روانسنجی بهزیستی ذهنی در مقیاس درمان نورولپتیک (SWN) در بیماران مبتلا به اسکیزوفرنی، بستگان و کنترل

Abstract The objective of this study was to evaluate the psychometric properties of the 20-item version of the Subjective Well-being Under Neuroleptic Treatment Scale (SWN) in patients, their siblings and parents and in healthy controls. In order to study heritability of subjective well-being, assessment in unaffected relatives and healthy controls is necessary. Data were obtained from the Dutch GROUP study (Genetic Risk and Outcome of Psychosis), a large cohort study on non-affective psychotic disorders incorporating patients, their relatives and healthy controls. The SWN scale and other relevant assessments were completed by 545 schizophrenia patients, 541 siblings, 75 parents, and 280 healthy controls. Reliability within the four groups ranged between Cronbach's alpha 0.88 and 0.92. Factor analysis indicated a single factor structure of the SWN scale, which makes only SWN total scores relevant.The WHO-Quality of Life psychological domain correlated highly with SWN total scores in all groups. Subclinical psychotic experiences were found to be associated with SWN total scores in relatives and healthy controls, supporting the psychosis continuum concept. The 20-item SWN scale is a reliable measure for subjective well-being that can also be used in relatives and healthy controls to investigate genetic and psychological dispositions of subjective well-being.

1. Introduction Recent research has provided evidence that psychotic phenomena are distributed throughout the general population. The hypothesis of a ‘psychosis continuum’ was based on findings of odd convictions and attenuated psychotic experiences in general population samples. These psychotic experiences were traditionally only recognized in severe form in patients fulfilling criteria of schizophrenia spectrum diagnoses. Levels of subclinical psychotic phenomena were found to be higher in at-risk populations, such as relatives of schizophrenia patients and subjects using cannabis (Johns and van Os, 2001, Van Os et al., 2009, Van Os et al., 2010, Linscott and Van Os, 2010, Binbay et al., 2011 and Dominguez et al., 2011). Based on this concept, it is relevant to investigate phenomena associated with schizophrenia in other populations, especially in relatives. One such related phenomenon is subjective wellbeing. Subjective wellbeing has been shown to be closely related to dopaminergic neuromodulation in the striatum (de Haan et al., 2000, de Haan et al., 2003 and Mizrahi et al., 2009). As all efficacious antipsychotic agents are dopamine antagonists, subjective well-being is influ-enced by antipsychotics, with higher dosage leading to a decrease in well- being. Subjective well-being of patients with schizophrenia is therefore relevant for medication adherence and recovery (Naber et al., 2001, Karow et al., 2007 and Lambert et al., 2007). Early improvement of subjective well-being after the start of treatment has been shown to predict better clinical outcome in the long term (Kluge et al., 2005, van Nimwegen et al., 2008, de Haan et al., 2008, Lambert et al., 2009, Mizrahi et al., 2009, Schennach-Wolff et al., 2010 and Kinon et al., 2010). Since subjective well-being is only moderately associated with psychopathology and clinicians' evaluation (Hunter and Barry, 2011, Karow et al., 2007, Karow et al., 2012, Chen et al., 2011 and Maurino et al., 2012), it is considered an independent outcome domain. The Subjective Well-being Under Neuroleptic Treatment Scale (SWN, Naber et al., 1994 and Naber, 1995) is the most widely used self-rating scale in recent studies on subjective well-being in schizophrenia (Vothknecht et al., 2011). Originally, the SWN was developed to assess subjective experience in five subscales (Naber et al., 1994 and Naber et al., 2001). However, this subscale structure was not confirmed in later research (Siamouli et al., 2009 and Phillips et al., 2010). The psychometric properties are firmly established in schizophrenia patients (Naber, 1995, de Haan et al., 2002, Balestrieri et al., 2006 and Siamouli et al., 2009). The SWN has been shown to be sensitive to treatment adherence, medication change and remission (Naber, 1995, Naber et al., 2005 and de Haan et al., 2002). In neuroimaging studies the SWN was shown to be related to dopamine D2 receptor blockade. Several studies have demonstrated that a striatal dopamine D2 receptor occupancy between 60% and 70% correlates with positive subjective well-being of patients with schizophrenia. Higher occupancy of D2 receptors was associated with lower subjective well-being (de Haan et al., 2000, de Haan et al., 2003, Mizrahi et al., 2007 and Mizrahi et al., 2009). Therefore, a close relationship seems to exist between the psychological experience of subjective well-being as measured by the SWN and neurobiological mechanisms measured by differences in dopaminergic neurotransmission. Although the importance of dopamine D2 receptor occupancy has been stressed in relation to psychosis and antipsychotic medication, D2 receptor density was also found to be associated with personality traits, attachment, social behavior and quality of life not only in patients with schizophrenia but also in healthy controls (Farde et al., 1997, Jönnson et al., 2003, Couture et al., 2007 and Cervenka et al., 2010). Recently, research on genetic dispositions of emotional states has been expanded, e.g. by the GENEQOL consortium, a worldwide network to study genetic factors of quality of life (Sprangers et al., 2009 and Sprangers et al., 2010). From twin studies, the heritability of subjective well-being is estimated to be between 40% and 50% (Bartels and Boomsma, 2009). There is a large body of evidence indicating subjective well-being to be stable over the long term. This stability is mainly attributable to additive genetic factors, whereas susceptibility to change seems to be more related to temporary individual environmental factors (Nes et al., 2006). IIn a large study conducted by Bartels and Boomsma (2009), subjective well-being was assessed by measures of life satisfaction, happiness, quality of life and life fulfillment. They found that the four measures were explained by one underlying genetic factor. Therefore, distinct measures of subjective well-being are probably not distinct at a genetic level and represent overlapping constructs. They concluded that future research should focus on the complex interplay between genes and environment, regardless of which measure of subjective well-being is being used. To determine hereditary aspects of subjective well-being and to facilitate future research into subjective well-being in non-clinicalsamples, the psychometric properties of the SWN scale in relatives of patients with psychotic disorders and in healthy controls need to be investigated. Although the SWN was developed to address well-being under neuroleptic treatment in patients with psychosis (Naber et al., 1994, Naber, 1995 and de Haan et al., 2002), it can be used as an instrument to measure subjective well-being independent of treatment-specific aspects (Wolters et al., 2006 and Wolters et al., 2009a). It contains several items referring to subjective well-being in general. Although some items of the SWN may seem to apply to clinical populations, we hypothesized that we would find variations associated with subclinical psychotic phenomena and other subclinical psychopathology in non-clinical samples. The content of the SWN was found to correlate highly with Quality of Life (QoL) scales (Wolters et al., 2006 and Wolters et al., 2009a). Therefore, we compared SWN measurements in non-clinical populations with the WHO-QoL Bref (The WHOQOL Group, 1998), a well-accepted standard of Quality of Life scales, as an indication of convergent validity. 1.1. Goal of the present study The present study has two related objectives: 1. To investigate and compare the psychometric properties of the SWN 20-item scale in patients, their unaffected relatives and healthy controls in terms of factor structure, internal consistency and reliability. 2. To investigate the convergent validity of the SWN compared to the WHO-QoL Bref in patients, their unaffected relatives and healthy controls. We hypothesized that the SWN scale is a reliable and valid instrument for measuring subjective wellbeing in patients as well as in relatives and healthy controls.

3. Results 3.1. Sample characteristics Groups differed in demographic variables (Table 1). We analyzed the influence of gender, age, education, ethnicity and lifetime cannabis use on subjective well-being in the different study groups. All analyses were non-significant. Patients had moderate PANSS ratings. CAPE scores of relatives were comparable to those of controls. Patients reported significantly lower subjective wellbeing than siblings, parents and healthy controls (Table 2). Table 1. Demographic and psychopathological variables of patients, siblings, parents and controls. Patients Siblings Parents Controls (n=600) (n=594) (n=80) (n=295) Gender (% male) 80.0 45.5 40 55.6 Agea 27.3 (7.1) 27.9 (8.0) 54.7 (5.6) 29.7 (9.9) Educational level % Primary 14.3 8.0 14.7 4.7 % Secondary 60.1 51.7 38.6 40.8 % Higher 25.5 40.4 58.7 54.4 Ethnicity (% Caucasian) 74.8 80.2 87.2 89.9 Lifetime cannabis dependence/abuse (%) 42.5 11.6 5.0 6.8 Duration of illnessa 4.5 (4.4) Age of onseta 22.3 (6.4) Number of psychotic episodea 1.6 (1.0) PANSS Positive scale 1.83 (0.73) Negative scale 2.16 (0.84) General scale 1.80 (0.52) CAPE Positive sympt. fr b 1.22 (0.20) 1.14 (0.12) 1.22 (0.19) Negative sympt. frb 1.56 (0.39) 1.54 (0.33) 1.51 (0.31) Depressive sympt. frb 1.65 (0.39) 1.66 (0.39) 1.58 (0.37) Positive sympt. d b 1.45 (0.47) 1.36 (0.45) 1.39 (0.41) Negative sympt. d b 1.70 (0.53) 1.68 (0.53) 1.67 (0.48) Depressive sympt. d b 1.22 (0.20) 1.96 (0.65) 1.90 (0.57) a Mean (S.D.) b sympt.=symptoms, fr=frequency; d=distress Table options Table 2. SWN total scores of patients, siblings, parents and controls. Total N Completed SWN N SWN missing SWN total score (SD) Patients 600 545 55 83.1 (15.4)⁎⁎ Siblings 594 541 53 98.6 (11.8)⁎⁎ Parents 80 75 5 96.9 (13.3)⁎⁎ Controls 295 280 15 100.8 (9.6)⁎⁎ Total 1569 1441 128 ⁎⁎ Significant at 0.01 level Table options 3.2. SWN evaluation 3.2.1. Factor analysis of SWN 20-item scale Firstly, we investigated correlations of the originally proposed five subscales, mental functioning, self control, emotional regulation, physical functioning and social integration. They were all highly intercorrelated: correlations varied from r=0.50 to 0.67 in patients, r=0.47 to 0.70 in relatives and r=0.37 to 0.60 in controls. All correlations were significant at the 0.01 level (two tailed). Secondly, we performed an exploratory factor analysis using principal component analysis. The Kaiser–Meyer–Olkin measure of sampling adequacy was 0.92. Bartlett's Test of Sphericity was highly significant (chi-square=4574; p<0.001). Both findings indicate that the variables entered were adequate for factor analysis ( Kaiser, 1974 and Bartlett, 1954). For the choice of numbers of factors, first Kaiser's criterion (eigenvalue>1) was inspected. There were five factors in patients and controls and four factors in relatives (siblings and parents) with eigenvalue>1. The first factor explained 37% of the total variance in patients, 38% in relatives and 34% in controls. Other factors explained less than 9% of the total variance in all three groups. Four- and five-factor models accounted for 58–61% of the total of common variance explained. Factor loadings for the three groups are presented in Table 3. Table 3. Factor loadings of the SWN 20-item scale in patients, relatives and controls. Factor loadings Patients Relatives Controls Powerless 0.651 0.709 0.635 Body is comfortable 0.721 0.647 0.597 Thinking is easy 0.456 0.400 0.420 No hope 0.697 0.724 0.641 Body feels familiar 0.708 0.678 0.669 Shy with new people 0.605 0.613 0.510 Imaginative 0.061 0.176 −0.011 Environment is friendly/familiar 0.682 0.702 0.590 Weak and exhausted 0.699 0.700 0.670 Emotions and sensations are dull 0.564 0.494 0.472 Thinking is difficult and slow 0.631 0.640 0.528 Feelings and behaviors are inappropriate 0.629 0.608 0.655 It is easy to keep in touch 0.612 0.647 0.626 Environment is perceived as changed 0.655 0.656 0.661 It is easy to draw a line 0.290 0.300 0.285 Body is a burden 0.676 0.636 0.657 Thought is flighty and undirected 0.703 0.670 0.656 Interested in surroundings 0.520 0.602 0.408 Feelings and behaviors are appropriate 0.569 0.606 0.680 Full of confidence 0.657 0.755 0.711 Extraction Method: Principal Component Analysis. 1 component extracted. Table options Thirdly, we inspected the scree plot, which showed a clear break after the first component indicating a single factor solution. As Kaiser's criterion (eigenvalues>1) has been shown to demonstrate a tendency to substantially overestimate the number of factors, a one-factor solution is most appropriate. This factor accounted for 34–39% of the variance in the different groups. 3.2.2. Internal consistency and reliability The internal consistency (Cronbach's alpha) of the 20-item SWN as measured in patients, siblings, parents and controls is high (Table 4). As an alternative indicator of reliability, we assessed the mean difference between the dual items (positively and negatively phrased items on the same topic). The mean difference positively and negatively formulated items was 0.53 (SD 0.68) in patients. Ninety-eight percent of the patients score a mean difference <2. In relatives and controls the mean differences were slightly larger (Table 4). Table 4. Internal consistency and reliability in patients, siblings, parents and controls. Completed SWN N Cronbach's alpha 20-item Mean difference between positive and negative items (SD; min−max) Patients 545 0.902 0.53 (0.68; −2.6−+2.8) Siblings 541 0.901 0.85 (0.57; −4.2−+1.3) Parents 75 0.919 0.92 (0.67; −3.2−+1.3) Controls 280 0.876 0.83 (0.47; −2.9−+0.4) All 1441 Table options 3.2.3. Convergent validity To assess convergent validity, we assessed associations between the SWN and quality of life. The SWN total score correlated highly with the four domains and two individual items (referring to general QoL and Health) of the WHO-QoL scale in all subjects (Table 5). Within all four groups highest correlations of the SWN total score were found with the psychological domain of the WHO-QoL, reflecting an overlap of these concepts. Table 5. Correlations of SWN total score with WHO-QoL items/domains. WHO-QoL QoL item Health item Physical domain Psychological domain Social domain Environmental domain Patients 0.501 0.423 0.568 0.706 0.460 0.448 Siblings 0.574 0.452 0.621 0.777 0.561 0.522 Parents 0.652 0.532 0.733 0.784 0.528 0.649 Controls 0.459 0.409 0.640 0.694 0.417 0.472 All significant at 0.01 level (two tailed) Table options 3.2.4. Correlations of SWN with PANSS and CAPE We compared the SWN total scores with PANSS and CAPE assessments as measures of psychopathological variables. SWN total scores and PANSS subscale scoreswere significantly correlated, varying from r=−0.20 to r=−0.30. Correlations between SWN total scores and subclinical psychotic experiences as assessed by the CAPE were significant as well, ranging from −0.36 to −0.64 in siblings and −0.27 to −0.63 in controls, with the strongest associations between SWN total scores and frequency of negative and depressive symptoms. In parents, frequency and distress of negative symptoms and frequency of depressive symptoms were significantly correlated with SWN total scores (r=−0.50 to r=−0.61).