Validation of the Portuguese version of the Social Adjustment Scale on Brazilian samples

  • Published on
    16-Sep-2016

  • View
    218

  • Download
    4

Embed Size (px)

Transcript

<ul><li><p>Journal of Affective Disorders 69 (2002) 167175www.elsevier.com/ locate / jad</p><p>Research report</p><p>Validation of the Portuguese version of the Social Adjustment Scaleon Brazilian samples</p><p>a,b , c c a c c* C. Gorenstein , R.A. Moreno , M.A. Bernik , S.C. Carvalho , S. Nicastri , T. Cordas ,c d b,cA.P.P. Camargo , R. Artes , L. Andrade</p><p>a Departamento de Farmacologia, Instituto de Ciencias Biomedicas, Universidade de Sao Paulo, Sao Paulo, SP, Brazil</p><p>b LIM-23, Departamento de Psiquiatria, Faculdade de Medicina da Universidade de Sao Paulo, Sao Paulo, SP, Brazil</p><p>c Instituto de Psiquiatria, Hospital das Clnicas, Faculdade de Medicina, Universidade de Sao Paulo, Sao Paulo, SP, Brazil</p><p>d Departamento de Estatstica, Instituto de Matematica e Estatstica, Universidade de Sao Paulo, Sao Paulo, SP, Brazil</p><p>Received 27 April 2000; accepted 4 January 2001</p><p>Abstract</p><p>Background: Social dysfunction is reported in several psychiatric diseases and its evaluation is becoming an importantmeasure of treatment outcome. The aim of this study was to obtain normative data, to test the validity and the ability of thePortuguese version of the Self-Report Social Adjustment Scale (SAS-SR) to detect different clinical conditions. Methods:The Portuguese version of the SAS-SR was applied to a carefully selected non-psychiatric sample, and to depressed, panic,bulimic and cocaine-dependent patients. Depressed and panic patients were evaluated in two different clinical conditions:acutely symptomatic and in remission. Results: SAS overall and sub-scale scores of the normal sample were consistentlylower than all patient groups, indicating better social adjustment in all areas. Panic patients were impaired to a lower levelthan depressed and cocaine-dependent patients in overall adjustment. Depressed patients in remission, although in bettercondition, were still impaired in relation to normal subjects in overall social functioning, leisure time and marital areas. Inpanic patients in remission, normalization was not achieved in overall functioning, work and marital areas. Limitations:Sample size was small in some groups and the evaluation was cross-sectional. Conclusions: The Portuguese version ofSAS-SR is a useful instrument for detecting differences between psychiatric patients and normal subjects and for theevaluation of different clinical conditions, recommending its use in outcome studies. 2002 Elsevier Science B.V. Allrights reserved.</p><p>Keywords: Social Adjustment Scale; Social functioning; Depression; Panic; Normals; Antidepressants</p><p>1. Introduction</p><p>*Corresponding author. LIM-23, Departamento de Psiquiatria, The importance of assessing the impact of aFMUSP, Caixa Postal 3671, CEP 01060-970, Sao Paulo, SP,</p><p>psychiatric disorder on patients social functioning inBrazil. Tel. / fax: 155-11-3069-6958.E-mail address: cgorenst@usp.br (C. Gorenstein). terms of the patients ability to perform roles and</p><p>0165-0327/02/$ see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S0165-0327( 01 )00300-7</p></li><li><p>168 C. Gorenstein et al. / Journal of Affective Disorders 69 (2002) 167 175</p><p>their adequacy in interpersonal relationships is now 1980; or Composite International Diagnostic Inter-well recognized (McDowell and Newell, 1996). view, Robins et al., 1988) and then in a semi-Scales that focus specifically on social adjustment, structured clinical interview (Schedules for Clinicalapart from the core symptoms of the disorder, have Assessment in Neuropsychiatry; SCAN, Wing et al.,been available since the 1960s. In her first review of 1990), applied by a trained psychiatrist. Subjectsthe techniques to assess social adjustment, Weissman were considered normal when they did not meet(1975) described the existence of 15 scales that met criteria for any lifetime psychiatric diagnosis accord-the criteria for utility and adequate psychometric ing to ICD-10 or DSM-III-R. The sample consistedproperties. This number increased to 27 in 6 years of 61 subjects of both genders (37 female), aged 20(Weissman et al., 1981), and new instruments, such to 68 (mean6S.D. 38.3612.8), with different levelsas the Social Adaptation Self-evaluation Scale of education: 8 years of school (n5 8), 11 years of(SASS; Bosc et al., 1997), are still being developed. school (n5 16), and college or further educationThe interest in social functioning as a measure of (n5 37).treatment outcome is increasing, particularly becauseof drug trials showing differential effects on social 2.1.2. Patientsfunctioning (e.g., Dubini et al., 1997). Patients came from specialized clinics of the</p><p>Several areas of social functioning, such as work, Institute of Psychiatry of the Clinical Hospital,social and leisure activities, and family relationships, School of Medicine, University of Sao Paulo.</p><p>are well covered by the Social Adjustment Scale(Weissman and Paykel, 1974), derived from the 2.1.2.1. Depressed patientsStructured and Scaled Interview to Assess Maladjust- Outpatients who met DSM-IV criteria for Re-ment (SSIAM; Gurland et al., 1972a,b). Originally current or Single Major Depressive Episode, treatedan interviewer-rated scale, the SAS was soon con- at the Mood Disorder Unit, were included. Althoughverted to a self-report form (SAS-SR; Weissman and they exhibited anxiety symptoms, they did not fulfillBothwell, 1976). The SAS-SR has the advantage of criteria for any anxiety disorder. They were inter-being simple to administer, inexpensive, waiving viewed using the Schedule for Affective Disordersinterviewers training programs, and its results are and Schizophrenia Lifetime Version (SADS-L,highly correlated with those obtained with the inter- Spitzer and Endicott, 1979), applied by a trainedview form of the scale (Weissman and Bothwell, psychiatrist.1976). SAS-SR scores of acute depressives, al- The first sample included 44 patients of bothcoholics, and schizophrenics were significantly dif- genders (26 female), aged 22 to 75 (mean6S.D.ferent from community sample data (Weissman et 39.1612.4) who completed the questionnaire duringal., 1978). an acute depressive episode, before receiving any</p><p>The objective of this study was to obtain norma- treatment (the acutely symptomatic depressive sam-tive data for the SAS-SR Portuguese version from ple). An independent sample of 28 patients of bothBrazilian psychiatric patients and normal subjects in genders (20 female), aged 22 to 54 (mean6S.D.order to assess the psychometric properties and the 35.869.0), who improved considerably in theirusefulness of the scale in a different culture and clinical conditions after receiving antidepressants forlanguage, and to test its ability to detect different at least 8 weeks formed the in remission depressiveclinical conditions. sample.</p><p>2.1.2.2. Panic disorder patients2. Methods Outpatients meeting DSM-III-R criteria for Panic</p><p>Disorder with or without Agoraphobia (American2.1. Subjects Psychiatric Association, 1987) treated at the Anxiety</p><p>Disorders Clinic were included. Although they ex-2.1.1. Normals hibited depressive symptoms, they did not fulfill</p><p>Normal subjects were first selected in a general criteria for any mood disorder.screening (Self-Report Questionnaire, Harding et al., Forty patients of both genders (30 female), aged</p></li><li><p>C. Gorenstein et al. / Journal of Affective Disorders 69 (2002) 167 175 169</p><p>18 to 60 (mean6S.D. 35.7610.2), answered the 2.2.2. Beck Depression Inventory (BDI)questionnaire before receiving any treatment (acutely The Portuguese version of the 21-item Revisedsymptomatic panic sample). An independent sample Form of the Beck Depression Inventory (Beck et al.,of 58 patients of both genders (36 female), aged 22 1979, 1982) was used. The scale consists of itemsto 58 (mean6S.D. 37.668.7), completed the ques- including symptoms and attitudes with intensitiestionnaire after being treated with clomipramine (for ranging from neutral to a maximum level of severity,at least 1 year), and were fully remitted of panic rated 0 to 3 (overall score varies from 0 to 63). Theattacks and phobic symptoms according to clinical scale has been validated in Brazilian samplescriteria (in remission panic sample). (Gorenstein and Andrade, 1996; Gorenstein et al.,</p><p>1995, 1999). For samples of patients with affectivedisorders the BDI cut-off scores (Beck et al., 1988)</p><p>2.1.2.3. Other patients are: ,10, none or minimal depression; 1018, mildFourteen male inpatients from the Alcohol and to moderate depression; 1929, moderate to severe</p><p>Drug Dependence Clinic, aged 20 to 34 (mean6S.D. depression; 3063, severe depression. The scale was24.665.1), matching ICD-10 criteria for cocaine completed by all samples, except the bulimic pa-dependence, were included. They were first inter- tients.viewed by the SCAN, and completed the ques-tionnaire before receiving any treatment. 2.2.3. Hamilton Depression Inventory (HAM-D)</p><p>Fifteen outpatients from the Eating Disorder The Portuguese version of the 21-item scaleClinic, mainly women (14 females), aged 22 to 48 (Hamilton, 1960), evaluating mood and somatic(mean6S.D. 30.967.7), matching DSM-IV criteria aspects of depression in relation to their prevalencefor bulimia nervosa before receiving any treatment and severity, was used. The overall score varied fromwere included. They did not evidence any current 0 to 62.Axis I psychiatric comorbidity.</p><p>2.2.4. Clinical Global Impression Scale (CGI)The items global improvement and severity of</p><p>2.2. Instruments illness of this global scale (Guy, 1976) were used toevaluate the two samples of depressed patients.</p><p>2.2.1. Social Adjustment Scale (SAS)The Self-Report Social Adjustment Scale (Weis- 2.3. Statistical analysis</p><p>sman and Bothwell, 1976) is a 42-item scale,measuring either affective or instrumental perform- The concurrent validity, i.e. the capability of theance over the past 2 weeks in seven major areas of scale to discriminate different populations, was test-functioning: work (as a worker, housewife or stu- ed by ANOVA followed by Tukeys multiple com-dent); social and leisure activities; relationship with parison method.extended family; marital role as a spouse; parental In order to evaluate the ability to detect differentrole; membership in the family unit; and economic clinical states, SAS means of acutely symptomaticadequacy. The overall score is obtained by summing and in remission sub-samples (depressed and panicup the scores of all the items actually assessed and patients) were compared using the t-test. In remis-dividing that sum by the total number of items. Each sion sub-sample mean scores were compared toitem is scored on a five-point scale, from which the normal group scores by ANOVA, followed byrole-area means are obtained, the higher scores being Tukeys multiple comparison method. When theindicative of greater impairment (1, normal; 5, assumptions of the tests were not valid, data weresevere maladjustment). log transformed.</p><p>Two mental health professionals (one psychiatrist) Correlation between SAS and other scales wastranslated the SAS-SR into Portuguese and a native evaluated by the Pearson product moment correlationEnglish speaker reviewed this translation. A bilingu- coefficient.al speaker back-translated the Portuguese version to The influence of demographic variables was testedEnglish. on the normal sample data using ANCOVA consider-</p></li><li><p>170 C. Gorenstein et al. / Journal of Affective Disorders 69 (2002) 167 175</p><p>ing two factors gender and educational level atric samples was tested by ANOVA. Table 1 showsand age as covariable. overall and sub-scores of the seven major areas of</p><p>To compare Brazilian scores with those in the the Social Adjustment Scale in different populationsliterature, mean scores obtained by Weissman et al. and the statistical results of ANOVA post-hoc com-(1978) were considered as reference values and parisons (Tukeys test).compared by Students t-test with the Brazilian ANOVA showed a significant difference for thesample scores. SAS overall score (P, 0.001) where normal sample</p><p>scores were significantly lower than all other groups.Differences between patient groups were evidenced</p><p>3. Results between depressed and panic, and cocaine-dependentand panic patients.</p><p>3.1. Psychometric properties of the Brazilian Significant differences were also found for sub-version scales [P, 0.001 for all sub-scales, except for</p><p>parental and family unit (P, 0.05); Table 1]. Nor-Considering the normal sample, there was no mals were different from all other groups on the</p><p>overall effect of demographic variables on SAS (P. sub-scales work, leisure time and extended family.0.70) and on individual subscales (P$ 0.14) except Cocaine-dependent patients were different from de-for an age effect on the parental subscale (P5 0.04). pressed, panic and bulimic patients on the sub-scaleThe SAS subscales reliability estimated by Cron- work; the cocaine-dependent group was also sig-bachs alpha coefficient was 0.85. nificantly different from panic patients on the sub-</p><p>scale extended family. For family unit, the only3.2. Concurrent validity significant difference was between normals and</p><p>depressed patients. For economic conditions, normalsThe capability of the scale to discriminate between were different from depressed and bulimic patients.</p><p>the normal sample and acutely symptomatic psychi- Post-hoc tests could not be performed for marital and</p><p>Table 1aSocial Adjustment Scale scores in different samples</p><p>Work Social and leisure Extended family Marital Parental Family unit Economic condition Overall</p><p>1. Normal subjects (n5 61) 1.2460.28 a*** 1.7760.49 a*** 1.4660.38 a*** 1.5960.39 1.6460.71 1.5760.73 b** 1.6460.93 b***, c* 1.5660.36 a***2. Acutely symptomaticdepressed patients (n5 44) 2.5561.11 d*** 2.8560.68 d* 2.1460.64 d** 2.8060.83 d* 2.1660.94 d* 2.3261.11 2.5561.25 d* 2.5560.56 d***, e*</p><p>3. In remission depressedpatients (n5 28) 1.4460.64 2.4060.94 f*** 1.6460.52 2.2960.81 f*** 1.6060.56 1.7760.88 1.9260.91 2.0060.59 f***</p><p>4. Acutely symptomaticpanic patients (n5 40) 2.1060.90 g*** 2.5660.67 g*** 1.9160.52 g*** 2.1460.59 g* 1.7860.51 g** 1.7660.69 2.2961.33 g* 2.1760.43 g***</p><p>5. In remission panicpatients (n5 58) 1.4960.49 h** 1.9760.62 i* 1.5360.43 1.8060.55 h*, i* 1.4260.36 1.7160.54 1.7861.11 1.7160.38 h*, i*</p><p>6. Bulimic patients (n5 15) 2.2860.60 2.8160.87 2.4160.77 2.2860.70 2.6960.63 2.1360.84 3.2161.63 2.5560.637. Cocaine dependent (n5 14) 3.9961.53 j*** 2.7860.86 2.5961.06 k* 2.0060.00 1.5560.69 2.5761.60 2.7060.78 k*8. Community sample (n5 399;Weissman et al., 1978) 1.4060.46 1.8360.52 1.3460.33 1.7560.48 1.4060.42 1.4660.58 1.5960.33</p><p>9. Depressed patients (n5 172;Weissman et al., 1978) 2.4860.75 2.8560.66 2.1560.69 2.4560.55 2.2760.81 2.8360.89 2.5360.46</p><p>10. Alcoholic dependents (n5 26;Weissman et al., 1978) 1.8260.62 2.5060.79 2.0260.65 2.0760.64 1.7760.62 2.4661.14 2.2360.61</p><p>a Scores are expressed as mean6standard deviation. Results of post-hoc ANOVA comparisons (Tukey test) are shown by lettersconsidering group numbers as in the first column: a, 12, 4, 6, 7; b, 12; c, 16; d, 23...</p></li></ul>