were only registered from 2006, week 31 until 2008, week 26. Consequently 2 3 only percentages are displayed for patients assessed for eligibility and excluded patients. Before discharge/6 weeks. One patient in the risperidone and olanzapine groups missed the first follow-up visit, but were retested on later visits.

ACCEPTED MANUSCRIPT Figure 3: Plot of predictors of depression level and change extracted from the LGCM-analyses

Other psychoses

Captions:

TE D

M AN U

SC

RI PT

Schizophrenia spectrum

Level and change in CDSS (Calgary Depression Scale for Schizophrenia) predicted by level and change in the PANSS (Positive And Negative Syndrome Scale for Schizophrenia), being

EP

medication naïve (no former use of antipsychotic medication) versus previous use of antipsychotic medication and schizophrenia versus other psychosis diagnoses. Due to the

AC C

sample size, all predictor variables were entered in a one sample analysis (N = 224).

ACCEPTED MANUSCRIPT Figure 4: Depression trajectories (Growth mixture model)

16 14 High depression level group (14.7%)

12

RI PT

Early depression response group (15.7%)

8 6 4

Low depression level group (69.6%)

2 0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28

M AN U

Weeks

SC

CDSS

10

Captions:

AC C

EP

TE D

Level and change in Calgary Depression Scale for Schizophrenia (CDSS) in three subpopulations based on growth mixture models (GMM).