Neurological soft signs in first-episode schizophrenia: State and trait related

relationships to psychopathology, cognition and antipsychotic medication effects

Running title: Neurological soft signs in schizophrenia

Robin Emsley*,a, Bonginkosi Chilizaa, Laila Asmala, Sanja Kiliana, M. Riaan Oliviera,

Lebogang Phahladiraa, Akinsola Ojagbemib, Freda Schefflera, Jonathan Carrc, Martin Kiddd,

Paola Dazzane.

a Department of Psychiatry, Faculty of Medicine and Health Sciences, Stellenbosch

University, Tygerberg Campus, Cape Town, South Africa

b Department of Psychiatry, University of Ibadan, Nigeria

c Division of Neurology, Faculty of Medicine and Health Sciences, Stellenbosch University,

Tygerberg Campus, Cape Town, South Africa

d Centre for Statistical Consultation, Stellenbosch University, South Africa

e Psychosis Studies Department, Institute of Psychiatry, Psychology and Neuroscience,

King’s College London, London, UK.

*To whom correspondence should be addressed: Department of Psychiatry, Faculty of

Medicine and Health Sciences, Stellenbosch University, PO Box 19063, Tygerberg Campus,

Cape Town 7505, South Africa; tel: +27 21 9389227; fax:+27 21 9389738; e-mail:

rae@sun.ac.za

Word count:

Abstract: 245

Text body: 3146

1

Abstract

Background: Neurological soft signs (NSS) are proposed to represent both state- and trait-

related features of schizophrenia.

Method: We assessed the course of NSS with the Neurological Evaluation Scale (NES) over

12 months of standardised treatment in 126 patients with first-episode schizophrenia,

schizophreniform or schizoaffective disorder, and evaluated their state- and trait-related

associations with psychopathology, functionality, cognition and antipsychotic treatment. We

considered change scores from baseline to be state-related and endpoint scores to be trait-

related. Results: Significant effects for time were recorded for all NSS domains. For state-

related change-scores greater improvements in sensory integration were predicted by more

improvement in working memory (p=0.01); greater improvements in motor sequencing

scores were predicted by more improvement in working memory (p=0.005) and functionality

(p=0.005); and greater improvements in NES Total score were predicted by more

improvement in disorganised symptoms (p=0.02). There were more substantial associations

between trait-related endpoint scores than for state-related change scores. For endpoint

scores lower composite cognitive score predicted poorer sensory integration (p=0.001);

higher Parkinsonism score predicted poorer motor co-ordination (p=0.0001); lower

composite cognitive score (p=0.001) and higher Parkinsonism score (p=0.005) predicted

poorer motor sequencing; higher Parkinsonism score (p=0.0001) and disorganised

symptoms (p=0.04), and lower composite cognitive score (p=0.0007) predicted higher NES

total score.

Conclusions: NSS improved with treatment, but were weakly associated with improvements

in psychopathology. Studies investigating NSS as trait-markers should ensure that patients

have been optimally treated at the time of testing, and should take possible effects of

extrapyramidal symptoms into account.

Keywords: neurological; schizophrenia; psychosis; outcome; depot; flupenthixol

2

1. Introduction

Neurological soft signs (NSS) are a well-established component of the symptom

expression of schizophrenia spectrum disorders (Chan et al., 2010). As such, they have

been proposed as an endophenotype for these conditions (Chan and Gottesman, 2008).

NSS are present at the time of the first psychotic episode (Dazzan and Murray, 2002), and

prior to first initiation of antipsychotic treatment (Gupta et al., 1995;Peralta et al., 2011). NSS

are not static over time, although findings are inconsistent. While several studies reported

that NSS remain stable, or worsen over the course of the illness (Madsen et al., 1999;Smith

et al., 1999;Chen et al., 2000;Chen et al., 2005) a recent meta-analysis reported that 14 of

17 longitudinal studies found reductions in NSS in parallel with symptom improvement

(Bachmann et al., 2014). NSS improvements were greater in patients achieving remission

(Bachmann et al., 2014), and worsened in patients not attaining remission (Prikryl et al.,

2012). However, even in patients achieving remission, some NSS persisted. This has led to

the proposal that NSS represent both state- and trait- related features of the illness

(Bachmann et al., 2014). Which NSS are state or trait markers is not clear (Cuesta et al.,

2012). Many previous studies are limited by methodological shortcomings including small

samples, chronic samples, varying degrees of antipsychotic exposure, different follow-up

durations and non-standardisation of treatments (Chan et al., 2015). Another critical

consideration is timing of baseline assessments (Bachmann et al., 2014). For assessing

state-related NSS, baseline assessments should be conducted during the acute psychosis,

before treatment. Similarly, trait-related NSS should be investigated after optimal treatment-

response.

We conducted a study addressing several of the methodological inconsistencies.

Firstly, we selected first-episode patients to avoid the effects of disease chronicity. Secondly,

by selecting treatment-naïve or minimally treated patients we minimised contaminating

effects of prior medication. This also allowed us to accurately document treatment-emergent

extrapyramidal symptoms (EPS) and their relationship to NSS. Thirdly, by treating patients

3

with a single antipsychotic and according to a fixed protocol, we avoided differential effects

of various antipsychotics. Fourth, we used a low-dosing strategy to minimise medication

effects on NSS. Finally, using a depot antipsychotic formulation eliminated the confounding

effect of non-adherence which may be substantial, given the very high rates of non-

adherence reported in the early stages of illness (Coldham et al., 2002). Our aim was to

investigate the course of NSS over the first 12 months of standardised treatment and to

assess how both state- and trait-related NSS are related to other symptoms, functionality,

cognition, antipsychotic dose and emergent EPS.

4

2. Methods

This longitudinal study assessed the outcome of patients with a first-episode of

schizophrenia or related disorder treated with the lowest effective dose of flupenthixol

decanoate according to a standard protocol over 12 months. Ethics approval was obtained

from the Human Research Ethics Committee of Stellenbosch University Faculty of Medicine

and Health Sciences.

2.1 Participants

Participants were recruited from hospitals and community clinics in Cape Town and

vicinity. Written, informed consent was obtained from participants. Inclusion criteria were:

men and women; aged 16-45 years; experiencing a first psychotic episode meeting

Diagnostic and Statistical Manual of Mental Diseases, Fourth Edition, Text Revisions (DSM-

IV TR) (American Psychiatric Association, 1994) criteria for schizophrenia, schizophreniform

or schizo-affective disorder. Exclusion criteria were: lifetime exposure >4 weeks

antipsychotic medication; previous treatment with a depot antipsychotic; mental retardation;

overt current substance abuse; unstable general medical condition; history of head injury or

epilepsy.

2.2 Treatment

Patients received oral flupenthixol 1-3mg/day for 7 days, followed by flupenthixol

decanoate injections 2-weekly for the study duration. Initiation dose was 10mg 2-weekly,

with 6-weekly increments of 10mg 2-weekly IMI permitted, to a maximum of 30mg 2-weekly

IMI. Additional oral flupenthixol was permitted at the discretion of the investigator, as was

lorazepam, anticholinergics, propranolol, antidepressants and medication for general

medical conditions. No benzodiazepines, propranolol or anticholinergics were permitted in

the 12 hours prior to assessments.

2.3 Assessments

5

Assessments were conducted at baseline (prior to treatment), 6 and 12 months.

2.3.1 Neurological soft signs

NSS were assessed with the Neurological Evaluation Scale (NES) (Buchanan and

Heinrichs, 1989). Thirteen of 26 items are assigned to three ‘functionally meaningful’ sub-

scales reflecting dysfunction in sensory integration, motor coordination and motor

sequencing, and NES total score (Buchanan and Heinrichs, 1989;Sanders et al., 2000). NES

assessments were performed by three psychiatrists (BC, LA, LP) who underwent initial

training (Inter-rater reliability >.9) and participated in ongoing reliability assessments to

ensure stability of rating over time.

2.3.2 Clinical evaluations

Patients were assessed with the Structured Clinical Interview for DSM-IV (SCID)

(First et al., 1994). Psychopathology was assessed with the Positive and Negative

Syndrome Scale (PANSS) (Kay et al., 1987). We used factor-analysis derived symptom

domains for positive, negative and disorganised symptoms (Emsley et al., 2003).

Functionality was assessed with the Social and Occupational Functioning Assessment Scale

(SOFAS) (American Psychiatric Association, 1994), depressive symptoms with the Calgary

Depression Rating Scale for Schizophrenia (CDSS) (Addington and Addington J., 1993) and

EPS with the Extrapyramidal Symptom Rating Scale (ESRS) (Chouinard and Margolese,

2005).

2.3.3 Cognitive evaluations

Cognitive performance was assessed by the MATRICS (Measurement and

Treatment Research to Improve Cognition in Schizophrenia) Cognitive Consensus Battery

(MCCB), developed to measure cognitive functioning in schizophrenia. The MCCB

6

measures seven cognitive domains (speed of processing; attention/vigilance; working

memory; verbal learning; visual learning; reasoning and problem solving; and social

cognition) and a composite score. (Nuechterlein and Green, 2006). The MCCB was

administered by trained psychologists.

2.4 State vs. trait assessments

NSS display both state and trait characteristics (Bachmann et al., 2014). We

assumed that NES scores in the acute state represent both state and trait components,

while those in clinically stable patients represent trait components, and the degree of change

from the acute state to stable state would represent state components. Therefore, for state-

related NSS we used change scores (endpoint - baseline score). (For ESRS change scores

we subtracted the baseline score from the highest score at any time point.) For trait-related

NSS we used endpoint scores, calculated by last observation carried forward.

2.5 Statistical analyses

All participants with completed baseline NES assessments were included in the

analyses. We assessed the distribution of the data by inspection of histograms and normal

probability plots. We employed linear mixed-effect models for continuous repeated measures

(MMRM) to assess the changes in NSS subscale and total scores over time, with age,

gender and previous antipsychotic exposure (yes/no) as covariates. Within analyses Fisher’s

Least Significant Difference (LSD) tests were used for post hoc multiple comparisons. We

conducted Pearson correlational coefficient analyses to investigate both state (change

scores) and trait-related (endpoint scores) associations between NSS and psychopathology,

cognition and EPS. We used general regression models to further explore these

correlations. NSS subscales and total score were dependent variables, covariates were age,

gender and level of education and for the predictor variables we were guided by effect sizes

>0.3 (moderate) in the correlational analyses. All tests were 2-tailed.

7

3. Results

Details of the 126 participants are provided in Table 1, and baseline and endpoint NSS,

clinical and cognitive scores are provided in Table 2. There were significant improvements in

NSS scores, as well as in psychopathology, functionality and cognitive performance. There

were significant increases in ESRS scores for Parkinsonism and dyskinesia subscales.

3.1 Longitudinal course of NSS

For the linear mixed effects models there was a significant effect for time for NES

total and all subscale scores (Figure 1). Post-hoc LSD test results indicated similar patterns

for the NES total and subscale scores – i.e. significant reductions from baseline to month 6

but no further reductions thereafter. Mean reductions (and confidence intervals) for baseline

to month 6 and from month 6 to month 12 were, respectively, NES total score=4.8(3.4-6.1),

p<0.0001 for 0-6 months and 0.8(-0.7-2.3), p=0.2 for 6-12 months; sensory

integration=1.3(0.8-1.8), p<0.0001 for 0-6 months and 0.1(-0.4-0.6), p=0.7 for 6-12 months;

motor coordination=0.5(0.1-0.8), p=0.007 for 0-6 months and 0.1(-0.2-0.4), p=0.5 for 6-12

months; motor sequencing=1.2(0.7-1.7), p<0.0001 for 0-6 months and 0.2(-0.4-0.7), p=0.5

for 6-12 months.

3.2 State-related change-scores

The change score correlations are provided in Supplementary Table 1. Mostly, effect

sizes for NSS correlations with psychopathology, functionality and cognition were small (0.1-

0.3) to moderate (0.3 – 0.5). For the general regression models, we entered gender as a

categorical factor and age, highest level of education, and change scores for PANSS

disorganised factor, SOFAS and MCCB working memory as continuous predictors. For the

sensory integration subscale model (R2=0.1, p=0.1) improved working memory (t=-2.6,

p=0.01) was the only significant predictor; for the motor coordination model (R2 0.07, p=0.2)

there were no significant predictors; for the motor sequencing model (R2=24, p=0.0001)

working memory (t=-2.9, p=0.005) and SOFAS score (t=-2.8, p=0.005) were significant

8

predictors; and for the NES Total score model (R2 22, p=0.0001) PANSS disorganised factor

(t=2.3, p=0.02) was a significant predictor.

3.3 Trait-related endpoint analyses

The endpoint score correlations are provided in Supplementary Table 2. There were

considerably more correlations with a moderate (0.3-0.5) and even large (>.5) effect size

between the endpoint scores than was the case for the change scores. For the general

regression models, we entered gender as a categorical factor and age, highest level of

education, PANSS positive, negative and disorganised factors, SOFAS, MCCB composite

score and ESRS Parkinsonism score as continuous predictors. (We elected to enter the

MCCB composite score only, as most of cognitive domains showed moderate to strong

correlations with NSS.) For the sensory integration subscale model (R2 0.3, p=0.0001),

MCCB composite score was the only significant predictor (t=-3.4, p=0.001); for the motor co-

ordination subscale model (R2 0.35, p=0.001) ESRS Parkinsonism score was the only

significant predictor (t=4.9, p= 0.0001); for the motor sequencing subscale model (R2 0.47,

p=0.0001) MCCB composite score (t=-3.4, p=0.001) and ESRS Parkinsonism score (t=2.8,

p=0.005) were significant predictors. For the NES total score model (R2 0.53, p=0.0001)

ESRS Parkinsonism score (t=3.7, p=0.0001), PANSS disorganised factor (t=2.0, p=0.04)

and MCCB composite score (t=-3.5, p=0.0007) were significant predictors.

9

4. Discussion:

While there have been a number of studies of NSS in first-episode psychosis (Schroder

et al., 1998;Whitty et al., 2003;Chen et al., 2005;Bachmann et al., 2005;Emsley et al.,

2005;Boks et al., 2006;Prikryl et al., 2007;Mayoral et al., 2008;Prikryl et al., 2012;Mayoral et

al., 2012), this is the first to prospectively assess changes in NSS in first episode

schizophrenia in which patients were treated according to a standard protocol with assured

medication adherence. Our main finding was that significant reductions occurred in all NSS

domains. While some state-related NSS reductions were related to improvements in

components of cognition, psychopathology and functionality, there were stronger

associations between the trait-related endpoint NSS scores and these domains. While

antipsychotic treatment was associated with an overall beneficial effect on NSS, the

presence of Parkinsonian side-effects predicted higher trait-related NSS scores.

4.1 Course of NSS:

The improvements in NSS that we observed are consistent with most longitudinal

studies (Bachmann et al., 2014). Our results further indicate that the greatest reductions

occurred between baseline and month 6, with no additional improvements thereafter. The

timing of the initial assessments may be critically important when assessing longitudinal NSS

changes. Our baseline assessments were conducted in the acute psychotic state, before

treatment was initiated. Of the four studies reporting no longitudinal improvement, one

conducted baseline assessments in chronically hospitalised patients (Smith et al., 1999) and

one in stable, chronic patients (Chen et al., 2000). The two other studies did not specify

whether patients were initially assessed prior to initiation of treatment – one stated that they

were assessed “at the time of their first admission” (Madsen et al., 1999), and the other “at

initial assessment” in first-episode patients (Chen et al., 2005). We observed improvements

in all of the NES subscales, suggesting that the state-related component of NSS is

10

generalised. This differs slightly from a study reporting improvements in motor coordination

and motor sequencing but not in sensory integration (Whitty et al., 2003), but is similar

another in which improvements in all subscales were reported (Cuesta et al., 2012).

4.2 State-related associations between NSS, psychopathology and functionality:

Several previous studies have reported a relationship between changes in NSS and

changes in psychopathology. Chan et al.(Chan et al., 2015) examined the course of NSS in

145 patients with first-episode schizophrenia, and compared patients with prominent

negative symptoms with those without prominent negative symptoms, for up to a year. They

found that, despite general improvement in NSS, the patients with prominent negative

symptoms exhibited poorer motor coordination and higher overall levels of NSS. Whitty et

al., (Whitty et al., 2003) reported that reductions in NES total scores and motor coordination

subscales were associated with improvements in PANSS total and positive and negative

subscale scores; reductions in motor sequencing scores were associated with improvements

in PANSS positive and negative subscale scores; and reductions in sensory integration

scores were associated with improvements in PANSS positive subscale score. Our findings

in terms of change correlations differ from the above studies insofar as only the PANSS

disorganised factor independently predicted NSS changes, and only with NES total score. A

possible explanation for this is that most previous studies applied the originally described

subscales of the PANSS, i.e. positive, negative and general psychopathology. Later PANSS

factor-analyses reported a best-fit five factor model comprising positive, negative and

disorganised domains (as well as anxiety/depression and excitement/hostility domains)

(Emsley et al., 2003). The disorganised domain that we used includes two items from the

original PANSS negative subscale, i.e. conceptual disorganisation and difficulty in abstract

thinking, plus the items stereotyped thinking, disorientation, poor attention and

preoccupation (Emsley et al., 2003). Indeed, another study using factor-analysis derived

PANSS domains also found that NSS change scores were positively correlated with the

11

disorganised (and positive) domains (Cuesta et al., 2012). Our findings are therefore only

partly consistent with the proposal that NSS changes are linked to changes in

psychopathology, and can be attributed to treatment response (Bombin et al., 2005).

4.3 State-related associations between NSS and cognitive performance:

Associations between NSS and cognitive performance are well recognised in

schizophrenia (Mohr et al., 2003;Mellacqua et al., 2012), and have led to the proposal that

they are manifestations of the same underlying pathophysiological process (Mellacqua et al.,

2012). In chronic patients, deficits in sensory integration had the strongest association with

cognitive impairments (Arango et al., 1999), while motor sequencing deficits were associated

with poor planning and inhibition, spatial span and spatial memory (Arabzadeh et al., 2014).

Furthermore, associations were reported between NSS and cognitive function in healthy

volunteers, although there are several associations that appear to be specific to psychotic

illness (Mellacqua et al., 2012). Higher scores in all NES subscales have been associated

with poorer general cognitive function, while poorer performances on the sensory integration

and motor sequencing tests are associated with deficits in specific cognitive domains

(Mellacqua et al., 2012). However, these associations were reported in cross-sectional

studies. The association between state-related NSS and cognition is less clear. We found

only one cognitive domain to be a significant predictor of NES scores - impaired working

memory predicted poorer sensory integration, motor sequencing and NES total scores -

suggesting that the state related component of NSS is related specifically to this cognitive

domain.

4.4 Trait-related associations between NSS, psychopathology and functionality:

Associations between NSS and psychopathology and functionality were stronger for

endpoint scores than for change scores, and the endpoint prediction models explained more

12

of the variation than was the case with the change scores. Nevertheless, the only

independent predictor of endpoint NSS that retained significance in the regression models

was the PANSS disorganised factor, and only for NES total score. The strong association at

endpoint between motor sequencing difficulties and persistent psychopathology and poorer

functional outcome is consistent with a report that motor-sequencing difficulties remained

relatively unchanged across the course of illness, suggesting that they are the trait-marking

NSS of schizophrenia (Ojagbemi et al., 2015).

4.5 Trait-related associations between NSS and cognitive performance:

As was the case for psychopathology and functionality, there were more substantial

associations between NSS and cognitive function in terms of the endpoint scores than for

change scores. This suggests a closer association between the trait-related components of

NSS and cognitive function than those related to the acute psychotic state. Also, the fact that

cognitive impairment significantly predicted poorer performance on NES sensory

integration, motor sequencing and NES total scores suggests that associations between

trait-related NSS and cognitive function are global rather than specific. That social cognition

was the only cognitive domain not to be significantly correlated with NSS may be due to this

cognitive domain being less suitable for non-Western cultural settings (Karim and Weisz,

2010).

The importance of these trait-related NSS and their clinical and cognitive

concomitants is that they likely reflect the persistent underlying neurodevelopmental

impairments associated with schizophrenia. Given the reported specificity and sensitivity of

NSS in schizophrenia spectrum disorders (Chan et al., 2016) and evidence linking them to

specific brain regions Mittal (Mittal, 2016), future studies investigating the

neurodevelopmental impairments in schizophrenia would do well to focus on endpoint NSS

specifically.

13

4.6 Effects of antipsychotic treatment on NSS:

Most evidence to date indicates that NSS improve with antipsychotic treatment, and

that this coincides with improvements in accompanying symptoms (Bachmann et al., 2014).

Our findings partly confirm this, although the effect sizes for change correlations between

NSS and psychopathology were only small to moderate. This suggests that NSS

improvements are not a direct consequence of improvements in psychopathology. Also, the

limited variation explained by the change score predictor models suggests that other factors

are involved. Whether the treatment effect on NSS is non-specific and secondary to

improvements in e.g. attention, concentration and cooperativeness remains to be

determined.

NSS have been reported to not be exacerbated by antipsychotic medication (Browne

et al., 2000). However, our finding that higher endpoint ESRS Parkinsonism and total scores

independently predicted greater motor coordination and motor sequencing difficulties is

consistent with a previous report of an association between NSS and EPS (Jahn et al.,

2006) and suggests that emergent EPS may negatively affect performance on the NES

scale – i.e. that some NSS may be iatrogenic. Even subtle EPS may have an effect, as we

prescribed the lowest effective antipsychotic dose and EPS were generally very mild (Chiliza

et al., 2016). Antipsychotic-induced EPS cannot, however, entirely account for these NSS,

as abnormalities in motor coordination and motor sequencing have also been reported in

antipsychotic-naïve individuals (Dazzan et al., 2008).

Strengths of this study include the selection of first episode patients who were

treatment naïve or only minimally medicated, standardised treatment regime with assured

adherence, and repeated, comprehensive assessments of psychopathology, functionality,

cognition and EPS. There are also limitations to our study. First, the absence of a healthy

14

control group prevented us from assessing practice effects, and from inferring that our

findings are specific to schizophrenia. Second, study investigators were not blinded as to

visit status, and as far as possible patients were assessed at follow-up visits by the same

investigator to minimise inter-rater variations. This introduced the possibility of observer bias.

Third, IQ was not measured and therefore could not be controlled for. This may be

important, as both premorbid and current IQ may influence NES performance (Dazzan et al.,

2008). We used highest education attained as a proxy for IQ which is not ideal, as other

factors such as socioeconomic circumstances and age of illness onset could have influenced

educational status. Finally, our findings cannot necessarily be generalised to patients treated

with other antipsychotics. However, we consider it unlikely that our findings are restricted to

treatment with flupenthixol, or to the conventional antipsychotics, for the following reasons:

Conventional and atypical antipsychotics are not distinct classes and it has been

recommended that this categorisation be abandoned (Leucht et al., 2009); flupenthixol has

receptor binding profile that is similar to some atypical antipsychotics and has been referred

to as a “partially atypical” antipsychotic (Gatazz et al., 2004); and because we used the

lowest effective dose we were able to minimise the occurrence of EPS. Indeed, the levels of

EPS reported in our study are similar to those reported with atypical antipsychotics in a

similar patient sample (Kahn et al., 2008).

In conclusion, our study provides further evidence that NSS improve with treatment

and that improvements are only partly associated with improvements in psychopathology,

functionality and cognition. The stronger endpoint associations between NSS and those of

psychopathology, functionality and cognition suggest that trait-related components of these

phenomena are more closely related to one another than state-related components. Future

studies should take care to distinguish between NSS present in acute psychotic states and

those present in optimally treated, stable patients. The latter have a different set of clinical

and cognitive correlates, and are better suited as trait endophenotypes.

15

References

Addington,D. and Addington J. Assessing depression in schizophrenia: the Calgary

Depression Scale. Br J Psychiatry. 163, S39-S44. 1993.

American Psychiatric Association, 1994. Diagnostic and Statistical Manual of Mental

Disorders., 4th edn. Washington, D.C..

Arabzadeh,S., Amini,H., Tehrani-Doost,M., Sharifi,V., Noroozian,M. and Rahiminejad,F.,

2014. Correlation of neurological soft signs and neurocognitive performance in first

episode psychosis. Psychiatry Res. 220, 81-88.

Arango,C., Bartko,J.J., Gold,J.M. and Buchanan,R.W., 1999. Prediction of

neuropsychological performance by neurological signs in schizophrenia. Am J

Psychiatry. 156, 1349-1357.

Bachmann,S., Bottmer,C. and Schroder,J., 2005. Neurological soft signs in first-episode

schizophrenia: a follow-up study. Am J Psychiatry. 162, 2337-2343.

Bachmann,S., Degen,C., Geider,F.J. and Schroder,J., 2014. Neurological soft signs in the

clinical course of schizophrenia: results of a meta-analysis. Front Psychiatry. 5, 185.

Boks,M.P., Selten,J.P., Leask,S. and van den Bosch,R.J., 2006. The 2-year stability of

neurological soft signs after a first episode of non-affective psychosis. Eur Psychiatry.

21, 288-290.

Bombin,I., Arango,C. and Buchanan,R.W., 2005. Significance and meaning of neurological

signs in schizophrenia: two decades later. Schizophr Bull. 31, 962-977.

Browne,S., Clarke,M., Gervin,M., Lane,A., Waddington,J.L., Larkin,C. and O'Callaghan,E.,

2000. Determinants of neurological dysfunction in first episode schizophrenia.

Psychol Med. 30, 1433-1441.

16

Buchanan,R.W. and Heinrichs,D.W., 1989. The Neurological Evaluation Scale (NES): a

structured instrument for the assessment of neurological signs in schizophrenia.

Psychiatry Res. 27, 335-350.

Chan,R.C., Geng,F.L., Lui,S.S., Wang,Y., Ho,K.K., Hung,K.S., Gur,R.E., Gur,R.C. and

Cheung,E.F., 2015. Course of neurological soft signs in first-episode schizophrenia:

Relationship with negative symptoms and cognitive performances. Sci Rep. 5,

11053.

Chan,R.C. and Gottesman,I.I., 2008. Neurological soft signs as candidate endophenotypes

for schizophrenia: a shooting star or a Northern star? Neurosci Biobehav Rev. 32,

957-971.

Chan,R.C., Xu,T., Heinrichs,R.W., Yu,Y. and Wang,Y., 2010. Neurological soft signs in

schizophrenia: a meta-analysis. Schizophr Bull. 36, 1089-1104.

Chan,R.C., Xie,W., Geng,F.L., Wang,Y., Lui,S.S., Wang,C.Y., Yu,X., Cheung,E.F. and

Rosenthal,R., 2016. Clinical Utility and Lifespan Profiling of Neurological Soft Signs

in Schizophrenia Spectrum Disorders. Schizophr Bull. 42, 560-570.

Chen,E.Y., Kwok,C.L., Au,J.W., Chen,R.Y. and Lau,B.S., 2000. Progressive deterioration of

soft neurological signs in chronic schizophrenic patients. Acta Psychiatr Scand. 102,

342-349.

Chen,E.Y., Hui,C.L., Chan,R.C., Dunn,E.L., Miao,M.Y., Yeung,W.S., Wong,C.K., Chan,W.F.

and Tang,W.N., 2005. A 3-year prospective study of neurological soft signs in first-

episode schizophrenia., 2005. Schizophr Res. 75, 45-54.

Chiliza,B., Ojagbemi,A., Esan,O., Asmal,L., Oosthuizen,P., Kidd,M., Gureje,O. and

Emsley,R., 2016. Combining depot antipsychotic with an assertive monitoring

programme for treating first-episode schizophrenia in a resource-constrained setting.

Early Interv Psychiatry. 10, 54-62.

17

Chouinard,G. and Margolese,H.C. Manual for the extrapyramidal symptom rating scale

(ESRS). Schizophr Res 76, 247-265. 2005.

Coldham,E.L., Addington,J. and Addington,D., 2002. Medication adherence of individuals

with a first episode of psychosis. Acta Psychiatr Scand. 106, 286-290.

Cuesta,M.J., Campos,M.S., Garcia-Jalon,E., Sanchez-Torres,A.M. and Peralta,V., 2012.

Treatment response of neurological soft signs in drug-naive patients with a first

psychotic episode. Schizophr Res. 139, 144-150.

Dazzan,P., Lloyd,T., Morgan,K.D., Zanelli,J., Morgan,C., Orr,K., Hutchinson,G., Fearon,P.,

Allin,M., Rifkin,L., McGuire,P.K., Doody,G.A., Holloway,J., Leff,J., Harrison,G.,

Jones,P.B. and Murray,R.M., 2008. Neurological abnormalities and cognitive ability in

first-episode psychosis. Br J Psychiatry. 193, 197-202.

Dazzan,P. and Murray,R.M., 2002. Neurological soft signs in first-episode psychosis: a

systematic review. Br J Psychiatry Suppl. 43, s50-s57.

Emsley,R., Rabinowitz,J. and Torreman,M., 2003. The factor structure for the Positive and

Negative Syndrome Scale (PANSS) in recent-onset psychosis. Schizophr Res. 61,

47-57.

Emsley,R., Turner,H.J., Oosthuizen,P.P. and Carr,J., 2005. Neurological abnormalities in

first-episode schizophrenia: temporal stability and clinical and outcome correlates.

Schizophr Res. 75, 35-44.

First,M.B., Spitzer,R.L.G.M. and Williams,L.B.W. Structured Clinical Interview for DSM-IV

Axis I Disorders, Patient Edition (SCID-P). 2nd. 1994. New York, New York State

Psychiatric Institute, Biometrics Research.

Gattaz WF, Diehl A, Geuppert MS, Hubrich P, Schmitt A, Linde I, et al (2004): Olanzapine

versus flupenthixol in the treatment of inpatients with schizophrenia: a randomized

double-blind trial. Pharmacopsychiatry 37:279-285.

18

Gupta,S., Andreasen,N.C., Arndt,S., Flaum,M., Schultz,S.K., Hubbard,W.C. and Smith,M.,

1995. Neurological soft signs in neuroleptic-naive and neuroleptic-treated

schizophrenic patients and in normal comparison subjects. Am J Psychiatry. 152,

191-196.

Jahn,T., Hubmann,W., Karr,M., Mohr,F., Schlenker,R., Heidenreich,T., Cohen,R. and

Schroder,J., 2006. Motoric neurological soft signs and psychopathological symptoms

in schizophrenic psychoses. Psychiatry Res. 142, 191-199.

Kahn RS, Fleischhacker WW, Boter H, et al. Effectiveness of antipsychotic drugs in first-

episode schizophrenia and schizophreniform disorder: an open randomised clinical

trial. Lancet 2008;371: 1085-1097

Karim,J. and Weisz,R. Cross-Cultural Research on the Reliability and Validity of the Mayer-

Salovey-Caruso Emotional Intelligence Test (MSCEIT). Cross-Cultural Research

44[4], 374-404. 2010.

Kay,S.R., Fizbein,A. and Opler,L.A. The Positive and Negative Syndrome Scale (PANSS) for

Schizophrenia. Schizophr Bull 13, 261-267. 1987.

Leucht S, Corves C, Arbter D, Engel RR, Li C, Davis JM (2009): Second-generation versus

first-generation antipsychotic drugs for schizophrenia: a meta-analysis. Lancet

373:31-41.

Madsen,A.L., Vorstrup,S., Rubin,P., Larsen,J.K. and Hemmingsen,R., 1999. Neurological

abnormalities in schizophrenic patients: a prospective follow-up study 5 years after

first admission. Acta Psychiatr Scand. 100, 119-125.

Mayoral,M., Bombin,I., Zabala,A., Robles,O., Moreno,D., Parellada,M., Ruiz-Sancho,A. and

Arango,C., 2008. Neurological soft signs in adolescents with first episode psychosis:

two-year followup. Psychiatry Res. 161, 344-348.

Mayoral,M., Bombin,I., Castro-Fornieles,J., Gonzalez-Pinto,A., Otero,S., Parellada,M.,

Moreno,D., Baeza,I., Graell,M., Rapado,M. and Arango,C., 2012. Longitudinal study

19

of neurological soft signs in first-episode early-onset psychosis. J Child Psychol

Psychiatry. 53, 323-331.

Mellacqua,Z., Eyeson,J., Orr,K.D., Morgan,K.D., Zanelli,J., Lloyd,T., Morgan,C., Fearon,P.,

Hutchinson,G., Doody,G.A., Chan,R.C., Harrison,G., Jones,P.B., Murray,R.M.,

Reichenberg,A. and Dazzan,P., 2012. Differential relationship between neurological

and cognitive dysfunction in first episode psychosis patients and in healthy

individuals. Schizophr Res. 142, 159-164.

Mittal,V.A., 2016. Cross-Cutting Advancements Usher in a New Era for Motor Research in

Psychosis. Schizophr Bull. 42, 1322-1325.

Mohr,F., Hubmann,W., Albus,M., Franz,U., Hecht,S., Scherer,J., Binder,J. and Sobizack,N.,

2003. Neurological soft signs and neuropsychological performance in patients with

first episode schizophrenia. Psychiatry Res. 121, 21-30.

Nuechterlein,K.H. and Green,M.F. MATRICS Consensus Battery Manual. 2006. Los

Angeles, MATRICS Assessment Inc.

Ojagbemi,A., Esan,O., Emsley,R. and Gureje,O., 2015. Motor sequencing abnormalities are

the trait marking neurological soft signs of schizophrenia. Neurosci Lett. 600, 226-

231.

Peralta,V., de Jalon,E.G., Campos,M.S., Basterra,V., Sanchez-Torres,A. and Cuesta,M.J.,

2011. Risk factors, pre-morbid functioning and episode correlates of neurological soft

signs in drug-naive patients with schizophrenia-spectrum disorders. Psychol Med. 41,

1279-1289.

Prikryl,R., Ceskova,E., Kasparek,T. and Kucerova,H., 2007. Neurological soft signs and their

relationship to 1-year outcome in first-episode schizophrenia. Eur Psychiatry. 22,

499-504.

Prikryl,R., Ceskova,E., Tronerova,S., Kasparek,T., Kucerova,H.P., Ustohal,L., Venclikova,S.

and Vrzalova,M., 2012. Dynamics of neurological soft signs and its relationship to

20

clinical course in patients with first-episode schizophrenia. Psychiatry Res. 200, 67-

72.

Sanders,R.D., Keshavan,M.S., Forman,S.D., Pieri,J.N., McLaughlin,N., Allen,D.N., van

Kammen,D.P. and Goldstein,G., 2000. Factor structure of neurologic examination

abnormalities in unmedicated schizophrenia. Psychiatry Res. 95, 237-243.

Schroder,J., Silvestri,S., Bubeck,B., Karr,M., Demisch,S., Scherrer,S., Geider,F.J. and

Sauer,H., 1998. D2 dopamine receptor up-regulation, treatment response,

neurological soft signs, and extrapyramidal side effects in schizophrenia: a follow-up

study with 123I-iodobenzamide single photon emission computed tomography in the

drug-naive state and after neuroleptic treatment. Biol Psychiatry. 43, 660-665.

Smith,R.C., Hussain,M.I., Chowdhury,S.A. and Stearns,A., 1999. Stability of neurological

soft signs in chronically hospitalized schizophrenic patients. J Neuropsychiatry Clin

Neurosci. 11, 91-96.

Whitty,P., Clarke,M., Browne,S., McTigue,O., Kamali,M., Feeney,L., Lane,A., Kinsella,A.,

Waddington,J.L., Larkin,C. and O'Callaghan,E., 2003. Prospective evaluation of

neurological soft signs in first-episode schizophrenia in relation to psychopathology:

state versus trait phenomena. Psychol Med. 33, 1479-1484.

21

Figure legend:

Figure 1. Mixed model repeated measures for the NES total and subscale scores over 12 months.

22

Table 1. Characteristics of the 126 participants with first-episode schizophrenia

23

Gender, N (%)

Male 93 (74%)

Female 33 (26%)

Age, Years, Mean (SD) 24.1 (6.6)

Highest level of education, N (%)

Elementary school 9 (7%)

Secondary school 76 (60%)

Completed secondary school 25 (20%)

Tertiary education 14 (11%)

Technical 2 (2%)

DSM Diagnosis, N (%)

Schizophrenia 84 (66%)

Schizophreniform 40 (32%)

Schizoaffective 2 (2%)

Race, N (%)

Mixed ancestry 98 (78%)

Black 18 (14%)

White 10 (8%)

Substance abuse, N (%) 59 (47%)

Previous antipsychotic exposure, N (%) 56 (45%)

Duration of exposure, Days, Mean (SD) 11 (6.8)

Completed the study, N (%) 84 (67%)

Achieved remission N (%) 61 (48%)

Table 2. Baseline and endpoint clinical and cognitive scores for the 126 patients

Baseline score Endpoint score

Mean SD Mean SD t P

NES:

Total 15.02 7.99 10.72 6.71 4.6 <0.0001

Sensory Integration 2.79 2.52 1.62 2.13 4.0 <0.0001

Motor Coordination 1.41 1.60 0.93 1.36 2.6 <0.01

Motor sequencing 2.94 2.47 1.90 2.18 3.5 <0.001

PANSS:

Total 94.78 16.48 52.85 17.92

19.3 <0.0001

Negative 17.57 4.80 10.68 4.56 11.7 <0.0001

Positive 13.59 2.76 5.47 3.36 21.0 <0.0001

Excitement hostility 8.57 3.83 5.24 2.33 8.3 <0.0001

Disorganised 18.26 4.46 10.69 4.40 13.6 <0.0001

CDSS Total 3.37 4.07 1.26 2.74 4.8 <0.0001

CGI SOFAS 44.04 11.55 60.32 13.27

-10.4 <0.0001

MCCB:*

Speed of processing 17.25 16.12 25.28 15.88

-3.6 <0.001

Attention and vigilance 25.61 11.72 33.62 11.27

-4.9 <0.0001

Working memory 23.29 15.47 30.95 13.85

-3.7 <0.001

Verbal learning 34.36 8.41 36.93 7.69 -2.3 0.02

Visual learning 29.07 15.40 35.92 13.65

-3.4 <0.001

Reasoning and problem solving 31.87 8.83 36.90 10.25

-3.8 <0.001

Social cognition 28.92 11.41 32.17 12.48

-1.9 0.06

Composite score 12.89 15.38 21.61 15.19

-3.9 <0.0001

ESRS:**

Parkinsonism 1.76 3.74 4.51 6.67 -6.34 <0.0001

Dystonia 0 0 0.11 0.88 -1.4 0.16

Dyskinesia 0.04 0.27 0.22 0.72 -2.99 0.003

24

Total 2.69 5.52 7.34 9.30 -8.11 <0.0001

NES, Neurological Evaluation Scale; PANSS, Positive and Negative Syndrome Scale;

CDSS, Calgary Depression Scale for Schizophrenia; SOFAS, Social and Occupational

Functioning Assessment Scale; MCCB, MATRICS Consensus Cognitive Battery

* For the cognitive assessments there were 100 participants with complete baseline data

** For ESRS the change scores are from baseline to the highest score ever achieved

25

Figure 1. Mixed model repeated measures for the NES total and subscale scores over 12 months.

Vertical bars denote 95% confidence intervals

26

Supplementary Table 1. Pearson correlation coefficients for the change scores† between

NES total and subscales and clinical and cognitive measures.

Sensory

integration

Motor co-

ordination

Motor

sequencing

NES Total

PANSS:

Positive factor 0.15 0.12 0.22* 0.24*

Negative factor 0.09 0.11 0.24* 0.23*

Disorganised factor 0.18* 0.22* 0.32*** 0.40*** Total 0.12 0.15 0.20* 0.26*

CDSS Total -0.02 0.04 -0.01 -0.00

SOFAS -0.18* -0.17 -0.39*** -0.35***

MCCB:

Speed of processing -0.23* -0.16 -0.22* -0.24*

Attention and vigilance -0.14 -0.18* -0.28* -0.23*

Working memory -0.35*** -0.22* -0.41*** -0.41*** Verbal learning -0.18* -0.25* -0.18* -0.18*

Visual learning -0.18* -0.20* -0.12 -0.24*

Reasoning and problem

solving-0.11 0.05 -0.22* -0.20*

Social cognition 0.03 -0.11 0.03 0.08

Composite score -0.32*** -0.27* -0.36*** -0.39***

ESRS:

Parkinsonism -0.03 -0.03 -0.05 -0.01

Dystonia -0.07 -0.05 -0.14 -0.09

Dyskinesia -0.18 -0.07 0.08 -0.05

Total score -0.06 -0.07 -0.06 -0.05

NES, Neurological Evaluation Scale; PANSS, Positive and Negative Syndrome Scale;

CDSS, Calgary Depression Scale for Schizophrenia; SOFAS, Social and Occupational

Functioning Assessment Scale; MCCB, MATRICS Consensus Cognitive Battery

27

† For NES, PANSS, CDSS, SOFAS and MCCB change scores are from baseline to endpoint;

for ESRS change scores are from baseline to the highest score ever achieved

*p<0.05; **p<0.001; ***p<0.0001

Supplementary Table 2. Pearson correlation coefficients for the endpoint scores between NES total and subscales and clinical and cognitive measures, extrapyramidal symptom scores and antipsychotic dose.

Sensory integration

Motor co-ordination

Motor sequencing NES Total

PANSS:

Positive factor 0,26* 0,11 0,31** 0,30**

Negative factor 0,22* 0,27* 0,37*** 0,40***

Disorganised factor 0,42*** 0,32*** 0,53*** 0,58***

Total 0,34*** 0,27* 0,45*** 0,49***

CDSS TOTAL 0,05 0,08 0,01 0,04

SOFAS -0,37*** -0,20* -0,45*** -0,45***

MCCB:

Speed of processing -0,48*** -0,35*** -0,55*** -0,55***

Attention and vigilance -0,39*** -0,23* -0,45*** -0,45***

Working memory -0,46*** -0,29* -0,51*** -0,48***

Verbal learning -0,40*** -0,31* -0,43*** -0,46***

Visual learning -0,42*** -0,36*** -0,43*** -0,50***

Reasoning and problem solving -0,38*** -0,08 -0,36*** -0,31*

Social cognition -0,07 -0,19 -0,19 -0,15

Composite score -0,49*** -0,34** -0,55*** -0,55***

ESRS:

Parkinsonism 0,25* 0,48*** 0,45*** 0,49***

Dystonia -0,08 -0,07 -0,09 -0,12

Dyskinesia -0,05 0,01 -0,01 -0,04

28

Total score 0,23* 0,45*** 0,39*** 0,44***

Antipsychotic dose -0,13 -0,02 -0,16 -0,13

NES, Neurological Evaluation Scale; PANSS, Positive and Negative Syndrome Scale;

CDSS, Calgary Depression Scale for Schizophrenia; SOFAS, Social and Occupational

Functioning Assessment Scale; MCCB, MATRICS Consensus Cognitive Battery; ESRS,

Extrapyramidal Symptom Rating Scale

*p<0.05; **p<0.001; ***p<0.0001

29