Cognitive function and competitive employment in schizophrenia: relative contribution of insight and psychopathology
- First Online:
- Cite this article as:
- Giugiario, M., Crivelli, B., Mingrone, C. et al. Soc Psychiatry Psychiatr Epidemiol (2012) 47: 553. doi:10.1007/s00127-011-0367-7
- 314 Views
This study investigated the relationships among insight, psychopathology, cognitive function, and competitive employment in order to determine whether insight and/or psychopathology carried the influence of cognitive function to competitive employment.
We recruited 253 outpatients with stable schizophrenia and we further divided our sample into two groups of patients (unemployed and competitive employment subjects). Clinical and neuropsychological assessments were performed. All clinical variables significantly different between the two groups of subjects were subsequently analyzed using a binary logistic regression to assess their independent contribution to competitive employment in the two patients’ groups. On the basis of the regression results two mediation analyses were performed.
Verbal memory, general psychopathology, and awareness of mental illness were significantly associated with competitive employment in our sample. Both awareness of mental illness and general psychopathology had a role in mediating the verbal memory–competitive employment relationship.
Taken together, these findings confirmed the importance of cognitive function in obtaining competitive employment. Our results also highlighted the independent role of general psychopathology and awareness of illness on occupational functioning in schizophrenia. Thus, a greater attention must be given to the systematic investigation of insight and general psychopathology in light of an amelioration of vocational functioning in stable schizophrenia.
Unemployment rates for clinical samples of patients with schizophrenia range usually from 60 to 90% [1, 2]. The ability to work is markedly impaired in individuals with schizophrenia, especially in developed countries where jobs are, on average, more complicated than in less advanced societies . Despite these limitations, employment, and particularly competitive employment, has become a central goal of mental health treatment for people with serious mental illnesses . Indeed, competitive employment resonates with personal accounts and many definitions of recovery [5–8] and not only improves individual’s income, but it also enhances self-esteem and quality of life . Also, the majority of people with a history of mental illness want to work, consider themselves capable to work, and repeatedly express the need for job training, services, and supports .
Thus, a number of studies tried to investigate which factors could be considered predictors of competitive employment and researchers have focused their analyses on clinical factors that could be modified by a specific treatment.
Particularly, cognitive function has been the most intensively studied variable and it is now considered the most reliable predictor of occupational functioning in schizophrenia. Among retrospective, cross-sectional, or prospective studies of general samples of people with schizophrenia or other severe mental illnesses, 100% of the studies examining cognitive functioning, not including consumers in vocational rehabilitation, found it predicted work  and an analysis of baseline data from the large Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study (N = 1,438) provided additional evidence for the negative effect of cognitive impairment on competitive employment in schizophrenia . Furthermore, Lysaker and colleagues  reported that consumers with more cognitive impairments at the beginning of participating in the vocational program had fewer improvements in work performance over time than did the less impaired consumers.
Regarding psychopathology, negative symptoms have been considered more reliable predictors of vocational functions than positive symptoms. McGurk and colleagues , analyzing patients not included in vocational rehabilitation programs, found that, in 87% of studies examining negative symptoms and in 78% of studies evaluating psychotic symptoms, those symptoms predicted work. General psychopathology has been less intensively studied. A recent review  on vocational predictors observed that, since 1998, only three studies investigated general psychopathology while 19 and 16 studies, respectively, focused on negative and positive symptoms. Several cross-sectional studies have suggested that performance on neurocognitive tests is correlated with at least one of the three of the major symptom factors, positive, negative, or disorganization [16–22]. A meta-analysis of 73 published English language studies (total n = 6,519), using Sobel test of mediation, yielded fairly strong evidence that the total effects of neurocognition on outcome were at least partially mediated via an indirect path through negative symptoms .
Lack of insight (i.e. poor awareness of illness) is a cardinal feature of schizophrenia, and indeed of psychosis in general . The fundamental underpinnings of insight are the aptitude to label certain mental events as pathological, the recognition of being affected by a mental disorder, and the admission of the necessity of treatment and the consequent compliance with it . Lysaker and colleagues [26, 27] have reported that schizophrenia patients with poor insight have difficulties in adaptation to the worker role: working fewer weeks at the onset of a work program, exhibiting greater deficits in several areas of work performance, greater difficulties cooperating with co-workers, poorer work habits, tending to produce poorer quality work, and presenting to others in a less workmanlike style. Seeman and colleagues  highlighted that, in order to choose whether and to what extent disclose their disabilities to a potential employer, persons with psychiatric disabilities need to understand and acknowledge their disability. Finally, a body of literature deals with the possible relationships between specific neurocognitive functions and insight [28, 29]. It has been a matter of debate in recent years whether poor insight into psychotic disorders may, in part, be explained by neuropsychological dysfunction . In particular, it has been proposed that impaired functioning of the prefrontal cortex, which subserves mental flexibility, abstract reasoning, concept formation, and self-reflection, may lead to impaired insight . Recently, Beck and colleagues [31, 32] have coined the term ‘cognitive insight’, which is a form of cognitive flexibility that encompasses the capacity to distance themselves from their distorted beliefs and misinterpretations, reflect on them rationally, and recognize and correct distorted beliefs and misinterpretations. They contend that a crucial cognitive problem in the psychoses (including schizophrenia) is that individuals are unable to distance themselves from their cognitive distortions (e.g., ‘there is a conspiracy against me’), and are also impervious to corrective feedback . The mental operations underlying cognitive insight could be divided into two mechanisms: self-reflectiveness, which includes a willingness to acknowledge fallibility, corrigibility, and recognition of dysfunctional reasoning, and self-certainty, which refers to a tendency to be overconfident. Recent findings have highlighted the potential importance of cognitive insight as a mediator or moderator of response to cognitive behavioral therapy for psychosis  with increases in cognitive insight associated with reductions in positive, negative, and general symptomatology .
To investigate the differences between patients who were unemployed and those who had a competitive employment in terms of insight, psychopathology, and cognitive function. We expected that patients who were working at the time of assessment would have a better cognitive function, an higher level of insight, and a less severe psychopathology;
to investigate which significantly different variables between the two groups of subjects would predict vocational functioning. Given the previous studies, we expected a role for cognition but also for insight and psychopathology;
to investigate as to what extent cognitive function continued to predict occupational functioning when both insight and psychopathology were also considered. On the basis of the previous literature, it was expected that cognitive function continues to play a major role in influencing work but we also hypothesized that insight and psychopathology partially mediated the relationship between cognition and competitive employment.
The study has been conducted at the Department of Neuroscience, Psychiatric Section, Out-patient Clinics, and the Department of Mental Health ASL 1-Molinette, Italy. In the period between July 2008 and March 2010 we recruited 253 consecutive schizophrenic outpatients (n = 147, 58% men; n = 106, 42% women). They all fulfilled formal Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR)  diagnostic criteria for schizophrenia. The diagnosis has been confirmed by two expert clinicians (C.M., M.G.) using the Structured Clinical Interview for DSM-IV (SCID) . Prior to this study, interviewers received training sessions for SCID. At the time of study entry, patients had been clinically stable for at least 6 months as judged by treating psychiatrist, i.e. during this period all patients had to be treated as outpatients, treatment regimen had not been modified, and there was no essential change in psychopathology. In addition to medical records, patients were considered to be in stable phase as assessed from reports from patients themselves, and observations of the psychiatric staff, personnel in the community psychiatry, and relatives. Patients were evaluated using a semi-structured interview to assess demographic and clinical features. Data were collected to determine age, gender, education, vocational functioning, age at onset of schizophrenia (report of first contact with a psychiatric service), and length of illness. Subjects were excluded if they had a current disorder other than schizophrenia on Axis I of the DSM-IV-TR (screened with the SCID), a current or past codiagnosis of autistic disorder or another pervasive developmental disorder, a history of severe head injury (coma ≥48 h), a diagnosis of a psychiatric disorder due to a general medical condition, and a current employment other than competitive. All the patients were receiving antipsychotic medication at time of assessment. The two clinicians (C.M., M.G.) were aware of previous diagnosis and they could also review the previous clinical charts, available for all patients. On the basis of vocational functioning, our population was further divided into two groups (unemployed patients and patients with a competitive employment). Competitive employment was defined as a job that (1) was paid minimum wage or higher; (2) was located in a mainstream, socially integrated setting; (3) was not set aside for mental health consumers; and (4) was held independently (i.e. not contracted with a social service agency) . Written informed consent was obtained from all subjects after a complete description of the study.
The study was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. The present study was approved by a Local Research Ethics Committee (LREC).
All assessments were performed by two expert clinicians (C.M., M.S.). In an attempt to reduce inter-rater variability, all raters were trained to administer the psychometric tools according to common standards prior to study enrolment. Also, throughout the study, psychopathological rater training was performed regularly to establish a high inter-rater reliability. Finally, all the investigators conducted joint interviews before the commencement of the study, and joint ratings were made throughout the study to check inter-investigator agreement. Current levels of psychopathological symptoms were assessed using the Positive and Negative Syndrome Scale, PANSS , a rater-administered 30-item scale for measuring positive symptoms (PANSS-P), negative symptoms (PANSS-N), and general psychopathology (PANSS-G). Each item is scored on a seven-point severity scale (1 = absent; 7 = extreme), resulting in a range of possible scores ranging from 30 to 210. Insight into illness was assessed using the SUMD , which is a semi-structured interview designed to assess several dimensions of insight, including the following: (1) awareness of mental illness, (2) awareness of the need for treatment, (3) awareness of the social consequences of disorder, (4) awareness of specific signs and symptoms, and (5) attribution for the specific signs and symptoms of the disorder. Each of these domains is rated on a 6-point rating: 0 (not applicable), 1 (aware), 3 (somewhat aware/unaware), and 5 (severely unaware). An average score is then calculated for each one of them. A score ≥3 was significant of a poor individual level of insight.
Neuropsychological tests were administered by two trained psychologists (B.C.; M.S.) who were unaware of clinical characteristics and results of psychiatric rating scales. The battery was administered and scored according to standard instructions in the same way for all subjects on the day after the psychiatric assessments. The total testing time ranged from 1 to 2 h per patient (one or two sessions). None of the subjects were familiar with the tests.
Verbal learning and memory were assessed using the California Verbal Learning Test (CVLT) . The format of the test permits assessment of multiple aspects of verbal memory, including overall recall ability, rate of learning over repeated trials, ability to retain learned material over time, and subsequent items recognition. Four indices were used: the total number of items correctly recalled over five learning trials, the number of intrusion errors during immediate and long delay recall, and the number of false-positive errors.
The Stroop Test  and Part B of Trail Making Test  were used to estimate attentional functions and vigilance. We employed the traditional version of the Stroop Test based on card presentation as it permits one to evaluate the ability to inhibit an automatic response while performing a task based on conflicting stimuli. Two indices were used: the number of colors named in 30 s (Stroop C) and the number of colors named in the conflicting card (Stroop CW). This second index specifically assesses the sensitivity to interference and/or response inhibition. In order to index executive functions, we used the number of perseverative errors and the number of completed categories on the Wisconsin Card Sorting Test (WCST) .
The time measured in seconds on Part A of the Trail Making Test (TMT)  was used to evaluate perceptual and motor speed. The Trail Making Test, Part A, requires some form of attention, motor coordination, and visual tracking such as joining a series of numbers in increasing order as quickly as possible.
Statistical analyses were performed using the software Statistical Package for the Social Sciences, SPSS, Version 17 (SPSS Inc., 2008).
First, the demographic, clinical, and neuropsychological characteristics of patients with and without competitive employment were compared with One-way analysis of variance (ANOVA).
Second, all variables significantly associated with competitive employment at p < 0.05 level were subsequently analyzed using a binary logistic regression with a backward stepwise procedure to test their contribution to competitive employment. Backward elimination involves starting with all candidate variables, testing them one by one for statistical significance, and deleting any that are not significant.
Third, two mediational hypotheses were tested, using multiple regression analyses specified by Baron and Kenny . Mediation is said to occur when the following conditions are met: (1) the independent variable (IV) significantly predicts the dependent variable (DV); (2) the IV significantly predicts the potential mediator (M); (3) the M predicts the DV; (4) the effects of IV on DV is reduced when the M is included in the model. Sobel tests for indirect effects were employed to determine whether this attenuation was significant and whether M fully or partially mediated the relationship between the IV and the DV. Failure to conduct Sobel tests when examining mediating effects often leads to conclusions that are false positive or false negative . Sobel tests were computed with a SPSS macro developed by Preacher and Hayes , because SPSS does not provide the possibility of testing indirect effects with the Sobel test.
Two hundred-fifty-three consecutive outpatients who met the inclusion criteria were enrolled in the study. Of the patients included, 117 were unemployed (N = 69, 39% women and N = 108, 61% men) while 76 patients were competitively employed (N = 37, 49% women and N = 39, 51% men). The mean age (±SD) was 40.1 (±10.7) years. The mean average duration of illness (±SD) was 14.3 (±9.82) years. Mean years of schooling (±SD) were 10.9 (±3.50) years.
Means, p levels, and effect sizes for socio-demographic, clinical, and neuropsychological variables across unemployed and employed groups
Mean ± SD
Mean ± SD
21.0 ± 8.18
16.5 ± 7.63
15.1 ± 6.68
13.1 ± 6.36
37.7 ± 12.2
30.2 ± 11.4
2.50 ± 1.32
1.94 ± 1.30
2.95 ± 1.33
2.55 ± 1.20
3.12 ± 1.14
2.67 ± 1.22
3.04 ± 1.17
2.63 ± 1.18
19.61 ± 6.54
21.9 ± 7.82
CVLT Trial 1–5
36.12 ± 11.6
43.9 ± 12.2
CVLT intrusions during learning
1.72 ± 2.56
1.02 ± 2.19
CVLT false positive
3.03 ± 3.23
2.07 ± 3.01
Logistic regression with a backward stepwise procedure: contributors to competitive employment
95% C.I. for odds ratio
Regression models testing the mediation chains
Verbal memory → competitive employment
Verbal memory → general psychopathology
General psychopathology → competitive employment
Verbal memory → competitive employment excluding general psychopathology
Sobel test z = 0.0009 p = 0.020
Verbal memory → competitive employment
Verbal memory → awareness of mental disorder
Awareness of mental disorder → competitive employment
Verbal memory → competitive employment excluding awareness of mental disorder
Sobel test z = 0.009 p = 0.030
In our case each step the IV was verbal memory, the hypothetical Ms were general psychopathology or awareness of mental disorder and the outcome variable, the DV, was competitive employment. In the first analysis, we tested the effect of verbal memory on competitive employment (Fig. 1a). The effect of verbal memory was significant (β = 0.060, SE = 0.012, p = 0.000). In the second analysis, we inserted in the model general psychopathology and awareness of mental disorder as potential M of the effects of verbal memory (Fig. 1b). As can be seen in Fig. 1b and Table 3, verbal memory significantly predicted general psychopathology (β = 0.155, SE = 0.063, p = 0.015) and awareness of mental disorder (β = 0.018, SE = 0.007, p = 0.008). On the right side of the figure, it can be seen that general psychopathology (β = 0.059, SE = 0.014, p = 0.000) and awareness of mental disorder (β = 0.034, SE = 0.011, p = 0.003) significantly predicted competitive employment. The direct coefficient of verbal memory remained significant after inclusion of general psychopathology as mediating variable, while it changes somewhat in magnitude (β from 0.060 to 0.051). When we considered the direct path coefficient of verbal memory after inclusion of awareness of mental disorder as mediating variable, it remained significant, while it decreased somewhat in magnitude (β from 0.060 to 0.049). Thus, part of the overall effect of verbal memory on competitive employment appeared to be mediated both by general psychopathology and awareness of mental disorder (via indirect path mediated by general psychopathology and awareness of illness). Sobel tests for mediation demonstrated that both general psychopathology (z = 0.0009, p = 0.020) and awareness of mental disorder (z = 0.009, p = 0.030) significantly mediated the relation between verbal memory and competitive employment.
To the best of our knowledge, this is the only study that examines the cognitive-general psychopathology-competitive employment and the cognitive-insight-competitive employment mediation chains in patients with stable schizophrenia.
This study shows several key results relevant to clinical practice.
First, the overall employment rate in our sample was 48%.
As for the profile emerging from our research, none of the socio-demographic characteristics explored in the current study showed discriminative value in differentiating the two groups (unemployed patients and patients with a competitive employment). Our results are in accordance with a recent meta-analysis on participants in supported employment programs  showing that age and education were not associated with a better occupational functioning in schizophrenia, whereas some authors have identified a correlation between younger age and/or advanced education and better vocational functioning [51–57].
Second, employment was significantly associated with verbal memory, general psychopathology, and awareness of mental illness. Furthermore, both general psychopathology and awareness of mental illness had a role in mediating the verbal memory-competitive employment relationship. However, indirect effects were small.
These findings are in accordance with a broad range of studies showing that cognitive impairment could be considered a strong predictor of psychosocial functioning, including work, in schizophrenia. Particularly, a number of studies have identified a correlation between better vocational functioning and better verbal memory performances in schizophrenia and in severe mental illness [3, 58–67]. Dickerson and colleagues  have pointed out verbal memory as the strongest predictor of whether participants had any recent work. Finally, Bryson and Bell  have suggested that, while other cognitive abilities are crucial for initial success, verbal memory is necessary for sustained work improvement: verbal memory includes the ability to acquire new verbally presented information, and this aspect of memory is considered fundamentally related to the ability to perform everyday activities such as those in the work setting .
Regarding psychopathology, a large number of studies have previously identified a strong association between greater symptom severity and poorer employment [2, 15, 70]. Similarly, in our sample positive, negative symptoms and general psychopathology seemed to differentiate patients with competitive employment from unemployed ones. Our results would also suggest that general psychopathology partially mediated the verbal memory-competitive employment pathway. These symptoms seemed to play a major role in worsening vocational functioning, whereas the preponderance of previous studies focused on negative symptoms, showing their crucial role on employment. However, as shown by the systematic review of predictors of vocational outcomes among individuals with schizophrenia in the period from January 1998 to 2008 , only nine of the nineteen selected studies focusing on negative symptoms supported them as significant predictors of employment. Instead, among the three studies investigating general psychopathology, two of them have highlighted its significant role on vocational functioning.
Regarding the role of insight on employment, our findings are in agreement with previous studies [26, 27]. Particularly, Lysaker and colleagues [26, 27] speculated that among people with impaired insight, the reduced appreciation of one’s mental illness would make it even more difficult to compensate for deficits resulting in poorer performance on domains such as work habits and work quality. Dissonance between their belief that they are “not ill” and others’ perceptions that they are “ill” could derail relationships in both social and instrumental contexts, resulting in poorer performance on work domains such as social skills, cooperativeness, and personal presentation. Regarding our findings concerning the relationship between insight and cognition, verbal memory has been linked to a high resistance to modify one’s own judgments and difficulty in retrieving past memories . Lepage and collaborators  suggested that cognitive insight may rely selectively on verbal memory as it requires reflection and self-searching in memory whereby current experiences are appraised based on previous ones. Further, the magnitude of verbal learning and memory deficits corresponded with the degree of self-certainty. They also suggested that belief inflexibility may cause memories to be held with strong conviction, which may dissuade elaborate searches for previous experiences in memory. An important question that is yet to be investigated is whether the brain systems underlying cognitive insight in first episode psychosis overlap with those underlying verbal memory. However, our results would suggest that the effect of insight on competitive employment in schizophrenia is a more complex issue that could not be entirely determined by neurocognitive disturbances. This study has some strengths. First, our sample was homogeneous with regard to stage of illness. Second, all patients were diagnosed using a structured clinical interview (SCID) rather than clinical charts review. Third, we included in our study only patients with a diagnosis of schizophrenia rather than various psychotic disorders. Finally, few studies on vocational functioning have investigated earlier a sample of Italian subjects with schizophrenia and the majority of surveys have been conducted in the United States.
In spite of these strengths, some limitations should be noted before drawing inferences from present data. First, the cross-sectional nature of the present study could not provide definite support for the direction of the investigated relationships. It remains possible that the associations found here may be the result of other variable not assessed. Also, we excluded from our sample patients with a current employment other than competitive. Unfortunately, the groups of subjects with other types of employment were too scarce to be considered statistically significant and for this reason we decided to exclude them from the present study.
Taken together, these findings confirmed the importance of cognitive functioning, and particularly of verbal memory, in competitive employment. While medications targeting cognitive enhancement are yet to be approved, a few experimental studies that used cognitive remediation with vocational rehabilitation have reported good results [14, 73–77]. Our results also highlighted the independent role of general psychopathology and awareness of illness on occupational functioning in schizophrenia, suggesting that greater attention should be paid to their systematic investigation in light of an amelioration of vocational functioning in stable schizophrenia. Concerning awareness of mental illness, recent studies have shown cognitive remediation therapy to be of benefit in the treatment of poor insight in schizophrenia , while atypical antipsychotics seem not to be specifically beneficial .
This study was supported by a research grant 2006–2167 by Compagnia di San Paolo, Torino, Italy, and a research grant from REGIONE PIEMONTE “Bando Regionale per il finanziamento di Progetti di Ricerca Sanitaria Finalizzata, 2007”.