Abstract
The study is aimed to examine the psychometric characteristics of Duke Social Support Scale (DSSI) in young rural Chinese individuals (379 suicides, 411 controls) aged 15–34 years. Social support was measured by 23-item DSSI which included Social Interaction Scale, Subjective Social Support and Instrumental Social Support. DSSI had high internal consistency (alphas all over .79), and correlated with hopelessness and anxiety in both samples. Confirmatory factor analysis showed that the structure models of DSSI were basically suitable for the original structure of DSSI but some items should be modified or deleted. Altogether, these findings support that DSSI has high reliability and validity, which makes it an acceptable measure for social support in young Chinese populations. However, further model tests should be carried out by deleting or modifying some items or being used in different populations.
Suicide is an important public health problem in the world (Mercy & Rosenberg, 2000). 814,000 suicides happened in the year of 2000 (World Health Organization, 2001). In China, suicide was the 5th cause of death for general population and the leading cause for those aged 15–34 years (Phillips, Li, & Zhang, 2002). Rural rates of suicide are 3 times higher than urban rates and female rates are 25% higher than male rates, which are different from that reported in Western countries (Phillips, Li, & Zhang, 2002). In 2004, disability adjusted life year (DALY) of suicide and self-inflicted injuries was 4,045,000, 2.02% of all DALYs of all diseases in China (World Health Organization, 2009). According to burden of disease measured by years of life lost (YLL), rural suicide completers share 90% of total YLL and rural women aged 25–39 years have the largest share of YLL (Yip, Liu, & Law, 2008). Rural suicides, especially of young women, have brought severe disease burden in China.
Suicide is the combined effect of multiple factors, such as biological, psychological and socioeconomic (Phillips et al., 2002; Zhang, Conwell, Zhou, & Jiang, 2004). Social support is one important factor (Beautrais, 2002; Compton, Thompson, & Kaslow, 2005; Zhang et al., 2004). Social support refers to the process through which social relationships promote health and well-being (Cohen, Gottlieb, & Underwood, 2000 ). Beautrias (2002) found that limited social network was positively associated with serious suicidal behavior. Compton et al. (2005) also found that social support was strongly associated with suicide attempts among low-income African Americans seeking treatment in a large, urban hospital. A study of risk factors of completed suicide in China found that social support was significantly and negatively associated with completed suicide (Zhang et al., 2004). These results suggest that social support is an independent factor in suicide or suicide attempts. Enhancing social support networks is one of four steps in the United Nations’ suicide prevention guidelines (United Nations, 1996), so measuring the level of social support of suicides is important in suicide prevention program.
In China mainland, social support is often measured bythe Social Support Rate Scale (SSRS; Xiao, 1994). SSRS has 10 items in three dimensions: subjective support, objective support, and utility of support. Due to its few items, it has been used in different populations in China, to study the relationship between social support and suicide ideation (Chen et al., 2008) or depression and anxiety(Fan & Chen, 2007) in college students, railroad workers (Dang, Dou, Dong, & Liu, 2008), and rural young women aged 15–39 (Sun, Ye, Fan, & Pan, 2007). However, it is only widely used in China mainland and the information of social support is provided by study participants themselves, there are no reports about its use in psychological autopsy (PA) studies, and the results about social support measured by SSRS are not likely to be compared with results from Western countries.
The Duke Social Support Index (DSSI) has also been used in China for studying the relationships between social support and suicide (Hu et al., 2005; Zhang et al., 2004; Zhang & Zhou, 2009). Zhang et al. (2003) analyzed the validity and reliability of the DSSI by comparing the responses of informants and controls, and found that the DSSI was reliable. However, the sample size was only 66 suicide cases, so data from a large sample in China would provide more information.
Method
Sample and design
Established PA method and case-control design were used to investigate rural young suicides and community living controls. In Liaoning, Hunan, and Shandong provinces 16 rural counties (6 in Liaoning, 5 in Hunan, and 5 in Shandong) were randomly selected. Suicides aged 15–34 years were consecutively sampled from October 2005 to June 2008. Community living controls also aged 15–34 years were recruited in the same counties for the same time periods. Altogether the information of 392 suicides (178 young women, 214 young men) and 416 living controls (214 young women, 202 young men) was obtained. In this study, data of 379 (96.7%) suicides and 411 (98.8%) controls were used to analyze psychometric characteristics of DSSI due to that they had no missing values on the DSSI.
Measures
Interviewers collected demographic information including age, gender, education, marital status, religion, party membership, and annual income. Physical illness status was obtained with this question, “Do you know whether s/he ever had serious or chronic illness?” Mental disorder was diagnosed by the psychiatrists based on all responses from two informants to the Chinese version of the Structured Clinical Interview for the DSM-III-R (SCID) (Gu & Chen, 1993; Spitzer, Williams, Gibbon, & First, 1988). The Chinese version of SCID has been shown to be reliable and valid in different Chinese samples (Zhang et al., 2003).
Social support was measured by the DSSI (Koenig et al., 1993), which is 23 items in three subscales: Social Interaction (SIS, 4 items), Subjective Social Support (SSS, 7 items) and Instrumental Social Support (ISS, 12 items). One example is, “How many family members could this person depend on or feel close to?” SIS and SSS subscales are answered on a 3-point Likert format; the ISS subscale is dichotomous. The DSSI total score is the sum of scores of SIS, SSS, and SIS, and it can range from 11 to 45, which a higher score of DSSI means more social support.
Hopelessness was measured by the Beck Hopelessness Scale (BHS; Beck, Weissman, Lester, & Trexler, 1974) which has 20 Likert responses from 1 (extremely low) to 5 (extremely high) with total scores ranging from 20 to 100. The BHS has high reliability and validity in Chinese samples., and evidence of validity is that total BHS correlated positively with total suicide ideation (Kong et al., 2007).
Trait anxiety was measured by the 20-item of Spielberger State-Trait Anxiety Inventory (STAI; Spielberger, 1983). Each item has a 4-point Likert response from 1 (never) to 4 (almost always); total scores can range from 20 to 80. In this study, the Cronbach’s alpha coefficient was .91 for suicide sample and .85 for the control sample.
The majority of responses by the two proxies was the same or similar. To the different responses of two informants on demographic variables or life events, we relied on the response provided by the informant who should best know the target. To determine a diagnosis with the SCID, we selected the positive response reasoning that the other informant might not have had an opportunity to observe the specific characteristic.
Procedure
An interview was scheduled between 2 and 6 months after the suicide. The local health agency or the village administration or doctor first approached informants of suicides and controls. Informants of suicides were recommended by the village head or the village doctor; informants of controls were recommended by the controls themselves. For each suicide or control, informant #1 was always a parent, spouse, or another important family member; informant #2 was always a friend, coworker, or a neighbor. After getting written informed consent, two informants were interviewed by our trained interviewers. Each informant was interviewed separately in a private place in a hospital/clinic or the informant’s home. The interview time averaged 2.5 hours.
Results
Of 379 suicides, 63.9% were aged 25–34 years old, 54.4% were men, 58.6% were married, 28.5% had some religion, 23.0% were Party /League member, 35.6% had physical illness, and 45.8% had mental disorders. Compared to controls, suicidal individuals were significantly less educated, less likely to be married, less likely have religion or Party/League membership, more likely to report physical illness, mental disorder, hopelessness, and anxiety. However, there were no significant differences between these two samples in age, gender, or personal annual income. See Table 1.
Table 1.
Characteristics of general information of suicide and control samples
| Variables | Suicide Sample | Control Sample | t/χ2 |
|---|---|---|---|
| N (%) / Mean(SD) | N (%) / Mean(SD) | ||
| Age | |||
| 15–24 | 137 (36.1) | 173 (42.1) | 2.923 |
| 25–34 | 242 (63.9) | 238 (57.9) | |
| Gender | |||
| Male | 206 (54.4) | 199 (48.4) | 2.780 |
| Female | 173 (45.6) | 212 (51.6) | |
| Education Years | 7.4 (2.7) | 9.1 (2.3) | −9.340*** |
| Personal annual income | 5611 (14179) | 7394 (13190) | −1.773 |
| Marriage | |||
| Yes | 222 (58.6) | 271 (65.9) | 4.555* |
| No | 157 (41.4) | 140 (34.1) | |
| Religion | |||
| Yes | 108 (28.5) | 68 (16.5) | 16.363*** |
| No | 267 (70.4) | 339 (82.5) | |
| Missing | 4 (1.1) | 4 (1.0) | |
| Party /League member | |||
| Yes | 87(23.0) | 183(44.5) | 40.606*** |
| No | 286 (75.5) | 223 (54.3) | |
| Missing value | 6 (1.5) | 5(1.2) | |
| Physical illness | |||
| Yes | 135 (35.6) | 56 (13.6) | 52.363*** |
| No | 243 (64.1) | 355 (86.4) | |
| Missing | 1 (0.3) | ||
| Mental disorder | |||
| Yes | 182 (45.8) | 16 (3.9) | 204.400*** |
| No | 197 (54.2) | 395 (96.1) | |
| Hopelessness | 69.8 (13.5) | 46.8 (8.0) | 28.387*** |
| Anxiety | 53.5 (10.6) | 40.7 (6.6) | 19.938*** |
p < 0.05,
p < 0.001
Mean (SD) was used for variables of personal annual income, hopelessness, and anxiety.
Multiple linear regression models were used to measure the relationship between social support and demographic and psychological factors in suicide and control samples. All variables in Table 1 were used with forward method with the probability of F being .05 (entry) and .10 (removal). The results showed that low anxiety and personal annual income were common factors related to DSSI score in both suicide and control samples. However, education years and low hopelessness related to DSSI score in the suicide sample whereas gender (being men), marriage, and Party/League member were related to DSSI scores in the control sample.
In the suicide sample, the DSSI alpha was .84. In the control sample, alpha was .79. In both suicide and control samples, total DSSI significantly and negatively correlated with total scores of BHS and STAI.
A confirmatory factor analysis yielded goodness-of-fit index (GFI) and adjusted goodness-of-fit (AGFI) values that were close to 1, however, the p values of chi-square were all less than .001 and the values of Root Mean Square Error of Approximation were also higher than .05, which indicated the confirmatory factor analyses just basically confirmed the original structure.
Discussion
In this study, proxies for suicidal individuals and living controls completed the DSSI to measure social support. The suicide sample was different from control sample in demographic and psychological characteristics, which is not surprising and is similar to the previous findings (Zhang et al., 2004). In both groups, personal annual income and low anxiety related to social support. In addition, some specific factors correlated with social support in the suicide sample and control sample. In the suicide sample, lack of education and hopelessness related to low social support. In controls, in contrast, gender, marriage, and party affiliation related to social support.
In this study, regardless of whether proxies or individuals themselves completed it, the DSSI was internally consistent. The implication is that it has broad usefulness in measuring social support in Chinese samples.
In the suicide sample, however, there was one item (item 4) with corrected item-total correlation coefficient of only .18. This item asked about the number of times in past week s/he attended meetings of club, religious groups, or other groups that s/he belong to (other than at work). In rural China, individuals may not understand what clubs are, and they seldom attend meetings, which is different from that of Western countries. Item 4 may be not suitable for a rural Chinese population. Also, item 11 had the lowest loading values in both samples (suicide: .39; control: .30). Item 11 refers to how satisfied the person was with relationships with family and friends. Informants may have had difficulty answering this question due to understanding capability and recall error. Deleting items 4 and 11 is recommended in future research.
In this study, hopelessness and anxiety significantly and negatively related to social support measured by DSSI. This finding supports the validity of the DSSI in measuring social support in young rural Chinese individuals.
This is the first study obtaining detailed information on reliability and validity of the DSSI when completed by loved ones of Chinese suicide and control samples. Its results provide some evidence that the DSSI would be suitable in measuring social support in Chinese youth populations by psychological autopsy method. As social support is an important predicator of suicide (Beautrais, 2002; Compton et al., 2005; Zhang et al., 2004), measuring social support of suicide behaviors by DSSI should help doctors and health policy makers in formulating suicide intervention programs. The results on social support measured by DSSI in Chinese culture should also provide a chance to compare findings on social support in China with that from Western countries.
Some limitations of this study should be mentioned here. First, the information of suicide cases and controls were obtained from their proxies, information bias might exist due to proxies’ bad memory or careless observation. Second, integrating the information from different sources had some arbitraryrules, and the informants might not be accurate in what they know about the target. Third, our sample was rural individuals aged 15–34 years old in China, which limits their representativeness. Fourth, we did not use the SSRS simultaneously because of cost and interview time considerations. Using both scales would have given us the chance to compare SSRS and DSSI in same the population.
Table 2.
Demographic factors related to DSSI in suicide sample and control sample by multiple linear regression analysis
| Variable | Suicide Sample a | Control Sample b | ||||||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| β | SE (β) | t | p | β | SE (β) | t | p | |
| Constant | 43.466 | 1.909 | 22.768 | 0.000 | 49.153 | 1.579 | 31.135 | 0.000 |
| Gender | 1.062 | 0.427 | 2.485 | 0.013 | ||||
| Marriage | 1.102 | 0.473 | 2.332 | 0.020 | ||||
| Education years | 0.312 | 0.106 | 2.955 | 0.003 | ||||
| Personal annual income | 4.801E-5 | 0.000 | 2.413 | 0.016 | 4.583E-5 | 0.000 | 2.803 | 0.005 |
| Party/League member | 1.554 | 0.443 | 3.505 | 0.001 | ||||
| Hopelessness | −0.123 | 0.027 | −4.554 | 0.000 | ||||
| Anxiety | −0.143 | 0.035 | −4.097 | 0.000 | −0.266 | 0.036 | −7.302 | 0.000 |
R = 0.524, R2 = 0.275, adjusted R2 = 0.266, F = 32.127, p < 0.001
R = 0.527, R2 = 0.278, adjusted R2 = 0.266, F = 23.175, p < 0.001
Table 3.
Mean (SD), item-total correlation and Cronbach’s alpha coefficients if item deleted of DSSI in suicide and control samples
| Item | Suicide Sample | Control Sample | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| Mean (SD) | Item-total correlation if item deleted (r) | Cronbach’s alpha if item deleted | Mean (SD) | Item-total correlation if item deleted (r) | Cronbach’s alpha if item deleted | |
| Social interaction | ||||||
| Item1 | 1.80 (0.66) | 0.336 | 0.837 | 2.31 (0.66)*** | 0.349 | 0.784 |
| Item2 | 1.64 (0.72) | 0.268 | 0.841 | 1.96 (0.74)*** | 0.343 | 0.786 |
| Item3 | 1.56 (0.73) | 0.333 | 0.838 | 2.00 (0.74)*** | 0.365 | 0.785 |
| Item4 | 1.20 (0.45) | 0.176 | 0.842 | 1.33 (0.59)** | 0.311 | 0.786 |
| Total score | 6.20 (1.80) | 7.60 (1.88)*** | ||||
|
| ||||||
| Subjected support | ||||||
| Item 5 | 1.91 (0.72) | 0.470 | 0.831 | 2.57 (0.55)*** | 0.469 | 0.775 |
| Item 6 | 2.04 (0.74) | 0.521 | 0.829 | 2.74 (0.48)*** | 0.510 | 0.773 |
| Item 7 | 2.29 (0.68) | 0.435 | 0.833 | 2.71 (0.49)*** | 0.409 | 0.779 |
| Item 8 | 2.21 (0.67) | 0.519 | 0.829 | 2.63 (0.61)*** | 0.340 | 0.784 |
| Item 9 | 2.16 (0.70) | 0.468 | 0.831 | 2.69 (0.55)*** | 0.425 | 0.778 |
| Item 10 | 1.80 (0.73) | 0.499 | 0.830 | 2.61 (0.54)*** | 0.448 | 0.776 |
| Item 11 | 2.02 (0.74) | 0.348 | 0.838 | 2.67 (0.61)*** | 0.273 | 0.789 |
| Total score | 14.42 (3.36) | 18.62 (2.44)*** | ||||
|
| ||||||
| Instrumental support | ||||||
| Item 12 | 0.94 (0.23) | 0.263 | 0.839 | 0.99 (0.11)** | 0.376 | 0.788 |
| Item 13 | 0.87 (0.34) | 0.477 | 0.834 | 0.98 (0.15)*** | 0.292 | 0.788 |
| Item 14 | 0.55 (0.50) | 0.426 | 0.833 | 0.78 (0.41)*** | 0.334 | 0.784 |
| Item 15 | 0.82 (0.39) | 0.372 | 0.836 | 0.90 (0.30)** | 0.318 | 0.785 |
| Item 16 | 0.80 (0.40) | 0.408 | 0.835 | 0.87 (0.33)** | 0.378 | 0.783 |
| Item 17 | 0.82 (0.39) | 0.416 | 0.835 | 0.90 (0.30)** | 0.354 | 0.784 |
| Item 18 | 0.68 (0.47) | 0.443 | 0.833 | 0.86 (0.35)*** | 0.350 | 0.783 |
| Item 19 | 0.78 (0.42) | 0.484 | 0.832 | 0.95 (0.22)*** | 0.276 | 0.788 |
| Item 20 | 0.64 (0.48) | 0.473 | 0.832 | 0.91 (0.29)*** | 0.310 | 0.786 |
| Item 21 | 0.74 (0.44) | 0.466 | 0.833 | 0.93 (0.26)*** | 0.311 | 0.786 |
| Item 22 | 0.74 (0.44) | 0.444 | 0.833 | 0.90 (0.30)*** | 0.378 | 0.783 |
| Item 23 | 0.89 (0.32) | 0.377 | 0.836 | 0.94 (0.24)** | 0.341 | 0.786 |
| Total score | 9.25 (3.03) | 10.91 (1.88)*** | ||||
|
| ||||||
| DSSI Total score |
29.87 (6.06) | 37.13 (4.51)*** | ||||
p < 0.01,
p < 0.001
The results were that means of items of DSSI in control group compared with that of suicide group.
Table 4.
Correlation between scores of hopelessness (BHS), anxiety (STAI) and social support (DSSI) in suicide and control samples by gender
| Social support | Hopelessness
|
Anxiety
|
||||
|---|---|---|---|---|---|---|
| n | r | p | n | r | p | |
| Suicide sample | 369 | −0.452 | 0.000 | 374 | −0.426 | 0.000 |
| Male | 201 | −0.474 | 0.000 | 202 | −0.488 | 0.000 |
| Female | 168 | −0.403 | 0.000 | 172 | −0.319 | 0.000 |
| Control sample | 410 | −0.332 | 0.000 | 410 | −0.404 | 0.000 |
| Male | 199 | −0.373 | 0.000 | 199 | −0.473 | 0.000 |
| Female | 211 | −0.292 | 0.000 | 211 | −0.341 | 0.000 |
Table 5.
Loadings of each item in Social Interaction Subscale (SIS), Subjective Support Subscale (SSS), and Instrumental Support Subscale (ISS) of DSSI by confirmatory factor analysis in suicide and control samples
| Item | Suicide Sample | Control Sample | ||||
|---|---|---|---|---|---|---|
|
| ||||||
| SIS | SSS | ISS | SIS | SSS | ISS | |
| Item 1 | 0.591 | 0.494 | ||||
| Item 2 | 0.626 | 0.599 | ||||
| Item 3 | 0.585 | 0.593 | ||||
| Item 4 | 0.479 | 0.491 | ||||
|
| ||||||
| Item 5 | 0.583 | 0.629 | ||||
| Item 6 | 0.732 | 0.591 | ||||
| Item 7 | 0.568 | 0.558 | ||||
| Item 8 | 0.712 | 0.533 | ||||
| Item 9 | 0.659 | 0.621 | ||||
| Item 10 | 0.554 | 0.624 | ||||
| Item 11 | 0.388 | 0.296 | ||||
|
| ||||||
| Item 12 | 0.456 | 0.502 | ||||
| Item 13 | 0.575 | 0.391 | ||||
| Item 14 | 0.496 | 0.506 | ||||
| Item 15 | 0.579 | 0.487 | ||||
| Item 16 | 0.592 | 0.659 | ||||
| Item 17 | 0.583 | 0.662 | ||||
| Item 18 | 0.590 | 0.529 | ||||
| Item 19 | 0.607 | 0.434 | ||||
| Item 20 | 0.599 | 0.435 | ||||
| Item 21 | 0.696 | 0.433 | ||||
| Item 22 | 0.602 | 0.539 | ||||
| Item 23 | 0.563 | 0.514 | ||||
|
| ||||||
| Goodness-of–fit index (GFI) | 0.888 | 0.873 | ||||
|
| ||||||
| GFI adjusted for degree of freedom (AGFI) | 0.864 | 0.847 | ||||
| Chi-Square | 546.195 | 715.368 | ||||
| Chi-square d.f. | 229 | 229 | ||||
| p | < 0.0001 | < 0.0001 | ||||
| RMSEA estimate | 0.061 | 0.072 | ||||
Acknowledgments
This research was supported by a grant of US NIMH: R01 MH068560. We acknowledge our research collaborators and interviewers in Liaoning, Hunan, and Shandong of China. We also appreciate all interviewees for their unique contribution to the study.
Contributor Information
Cunxian Jia, Email: jiacunxian@sdu.edu.cn, jiacx001@gmail.com, Shandong University School of Public Health, Postdoc. Program in School of Economics, 44 Wenhuaxi Road, Jinan, Shandong, 250012, China, Tel: 86-531-88382141-8803
Jie Zhang, Email: zhangj@buffalostate.edu, Shandong University School of Public Health, 44 Wenhuaxi Road, Jinan, Shandong, 250012, China, State University of New York College at Buffalo, 1300 Elmwood, Buffalo, New York, 14222, USA, Tel: 1-716-878-6425; Fax: 1-716-878-4009
References
- Beautrais AL. A case-control study of suicide and attempted suicide in older adults. Suicide and Life-Threatening Behavior. 2002;32:1–9. doi: 10.1521/suli.32.1.1.22184. [DOI] [PubMed] [Google Scholar]
- Beck AT, Weissman A, Lester D, Trexler L. The measurement of pessimism: the hopelessness scale. Journal of Consulting and Clinical Psychology. 1974;42:861–865. doi: 10.1037/h0037562. [DOI] [PubMed] [Google Scholar]
- Chen LF, Ma SB, Fan CX, Jing CX, Luo HY, Yan HH, et al. Social support and it’s effects on suicidal ideation among undergraduates. Chinese Journal of Public Health. 2008;24:925–926. (in Chinese) [Google Scholar]
- Cohen S, Gottlieb B, Underwood L. Social relationships and health. In: Cohen S, Underwood L, Gottlieb B, editors. Measuring and intervening in social support. New York: Oxford University Press; 2000. pp. 3–25. [Google Scholar]
- Compton MT, Thompson NJ, Kaslow NJ. Social environment factors associated with suicide attempt among low-income African Americans: the protective role of family relationships and social support. Social Psychiatry & Psychiatric Epidemiology. 2005;40:175–185. doi: 10.1007/s00127-005-0865-6. [DOI] [PubMed] [Google Scholar]
- Dang TM, Dou HZ, Dong YJ, Liu LQ. The relationship between anxiety, depression and personality and social support in railroad workers. Medical Journal of Chinese People’s Health. 2008;20:1990–1992. (in Chinese) [Google Scholar]
- Fan RQ, Chen WQ. Association of social support and coping styles with depression and anxiety among undergraduates. Chinese Journal of School Health. 2007;28:620–621. (in Chinese) [Google Scholar]
- Gu J, Chen Y. Instruction Manual for the Structured Clinical Interview for DSM-III-R (Chinese Edition) Kaohsiung: Kaohsiung Medical College; 1993. [Google Scholar]
- Hu DM, Jiang C, Liu QG, Sun YJ, Jia SH, Zhou L, et al. A psychological autopsy study on rural studies using conditional logistic regression model. Chinese Mental Health Journal. 2005;19:131–133. (in Chinese) [Google Scholar]
- Koenig HG, Westlund RE, George LK, Hughes DC, Blazer DG, Hybels C. Abbreviating the Duke Social Support Index for use in chronically ill elderly individuals. Psychosomatics. 1993;34:61–69. doi: 10.1016/S0033-3182(93)71928-3. [DOI] [PubMed] [Google Scholar]
- Kong Y, Zhang J, Jia S, Zhou L. Reliability and Validity of the Beck Hopelessness Scale for Adolescents. Chinese Journal of Mental Health. 2007;21:686–689. (in Chinese) [Google Scholar]
- Mercy JA, Rosenberg ML. Building a foundation for suicide prevention: the contributions of Jack C. Smith. American Journal of Preventive Medicine. 2000;19 (1 Suppl):26–30. doi: 10.1016/s0749-3797(00)00170-7. [DOI] [PubMed] [Google Scholar]
- Pfeiffer PN, Ganoczy D, Ilgen M, Zivin K, Valenstein M. Comorbid Anxiety as a Suicide Risk Factor among Depressed Veterans. Depression and Anxiety. 2009;26:752–757. doi: 10.1002/da.20583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Phillips MR, Li XY, Zhang YP. Suicide rates in China 1995–99. Lancet. 2002;359:835–840. doi: 10.1016/S0140-6736(02)07954-0. [DOI] [PubMed] [Google Scholar]
- Phillips MR, Yang G, Zhang Y, Wang L, Ji H, Zhou M. Risk factors for suicide in China: a national case-control psychological autopsy study. Lancet. 2002;360:1728–1736. doi: 10.1016/S0140-6736(02)11681-3. [DOI] [PubMed] [Google Scholar]
- Spielberger CD. Manual for the State-Trait Anxiety Inventory (Form Y) Palo Alto, CA: Consulting Psychologists Press, Inc; 1983. [Google Scholar]
- Spitzer RL, Williams JBW, Gibbon M, First AB. Instruction Manual for the Structured Clinical Interview for DSM-III-R (SCID, 6/1/88 Revision) New York: Biometrics Research Department, New York State Psychiatric Institute; 1988. [Google Scholar]
- Sun ML, Ye DQ, Fan YG, Pan FM. A correlation study on social support and suicidal ideation among rural females aged 15–39. Anhui Medical and Pharmaceutical Journal. 2007;11:624–626. (in Chinese) [Google Scholar]
- United Nations. Prevention of Suicide: Guidelines for the Formulation and Implementation of National Strategies. New York: United Nations; 1996. [Google Scholar]
- World Health Organization. World Health Report 2001. Geneva: Author; 2001. [Google Scholar]
- World Health Organization. Age-standardized DALYs per 100,000 by cause, and Member State, 2004. 2009 from http://www.who.int/healthinfo/global_burden_disease/estimates_country/en/index.html.
- Xiao SY. The social support rating scale and its applications. Journal of Clinical Psychological Medicine. 1994;4:98–100. (in Chinese) [Google Scholar]
- Yip PSF, Liu KY, Law CK. Years of life lost from suicide in China, 1990–2000. Crisis. 2008;29:131–136. doi: 10.1027/0227-5910.29.3.131. [DOI] [PubMed] [Google Scholar]
- Zhang J, Conwell Y, Wieczorek WF, Jiang C, Jia S, Zhou L. Studying Chinese suicide with proxy-based data: Reliability and validity of the methodology and instruments in China. Journal of Nervous and Mental Disease. 2003;191:450–457. doi: 10.1097/01.NMD.0000081613.03157.D9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang J, Conwell Y, Zhou L, Jiang C. Culture, risk factors and suicide in rural China: a psychological autopsy case control study. Acta Psychiatrica Scandinavica. 2004;110:430–437. doi: 10.1111/j.1600-0447.2004.00388.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang J, Zhou L. A case control study of suicides in China with and without mental disorder. Crisis. 2009;30:68–72. doi: 10.1027/0227-5910.30.2.68. [DOI] [PMC free article] [PubMed] [Google Scholar]