Association of negative symptom domains and other clinical characteristics of schizophrenia on long-term hospitalization.

Background: Recent studies suggest five domains for negative schizophrenia symptoms: anhedonia, asociality, avolition, blunted affect, and alogia. Avolition has been considered a characteristic symptom in community-dwelling patients. However, few studies have explored the association of these symptoms with long-term hospitalization. Purpose: This study explored the relative association of each of the five domains of negative symptoms in two groups of patients with schizophrenia: long-term hospitalized and community-dwelling patients.
Methods: Participants included 56 long-term inpatients and 111 community-dwelling patients at Nasukougen Hospital in Japan. The nearest neighbor matching within caliper was used. After matching participants by age, sex, disease duration, and years of education, each group was assigned 30 participants. Model 1 was analyzed with a logistic regression analysis with 5 subdomains as independent variables. Model 2 was analyzed after adding positive symptoms, cognitive function, functional skills, and functional outcomes to the subdomains that were significant in model 1.
Results: The results indicated that asociality was significantly associated with long-term hospitalization. When the characteristic clinical factors of schizophrenia were added, asociality, daily living skills, and social and role functions were found to be characteristic of long-term hospitalization. Among the negative symptoms, lack of social motivation was more characteristic among the group with a poor prognosis. Conclusions: Of the negative symptoms associated with long-term hospitalization, asociality, lack of social motivation, rather than avolition was found to be most strongly associated with community-dwelling patients' functional outcomes. Further studies are required to establish a causal association as it may have therapeutic implications. Copyright:

INTRODUCTION

Schizophrenia is considered a chronic, relapsing disease process.[1] It has an extremely low recovery rate and occurs in at least 0.7% of the global population; it reportedly affected 23.6 million people worldwide in 2013.[1] Globally, severe social dysfunction after the onset of schizophrenia was ranked 11th among causes of disability in such patients afflicted with this disease. A recent systematic review showed that the life expectancy of patients with schizophrenia is 15–20 years shorter than that of the average general population[2] and that the mortality risk of patients with schizophrenia is twice that of the general population (risk ratio of 2.54).[3] Thus, there is an urgent clinical and societal need for improving outcomes of schizophrenia.

However, clinical improvements in outcomes have barely changed since the early 1950s when the first antipsychotic treatments for schizophrenia were introduced.[4] Positive symptoms of schizophrenia can be improved with antipsychotic treatment, but it rarely improves cognition, negative symptoms, and outcomes.[4] Therefore, the fact that one of the worst outcomes is long-term hospitalization remains unchanged.

Although several studies have been conducted over the past three decades to prevent and eliminate long-term hospitalization, it remains an issue, particularly in Asian countries.[5]

Research on long-term inpatients conducted over the years shows a considerable increase in positive symptoms and cognitive impairment.[6] Moreover, long-term hospitalized patients with schizophrenia exhibit hostile and uncooperative behavior, delusions of grandeur, dysfunction,[7] and a significant amount of aggressive and unpredictable behavior compared to discharged older patients with schizophrenia[8]; these findings have been observed in many countries.[910] Additionally, long-term hospitalized patients report greater cognitive impairment and negative symptoms than younger patients and community-dwelling patients.[61112] Negative symptoms of schizophrenia have garnered a lot of attention in recent years given the condition’s relationship with aggression, which prevents hospital discharge,[13] and as factors that worsen social function.[14] As described in a National Institute of Mental Health (NIMH) report,[15] negative symptoms can be divided into five domains: anhedonia (decreased ability to experience pleasure), asociality (decreased interest in communicating with others), avolition (lack of motivation to carry out goal-oriented behavior), blunted affect (inability to engage in non-verbal communication), and alogia (poverty of speech). A recent study also suggested that total symptom severity, deterioration of social skills, and asocial behavior were predictors of long-term hospitalization.[16] However, it should be noted that many of these studies did not consider the potentially different types and courses of chronically hospitalized patients or only included community-dwelling experimental groups.[1718]

Furthermore, even though negative symptoms are divided into two-[19] or five-factor solutions,[2021] previous studies of long-term hospitalization considered negative symptoms to be one-dimensional. Recently, it was suggested that studies should use a five-factor instead of a two-factor solution because of the function of negative symptoms and the quality of prognosis.[2021] Therefore, to solve this problem, Kirkpatrick et al.[22] developed the Brief Negative Symptom Scale (BNSS), which adequately assesses the five domains of negative symptoms of schizophrenia.

In a study dealing with community-dwelling patients, avolition was reported to have a significant impact on clinical outcomes after 10 years and a 30% distribution of functional outcome among domains.[2324] However, no studies have clarified which domain of negative symptoms prevents hospital discharge. Thus, the relative effects of the five domains of negative symptoms on long-term hospitalization should also be assessed.

A better understanding of the factors associated with long-term hospitalization may help health practitioners develop preventive interventions for patients dealing with long-term hospitalization. Thus, the present study aims to examine the differences in the negative symptom profile of people with schizophrenia who are in long-term hospital-based care and of those living in the community. To achieve this goal, two aspects were considered in this study. First, many previous studies on long-term hospitalization did not consider the different courses of schizophrenic patients, so we matched the demographic features of two groups of patients with schizophrenia: long-term hospitalized patients and community-dwelling patients. Second, for the evaluation of negative symptoms, which is the main purpose of this study, we attempted to characterize negative symptoms in long-term inpatients using the BNSS. Furthermore, based on prior research, we sought to explore and clarify the association of clinical factors, such as positive symptoms, cognitive function, functional skills, and social and role functioning, with long-term hospitalization.

MATERIALS AND METHODS

Participants

Participants of this study were inpatients (n = 56) and community-dwelling patients (n = 111) at Nasukougen Hospital in Japan. In this study, it was difficult to predict how many subjects would be analyzed after matching; therefore, we recruited as many participants as possible. Patients were selected on the basis of having an International Statistical Classification of Diseases and Related Health Problems – version 10 (ICD-10) diagnosis of schizophrenia or schizoaffective disorder. These diagnoses were made by the attending physician from normal medical examinations and medical records. Subjects were excluded from the study in case of history of neurological disorders, such as seizure disorder, stroke, head injury, brain surgery, mental retardation, or severe recurrent headaches. Long-term inpatients were those who had been hospitalized for more than one year. Community-dwelling participants were those who had not been hospitalized or readmitted to a psychiatric ward in the preceding year. The status of community-dwelling patients was defined as patients living outside of any healthcare setting, including nursing homes.

This study was approved by the Institutional Review Board of the International University of Health and Welfare and Nasukougen hospital. This study occurred in the part of the Nasukougen Hospital Negative Symptom Study. Therefore, we used the part of community-dwelling patients’ data set from the past study.[25] All participants provided written informed consent after a complete description of the study was provided.

Procedure

After assessing potential participants for eligibility, consenting participants underwent a series of structured clinical assessments and evaluations of symptoms and functioning. All symptoms were evaluated by the researchers and the attending physician, while functional skills and social and role functioning were evaluated by an attending nurse and a care worker.

Measures

Positive symptom severity was assessed using a subset of the Brief Psychiatric Rating Scale (BPRS), which is based on a recent factor analysis[26] and includes the following items: grandiosity, suspiciousness, hallucinations, unusual thought content, bizarre behavior, disorientation, and conceptual disorganization. Negative symptom severity was assessed using the BNSS. Neurocognitive function was measured using the Schizophrenia Cognition Rating Scale (SCoRS). Functional skills were evaluated using the Life Assessment Scale for the Mentally Ill (LASMI),[27] and the Social Functioning Scale (SFS) was used to measure social and role functioning. All these scales except the LASMI were translated into Japanese. The reliability and validity of these scales have been verified in previous studies.[28293031]

BPRS

The BPRS was developed by Overall and Gorham to evaluate a wide range of mental symptoms. We used the Japanese version of the BPRS translated by Miyata et al.[28] In this study, seven items related to positive symptoms were used.

BNSS

The BNSS was developed based on the NIMH consensus statement to identify negative symptoms accurately. The scale consists of six items: anhedonia, asociality, avolition, blunted affect, alogia, and distress. However, since distress is not included in the experience or the expression factor, it was excluded from this study. We used the Japanese version of the BNSS translated by Hashimoto et al.[31]

SCoRS

The SCoRS, which was developed based on the recommendations from the Measurement and Treatment Research to Improve Cognition of Schizophrenia (MATRICS) project, measures cognition. It consists of seven cognitive domains: vigilance, working memory, processing speed, language learning and memory, visual learning and memory, reasoning and problem solving, and social cognition. We used the Japanese version of the SCoRS, which was translated by Kaneda et al.[30]

LASMI

The LASMI was developed to assess disability in daily living and lack of functional skills.[26] It comprises of the following five categories: (i) daily living; (ii) interpersonal; (iii) work; (iv) endurance and stability; and (v) self-recognition. The items in each category are rated on a five-point scale (where, 0 = no problem; 4 = a serious problem). Lower scores indicate higher degrees of independent living in the community. In this study, we used categories like “daily living,” “interpersonal,” and “work” to evaluate skills.

SFS

The SFS is a self-reported measure of social and role functioning that is sensitive to functional impairment across the psychosis spectrum in the real world. SFS outcomes include a total score and seven subscale scores of social functioning (withdrawal, interpersonal behavior, and prosocial activities) and role functioning (recreation, independence competence, independence-performance, and occupation). We used the Japanese version of the SFS translated by Nemoto et al.[29] Because higher score on the SFS indicate higher outcomes, scores in Table 1 were reversed to maintain consistency in the presentation of results.

Table 1

Basic attributes of all patients before and after matching, scores on each evaluation item, and comparison results

Item	All patients		P	Propensity-Matched patients		P
	Long-term hospitalization group n=56	Group of community-dwelling n=111		Long-term hospitalization group n=30	Group of community-dwelling n=30
Age (years)	58.2±13.6	49.2±13.9	<0.001	52.9±12.3	53.8±13.7	0.805
Sex (male)	30 (55%)	46 (42.9%)	0.137※	14 (46%)	14 (46%)	1.000※
Disease duration (years)	35.7±14.7	28±14.4	<0.001	27.6±14.1	27.5±13.9	0.963
Years of education	10.6±1.4	12.7±1.8	<0.001	11.5±2.1	11.6±1.9	0.844
BPRS positive	8.9±6.1	7.3±5.2	0.094	8.6±6.9	6.6±5.5	0.195
BNSS
Anhedonia	7.4±3.7	6.1±4.4	0.058	7.2±3.8	6±4.3	0.199
Asociality	6.3±2.9	4.4±2.7	<0.001	6.5±1.9	4.3±2.4	<0.001
Avolition	5±2.7	3.6±2.4	<0.001	5±2.5	3.3±2.2	0.007
Blunted affect	6.2±4.5	5.1±4.1	0.135	6.5±4.4	4.4±3.4	0.046
Alogia	4.2±3.4	3.3±3.4	0.110	4.8±3.3	2.9±3.3	0.028
SCoRS-J	45.7±9	37.8±10.8	<0.001	45.9±9.5	37.9±8.5	0.001
LASMI
Daily living skill	23.7±8	14.4±11	<0.001	24.5±8.1	12.5±10.5	<0.001
Interpersonal skill	19.7±8.8	15.2±9.9	0.006	18.6±8.5	14.9±8.9	0.108
Work skill	22.6±7.1	17.6±10.5	0.002	20.2±8.1	16.1±9.7	0.065
SFS
Social functioning	16.5±6.2	12.9±4.1	<0.000	17±5.1	10.1±3.3	<0.001
Role functioning	33.6±11.6	28.6±12.9	0.032	39.6±9.3	28.6±11.6	<0.001

Representative value: Mean value, standard deviation test: t-test, ※: χ2 test. BPRS=Brief Psychiatric Rating Scale, BNSS=Brief Negative Symptom Scale, SCoRS=Schizophrenia Cognition Rating Scale, LASMI=Life Assessment Scale for the Mentally Ill, SFS=Social Functioning Scale

Statistical analysis

Both long-term hospitalized patients (n = 56) and community patients (n = 111) have demographic characteristics such as duration of illness and age that may influence their illness [Table 1]. Therefore, we tried to match these characteristics in both participant groups using propensity scores to account for the complex course of each patient’s illness and to analyze the influence of the independent variable with less bias. To estimate the propensity score, we fitted a logistic regression model for community-dwelling patients and inpatients as a function of patients’ demographic factors including age, sex, disease duration, and years of education. Then, we performed case–control for each group using nearest neighbor matching within a caliper. This caliper value was 0.2 times the standard deviation of the propensity score. Only nearest neighbor matching within calipers was performed. The patients with the closest propensity scores to the target patients were matched. Cases with dissimilar propensity scores were excluded. When there were two or more subjects, a random number was generated after calculating the propensity score, and matching was performed at random. Moreover, we used the C-statistic to confirm the discriminant ability of the estimated propensity score.

We then examined the differences between the items using a t-test, except gender, which we examined via the Χ2 test. Subsequently, in model 1, we performed a logistic regression analysis (stepwise method) with the BNSS domains (anhedonia, asociality, avolition, blunted affect, and alogia) as independent variables and the long-term hospitalization group (coded as 1) and the group of residents (coded as 0) as the dependent variable.

In model 2, we performed another logistic regression analysis (stepwise method) after adding BPRS, SCoRS, LASMI (daily living, interpersonal relations, and work), and SFS (social and role functioning) items to the significant BNSS domains from the regression analysis in model 1. As this was an exploratory study, no statistical correction was applied for multiple hypotheses testing. All analyses were performed using Statistical Package for the Social Sciences (SPSS) 24.0 for Windows.

RESULTS

Each group had a total of 30 participants after calculating the propensity scores and matching participants by their clinical and demographic characteristics. Long-term hospitalized patients (n = 26) and community-dwelling patients (n = 81) were excluded from the analysis by nearest neighbor matching. Figure 1 shows the propensity scores before and after matching and demonstrates that the covariates match. The C-statistic at the time of matching showed a good value of 0.756. Table 1 shows the comparison of the basic attributes for both groups and the results from each evaluation scale based on the t-test and Χ2 test. After matching, there was no significant difference in the basic attribute items.

Figure 1

Propensity score comparison before and after matching

Table 2 shows the results of the logistic regression analysis (stepwise method) with the five domains of the BNSS. Asociality had a significant contribution, while anhedonia, avolition, blunted affect, and alogia were not significant (P > 0.05). Compared to community-dwelling patients, hospitalized patients were found to have greater asociality within the five negative domains. The Hosmer–Lemeshow test showed a good value of Χ2 = 4.995 and a P = 0.661. The variance values were as follows: Cox-Snell R2 = 0.318, Nagelkerke R2 = 0.424.

Table 2

Final version of the stepwise multiple regression model for the five negative symptom domains

Model 1 Variable	B	P	Exp (B)	95% CI interval EXP (B)
				Lower	Upper
Asociality	0.602	0.000	1.826	1.349	2.573

−2 log likelihood=59.199. Anhedonia score=−1.995 P=0.143. Avolition score=1.702 P=0.198. Blunted affect score=0.670 P=0.221. Alogia Score=−0.336 P=0.423

In model 2, BNSS asociality and BPRS, SCoRS, LASMI, and SFS scores were included in the logistic regression analysis. The results of the logistic regression analysis with asociality, BPRS, SCoRS, LASMI, and SFS are shown in Table 3. Asociality, the “daily living” domain of the LASMI, and SFS scores were included in the model. BPRS, SCoRS and LASMI (interpersonal, work) were not statistically significant (P > 0.05). SCoRS and LASMI (work), which were significant in the t-test, were also not included in the model. These results show that in addition to asociality, daily living skills and social and role functioning were characteristic of long-term hospitalization. The Hosmer–Lemeshow test showed a good value of Χ2 = 3.617 and a P = 0.890. The variance values were as follows: Cox-Snell R2 = 0.587, Nagelkerke R2 = 0.782.

Table 3

Final version of the stepwise multiple regression model with different facets of clinical factors

Model 2 Variable	Ba	P	Exp (B)	95% CI interval EXP (B)
				Lower	Upper
Asociality	0.604	0.045	1.829	1.012	3.305
LASMI Daily living	0.182	0.007	1.225	1.052	1.368
SFS social	−0.466	0.002	0.627	0.469	0.840
SFS role	−0.150	0.004	0.861	0.776	0.954

-2 log likelihood=30.156. BPRS positive score=0.011 P=0.773. SCoRS score=0.020 P=0.684. LASMI interpersonal score=0.006 P=0.435. LASMI work score=2.263 P=0.852

DISCUSSION

This study aimed to identify the characteristics of negative symptoms in long-term hospitalized patients with schizophrenia. The results of the logistic regression analysis showed that as opposed to community-dwelling patients, asociality characterized long-term inpatients. Therefore, it can be suggested that long-term inpatients tend to demonstrate a lesser interest in forming intimate relationships and lack the motivation to interact with others. This is a novel and interesting discovery as we did not find the same to be true about avolition, which has historically been proven as the most powerful predictor of functional outcome in community-dwelling patients.[2324]

Patients with schizophrenia tend to experience less joy in positive social cues, such as a smile or a friendly attitude.[32] Lack of positive emotions while communicating with others is considered to be a feature of asociality[33] that might be present in patients with schizophrenia and may reduce their motivation to form intimate relationships. Moreover, a recent study on community-dwelling residents with schizophrenia suggested that patients make less of an effort to be involved with others.[34] Another study reported that such patients tend to spend time on activities that are related to “disengagement desires,” such as watching television, which avoids involvement with others.[35] Based on the present findings, which showed that asociality is characteristic of individuals hospitalized with schizophrenia, it can be said that the desire to stay away from society may be stronger in hospitalized patients than in community-dwelling individuals with schizophrenia. Previously, Blanchard et al.[36] compared schizophrenia and social anhedonia trends in people with major depression and found that their ability to experience pleasure in interpersonal relationships was restored after being hospitalized. However, in the case of patients with schizophrenia, patients faced anhedonia even after one year of hospitalization.[36] Given that asociality is associated with a reduction in positive emotions felt when communicating with others, Blanchard et al.’s[36] study concurred with the results of this study, which reported characteristic asociality among patients under long-term hospitalization.

In addition to asociality, this study analyzed whether other factors characteristic of schizophrenia, including positive symptoms, cognitive function, functional skills, and social and role functioning, were also characteristic of long-term hospitalized patients. Regarding social and role functioning and the three functional skills (daily living, interpersonal, and work), results showed that lack of daily living ability is a characteristic of long-term hospitalized patients compared to community-dwelling patients. Given that the patient is in a restricted hospital environment, it is easy to imagine that social and role functioning, which evaluates the performance of functional skills, is more severely impaired in such patients when compared to community dwellers. However, regarding the three functional domains for evaluating ability (daily living, interpersonal, and work), the finding that daily living skills were the most impaired is novel. A recent study that examined the social competence of long-term hospitalized patients with schizophrenia[16] revealed that these patients showed poorer functional skills across all domains, including daily living and interpersonal skills when compared with community-dwelling patients; however, it did not examine the relative influence of the three skills. Therefore, the findings of this study showed that compared to interpersonal and labor skills, lack of daily living skills such as self-care, medication management, and lack of management among each functional domain may result in an increased risk of hospitalization.

Meanwhile, unlike asociality and lack of functional skills, positive symptoms and cognitive dysfunction were not included in the regression analysis despite being characteristic of long-term hospitalization in previous studies.[6] This indicates that asociality is more characteristic of long-term hospitalized patients than positive symptoms. Lack of interest in others can hinder actual interpersonal relationships and social participation more than hallucinations and delusions, making life in the community more difficult.[37] In fact, the results indicated that impaired functional outcomes such as social and role functioning are also characteristic of major long-term hospitalized patients. According to a recent large study, negative symptoms are reported to have a stronger impact on functional outcomes than positive symptoms.[38] Therefore, the ability to live in the community may be more influential in terms of the severity of negative symptoms, like asociality, than in positive symptoms, such as delusions and hallucinations. Cognitive dysfunction may also be less characteristic than asociality for long-term hospitalized patients with negative symptoms.

This study had limitations. Research strengths included accounting for basic attributes, such as age and disease duration, which reflect the course of schizophrenia; however, as a result, only a small sample was analyzed in this study. Therefore, caution should be exercised when generalizing the results of the regression analysis. Further, though this study has used matching methods, it is not possible to match for all confounders and possibility of residual confounding exists. In the future, sufficient sample size and inclusion of more clinical factors are needed to conduct more comprehensive research.

CONCLUSION

This study aimed to identify the factors characteristic of long-term hospitalized patients with schizophrenia among the five domains of negative symptoms (anhedonia, asociality, avolition, blunted affect, and alogia) compared to community-dwelling patients. This study is one of the first to examine the aspects of negative schizophrenia symptoms associated with long-term hospitalization. Our findings show that among the negative symptoms, severe asociality is the most characteristic of long-term inpatients. In addition, the exploratory analysis suggests that asociality may have a stronger association with long-term hospitalization than positive symptoms or cognitive function. Long-term patients are more likely to show asocial traits, that is, avoid social interactions, and may have very low motivation to engage in social activities. Among community-dwelling patients, those with a stronger tendency toward asociality may be more likely to move to long-term hospitalization in the future.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form the patient(s) has/have given his/her/their consent for his/her/their images and other clinical information to be reported in the journal. The patients understand that their names and initials will not be published and due efforts will be made to conceal their identity.

Ethical considerations

This study was approved by the Institutional Review Board of the International University of Health and Welfare and Nasukougen Hospital. All participants provided written informed consent after being given a complete explanation of the study.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.