Relationship of psychopathological symptoms and cognitive function to subjective quality of life in patients with chronic schizophrenia

Abstract

Aims:  The purpose of the present study was to examine the extent of the effects of psychopathological symptoms and cognitive function on quality of life (QOL) in patients with chronic schizophrenia.

Methods:  Data were obtained using the Japanese Schizophrenia Quality of Life Scale (JSQLS), Positive and Negative Syndrome Scale (PANSS), Wisconsin Card-Sorting Test (WCST) Keio version, and Continuous Performance Test (CPT) for 52 schizophrenia patients.

Results:  Stepwise regression analysis showed that PANSS depression/anxiety factors predicted JSQLS psychosocial conditions and motivation/energy, and that WCST Categories Achieved predicted JSQLS symptoms/side-effects.

Conclusions:  Psychopathological symptoms and cognitive function affect subjective QOL in patients with schizophrenia. If the final goal is treatment that improves QOL in a manner that patients themselves are aware of, clinicians probably need to consider a treatment strategy that improves depression/anxiety symptom.

IN ADDITION TO positive and negative symptoms, patients with schizophrenia have reduced cognitive function and are consequently impaired in everyday social functioning. In the past, the first goal of schizophrenia treatment was to reduce psychological symptoms, mainly positive symptoms,¹ rather than recovering social functioning. Recently, as a result of an emphasis on patient needs, the concept of quality of life (QOL) has been brought into the treatment of somatic illness, particularly chronic illness such as chronic heart failure.² The goal of treatment has therefore changed from the alleviation of symptoms to improvement of the patient's own satisfaction with social activities. Because of this trend, attempts to evaluate the effects of treatment using QOL as an indicator have occurred in the field of clinical psychiatry, including treatments and rehabilitation for schizophrenia.

Essentially, the basic concept of QOL places importance on subjectivity in terms of patients' self-appraisal of their own satisfaction. Self-evaluations by people with schizophrenia were previously thought to lack reliability because of the presence of psychopathological symptoms and poor awareness of the disease.³ Hence many trials have used objective QOL evaluations, such as the Quality of Life Scale (QLS),⁴ which rely on interviews with psychiatrists or other trained interviewers. The importance of evaluating the satisfaction of patients themselves, however, has been recognized in schizophrenia. Reporting that patients with schizophrenia were aware of and could express their social dysfunction, Skantze et al. supported the view that QOL could be ascertained only on subjective evaluation.⁵ Lehman demonstrated that QOL data from patients with chronic mental illness were reliable and concluded that subjective QOL evaluation was applicable to such patients.^6,7 QOL is considered important in research on treatment outcome for schizophrenia, and researchers have argued strongly for development of a robust QOL scale specific to schizophrenia, based on the subjective judgment of patients.⁸

The Schizophrenia Quality Life of Scale (SQLS), which is a practical and simple self-administered evaluation, was developed for the purpose of measuring patient-specific QOL in patients with schizophrenia. It is primarily intended for use in clinical trials and has been reported to have high levels of reliability and validity.⁹ Kaneda et al. translated the SQLS into Japanese, and this version also yields high reliability (Japanese Schizophrenia Quality of Life Scale [JSQLS]).¹⁰ With the spread of QOL evaluations for patients with schizophrenia, there has been active research concerning factors related to QOL, which represents the degree to which patients are satisfied with their lives. First of all, in research examining the relationship between psychopathological symptoms and QOL, it has been repeatedly reported that symptoms such as depression and anxiety have a strong effect on subjective QOL,^11–13 but no consistent view on the relationship between QOL and positive symptoms, or that between QOL and negative symptoms has been obtained.^14–17 In addition, QOL evaluation measures used in those studies have been a mixture of subjective and objective ones.

Specific cognitive functions are significantly impaired in patients with schizophrenia when compared to healthy persons.^4,18 Green analyzed the influence of cognitive deficits on the daily lives of patients with schizophrenia, and reported that vigilance (sustained attention) was associated with social skill and that executive functioning was related to community functioning.¹⁹ In the field of schizophrenia research, Heinrichs reported that the Continuous Performance Test (CPT) for sustained attention and Wisconsin Card-Sorting Test (WCST) for executive functioning were powerful and reliable tool, respectively.²⁰ Relationships between executive functioning and QOL could not be confirmed.^21–23 In addition, only Wegener et al. have reported a significant relationship between sustained attention and QOL.²⁴

A few studies have examined both aspects of the relationship between psychopathological symptoms and QOL and that between cognitive function and QOL. These studies reported that psychopathological symptoms, particularly negative symptoms,^25,26 have a stronger effect than cognitive function on QOL.²⁷ In contrast, one report showed that cognitive function and psychopathological symptoms affect each other.²⁴ Because studies examining the relationship of both psychopathological symptoms and cognitive function to subjective QOL are scarce, and different aspects of cognitive function are measured in each study, a consistent view has not been obtained.

In light of these reports, we verified the relationship between (i) subjective QOL, as measured by the JSQLS, and psychopathological symptoms, as measured by the Positive and Negative Syndrome Scale (PANSS); and (ii) subjective QOL and cognitive function, as measured by the CPT (sustained attention) and the WCST (executive functioning). The ultimate aim of the present study was to identify an objective predictor for treatment that is compatible with the needs of patients and reflects patient satisfaction.

METHODS

Subjects

Subjects were inpatients or outpatients diagnosed with schizophrenia according to DSM-IV.²⁸ They provided written consent to participate in this research. Diagnosis was performed using the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID). Patients fulfilling all of the following three criteria were enrolled in the present study: (i) presence of chronic illness without acute exacerbation; (ii) PANSS total score >50 points; and (iii) absence of other axis I disorders, including major depressive episodes or anxiety disorders. Demographic data, including age, sex, disease subtype, living situation (outpatients/ inpatients), onset age, duration of disorder, number of hospital admissions for schizophrenia, and dose of antipsychotics, were obtained from medical records. A total of 52 patients were enrolled. Table 1 lists the subjects' demographic characteristics. The most common schizophrenia subtype was paranoid type (62%). With regard to the administration of antipsychotic drugs, an atypical antipsychotic drug was prescribed to 41 patients (79%), and more than two kinds of atypical or typical antipsychotic drugs were prescribed to other patients (21%). Average antipsychotic drug dose was 11.2 mg haloperidol-equivalent dose;²⁹ this was a low–average dose compared with other studies.^24,30 With regard to other psychopharmaceutics, nine patients (17%) received a mood stabilizer and none received antidepressant. The present study was approved by the Ethics Committee of the Nagoya University School of Medicine.

Table 1.  Patient characteristics (n = 52; mean ± SD)

Age (years)	37.7 ± 11.6
^† Haloperidol equivalent.
Sex
Male	29
Female	23
Schizophrenia subtypes
Paranoid type	32
Disorganized type	6
Residual type	13
Undifferentiated type	1
Outpatients/inpatients	30/22
Onset age (years)	24.9 ± 8.5
Duration of disorder (years)	12.8 ± 10.5
No. admissions	2.1 ± 3.5
Dose of antipsychotics^† (mg)	11.2 ± 9.4

Evaluation

Evaluation of psychopathological symptoms

Evaluation of psychopathological symptoms used the PANSS.³¹ The PANSS was administered by trained psychiatrists or psychologists. According to the Lindenmayer et al. model, this is classified into the following five areas: (i) negative factors; (ii) excitement factors; (iii) positive factors; (iv) cognitive factors; and (v) depression/anxiety factors,^32,33 and mean scores for each area were calculated.

Subjective QOL evaluation

For subjective QOL evaluation, we used the JSQLS developed by Wilkinson et al.⁹ and translated by Kaneda et al.¹⁰ As proposed by Wilkinson et al. the 30 items on the JSQLS were classified into the following three areas: (i) psychosocial conditions; (ii) motivation/energy; and (iii) symptoms/side-effects. Each area scale is transformed to have a range from 0 (the best status as measured on the JSQLS) to 100 (the worst status as measured on the JSQLS), with each scale calculated as follows: the scale score (SS) equals the total of raw scores of each item in the scale (RS_tot), divided by the maximum possible raw scores of all the items in the scale (RS_max), all multiplied by 100: SS = (RS_tot/RS_max) × 100. The ‘psychosocial conditions’ area addresses various emotional conditions such as loneliness, hopelessness, difficulty in social situations, and worries about the future. The ‘motivation/energy’ area addresses various problems of motivation and activity, such as the lack of will or drive to do things. The ‘symptoms/side-effects’ area addresses issues such as muscle twitches and dry mouth, which can be caused by medication.

The JSQLS was rated within 2 weeks of the evaluation of psychopathological symptoms.

Examination of cognitive function

Executive functioning was evaluated using the WCST (Keio version),³⁴ the computerized version of which was developed by Kobayashi.³⁵ The patient classifies a single card shown at the bottom of a computer screen in terms of color, shape and number and selects one type of card from four basic types of cards shown at the top. Without letting the patient know the correct category, the computer gives feedback as to whether it was a correct or incorrect selection. If the patient makes six continuous correct selections, the categories in the computer are changed, and the patient must select another category to make a correct selection. This test is carried out for up to 48 selections. In the present study Categories Achieved (CA) and Perseverative Errors of Nelson (PEN) were calculated.³⁴

Sustained attention was evaluated using the CPT–Identical Pairs.³⁶ A four-digit number is displayed on a computer screen as a single stimulus, and the patient must click the mouse as quickly as possible while exactly the same stimulus continues. One stimulus is shown for 50 ms, and the interval between stimuli is 950 ms. There are a total of 150 trials, 30 of which involve the target. In the present study d', which is a discrimination index calculated from the number of correct and incorrect answers, was measured.³⁶

All tests were administered by experienced examiners within 2 weeks of the evaluation of psychopathological symptoms.

Statistical analysis

In order to study the relationship between subjective QOL and clinical variables (age, living situation, duration of disorder, number of hospital admissions for schizophrenia, type of antipsychotics (one type of atypical antipsychotics or more than two types of atypical or typical antipsychotics), dose of antipsychotics, scores on each of the five PANSS areas, CA and PEN on WCST, and d' on CPT), Spearman rank correlation coefficients were calculated. Because the range of each PANSS subscore was narrow and the SD was small, we used non-parametric analysis.

In order to examine the extent of the effect of clinical variables on subjective QOL, multiple regression analysis using a stepwise forward selection method was performed. Clinical variables that were statistically significant or nearly significant (P < 0.1) were regarded as independent variables, and scores on each of the three JSQLS areas were considered dependent variables.

Kruskal–Wallis H-test was used to analyze psychopathological characteristics of samples, and a post-hoc analysis was performed using the Mann–Whitney U-test with Bonferroni correction.

SPSS version 10.0 (SPSS, Chicago, IL, USA) was used for the analysis, and the level of significance was set at 5%.

RESULTS

Results of subjective QOL evaluation, psychopathological symptom evaluations, and cognitive function examination are given in Table 2. According to the Lindenmayer et al. five-factor model^32,33, the score for the excitement factors was significantly lower than the scores for other factors in the present participants.

Table 2.  QOL, psychopathological symptoms and cognitive functioning

		Mean	SD
CA, Categories Achieved; CPT, Continuous Performance Test; JSQLS, Japanese Schizophrenia Quality Life of Scale; PANSS, Positive and Negative Syndrome Scale; PEN, Perseverative Errors of Nelson; QOL, quality of life; WCST, Wisconsin Card-Sorting Test.
JSQLS	Psychosocial	48.7	19.8
	Motivation/energy	48.8	17.1
	Symptoms/side effect	34.0	18.6
PANSS	Total score	83.2	17.6
	Negative	2.9	1.1
	Excitement	2.1	0.8
	Cognitive	2.7	0.8
	Positive	3.1	0.8
	Depression/anxiety	2.8	1.0
WCST	CA	4.4	1.4
	PEN	3.5	3.8
CPT	d′	1.4	0.8

The correlation matrix of the scores for each of the three JSQLS areas and clinical variables is given in Table 3. To determine the extent of the effects of clinical variables on the three JSQLS areas, multiple regression analysis was performed using the stepwise forward selection method. As a result, the models had a good fit with the data (psychosocial conditions area, F = 10.548, P < 0.001; motivation/energy area, F = 9.285, P < 0.01; symptoms/side-effects area, F = 4.239, P < 0.05). Excluded variables are not reported herein. The psychosocial conditions area of the JSQLS was predicted independently on the basis of the duration of disorder and the PANSS depression/anxiety factors. The motivation/energy area of the JSQLS was predicted by the PANSS depression/anxiety factors. The symptoms/side-effects area of the JSQLS was predicted by the WCST CA (Table 4).

Table 3.  JSQLS scores and clinical variables

	JSQLS
	Psychosocial	Motivation/energy	Symptoms/side-effect
* P < 0.10; ** P < 0.05; *** P < 0.01, Spearman correlations ^† Outpatients = 0, Inpatient = 1. ^‡ One type of atypical antipsychotic = 0; more than two kinds of atypical or typical antipsychotics = 1. CA, Categories Achieved; CPT, Continuous Performance Test; JSQLS, Japanese Schizophrenia Quality Life of Scale; PANSS, Positive and Negative Syndrome Scale; PEN, Perseverative Errors of Nelson; WCST, Wisconsin Card-Sorting Test.
PANSS
Negative	0.101	0.315**	0.145
Positive	−0.153	−0.034	0.035
Cognitive	0.031	0.032	0.090
Excitement	0.111	0.162	0.068
Depression/anxiety	0.407***	0.391***	0.088
WCST
CA	0.120	−0.123	0.268*
PEN	−0.122	0.120	−0.279**
CPT
d′	0.060	0.138	0.079
Age	−0.348**	−0.069	−0.071
Living situation^†	−0.286**	−0.195	−0.056
Duration of disorder	−0.334**	−0.024	−0.201
No. admissions	−0.164	−0.360***	−0.016
Type of antipsychotics^‡	−0.049	0.183	0.071
Dose of antipsychotics	−0.100	−0.078	0.163

Table 4.  Multiple regression of psychopathological symptoms and cognitive functioning

Outcome variable: JSQLS	Predictor	Adjusted R^2	β
* P < 0.05, ** P < 0.01. CA, Categories Achieved; JSQLS, Japanese Schizophrenia Quality Life of Scale; PANSS, Positive and Negative Syndrome Scale; WCST, Wisconsin Card-Sorting Test.
Psychosocial	PANSS: Depression/anxiety	0.272	0.390**
	Duration of disorder		−0.391**
Motivation/energy	PANSS: Depression/anxiety	0.140	0.396**
Symptoms/side effect	WCST: CA	0.060	0.280*

DISCUSSION

In the present study, depression/anxiety factors, age, living situation, and duration of disorder correlated with the score for the psychosocial conditions area. Stepwise regression analysis indicated that the psychosocial condition worsens with an aggravation of the depression/anxiety factors and improves with an increase in the duration of disorder. Negative factors, depression/anxiety factors, and number of admissions for schizophrenia correlated with the scores for motivation/energy area. Stepwise regression analysis also indicated that with an increase in the depression/anxiety factors, the scores for motivation/energy area deteriorate.

With regard to psychopathological symptoms, some areas of subjective QOL were not influenced by positive factors or negative factors but were significantly affected by depression/anxiety factors. This finding supports those of other reports on the relationship between depression/anxiety symptoms and subjective QOL.^11–13 Because the psychosocial condition area of JSQLS addresses various emotional problems, patients with schizophrenia appear to be able to validly express their emotions. The motivation/energy area of JSQLS addresses various problems of activity rather than emotion, and such issues might be associated with negative factors, but depression/anxiety factors rather than negative factors affect this area. It is suggested that the better that emotional problems are controlled, the more energy/motivation patients with schizophrenia feel, even if their activity levels are actually poor. Several studies have reported that objective QOL, which is evaluated with QLS, has a close relationship with negative symptoms.^25,37 The fact that QLS was developed for measuring defect symptoms in schizophrenia might explain this relationship with negative symptoms. Subjective QOL, however, is not determined by a therapist's evaluations but by how the patient with schizophrenia feels.

Until recently, depression and anxiety have tended not to be seen as important as treatment targets for patients with schizophrenia. Ginsberg et al., however, reported that 50% of patients with schizophrenia suffered from depression, and that this was a major risk factor for suicide.³⁸ Given that subjective QOL correlated with depression/anxiety factors rather than other factors of the PANSS, which is an objective symptom evaluation, we may have to target improvements in depression/anxiety factors in order to improve subjective QOL. Consequently, it is possible that the patients feel the effects of treatments, leading to improvements in adherence. Taniguchi et al. reported that replacement of antipsychotic drugs with quetiapine improved clinical symptoms, including depression/anxiety, and the psychosocial conditions score on JSQLS.³⁹ Treatment plans focusing on the improvement of depression/anxiety will lead to patients feeling the effects of treatment and, consequently, to increased adherence to treatment. In the future, trends in areas such as objective psychopathological symptoms and subjective QOL, as well as treatment adherence, must be examined before and after drug therapy or psychosocial treatments such as cognitive behavioral therapy. This could help identify treatments that are compatible with patient needs and could lead to increased adherence.

The present findings suggested that the longer the duration of disorder, the better the psychosocial condition. As the disorder progresses, patients with schizophrenia might become acclimated to their condition and may not be troubled by their emotional problems. Yamauchi et al., however, reported a non-significant correlation between the psychosocial conditions of the JSQLS and the duration of illness,⁴⁰ therefore further investigations are necessary to clarify this aspect.

With regard to the relationship between the cognitive function and subjective QOL, the correlation of WCST CA and PEN with the symptoms/side-effects area was evaluated. Stepwise regression analysis suggested that the worse the executive functioning, the better the score for the symptoms/side-effects area. Most of the items in the JSQLS symptoms/side-effects area concern side-effects of drug therapy. The lower the executive functioning, the more indifferent patients are to side-effects and, as a result, patients might rate their QOL higher. We did not assess the objective side-effects. Yamauchi et al. reported that objective side-effects predicted the symptoms/side-effects area of JSQLS.⁴⁰ It might be necessary to investigate the correlation between the objective QOL and executive functioning and how these factors predict subjective QOL. Matsui et al. reported that there was no significant relationship between executive functioning and subjective QOL using the abbreviated version of SQLS.²² Hofer et al. used the same cognitive function survey, and reported no relationship between executive functioning and subjective QOL.³⁰ The fact that these results are inconsistent with the present results might be explained by the fact that executive functioning in the Matsui et al. study was not measured using the WCST and that subjective QOL in the Hofer et al. study was measured with the World Health Organization Quality of Life Assessment–Short Form (WHOQOL-Bref),⁴¹ which is not a QOL scale specific to schizophrenia. Some insight measure might be useful to investigate in this area. Patients with schizophrenia exhibit significantly impaired sustained attention.^18,20 Cornblatt et al. reported that attentional deficits using CPT-IP resulted in a schizophrenia spectrum with a sensitivity of 67% and specificity of 79%,⁴² and that the mean d' in normal adults was 1.720 (SD = 0.778).³⁶ In the present study sustained attention in subjects would be lower than that in the normal population, and this did not affect subjective QOL. Prouteau et al. reported that poorer sustained attention predicted better subjective QOL,⁴³ and Wegener et al. reported that sustained attention had a negative effect on subjective QOL.²⁴ The inconsistency of these findings with the present findings might result from the fact that each study used different instruments to measure subjective QOL assessment and sustained attention. In the future, there is a need for methodology to be standardized in further investigations into the relationship between cognitive function and subjective QOL.

The present study had several limitations. First, the average total PANSS score for the subjects in the present study was 85.2 ± 19.3; thus, psychopathological symptoms were relatively mild. In particular, excitement symptoms had subsided. Moreover, the subjects were chronically ill patients who were not in acute exacerbation. Verification in severely ill and acute patients is insufficient, therefore it is difficult to assume that these results can be generalized to schizophrenia patients as a group. If possible, future investigations should examine subject groups that include the severely and acutely ill.

If we include improvement of subjective QOL as well as reduction of psychopathological symptoms in treatment goals for schizophrenia, the present findings indicate a need to develop treatments that focus on symptoms of depression/anxiety. Such treatments lead to patients really feeling the effects of treatments and can improve treatment adherence. In the future, longitudinal research is needed into how psychopathological symptoms and cognitive function affect subjective QOL.