Reliability and validity of the multidimensional scale of perceived social support in Chinese mainland patients with methadone maintenance treatment

Available online at www.sciencedirect.com

ScienceDirect Comprehensive Psychiatry 60 (2015) 182 – 188 www.elsevier.com/locate/comppsych

Reliability and validity of the multidimensional scale of perceived social support in Chinese mainland patients with methadone maintenance treatment Kaina Zhou a, b , Hengxin Li c , Xiaoli Wei c , Juan Yin b , Peifeng Liang b , Hongmei Zhang b , Lingling Kou b , Mengmeng Hao b , Lijuan You b , Xiaomei Li a , Guihua Zhuang b,⁎ a

b

School of Nursing, Xi’an Jiaotong University Health Science Center, No. 76 Yanta Western Road, Xi’an, Shaanxi 710061, China Department of Epidemiology and Biostatistics, School of Public Health, Xi’an Jiaotong University Health Science Center, No. 76 Yanta Western Road, Xi’an, Shaanxi 710061, China c Xi’an Center for Disease Control and Prevention, No. 599 Xiying Road, Xi’an, Shaanxi 710054, China

Abstract Introduction: The multidimensional scale of perceived social support (MSPSS) is a valid tool for assessing perceived support from family, friends and significant others. However, evidence about reliability and validity of the MSPSS in Chinese mainland patients with methadone maintenance treatment (MMT) is lacking. Methods: The patients (n = 1212) being admitted to the first two largest MMT clinics in Xi’an were recruited in the study. Reliability was estimated with Cronbach's α and intra-class correlation (ICC). Convergent and discriminant validity was assessed using item-subscale correlation. Factorial validity was examined using exploratory and confirmatory factor analysis. The patients answered the questions of MSPSS at baseline and re-test after 6 months, respectively. Results: Cronbach's α of the overall MSPSS was 0.92 (subscales range: 0.84–0.89). ICC of the overall MSPSS was 0.65 (subscales range: 0.57–0.64). Better convergent validity (r ≥ 0.40) was demonstrated by the satisfactory hypothesized item-subscale correlation. All of the hypothesized item-subscale correlations were higher than the correlations between the hypothesized items and other subscales, indicating better discriminant validity. Two factors were extracted from the 12 items, with factor 1 mainly covering friends and significant others subscales (explained 55.56% variance) and factor 2 mainly covering family subscale (explained 11.77% variance). In comparison with the proposed three-subscale model, the two-factor observed model did not fit well in this sample according to model fit indices. Conclusions: The MSPSS has acceptable reliability and convergent/discriminant validity in Chinese mainland MMT patients. The proposed three-factor model of MSPSS is much better fit than the two-factor observed model in this study. Findings of the study will provide evidence of psychometric properties of the MSPSS in MMT patient population and expand the use of the MSPSS in clinical MMT context. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Opioid dependence is a chronic maladaptive pattern of heroin or opioid use, which poses adverse effects on individuals and society [1]. Methadone maintenance treatment (MMT) is a long-term opioid replacement therapy with daily methadone administration [2]. In China, it is estimated that about 190,000 drug users are being treated in the total

767 MMT clinics at the end of 2014 [3]. Due to the alternative characteristic and long-term treatment requirement of MMT, the majority of MMT patients cannot sustain the long-term treatment due to various factors, among which the most important one is social support [4–8]. Social support is a multidimensional and complex construct that has both structural and functional components [9]. The structural component includes quantitative properties of the

⁎ Corresponding author. Tel.: +86 29 82655103; fax: +86 29 82655032. E-mail addresses: [email protected] (K. Zhou), [email protected] (H. Li), [email protected] (X. Wei), [email protected] (J. Yin), [email protected] (P. Liang), [email protected] (H. Zhang), [email protected] (L. Kou), [email protected] (M. Hao), [email protected] (L. You), [email protected] (X. Li), [email protected] (G. Zhuang). http://dx.doi.org/10.1016/j.comppsych.2015.03.007 0010-440X/© 2015 Elsevier Inc. All rights reserved.

K. Zhou et al. / Comprehensive Psychiatry 60 (2015) 182–188

social network (e.g., size, proximity, range, and accessibility), frequency of contact with members of the network, reciprocal support and quality of the support [10]; while the functional component refers to perceived level of received support, such as emotional support, instrumental support (practical support), informative support and appraisal support [11]. These two components of social support can be broadly distinguished in terms of received and perceived support. Published studies have documented that social support has a buffering effect on stressful life events and depression [12,13] as well as great impacts on treatment outcomes [14–17], especially the perceived social support [18–20]. Perceived social support has been conceptualized as a function of beliefs about self-worth and the availability and responsiveness to others [21]. A promising scale that aims to measure perceived social support is the Multidimensional Scale of Perceived Social Support (MSPSS). The MSPSS is a short 12-item scale assessing both perceived availability and adequacy of emotional and instrumental support from three sources, i.e., family, friends and significant others [22]. The significant others subscale is a unique aspect of the MSPSS, which leaves the definition of who the “significant others” are to the respondent [23]. The MSPSS has been found to be reliable and valid both in its original English version [24] and other language versions across different subgroups, such as university students [25–27], nursing students [11], females [28,29], adolescents [30,31], South Asian migrants [32], schizophrenia patients [21], myocardial infarction patients [33], and psychiatric patients [34]. Being a specific population, MMT patients are characterized by chronic brain dysfunction due to drug addiction, which needs long-term methadone intake and duration for rehabilitation [1]. Thus, social support is very important for these patients maintaining in the treatment [4–8]. However, evidence regarding MSPSS application in MMT patient population is lacking. With respect to the MSPSS validation, most of the published studies are conducted using western samples and report acceptable internal consistency reliability, test–retest reliability and construct validity of the translated MSPSS. Only few studies replicate the three-factor model (i.e., family, friends and significant others) of the MSPSS; other two-factor models of MSPSS include friends and combination of family and significant others [35], or family and combination of friends and significant others [36]. Additionally, Akhtar et al. [29] found a single-factor model of the MSPSS in a group of antenatal women. Huang et al. [37] translated the MSPSS into Chinese and validated this instrument in cancer patients with sound psychometric properties. The Chinese MSPSS also has been validated in general populations such as medical students and adolescents [23,36]. However, psychometric properties of the MSPSS in MMT patient population are still unclear. The purpose of the study was to examine reliability and validity of the Chinese version MSPSS in MMT patients of mainland China. The hypotheses are (1) the items have

183

satisfied internal consistency and reproducibility; (2) the items consisting of a certain subscale have higher correlations with this subscale and lower correlations with other subscales; (3) the items reflect the three-factor model, i.e., family, friends and significant others. To our knowledge, this is the first study to evaluate MSPSS in Chinese mainland MMT patient population. Findings of the study will provide evidence of psychometric properties of the MSPSS in a specific population and expand the use of the MSPSS in clinical MMT context.

2. Methods 2.1. Ethics statement The study protocol was reviewed and approved by the Human Research Ethics Committee of Xi’an Jiaotong University. The written informed consent was obtained from each recruited patient before the questionnaire survey. 2.2. Subjects and data collection The subjects were admitted patients of the first two largest MMT clinics in Xi’an, China. One is privately funded and the other is publicly funded. Inclusion criteria were aged 18 years or over and Chinese-speaking. If the patients had cognitive disorders or refused to give written informed consent, they were excluded. Data were collected from March to September, 2014. In this study, the patients came to the clinics at different time and they took methadone one by one under the supervision of the clinical staff. Considering different education levels of the patients, individual face-to-face interview was used to collect data in order to ensure the data quality. After methadone taking, the patients received individual face-toface interview in a quiet and well-lit room, which was provided by the researchers and data collectors with special training. The data collection work did not interfere with the busy clinical settings. In the face-to-face interview, the patients understood the questions of MSPSS well and finished the questionnaire completely. Published studies show that the time frame of test–retest reliability of the MSPSS ranges from 1 week to 2 months in different populations [26,27,33,34]. However, few studies focus on test–retest reliability of the MSPSS with time frame over 2 months. According to the recommended longest time frame of 6 months in test–retest reliability evaluation and considering the long-term treatment process of MMT, the 6-month time frame was selected between test and retest in this study. 2.3. Multiple scale of perceived social support (MSPSS) The MSPSS is a 12-item instrument measuring perceived support from three subscales, i.e., family (FA), friends (FR) and significant others (SO) [22,24]. Each item is rated on a 7-point Likert-type response format ranging from 1 (very

184

K. Zhou et al. / Comprehensive Psychiatry 60 (2015) 182–188

strongly disagree) to 7 (very strongly agree). A total score is calculated by summing the results for all items, while the individual subscale can be used by summing the responses from the items in each of the three dimensions. The possible scores range from 4 to 28 for each subscale and from 12 to 84 for all items, with higher scores representing higher perceived social support. 2.4. Data analyses Internal consistency reliability was measured using Cronbach's α (desired value N 0.70) [38]. Test–retest reliability was measured by intra-class correlation (ICC) based on the re-test after 6 month. Floor and ceiling effects were calculated as the number and percentage of the total patients at the lowest and highest possible scores. This should be less than 15% regarding floor or ceiling effect, respectively [39]. Validity analyses included convergent, discriminant and factorial validity. Convergent and discriminant validity was examined using the hypothesized item-subscale correlation. A correlation ≥0.40 was considered as better convergent validity, while discriminant validity was supported whenever the hypothesized item-subscale correlation was higher than the correlations between the hypothesized items and other subscales [40]. In this study, the items consisting of the family, friends or significant others subscale were expected having higher correlations with their corresponding subscale but lower correlations with other subscales. Factorial validity was tested using the exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) respectively based on two equal randomly allocated sample groups. EFA was employed to determine whether the MSPSS has two or three underlying factors. The principal component factor analysis with oblique rotation was used, in that oblique rotation allows various amounts of correlation among the factors and provides a more compelling solution than orthogonal rotation [27]. The predictive items were considered to be relevant with factor loading coefficient ≥0.40 and the extracted factors with an eigenvalue ≥1.0 [40]. CFA is a modeling technique used to confirm whether a particular factor structure is consistent with the correlations or covariance of a set of observed variables [41]. In this study, the observed model of EFA was confirmed and the proposed three-factor model (i.e., family, friends and significant others) [24] was fitted using CFA. The model fit indices included goodness-of-fit index (GFI; desired value ≥ 0.90), adjusted goodness-of-fit index (AGFI; desired value ≥ 0.90), root mean square error of approximation (RMSEA; desired value b 0.08), comparative fit index (CFI; desired value ≥ 0.90), Tacker–Lewis index (TLI; desired value ≥ 0.90), parsimony goodness-of-fit index (PGFI; desired value N 0.50) and normed chi-square (i.e., the ratio of χ 2/df) (NC; desired value 1.0–3.0) [42]. The collected data were analyzed using SPSS 20.0 and AMOS 17.0 (SPSS Inc., Chicago, IL). A value of P b 0.05 (two-tailed) was considered as statistically significant.

3. Results A total of 1212 patients were recruited at baseline, with 851 (70.2%) in the privately funded clinic and 361 (29.8%) in the publicly funded clinic. One thousand and ten patients completed the re-test after 6 months. Two-hundred and two (16.7%) patients did not participate in the re-test because of transferring to other MMT clinics (n = 153, 75.7%), being admitted to hospital (n = 40, 19.8%), or losing contact (n = 9, 4.5%). The patients aged 42.48 ± 6.24 (range: 21–65) years, with 934 (77.1%) males. Over 80% patients had secondary education level (n = 987, 81.4%). More than half patients were married (n = 705, 58.2%) or employed (n = 618, 51.0%). Cronbach's α of the overall MSPSS was 0.92, with subscales of 0.88, 0.89 and 0.84 for family, friends and significant others. ICC of the overall MSPSS was 0.65, with subscales of 0.64, 0.61 and 0.57 for family, friends and significant others. Floor and ceiling effects of the overall MSPSS and subscales were all less than 15% (Table 1). Better convergent validity (r ≥ 0.40) was supported by the satisfactory hypothesized item-subscale correlation. The family (r = 0.83), friends (r = 0.86) and significant others (r = 0.88) subscales had higher correlation with overall MSPSS than that of the individual items (r b 0.80). All of the hypothesized item-subscale correlations were higher than the correlations between the hypothesized items and other subscales, indicating better discriminant validity (Table 2). Based on oblique rotation of EFA (n = 606, half randomly allocated MMT patients), two factors were extracted from the 12 items both in the pattern matrix and the structure matrix. The difference between the high and low loading factors was far more apparent in the pattern matrix, indicating that all items loaded most highly on their respective subscales with minimal cross-loading. Factor 1 explained 55.56% of the variance and covered the items of friends and significant others subscales, except for the item “I have a special person who is a real source of comfort to me (SO3)” of significant others subscale. Factor 2 explained 11.77% of the variance and covered the items of family subscale and the item SO3 (Table 3).

Table 1 Internal consistency reliability, test-retest reliability, and floor/ceiling effect of the multidimensional scale of perceived social support (MSPSS) (N = 1212). MSPSS

Internal consistency reliability (Cronbach's α)

Test–retest Floor/ceiling effect reliabilitya [n (%)] b (intra-class Floor effect Ceiling effect correlation, ICC)

Family (FA) Friends (FR) Significant others (SO) Overall MSPSS

0.88 0.89 0.84

0.64 0.61 0.57

12 (1.0) 29 (2.4) 18 (1.5)

34 (2.8) 14 (1.2) 9 (0.7)

0.92

0.65

6 (0.5)

5 (0.4)

a

Test–retest reliability: n = 1010. Floor/ceiling effect [n (%)]: the number and percentage of the patients at the lowest or highest possible score. b

K. Zhou et al. / Comprehensive Psychiatry 60 (2015) 182–188 Table 2 Item-subscale correlation (Spearman r): convergent validity a and discriminant validity b (N = 1212). Items

Subscale FA

Family (FA) 3 (FA1). My family really tries to help me. 4 (FA2). I get the emotional help and support I need from my family. 8 (FA3). I Can talk about my problems with my family. 11 (FA4). My family is willing to help me make decisions. Friends (FR) 6 (FR1). My friends really try to help me. 7 (FR2). I can count on my friends when things go wrong. 9 (FR3). I have friends with whom I can share my joys and sorrows. 12 (FR4). I can talk about my problems with my friends. Significant others (SO) 1 (SO1). There is a special person who is around when I am in need. 2 (SO2). There is a special person with whom I can share joys and sorrows. 5 (SO3). I have a special person who is a real source of comfort to me. 10 (SO4). There is a special person in my life who cares about my feelings.

FR

Overall

SO

0.83 0.85 0.42 0.54 0.67 0.86 0.45 0.59 0.71 0.85 0.54 0.57 0.73 0.85 0.53 0.58 0.74 0.86 0.53 0.87 0.61 0.79 0.41 0.88 0.50 0.70 0.49 0.85 0.59 0.75 0.55 0.81 0.58 0.75 0.88 0.48 0.47 0.81 0.66 0.46 0.49 0.81 0.66 0.66 0.57 0.81 0.77 0.62 0.64 0.80 0.78

a Convergent validity: the hypothesized item-subscale correlations ≥0.40 are in bold. b Discriminant validity: the hypothesized item-subscale correlation is higher than the alternative ones.

Table 3 Exploratory factor analysis of the multidimensional scale of perceived social support (MSPSS) (N = 606). Items

Factor Factor 1 2

1 (SO1). There is a special person who is around when I am 0.50 in need. 2 (SO2). There is a special person with whom I can share 0.50 joys and sorrows. 3 (FA1). My family really tries to help me. 4 (FA2). I get the emotional help and support I need from my family. 5 (SO3). I have a special person who is a real source of comfort to me. 6 (FR1). My friends really try to help me. 0.86 7 (FR2). I can count on my friends when things go wrong. 0.98 8 (FA3). I Can talk about my problems with my family. 9 (FR3). I have friends with whom I can share my joys and 0.92 sorrows. 10 (SO4). There is a special person in my life who cares 0.51 about my feelings. 11 (FA4). My family is willing to help me make decisions. 12 (FR4). I can talk about my problems with my friends. 0.79 Eigenvalue 6.67 Explained variance (%) 55.56

0.98 0.97 0.55

0.72

0.75 1.41 11.77

Kaiser–Meyer–Olkin measure of sampling adequacy: 0.92. Bartlett's test for sphericity: P b 0.001.

185

The two-factor observed model, i.e., friends combined with significant others and family combined with significant others (FR + SO-FA + SO), was tested using CFA (n = 606, the remaining half randomly allocated MMT patients). However, the model fit indices were not acceptable except PGFI (0.56). The proposed three-factor model was also fitted in the study based on the remaining 606 MMT patients, with eligible fit indices in GFI (0.90), CFI (0.90) and PGFI (0.58). Although the value of NC (χ 2/df = 8.39) was higher than 3.0, it was significantly lower than that of the two-factor observed model (χ 2/df = 14.20) in this study (Table 4) (Fig. 1).

4. Discussion Findings of the study provide evidence of reliability and validity of the Chinese version MSPSS in patients with MMT of mainland China. Both of the overall MSPSS and three subscales showed good internal consistencies, with Cronbach's α over 0.70. ICC of the overall MSPSS and subscales were over 0.60, except for significant others subscale (ICC = 0.57), indicating moderate (ICC range: 0.41–0.60) to good (ICC range: 0.61–0.80) test–retest reliability [43]. The similar result is also found in a sample of medical and dental students, which shows excellent internal consistency (Cronbach's α = 0.92) and high one-week test–retest reliability (ICC = 0.71) [26]. The consistent findings confirm that the MSPSS has acceptable reliability. Percentages of the floor and ceiling effects regarding the overall MSPSS and three subscales were all less than 15%, indicating that the instrument can capture the full range of potential responses in MMT patients during the treatment process. Convergent validity was better between the 12 items and their hypothesized subscales (r N 0.80); meanwhile, the three subscales had higher correlation with overall MSPSS (r N 0.80), which is consistent with the three-subscale model [22,24] and the higher-order three-factor model (i.e., the three individual factors of perceived social support are nested within a single higher order factor of global social support) [44]. The finding supports that the items of each subscale and the three individual subscales can reflect major characteristics consistently. All of the hypothesized item-subscale correlations were higher than the correlations between the hypothesized items and other subscales. It demonstrates that the items of different subscales measure different aspects of perceived social support, indicating better discriminant validity of the MSPSS. Different from the proposed three-factor model [22,24], two dominant factors were extracted based on oblique rotation and all items had unique loading on the factors. Factor 1 mainly covered friends and significant others subscales except the item SO3, while factor 2 mainly covered family subscale combined with the single item SO3. The finding is consistent with the conclusion that two individual factors of perceived social support consisting of

186

K. Zhou et al. / Comprehensive Psychiatry 60 (2015) 182–188

Table 4 Summary of model fit statistics from confirmatory factor analysis for the MSPSS factorial structure in patients with MMT (n = 606). Model specification

GFI

AGFI

RMSEA

CFI

TLI

PGFI

NC

(Criteria) Two-factor model (FR + SO-FA + SO) a Three-factor model (FA-FR-SO)

≥0.90 0.82 0.90

≥0.90 0.73 0.83

b0.08 0.15 0.11

≥0.90 0.85 0.92

≥0.90 0.81 0.90

N0.50 0.56 0.58

1.0–3.0 14.20 8.39

GFI = goodness-of-fit index; AGFI = adjusted goodness-of-fit index; RMSEA = root mean square error of approximation; CFI = comparative fit index; TLI = Tacker-Lewis index; PGFI = parsimonious-fit index; NC = normed chi-square. FA = family; FR = friends; SO = significant others; MSPSS = multiple scale of perceived social support. a The items in the two-factor model were factor 1 (FR1 + FR2 + FR3 + FR4 + SO1 + SO2 + SO4) and factor 2 (FA1 + FA2 + FA3 + FA4 + SO3).

family and combination friends and significant others are supported in the sample of older adults suffering from generalized anxiety disorder as well as students and depressed patients [23,35,36]. The probable explanation is that the respondents may not have been able to differentiate between significant others and other supportive people such as friends and family, which leads to the subscale friends and significant others merged into one. Specifically, the item SO3 (“I have a special person who is a real source of comfort to me”) was loaded on factor 2 (i.e., family subscale) in MMT patients of this study. It is probably due to majority of the patients regarded their family member as the special person who is a real source of comfort to them. The finding suggests that the support sources should be explicitly defined, especially for the specific population. The two-factor observed model (i.e., FR + SO-FA + SO) in this study was further confirmed using CFA. However, the

model fit indices were not eligible except PGFI (0.56). For the proposed three-factor model (i.e., FA-FR-SO), the model fit indices were eligible in GFI (0.90), CFI (0.90) and PGFI (0.58); additionally, the NC value was significant lower than that of the two-factor observed model in this study (8.39 vs. 14.20). The result supports that the three-factor model of MSPSS is much better fit than the two-factor model [44]. Since the interpersonal relationships that a person has with his or her family and the expectations that the individual has from them differ widely from that from their significant others, the three-factor model of MSPSS needs to be further examined in MMT patients in future work. There were some limitations of the study. First, the MSPSS was administered using a face-to-face interview, the performance of the instrument by self-completion will need to be confirmed in future study. Second, the external validity or predictive validity needs to be further explored in future

Fig. 1. Confirmatory factor analysis of the MSPSS (FA = family; FR = friends; SO = significant others; MSPSS = multiple scale of perceived social support).

K. Zhou et al. / Comprehensive Psychiatry 60 (2015) 182–188

work. Third, this study was conducted in Xi’an, which limited the generalization of the results to all MMT patients. The Chinese version MSPSS has been demonstrated as a valid tool for assessing perceived social support in patients with MMT. Both reliability and convergent/discriminant validity were supported to be strongly satisfactory. Twofactor model was observed in MMT patients of this study, but the related model fit indices were not acceptable. Further work should be carried out to evaluate the three-factor model of MSPSS in MMT patient population. By using MSPSS, MMT clinicians can evaluate social support status of the patients appropriately. It will be beneficial to clinicians developing target interventions to improve social support, which is helpful to enhance treatment retention and treatment outcomes for patients with MMT.

[14]

[15]

[16]

[17]

[18]

Acknowledgment [19]

Funding from the National T&S Major Project of China is gratefully acknowledged (grant no. 2012ZX10002001). [20]

References [1] Reist D. Methadone maintenance treatment in British Columbia, 1996–2008. Ministry of Health Living and Sport; 2010. [2] Faggiano F, Vigna-Taglianti F, Versino E, Lemma P. Methadone maintenance at different dosages for opioid dependence (Review). The Cochrane Collaboration. New York: John Wiley & Sons; 2008. [3] Du XZ. The progress of Community Methadone Maintenance Treatment in China. Available: http://www.jhak.com/jdwh/wczl/ 20141218/11895.html 2014 [Accessed 2014 December 31]. [4] Zhou K, Zhuang G. Retention in methadone maintenance treatment in mainland China, 2004–2012: a literature review. Addict Behav 2014;39:22-9. [5] de Dios MA, Stanton CA, Caviness CM, Niaura R, Stein M. The social support and social network characteristics of smokers in methadone maintenance treatment. Am J Drug Alcohol Abuse 2013;39:50-6. [6] Lin C, Wu Z, Detels R. Family support, quality of life and concurrent substance use among methadone maintenance therapy clients in China. Public Health 2011;125:269-74. [7] Yang L, Li J, Zhang Y, Li H, Zhang W, Dai F. Societal perception and support for methadone maintenance treatment in a Chinese province with high HIV prevalence. Am J Drug Alcohol Abuse 2008;34:5-16. [8] Panchanadeswaran S, El-Bassel N, Gilbert L, Wu E, Chang M. An examination of the perceived social support levels of women in methadone maintenance treatment programs who experience various forms of intimate partner violence. Womens Health Issues 2008;18:35-43. [9] Nausheen B, Gidron Y, Peveler R, Moss-Morris R. Social support and cancer progression: a systematic review. J Psychosom Res 2009;67:403-15. [10] Goebert D, Loue S. Social support, mental health, minorities and acculturative stress. In: Loue S, & Sajatovic M, editors. Determinants of Minority Mental Health and Wellness. New York: Springer; 2009. [11] Ekbäck M, Benzein E, Lindberg M, Arestedt K. The Swedish version of the multidimensional scale of perceived social support (MSPSS)-a psychometric evaluation study in women with hirsutism and nursing students. Health Qual Life Outcomes 2013;11:168-76. [12] Baek RN, Tanenbaum ML, Gonzalez JS. Diabetes burden and diabetes distress: the buffering effect of social support. Ann Behav Med 2014;48:145-55. [13] Giesbrecht GF, Poole JC, Letourneau N, Campbell T, Kaplan BJ, APrON Study Team. The buffering effect of social support on hypothalamic-

[21]

[22] [23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

187

pituitary-adrenal axis function during pregnancy. Psychosom Med 2013;75:856-62. Wang W, Lau Y, Chow A, Thompson DR, He HG. Health-related quality of life and social support among Chinese patients with coronary heart disease in mainland China. Eur J Cardiovasc Nurs 2014;13:48-54. Salonen P, Rantanen A, Kellokumpu-Lehtinen PL, Huhtala H, Kaunonen M. The quality of life and social support in significant others of patients with breast cancer—a longitudinal study. Eur J Cancer Care (Engl) 2014;23:274-83. Koetsenruijter J, van Lieshout J, Vassilev I, Portillo MC, Serrano M, Knutsen I. Social support systems as determinants of self-management and quality of life of people with diabetes across Europe: study protocol for an observational study. Health Qual Life Outcomes 2014;12:29-37. Dobkin PL, De CM, Paraherakis A, Gill K. The role of functional social support in treatment retention and outcomes among outpatient adult substance abusers. Addiction 2002;97:347-56. Khalil AA, Abed MA. Perceived social support is a partial mediator of the relationship between depressive symptoms and quality of life in patients receiving hemodialysis. Arch Psychiatr Nurs 2014;28:114-8. Eom CS, Shin DW, Kim SY, Yang HK, Jo HS, Kweon SS. Impact of perceived social support on the mental health and health-related quality of life in cancer patients: results from a nationwide, multicenter survey in South Korea. Psychooncology 2013;22:1283-90. Bekele T, Rourke SB, Tucker R, Greene S, Sobota M, Koornstra J. Direct and indirect effects of perceived social support on health-related quality of life in persons living with HIV/AIDS. AIDS Care 2013;25:337-46. Vaingankar JA, Abdin E, Chong SA. Exploratory and confirmatory factor analyses of the Multidimensional Scale of Perceived Social Support in patients with schizophrenia. Compr Psychiatry 2012;53:286-91. Zimet GD, Dahlem NW, Zimet SG, Farley GK. The multidimensional scale of perceived social support. J Pers Assess 1988;52:30-41. Cheng S, Chan AC. The Multidimensional Scale of Perceived Social Support: dimensionality and age and gender differences in adolescents. Pers Indiv Dif 2004;37:1359-69. Zimet GD, Powell SS, Farley GK, Werkman S, Berkoff KA. Psychometric characteristics of the Multidimensional Scale of Perceived Social Support. J Pers Assess 1990;55:610-7. Guan NC, Sulaiman AR, Seng LH, Ann AY, Wahab S, Pillai SK. Factorial validity and reliability of the Tamil version of multidimensional scale of perceived social support among a group of participants in university Malaya medical centre, Malaysia. Indian J Psychol Med 2013;35:385-8. Guan NC, Huai Seng L, Hway Ann AY, Ong Hui K. Factorial validity and reliability of the Malaysian simplified Chinese version of Multidimensional Scale of Perceived Social Support (MSPSS-SCV) among a group of university students. Asia Pac J Public Health 2013, http://dx.doi.org/10.1177/1010539513477684. Ng CG, Amer Siddinq AN, Aida SA, Zainal NZ, Koh OH. Validation of the Malay version of the Multidimensional Scale of Perceived Social Support (MSPSS-M) among a group of medical students in Faculty of Medicine, University Malaya. Asian J Psychiatr 2010;3:3-6. Aroian K, Templin TN, Ramaswamy V. Adaptation and psychometric evaluation of the Multidimensional Scale of Perceived Social Support for Arab immigrant women. Health Care Women Int 2010;31:153-69. Akhtar A, Rahman A, Husain M, Chaudhry IB, Duddu V, Husain N. Multidimensional scale of perceived social support: psychometric properties in a South Asian population. J Obstet Gynaecol Res 2010;36:845-51. Ramaswamy V, Aroian KJ, Templin T. Adaptation and psychometric evaluation of the multidimensional scale of perceived social support for Arab American adolescents. Am J Community Psychol 2009;43:49-56. Bruwer B, Emsley R, Kidd M, Lochner C, Seedat S. Psychometric properties of the Multidimensional Scale of Perceived Social Support in youth. Compr Psychiatry 2008;49:195-201.

188

K. Zhou et al. / Comprehensive Psychiatry 60 (2015) 182–188

[32] Tonsing K, Zimet GD, Tse S. Assessing social support among South Asians: the multidimensional scale of perceived social support. Asian J Psychiatr 2012;5:164-8. [33] Bagherian-Sararoudi R, Hajian A, Ehsan HB, Sarafraz MR, Zimet GD. Psychometric properties of the Persian version of the multidimensional scale of perceived social support in Iran. Int J Prev Med 2013;4:1277-81. [34] Wongpakaran T, Wongpakaran N, Ruktrakul R. Reliability and validity of the Multidimensional Scale of Perceived Social Support (MSPSS): Thai version. Clin Pract Epidemiol Ment Health 2011;7:161-6. [35] Stanley MA, Beck JG, Zebb BJ. Psychometric properties of the MSPSS in older adults. Aging Ment Health 1998;2:186-93. [36] Chou KL. Assessing Chinese adolescent's social support: the multidimensional scale of perceived social support. Pers Indiv Differ 2000;28:299-307. [37] Huang L, Jiang QJ, Ren WH. The correlation among coping mode, social support and psychosomatic symptoms in patients with cancer. Chin Ment Health J 1996;10:160-1. [38] Nunnally J, Bernstein I. Psychometric theory. New York: McGrawHill; 1994.

[39] Amiri P, Eslamian G, Mirmiran P, Shiva N, Jafarabadi MA, Azizi F. Validity and reliability of the Iranian version of the Pediatric Quality of Life Inventory TM 4.0 (PedsQL TM) Generic Core Scales in children. Health Qual Life Outcomes 2012;10:3-11. [40] Fayers PM, Machin D. Quality of life: the assessment, analysis and interpretation of patient-reported outcomes. 2nd ed. New York: John Wiley & Sons; 2007. [41] Jacob KS, Everitt BS, Patel V, Weich S, Araya R, Lewis GH. The comparison of latent variable models of non-psychotic psychiatric morbidity in four culturally diverse populations. Psychol Med 1998;28:145-52. [42] Wu ML. Structural equation modeling: AMOS operation and application. 2nd ed. Chongqing: Chongqing University Press; 2010. [43] Landis J, Kock G. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74. [44] Clara IP, Cox BJ, Enns MW, Murray LT, Torgurdc LJ. Confirmatory factor analysis of the multidimensional scale of perceived social support in clinically distressed and student samples. J Pers Assess 2003;81:265-70.