546

Chin Med J 2011;124(4):546-550

Original article
Correlations between self-reported symptoms and psychosocial factors of perpetrators with domestic violence in China: a population-based sample
CAO Yu-ping, ZHANG Ya-lin, Doris F. Chang, YANG Shi-chang and WANG Guo-qiang Keywords: domestic violence; perpetrator; symptom; correlation
Background Domestic violence (DV) is not only a devastating societal problem, but also a severe medical and mental health problem worldwide. Our previous study has shown that perpetrators were with higher prevalence of self-reported symptoms than that of controls. This study based on our former large scale population-based samples is aimed to further explore the correlations between the symptoms and psychosocial factors of the perpetrators with DV. It was helpful to provide some insight into possible strategies for clinicians to reduce the symptoms of the perpetrators with DV in China. Methods From our former population-based epidemiological samples, 1098 households with a history of DV in preceding year, 318 perpetrators with DV were randomly selected. Face-to-face interviews were conducted. Symptom Checklist-90 (SCL-90) was administrated to check and classify the symptoms of perpetrators, Eysencks personality questionnaire (EPQ), trait coping style questionnaire (TCSQ), life events scale (LES) and social supporting rating scale (SSRS) were administrated to evaluate the psychosocial factors of perpetrators. The correlation analysis was used to analyze the relationships between the symptoms and psychosocial factors of perpetrators of DV. Results The global and all subscale scores of SCL-90 were significantly positively correlated with EPQ-N, negative TCSQ and negative LES scores (P <0.01). The global score of SCL-90 was negatively correlated with both objective and subjective SSRS (P <0.01). The negative LES and negative TCSQ were significantly positively correlated with EPQ-N (P <0.01). Negative TCSQ was significantly positively correlated with negative LES and negatively correlated with subjective SSRS (P <0.01). Conclusions The self-reported symptoms of perpetrators with DV were strongly correlated with their psychosocial factors, such as the neurotic personality, negative coping style, more negative life events and less subjective social supports. It suggested bio-psycho-socially oriented interventions were necessary to buffer the symptoms of perpetrators with DV. Chin Med J 2011;124(4):546-550

omestic violence (DV) is not only a widespread and devastating societal problem, but also a severe medical and mental health problem.1-4 Over 10 million Americans experience DV each year,5,6 and DV was estimated to be the leading cause of serious injury and the second-leading cause of injury and death in the United States among women of child-bearing age.7 In China, in our former population-based epidemiological survey, it was reported that 16.2% of Chinese households had a history of experiencing DV.8 Extrapolating from this data combining with the census data,9 there may be more than 56 million households and 200 million people suffering from DV in China. Moreover, DV was the third-leading cause of completed suicide among young rural women in China.10 However, despite its seriousness and widespread prevalence in Chinese culture, DV has been ignored and minimized historically. Although, in recent years there has been an increasing public attention to DV issues in China, and some basic data have been gathered,8,11,12 there was little empirical information available especially about perpetrators in China. Even from western countries there was also sparse literature available which focus on perpetrators because it was difficult to gather the samples

in such studies. Moreover, the major problem in studying perpetrators of DV was that most researches obtained information about the perpetrators from the person who referred them to treatment. As direct assessment of non-forensic perpetrators is often hindered by lack of cooperation, assessment by informants may be a next-best,
DOI: 10.3760/cma.j.issn.0366-6999.2011.04.012 Mental Health Institute, Second Xiangya Hospital, Central South University, Changsha, Hunan 410011, China (Cao YP, Zhang YL, Yang SC and Wang GQ) Department of Psychology, the New School for Social Research, New York 10011, USA (Chang DF) Department of Psychology, Second Affiliated Hospital, Xinxiang Medical College, Xinxiang, Henan 453002, China (Yang SC) Wuxi Mental Health Center, Nanjing Medical University, Wuxi, Jiangsu 214151, China (Wang GQ) Correspondence to: Dr. ZHANG Ya-lin, Mental Health Institute, Second Xiangya Hospital, Central South University, Changsha, Hunan 410011, China (Tel: 86-731-85292159. Email: zhangyl69@vip.sina.com) This project was supported by grants from National Natural Science Foundation of China (No. 30670753); National Social Science Foundation of China (No. 06BSH043); American China Medical Board in New York, USA (No. 01-749).

Chinese Medical Journal 2011;124(4):546-550

547

but currently underused approach. However, results obtained in this manner, might not be entirely generalized to all perpetrators.13,14 To solve the sample bias, we used our former population-based epidemiological samples to get the information.8,15 In one of our case-control studies on DV,16 we have found perpetrators of DV had a higher prevalence of self-reported symptoms using Symptom Checklist-90 (SCL-90). SCL-90 was a widespread used scale in clinical study oversea and domestically. It was used for checking and classifying possible illness or distress of some persons.17 Based on these results, this study hypothesized the symptoms of perpetrators might be associated with their psychosocial factors, and these factors might bridge the symptoms and DV behaviors of the perpetrators. It may be helpful for providing some insight into possible strategies for clinicians to reducing the symptoms of perpetrators with DV in China. METHODS Context and participants The sample of this study was based on our previous research. We conducted the first population-based epidemiological survey to examine the prevalence of DV within diverse family constellations and various geographic settings including urban, rural and industry areas in central south of China. A multi-stage sample of residences was drawn following standard procedures for complex samples and found 1098 households had a history of DV in preceding year.8,15 A case-control study was conducted to examined the symptoms in perpetrators with DV.16 In this study, 310 households were randomly selected from the households with a history of DV in preceding year, using the table of random numbers. There were 318 perpetrators within the 310 households. It consisted of 216 (67.9%) males and 102 (32.1%) females. The average age was (42.011.0) years and the average amount of education was (8.94.7) years; 301 (94.6%) were married, 7 (2.2%) were single, 10 (3.2%) were divorced or lost the partner; 263 (82.7%) were employed, 31 (9.7%) were retired, and 24 (7.5%) were unemployed. Measures SCL-90 SCL-90 was built up by Derogatis et al in 197318 and widely used in clinical study. The Chinese version19 was applied in this study for checking and classifying the perpetrators symptoms. The SCL-90 is a self-report questionnaire with 90 items rated from 0 to 4 on the basis of the degree of illness or distress caused over the past week. Each item was scored: 0, not at all; 1, a little bit; 2, moderately; 3, quite a bit; 4, extremely. It consists of 10 subscales that assess somatization, obsessivecompulsiveness, interpersonal sensitivity, depression, anxiety, hostility, phobia, paranoid, psychosis and appendix symptoms. It provides a global score as a best single indicator of the global symptom. The SCL-90 was designed for use as a case-finding instrument to measure

symptom severity and as a descriptive measure of psychopathology.18 Eysencks personality questionnaire (EPQ) The Chinese version of EPQ was used to assess the personality of the respondent. It consists of 88 items.20 It presents to the respondent yes/no questions for each item, and it includes 4 factors such as neuroticism (N), psychoticism (P), extroversion/introversion (E) and lie (L). Higher scores for extroversion/introversion dimensionality represented extroversion and lower dimensionalities represented introversion; higher scores of neuroticism dimensionality suggested anxiety and worry. Higher scores for psychoticism suggested solitude and inadaptability to the outer environment. Trait coping style questionnaire (TCSQ) TCSQ was used to assess the respondents coping style. It consists of 20 items.19 It divides coping style into two dimensions: positive coping style and negative coping style. Life events scale (LES) LES was used for assessing exposure to psychological stress. It consists of 48 items.19 In this study, the questionnaire presents to the respondent positive/negative questions about the quality of the events, and also were rated from 0 to 4 on the basis of the degree of severity of events in past half year. Each item was scored 0, not at all; 1, a little bit; 2, moderately; 3, quite a bit; or 4, extremely. The LES provides a total score for life events, negative event score and positive event score. Social support rating scale (SSRS) SSRS is a 10-item scale and divides social support into three dimensions: objective social support (OS), subjective social support (SS) and the degree of social support utilization (US).19 Objective social support refers to material and tangible support; subjective social support refers to emotional support, which is the feelings of respect, support and being understood. Data collection In this study, DV was defined as physical, mental, or sexual abuse occurring among family members, including hitting, slapping, kicking, verbal insults, threats or intimidation, social isolation, deprivation, neglect, and sexual assault regardless of whether the abuse occurred inside or outside of the home. Social isolation included depriving another of freedom, or forbidding contact with individuals outside of the family was also considered as a form of DV.21 A perpetrator was defined as an individual who committed the acts described above towards another family member. Face-to-face interviews were conducted. All the interviewers participated in one day of systematic training which provided an overview of the goals, instruction in interviewing techniques and communication skills, as well as an explanation of the instruments used in this

548

Chin Med J 2011;124(4):546-550 Table 1. The correlations between scores of SCL-90 and EPQ, TSCQ, LES, SSRS (r)

Items SCL1 SCL2 SCL3 SCL4 SCL5 SCL6 SCL7 SCL8 SCL8 SCL10 Global EPQ P 0.07 0.21 0.20 0.16 0.13 0.11 0.11 0.15 0.13 0.16 E N 0.39 0.41 0.38 0.48 0.43 0.33 0.29 0.29 0.29 0.37 0.45 L 0.09 0.17 0.19 0.20 0.16 0.13 TCSQ NC 0.26 0.38 0.35 0.40 0.37 0.29 0.27 0.29 0.29 0.27 0.37 PC 0.13 LES Total 0.16 0.20 0.22 0.26 0.25 0.16 0.14 0.19 0.15 0.15 NLES 0.26 0.30 0.23 0.40 0.35 0.28 0.24 0.23 0.25 0.26 0.36 PLES 0.08 SSRS Total 0.10 0.14 0.22 0.23 0.14 0.15 0.26 0.16 0.18 OS 0.14 0.19 0.18 0.14 0.13 0.22 0.15 0.16 SS 0.08 0.20 0.22 0.13 0.16 0.21 0.14 0.16 US 0.13 This table only showed the significant variables. All P <0.05 or 0.01. SCL: symptom checklist; SCL1: somatization; SCL2: obsessive; SCL3: sensitive; SCL4: depression; SCL5: anxiety; SCL6: hostility; SCL7: phobia; SCL8: paranoid; SCL9: psychosis; SCL10: appendix; Global: SCL-90 global score; EPQ: Eysencks personality questionnaire; P: psychoticism; E: extroversion/introversion; N: neuroticism; L: lie; TCSQ: trait coping style questionnaire; NC: negative coping style; PC: positive coping style; LES: the life events scale; NLES: negative life event; PLES: positive life event; SSRS: social support rating scale; OS: objective social support; SS: subjective social support; US: social support utilization.

Table 2. The correlation within the scores of EPQ, LES, TSCQ, SSRS
Items P E N L PLES NLES PC NC LES PLES 0.13 1.00 NLES 0.18 0.39 0.13 1.00 TCSQ PC 0.35 0.14 1.00 NC 0.18 0.47 0.14 0.25 1.00 SSRS OS 0.14 0.14 SS 0.24 0.14 0.16 0.20 US 0.12 0.11 Total 0.24 0.13 0.17 0.14 This table only showed the significant variables. All P <0.05 or 0.01. LES: the Life events scale; PLES: positive life event; NLES: negative life event; TCSQ: Trait coping style questionnaire; PC: positive coping style; NC: negative coping style; SSRS: social support rating scale; OS: objective social support; SS: subjective social support; US: social support utilization.

study. Respondents were informed that the survey was entirely voluntary and confidential; after they had an opportunity to ask questions, individuals who agreed to participate provided both oral and written informedconsents. In the case with poor education who could not understand the contents of the instrument, the survey was explained and filled out by the interviewer according to respondents oral answers. This study was approved by Medical Ethic Committee of Second Xiangya Hospital of Central South University, China. Statistical analysis SPSS 15.0 statistic software (SPSS Inc., Chicago, USA) was used for analyzing the data. Partial correlation analyses were used between the symptoms and psychosocial factors of perpetrators by controlling age, gender and amount of education. Pearson correlation analysis was used between each subscale of the symptom. A two-sided P value of <0.05 was considered statistically significant.

RESULTS Correlations of SCL-90 and EPQ, TSCQ, LES, as well as SSRS After controlling for age, gender and the amount of education within the group, there was a significant positive correlation between all subscales scores of SCL-90 and scores of EPQ-N, negative TSCQ and negative LES (all P <0.01). The global score of SCL-90 was negatively correlated with the total, objective and subjective scores of SSRS (P <0.01, Table1). Pearson correlation analysis showed that there were significant positive correlations between each subscale score of SCL-90 (r=0.3690.762, all P <0.01). Correlations of within EPQ, LES ,TSCQ, SSRS After the same controlling method, EPQ-N was significantly positively correlated with negative LES and negative TSCQ (P <0.01); EPQ-E was strongly positively correlated with positive TSCQ and subjective SSRS (P <0.01). Negative TSCQ was strongly positively correlated with negative LES and negatively correlated

Chinese Medical Journal 2011;124(4):546-550

549

with subjective SSRS (P <0.01, Table 2). DISCUSSION This study was based on the population-based and standard random sample to explore the self-reported symptoms and their related psychosocial factors of the perpetrators with DV in China. From our former result, the perpetrators of DV endorsed various symptoms.16 In most cases, the levels of these symptoms reported were not sufficient to result in a diagnosis of any disorders.22 Perpetrators reported they were with more somatic symptoms, an increased sensitivity to environmental stimuli, increased anxiety and depression, more obsessive ideas and compulsive behaviors, disturbances in the control of fear conditioning, increased hostility to others, and diet and/or sleep problems as well. Hostility was a risk factor in predicting perpetrators of DV.16 This study showed hostility may escalate other symptoms of perpetrators. Prspero et al23 and Bitler et al24 reported perpetrators tend to be generally more hostile and angry when confronted with domestic conflict situations, and their somatic symptom may be escalated by the sense of anger or fear. The references available have given some potential biological explanations for the condition of perpetrators. George et al25,26 found that changes in levels of neurotransmitters seen after perpetrators committed acts of violence, may lead to a heightened sensitivity to environmental stimuli, anxiety, and conditioned fear. The decreased correlations in glucose metabolism between the cortex and the amygdala might impair the perpetrators ability to extinguish anxiety and/or conditioned fear and gives rise to either innate behaviors (e.g., fight and flight) or learned fear avoidant behaviors designed to avoid anxiety (e.g., alcohol consumption, self-injurious acts, and obsessive behaviors). George et al27 also found that perpetrators may not only feel fear and panic, but also rage in response to sodium lactate which in previous literature has been shown to increase anxiety. After the administration of sodium lactate, it often took the participants hours or days for the intense feelings to subside. Both low 5-hydroxytryptamine (5-HT) and high testosterone concentrations can modulate sensory stimuli that serve to activate the neuropathways which mediate fear-induced aggression. These changes predispose perpetrators to overreact to actual as well as perceived threats.28 It could be seen that the perpetrators not only committed the DV behavior but also suffered from various symptoms raised by his behavior.24 Medical intervention was needed for some perpetrators. Moreover, this results found that hostility was also strongly associated with perpetrators neurotic personality, negative coping style, negative life events and less subjective social supports. It suggested psychosocial intervention focus on these factors might be useful for relieving the hostility and other symptoms of perpetrators. The perpetrators were more likely to have neurotic personality.16 This study found all subscales of symptoms of perpetrators were strongly positively correlated with

their neurotic personality. With this type of personality, the person easily became nervous, anxious, irritable and depressed; this might lead him to react to his surroundings in a more volatile fashion and had increased difficulty remaining calm. Zhang et al29 found that the neurotic personality had a plastic feature after psychological treatment. It implied that perpetrators neurotic personality might be remodeled by psychological intervention; and therefore relieve their symptoms. All subscales of the symptoms of perpetrators were found to be significantly positively associated with negative coping style and negative life events. It was reported the perpetrators suffered from more negative life events.30 Negative life events often led to emotional symptoms, such as anxiety, depression, and hostility. Clemens et al31 found that stress or trauma was one of the bridges between psychological symptoms and DV behavior. Perpetrators possibly suffered from more symptoms when they were in the midst of a negative life events; conversely, perhaps the perpetrators negative coping styles and negative mindset and maladaptive methods for solving problems secondary to their symptoms might result in their creating new negative life events. It was unclear which one was the cause and which one was the consequence. Further longitudinal study is required. It suggested readjustment the negative coping styles and reducing the negative life events might buffer against the symptoms of perpetrators. Perpetrators had limited social support network,16,32 both objectively and subjectively.16 They were not adept in utilizing outside material and emotional support, and felt less respected, supported, and understood, and low subjective social support was a significant predictor of DV.16 The impact of social support depends on the type of support available. For example, Coohey and Braun33 found that mothers who abused their children received less emotional support but no difference in tangible support compared with nonabusing mothers. This study also found the more subjective social support, the less interpersonal sensitivity, depression, hostility and paranoid symptoms, and the less negative coping styles as well. It inferred the symptoms of perpetrators would be benefit from being given subjective social support during the psychotherapy, which meant to teach them how to feel and experience the support, the respect, and being understood. In conclusion, this study found that the self-reported symptoms of perpetrators with DV were strongly linked with their psychosocial factors, such as the neurotic personality, negative coping style, more negative life events and less subjective social supports. It may merit attention to clinical applications for establishing bio-psycho-socially oriented prevention and intervention strategies for reducing the symptoms of perpetrators with DV. Based on this cross-sectional study, future considerations are required to involve longitudinal studies to investigate causality between the symptoms and

550

Chin Med J 2011;124(4):546-550 psychiatric rating scale preliminary report. Psychopharmacol Bull 1973; 9: 13-28. Rating Scales for Mental Health. Chin Mental Health J (Chin) 1999; Suppl: 12: 13-28, 101-106, 120-122, 127-130. Gong YX. Revised Eysenck Personality Questionnaire (EPQ-R) (Chin). ChangshaHunan Map Publishing House; 1992. Zhang YL, Cao YP. Domestic violence and mental health. Chin J Clin Psychol (Chin) 2002; 10: 233-234, 240. Tolan P, Gorman-Smith D, Henry D. Family violence. Annu Rev Psychol 2006; 57: 557-583. Prspero M. The effect of coercion on aggression and mental health among reciprocally violent couples. J Fam Viol 2008; 23: 195-202. Bitler DA, Linnoila M, George DT. Psychosocial and diagnostic characteristics of individuals initiating domestic violence. J Nerv Men Dis 1994; 182: 583-585. George DT, Phillips MJ, Doty L, Umhau JC, Rawlings RR. A model linking biology, behavior and psychiatric diagnoses in perpetrators of domestic violence. Med Hypotheses 2006; 67: 345-353. George DT, Rawlings RR, Williams WA, Phillips MJ, Fong G, Kerich M, et al. A select group of perpetrators of domestic violence: evidence of decreased metabolism in the right hypothalamus and reduced relationships between cortically subcortical brain structures in position emission tomography. Psychiarty Res Neuroimaging 2004; 130: 11-25. George DT, Hibbeln JR, Ragan PW, Umhau JC, Phillips MJ, Doty L, et al. Lactate induced rage and panic in a select group of subjects who perpetrate acts of domestic violence. Biol Psychiatry 2000; 47: 804-812. George DT, Umhaua JC, Phillipsa MJ, Emmelaa D, Raganb PW, Shoaf SE, et al. Serotonin, testosterone and alcohol in the etiology of domestic violence. Psychiatry Res 2001; 104: 27-37. Zhang YL, Yang DS, Cao YP. A study on the plasticity of personality, behavior pattern of the patients with generalized anxiety disorder. Chin J Behavioral Med Sci (Chin) 2001; 10: 414-415. Huang GP, Zhang YL, Cao YP, Momartin S, Wei M. Relationship between recent life events, social supports, and attitudes to domestic violence: predictive roles in behaviors. J Interpers Violence 2010; 25: 863-867. Clemens P, Hietala JR, Rytier MJ, Schmidt RA, Reese DJ. Risk of domestic violence after flood impact: effects of social support, age, and history of domestic violence. Applied Behavior Sci Review 1999; 7: 199-206. Levitt HM, Swanger RT, Butler JB. Male perpetrators perspectives on intimate partner violence, religion, and masculinity. Sex Roles 2008; 58: 435-448. Coohey C, Braun N. Toward an integrated framework for understanding child physical abuse. Child Abuse Negl 1997; 21: 1081-1094.

psychosocial factors and DV behavior of the perpetrators.

REFERENCES 1. 2. Marwick C. Domestic violence recognized as world problem. JAMA 1998; 279: 1510. Garcia-Moreno C, Heise L, Jansen HA, Ellsberg M, Watts C. Public health: violence against women. Sci 2005; 25: 310: 1282-1283. Krug EG, Dahlberg LL, Mercy JA, Zwi AB, Lozano R, eds. World report on violence and health. Geneva: World Health Organization; 2002: 100-103. Kornblit AL. Domestic violence-an emerging health issue. Ser Sci Med 1994; 39: 1181-1188. Schafer J, Caetano R, Clark CL. Rates of intimate partner violence in the United States. Am J Public Health 1998; 88: 1702-1704. Sorenson SB, Upchurch DM, Shen H. Violence and injury in marital arguments: risk patterns andgender differences. Am J Public Health 1996; 86: 35-40. Grisso JA, Wishner AR, Schwarz DF, Weene BA, Holmes JH, Sutton RL. A population-based study of injuries in inner-city women. Am J Epidemiol 1991; 134: 59-86. Cao YP, Zhang YL, Sun SQ, Guo GY, Li YC, Yuan D, et al. An epidemiological study on domestic violence in Hunan, China. Chin J Epidemiol (Chin) 2006; 27: 200-203. Ma Y, ed. A census data of population in Hunan Province in 2002 (Chin). Beijing: Statistic Publishing House of China; 2002. Pearson V, Phillips MR, He FS, Ji HY. Attempted suicide among young rural women in the Peoples Republic of China: possibilities for prevention. Suicide Life Threat Behav 2002; 32: 359-369. Xu X, Zhu F, OCampo P, Koenig MA, Mock V, Campbell J. Prevalence of and risk factors for intimate partner violence in China. Am J Public Health 2005; 95: 78-85. Tang CS, Lai BP. A review of empirical literature on the prevalence and risk markers of male-on female intimate partner violence in contemporary China, 1987-2006. Aggress Violence Behav 2008; 13: 10-18 Stuart GL, Meehan JC, Moore TM, Morean M, Hellmuth J, Follansbee K. Examining a conceptual framework of intimate partner violence in men and women arrested for domestic violence. J Stud Alcohol 2006; 67: 102-112. Robinson GE. Current concepts in domestic violence. Prim Psychiatry 2003; 10: 48-52. Cao YP, Zhang YL, Chang DF. A Comparison study of domestic violence in Hunan, China. Yale-China Health J 2006; 4: 27-43. Cao YP, Zhang YL, Wang GQ, Yang SC, Zhang Y, Huang GP. A case-control study of mental health status of perpetrators with domestic violence. Chin J Psychiatry (Chin) 2008; 41: 37-40. Chen CH. Psychological assessments related to stress. In: Shen YC, eds. Psychiatry, 5th ed (Chin). Beijing: Peoples Medical Publishing House; 2009: 269-273. Derogatis LR, Lipman RS, Covi L. SCL-90: an outpatient 19. 20.

21. 22. 23.

3.

4. 5.

24.

6.

25.

7.

26.

8.

9.

27.

10.

28.

11.

29.

12.

30.

13.

31.

14. 15.

32.

16.

33.

17.

18.

(Received July 27, 2010) Edited by PAN Cheng