Skip Navigation


American Journal of Epidemiology Advance Access originally published online on October 10, 2006
American Journal of Epidemiology 2006 164(12):1199-1208; doi:10.1093/aje/kwj339
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
164/12/1199    most recent
kwj339v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (7)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Coid, J.
Right arrow Articles by Singleton, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Coid, J.
Right arrow Articles by Singleton, N.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

American Journal of Epidemiology Copyright © 2006 by the Johns Hopkins Bloomberg School of Public Health All rights reserved; printed in U.S.A.

ORIGINAL CONTRIBUTIONS

Violence and Psychiatric Morbidity in a National Household Population—A Report from the British Household Survey

Jeremy Coid1, Min Yang1, Amanda Roberts1, Simone Ullrich1, Paul Moran2, Paul Bebbington3, Traolach Brugha4, Rachel Jenkins2, Michael Farrell2, Glyn Lewis5 and Nicola Singleton6

1 Forensic Psychiatry Research Unit, Queen Mary College, University of London, London, United Kingdom
2 Institute of Psychiatry, King's College, London, United Kingdom
3 Department of Psychiatry and Behavioural Science, Royal Free and University College Medical School, London, United Kingdom
4 Department of Psychiatry, University of Leicester, Leicester, United Kingdom
5 Division of Psychiatry, University of Bristol, Bristol, United Kingdom
6 Home Office, London, United Kingdom

Correspondence to Dr. Min Yang, Forensic Psychiatry Research Unit, Queen Mary College School of Medicine, University of London, 61 Bartholomew Close, William Harvey House, London EC1A 7BE, United Kingdom (e-mail: m.yang{at}qmul.ac.uk).

Received for publication December 22, 2005. Accepted for publication May 9, 2006.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
This study measured the prevalence of self-reported violence and associations with psychiatric morbidity in a national household population, based on a cross-sectional survey in 2000 of 8,397 respondents in Great Britain. Diagnoses were derived from computer-assisted interviews, with self-reported violent behavior over the previous 5 years. The 5-year prevalence of nonlethal violence in Britain was 12% (95% confidence interval: 11, 13). The risk of violence was substantially increased by alcohol dependence (odds ratio = 2.72, 95% confidence interval: 1.85, 3.98), drug dependence (odds ratio = 2.63, 95% confidence interval: 1.45, 4.74), and antisocial personality disorder (odds ratio = 6.12, 95% confidence interval: 3.87, 9.66). Low prevalences of these conditions (7%, 4%, and 4%, respectively) contrasted with their relatively high proportions of attributed risk of violence (23%, 15%, and 15%). Hazardous drinking was associated with 56% of all reported violent incidents. Screening positive for psychosis did not independently increase risk (odds ratio = 3.20, 95% confidence interval: 0.35, 29.6). The study concluded that psychiatric morbidity makes a significant public health impact on violence exerted primarily by persons with any personality disorder, substance dependence, and hazardous drinking. Population interventions for violent behavior are appropriate for hazardous drinking as are targeted interventions for substance dependence and antisocial personality disorder. Despite public concern, the risks of violence from persons with severe mental illness were very low.

antisocial personality disorder; cross-sectional studies; morbidity; risk; substance-related disorders; violence


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
The public health impact of mental disorder on violence depends on the base rate of violence in the general population. This may ultimately influence whether targeted "high-risk" or large-scale "population" strategies are chosen for violence prevention (1). For example, international homicide rates show wide variations among different countries, but rates among individuals with mental disorders are similar between countries and within the same countries over time (2). In geographic locations with low violence rates, the proportion attributed to mentally disordered persons may appear high, and containing their violence will achieve public health and political prominence. In locations with high base rates, more relevant risk factors may include weapon availability, substance misuse, and gang violence. Nevertheless, there is consensus that mental disorder is related to violence (3–8) and increases risk of violence over the life span (9–13). However, patients with psychotic, affective, and anxiety disorders have only moderately increased risks compared with the general population, with considerably greater risks from personality disorder (5, 14), substance misuse (5), and where these conditions are comorbid with personality disorder and substance misuse (3, 6–8).

We measured the prevalence of self-reported violent behavior over the past 5 years, its association with individual categories of mental disorder, comorbidity, and the public health impact of psychiatric morbidity, using the population attributable risk in a two-phase survey of a representative sample of adults (aged 16–74 years) in households in Britain, conducted in 2000.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Sample
Subjects aged 16–74 years were sampled in the survey of Psychiatric Morbidity among Adults Living in Private Households in England, Wales, and Scotland in 2000. Details have been described previously (15). Computer-assisted interviews were carried out in this two-phase survey (16) by Office of National Statistics interviewers. The small-user Postcode Address File was the sampling frame, and the grid method of Kish (17) systematically selected one person in each household.

A total of 8,886 adults completed the first-phase interview, a response rate of 69.5 percent. Of these, 8,397 (94.5 percent) completed all questionnaire sections. Among nonrespondents, 24.0 percent were refusals and 6.5 percent were noncontacts. There was no information on the psychiatric status of nonrespondents to conclude whether attrition resulted in biased estimates in the prevalence of violence. However, weighting procedures applied throughout the analyses took into account the proportions of nonrespondents according to age, sex, and region to ensure a sample representative of the national population, compensating for sampling design and nonrespondents in the standard error of the prevalence and controlling for the effects of selecting one individual per household.

Measurement of psychiatric morbidity
Participants screened positive for psychosis if any two of four criteria were currently present from the Psychosis Screening Questionnaire (18). With criteria from the Diagnostic and Statistical Manual of Mental Disorders: DSM-IV, the screening questionnaire for the Structured Clinical Interview for Axis II (known as the "SCID-II screening questionnaire") (19) identified this personality disorder. Subjects gave "yes" or "no" responses to 116 questions on laptop computers. Ten categories of Axis II disorder were created by manipulating cutoff points to increase levels of agreement, measured by the kappa coefficient, between both individual criteria and clinical diagnoses. These were obtained using the Structured Clinical Interview administered by trained interviewers in a previous survey of prisoners (20). The same algorithms were used in the present survey. Ten categories of lifetime personality disorder could be derived from the screening instrument but were combined into a single category of "any" personality disorder for this study. For some analyses, participants with antisocial personality disorder were analyzed separately.

The revised version of the Clinical Interview Schedule (known as "CIS-R") (21) was used to obtain the prevalence of common mental disorders in the past week (affective and anxiety disorders), including generalized anxiety disorder, mixed anxiety and depression, depressive episodes, phobias, obsessive-compulsive disorder, and panic disorder. These six syndromes were combined into a single category of "any" affective or anxiety disorder. The principal instrument to assess alcohol misuse over the past year was the Alcohol Use Disorders Identification Test (known as "AUDIT"), which defines hazardous alcohol use (score of 8 or more) as an established pattern of drinking that brings risk of physical and psychological harm over the previous year (22). The Severity of Alcohol Dependence Questionnaire (known as "SADQ") (23) was included to measure alcohol dependence. A number of questions designed to measure drug use were included in the phase I interviews. Positive responses regarding a series of different substances (cannabis, amphetamines, cocaine, crack cocaine, ecstasy, tranquilizers, opiates, and volatile substances) to any of five questions measuring drug dependence over the past year were combined to produce a single category of "any" drug dependence (15).

A category of "no psychiatric disorder" was applied to respondents who did not have personality disorder, affective or anxiety disorder, drug or alcohol dependence, or possible psychosis.

Measurement of violent behavior
All subjects were asked questions about violent behavior in the first phase of the study in the context of establishing the diagnosis of antisocial personality disorder. These included questions from the conduct disorder section, including whether they had started fights and had threatened or hurt anyone with a weapon before the age of 15 years. In addition, they were asked if they had been in a fight and had used a weapon in a fight since the age of 15 years. As we intended to retain the diagnostic category antisocial personality disorder in subsequent analyses, we included an additional question used in previous surveys in New York (4) and Israel (5). Subjects were asked, "Have you been in a physical fight, assaulted, or deliberately hit anyone in the past 5 years?"

Statistical analysis
To estimate the prevalence of violent behavior in the population of Great Britain, we used weights to account for unequal selection of probabilities in the two-phase sample survey. Detailed procedures for constructing weighting variables were reported by Singleton et al. (24). Based on the second-phase sample, comparisons between unweighted and weighted prevalence of personality disorders showed considerable differences. Weighted results are a more accurate representation of the general population, and weighted analyses were therefore performed throughout.

To measure violent behavior by demography and diagnostic categories, we carried out cross-tabulation with weighting factors in SPSS, version 12, software (SPSS, Inc., Chicago, Illinois). Adjusted odds ratio estimates for violent behavior by demographic factors and clinical syndromes were ascertained by use of two-level hierarchical weighted logistic regression analyses in MLwiN software (25), which takes into account the clustering effects of violent behavior within survey areas. Adjustments were mainly for the "static" attributes of respondents, including age, gender, marital status, and social class. The adjusted model was fitted for each diagnostic category entered in comparison with the category "no psychiatric disorder." This established the magnitude of associations between violent behavior and each diagnostic category, controlling for static factors and other clinical syndromes.

Population attributable risk was calculated for each diagnostic category and some comorbid conditions (26). In the absence of an incidence of violent behavior due to the cross-sectional method, relative risk was approximated by the odds ratio together with the representative prevalence in the total population.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Prevalence of violent behavior
Weighted data excluding missing data included 8,397 respondents (4,179 men, 4,278 women), of whom 982 (12 percent) reported violent behavior in the last 5 years. The prevalence of self-reported violence among men (n = 749, 18 percent) was three times that among women (n = 233, 6 percent). As expected, the prevalence of reporting fighting across the life span since 16 years was higher in all respondents (n = 2,148, 26 percent) and, among males, was more than 4.5 times (n = 1,771, 42 percent) that of women (n = 377, 9 percent). A total of 155 (2 percent) reported using a weapon in a fight since 16 years of age: 137 (3 percent) men and 18 (0.4 percent) women. A total of 566 (7 percent) reported starting fights before the age of 15 years: 416 (10 percent) men and 150 (4 percent) women. A total of 267 (3 percent) reported threatening or hurting someone with a weapon before 15 years of age: 231 (6 percent) men and 36 (1 percent) women.

Table 1 demonstrates that being young, male, of lower socioeconomic status, single, separated, divorced, and in part-time work or economically inactive were independently associated with self-reporting violent behavior. Violence was less prevalent among persons of Asian origin and declined progressively with age.


View this table:
[in this window]
[in a new window]

  		TABLE 1 Prevalence of violence in the last 5 years by demographic characteristics and odds ratios adjusted for each other by logistic regression analysis, England, Wales, and Scotland, 2000

 
Diagnostic categories and violence
Of 982 respondents reporting violent behavior in the preceding 5 years, 644 (66 percent) met survey criteria for any psychiatric disorder, compared with 2,767 (37 percent) nonviolent respondents (odds ratio = 3.19, 95 percent confidence interval: 2.77, 3.67; p < 0.001). Table 2 demonstrates that violence among men was approximately twice that in women for most diagnostic categories, except antisocial personality disorder and psychosis where prevalences were similar in women. However, the prevalence for men was almost four times that of women with "no disorder."


View this table:
[in this window]
[in a new window]

  		TABLE 2 Prevalence of respondents reporting violent behavior in the last 5 years by number and categories of self-reported diagnoses, as well as adjusted odds ratios from logistic regression analysis, England, Wales, and Scotland, 2000*

 
Prevalences of any affective/anxiety disorder, any personality disorder, and screening positive for psychosis were more than twice as high in violent male respondents. However, in women, the prevalences of violence were higher among those with any affective/anxiety disorder, personality disorder, and those who screened positive for psychosis than among women with no disorder. Although the prevalences of hazardous drinking, alcohol dependence, drug dependence, and antisocial personality disorder were lower among women, the prevalence of reported violence among women with these conditions was higher than that in men.

Comorbidity and violence
The effect of multiple diagnoses on the risk of violence is demonstrated in table 2. There is a positive association between the number of diagnoses and the prevalence of violence, with prevalences approximately doubling as the number of diagnoses increases from none through three or more. Table 3 also shows the relation between specific combinations of diagnostic categories, suggesting that the effects of substance dependence on reported violence when comorbid with affective/anxiety disorder or any personality disorder were greater than those of comorbid combinations of these two conditions. However, the table also demonstrates that substance dependence did not entirely explain the association between mental illness and violence, as the prevalence of violence over the past 5 years was elevated among respondents with affective/anxiety disorder, personality disorder, or the combination of the two in the absence of substance dependence comorbidity. The elevated prevalence of reported violence in the combination of psychosis and substance dependence diagnosis was not significant because of the small number of respondents.


View this table:
[in this window]
[in a new window]

  		TABLE 3 Weighted logistic regression analysis of associations between psychiatric morbidity (self-reported diagnosis) and violence in the past 5 years, England, Wales, and Scotland, 2000

 
Multivariate analysis
Table 3 demonstrates the independent effects of psychiatric morbidity on self-reported violence in the past 5 years, controlling for demography and the effects of confounding from comorbid diagnoses. Compared with the no psychiatric disorder category, all diagnostic categories were associated with increased risks of reporting violence except screening positive for psychosis, which was not significant after adjustments. The highest risk of violence was associated with antisocial personality disorder. This independently increased the risks over six times compared with persons with no psychiatric disorder.

Public health impact of psychiatric morbidity on violent incidents/individuals
Table 4 demonstrates the potential public health impact of psychiatric morbidity on violent events and violent individuals. This is shown in two ways: 1) by the population attributable risk percentage, which is the proportion in the population that could be prevented by eliminating exposure to the risk factor of each category of mental disorder; 2) by the number of violent events reported by the respondents in each diagnostic category. This represents a simple measure of the proportion of violent incidents accounted for by respondents within each category or that might have been removed from the total had the sample not been exposed to each risk factor.


View this table:
[in this window]
[in a new window]

  		TABLE 4 Public health impact of diagnoses on violent incidents and violent individuals in the last 5 years, England, Wales, and Scotland, 2000*

 
The majority of respondents did not have a psychiatric diagnosis and accounted for almost a third of all violent incidents. Diagnoses with relatively high prevalences in the population, such as affective/anxiety disorder and personality disorder, accounted for relatively large proportions of all violent incidents. This was particularly the case for hazardous drinking, accounting for over half of all incidents. In contrast, subjects screening positive for psychotic illness constituted a very small percentage of respondents with psychiatric morbidity and accounted for a very small percentage of all violent incidents. However, respondents with alcohol dependence, drug dependence, and antisocial personality disorder, while of relatively low prevalence in the population, accounted for a relatively high proportion of all incidents.

The population attributable risk for each diagnostic category in table 4 demonstrates that eliminating psychosis as a risk factor for violence among persons with this diagnosis would have had an almost negligible impact on the percentage of individuals reporting violence in the past 5 years. However, eliminating hazardous drinking would have reduced it by almost one half. Eliminating affective/anxiety disorder or any personality disorder would have had a relatively low impact on individuals reporting violence despite these conditions having high prevalences. However, eliminating less prevalent risk factors, such as alcohol and drug dependence and antisocial personality disorder, would have had a moderate impact. Although the comorbidity of substance dependence, any personality disorder, and affective/anxiety disorder was associated with a high risk of violence, these diagnoses had moderate public health impact (table 4) because of their low prevalences.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
Comparison with previous surveys
The survey demonstrated that psychiatric morbidity among adults living in households in Britain increases the risk of violent behavior, replicating previous survey findings in the United States and Israel (3–5). However, it adds to the current body of knowledge in this area by quantifying the public health impact and indicates future approaches to intervention. Use of illicit drugs, hazardous drinking, personality disorders, and affective/anxiety disorders all increased risk, as in previous studies. However, screening positive for psychosis was not independently associated with violence after controlling for demographic factors and comorbidity. Alcohol and drug dependence and antisocial personality substantially increased the risk of reporting violence. In addition, diagnostic comorbidity substantially increased this risk, with a doubling of prevalence at each stage from no diagnosis to three diagnostic categories or more. This almost exactly replicated the earlier findings of Swanson et al. (3) in the United States who used different diagnostic categories derived from clinical interviews.

The survey also suggests that nonlethal violence is a problem similar in magnitude in Britain and the United States. A weighted national prevalence of 12.2 percent in persons aged 19–59 years is comparable to 15.1 percent in a predominately working-class population in Upper Manhattan using similar measures (4). In contrast, young persons in Israel reported a prevalence of only 5.2 percent (5), compared with 17.4 percent for respondents in the same age group in Britain. In the Epidemiological Catchment Area study at three US sites 15 years earlier, reported fighting before the age of 15 years was higher in US respondents but lower in adulthood than it was in the present survey (27). Similarities in the rates of nonlethal violence are supported by surveys of crime victims and police statistics, which demonstrated a fall in crime in the United States until 1996 compared with a rise in England and Wales (28). Thus, although the murder rate remains markedly higher in the United States, crime rates for assault and robbery are slightly higher in England and Wales (29). Taken together, these findings suggest that the base rate of violence is relatively high in Britain and that factors other than psychiatric disorder make the highest impact.

As in previous surveys, being young, male, single, separated, divorced, and economically inactive substantially increased the risk of reporting violence, irrespective of psychiatric illness. Belonging to an ethnic minority did not increase risk after controlling for social class and other confounders, although persons of Asian or Oriental origin were less likely to report violence, particularly women. However, the association between the measure of socioeconomic class used in this survey did not show the linear relation demonstrated by Swanson et al. (3) using a composite score based on occupation, education, and income ranking. This could reflect differences in measures but also different patterns of violence among young males in Britain, where those in social classes IIINM (skilled nonmanual) and IIIM (skilled manual) engage in levels of violence similar to those in social classes IV (semiskilled) and V (unskilled) and are violent in settings associated with heavy drinking.

Among subjects with mental disorder, women demonstrated lower prevalences of reported violence than did men in every diagnostic category. However, when compared with respondents with no disorder, women with psychiatric morbidity had increased risks of violence compared with men. Previous authors (10, 12) have also demonstrated that major mental disorder is associated with a greater increase in risk of criminal offenses for women than for men. Theoretical explanations include the possibility that women are more vulnerable to the effects of psychiatric illness on their behavior and the threshold/paradox hypothesis that argues that females who develop antisocial behavior surmount a threshold of risk higher than that of males and are therefore more severely afflicted (30).

Psychotic, affective, anxiety disorders and violence
We found that a combined category of affective and anxiety disorder independently increased the risk of reporting violence. This contrasted with a survey in Israel that also controlled for other disorders and demography (5). Violence attributable to persons screening positive for psychotic illness had almost negligible impact on the overall level of violence at the population level despite public concern over risks from seriously mentally ill persons, in both Europe and the United States. Furthermore, their risk was not increased independently above the risk of persons with no disorder in this study. This was unexpected, as previous population studies have found associations between psychotic disorders and violent behavior. This may have resulted from study limitations: first, using a screen to identify participants with psychosis; second, a low prevalence of respondents screening positive for psychosis compared with other diagnostic categories; third, the sampling frame exclusion of prisoners, the homeless, and psychiatric inpatients who have higher prevalences of violent behavior.

A strength of the study was elimination of potential confounders. This resulted in failure to demonstrate an independent association with psychotic illness (while confirming it for other diagnostic categories). Previous population surveys have demonstrated that psychiatric patients (4) and those with schizophrenia and manic-depressive psychosis (3, 5, 8) are more likely to report violence after controlling for demography. However, not all studies controlled for antisocial personality disorder (3, 4, 8) or substance abuse (4). A case register study in Australia found greatly reduced associations between schizophrenia and violent convictions after controlling for substance misuse (31). Four birth cohort studies have demonstrated associations among schizophrenia, schizophrenic spectrum disorder, manic-depressive psychosis, and violent offending (9–13). However, only one controlled for both personality disorder and substance abuse (12), and only Stueve and Link (5) controlled for additional confounding from other nonpsychotic, Axis I disorders. These studies confirmed independent associations, but in populations with low base rates of violence.

Antisocial personality disorder
A targeted approach to individuals with substance dependence and antisocial personality disorder was partly supported by a relatively large percentage of violence among people with these conditions, but the relatively small proportion of the population with these diagnoses. Although national surveys have demonstrated prevalence ranges from 0.6 to 4 percent, persons with antisocial personality disorder constitute a large number of individuals beyond the resources of the criminal justice system or mental health services. The finding that approximately half do not report violence indicates problems of accuracy in identifying those posing future risks. Furthermore, evidence on the effectiveness of mental health services to intervene once these conditions are established in adulthood remains limited.

Methodological limitations
There are several study limitations. Violent behavior within the last 5 years was assessed via self-report. This measure was restricted, and we did not include objective information such as arrests or convictions. Moreover, self-report may have underestimated true prevalence, as socially undesirable behaviors tend to be less frequently reported. Diagnoses of Axis I and Axis II mental disorders were also derived from self-report questionnaires in the first phase of the survey. Research diagnostic instruments were administered by clinically trained raters in the second phase. However, the second-phase sample was considerably smaller, and prevalences were insufficient for detailed statistical analyses. Specific limitations may lie in the use of self-report assessments of personality disorder (32), although it has been questioned whether clinicians' assessments are by definition superior (33). Some clinical reappraisal interviews may be of value. As described elsewhere (15), a stratified second-phase sample of over 600 respondents was assessed by clinical interviewers using the SCID-II screening questionnaire (34). Comparison of clinical and self-report diagnostic categories demonstrated very good specificity and sensitivity but poor positive predictive value for clinician-rated categories, indicating an area for future development of feasible epidemiologic assessment methods (35). The 31.5 percent of nonresponders were less likely to be White and more likely of lower social class and lower educational level. This could introduce bias through underestimating the true prevalence. As the differences in violent behavior among ethnic groups were not significant and the weighting procedure took into account certain nonresponse factors, the underestimation bias may not be severe, and our findings regarding risk factors should remain valid.

Prevalences of mental disorders in the only comparable previous survey in Great Britain (36) demonstrated close similarities, except substance misuse categories which had substantially increased. Data were weighted by poststratification to national population totals to compensate for known differences in response by age, sex, and geographic region. Social class was associated with prevalence of violence. However, distributions of occupational groupings for those who had ever worked in our survey were similar to those currently working in the 2001 national census. This suggested that there were no major biases with respect to social class within the survey data.

The dating of episodes of mental disorder proved difficult, and it was not identified whether violent incidents related to time periods when symptoms were present. Apart from antisocial personality disorder and drug dependence, the number of individuals reporting violence was relatively small. This might have complicated the statistical analyses and should be considered when interpreting results. However, the community-based design and large sample size allowed us to examine associations between different categories of mental disorders and violent behavior without introducing the selection bias associated with treated samples. Furthermore, the sample size provided sufficient statistical power to test complex models and to control for confounding from demographics and comorbidity.

Public health implications of alcohol, drug misuse, and violence
The public health approach to violence has generated more interest in the United States (37) than the United Kingdom (38). In the former, homicide is the second leading cause of death for persons aged 15–34 years, the leading cause of death for young African Americans, and where the average ages of both violent offenders and victims have become progressively younger (39, 40). However, the high annual medical and social costs of injury from deliberate harm are highlighted by measures from emergency rooms in the United Kingdom. These correlate with unemployment, poverty and, in particular, expenditure on alcohol (41, 42). Alcohol misuse and its relation to violence, particularly binge-drinking, have generated increasing public and political concern in the United Kingdom following a marked increase in licensed premises selling alcohol over the past 25 years (43) and legislative changes relaxing the selling of alcohol. Research into alcohol-related disorder highlights the concentration of violent and public order offenses in urban areas with high densities of licensed premises that peak at weekends. This has emerged within planned regeneration of certain inner-urban areas in the United Kingdom but, where there is competition among licensed premises designed to accommodate large numbers of drinkers, resulting in cheaper alcohol, the financial resources available to the alcohol industry in the United Kingdom (facilitating more effective litigation to overcome objections of residents and regulations of local authorities), coupled with an inherent culture of binge-drinking, have compounded these problems (44).

In this survey, the highest percentage of incidents and the highest population attributable risk were explained by individuals engaging in hazardous drinking, followed by drug misuse. Measures of the public health impact used in this study assume a direct association between diagnostic categories and violent behavior, which cannot be verified in a cross-sectional survey. Furthermore, associations between drug misuse and violence may result more from involvement in the illegal economy of drug markets than from the effects of intoxication (45), and where criminal justice control of drug use through law enforcement outweighs public health interventions. However, studies of alcohol use and violence have confirmed strong, if complex, associations with hazardous drinking (46). As the proportion of respondents reporting hazardous drinking in Britain was substantial, particularly among younger men, this indicates that "population" approaches involving risk-reduction programs to encourage healthy drinking and control of outlets, particularly those associated with drunken disorder, many within the "night-time economy" (44), are more appropriate preventive interventions (46). By use of Rose's model (1), a relatively small reduction in exposure to the risk factor of hazardous drinking at the individual level (which affects a relatively large proportion of the population) could result in a relatively large overall impact on the population's behavior in association with drinking.


   ACKNOWLEDGMENTS
 
Funded by the Department of Health (England), with contracts to the Office for National Statistics, Social Survey Division, London, which carried out fieldwork.

Conflict of interest: none declared.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Rose G. (1992) The strategy of preventive medicine. (Oxford University Press, Oxford, United Kingdom).
  2. Coid J. (1983) The epidemiology of abnormal homicide and murder followed by suicide. Psychol Med 13:855–60.[Web of Science][Medline]
  3. Swanson JW, Holzer CE, Ganju VK, et al. (1990) Violence and psychiatric disorder in the community: evidence from the Epidemiologic Catchment Area surveys. Hosp Community Psychiatry 41:761–70.[Abstract/Free Full Text]
  4. Link BG, Andrews HA, Cullen FT. (1992) The violent and illegal behavior of mental patients reconsidered. Am Sociol Rev 57:275–92.[CrossRef][Web of Science]
  5. Stueve A and Link BG. (1997) Violence and psychiatric disorders: results from an epidemiological study of young adults in Israel. Psychiatr Q 68:327–42.[CrossRef][Web of Science][Medline]
  6. Wallace C, Mullen P, Burgess P, et al. (1998) Serious criminal offending and mental disorder. Case linkage study. Br J Psychiatry 172:477–84.[Abstract/Free Full Text]
  7. Mullen P, Burgess P, Wallace C, et al. (2000) Community care and criminal offending in schizophrenia. Lancet 355:614–17.[CrossRef][Web of Science][Medline]
  8. Corrigan PW and Watson AC. (2005) Findings from the National Comorbidity Survey on the frequency of violent behaviour in individuals with psychiatric disorders. Psychiatry Res 136:153–62.[CrossRef][Web of Science][Medline]
  9. Hodgins S. (1992) Mental disorder, intellectual deficiency, and crime. Evidence from a birth cohort. Arch Gen Psychiatry 49:476–83.[Abstract/Free Full Text]
  10. Hodgins S and Janson CG. (2002) Criminality and violence among the mentally disordered. (Cambridge University Press, Cambridge, United Kingdom).
  11. Tiihonen J, Isohanni M, Rasanen P, et al. (1997) Specific major mental disorders and criminality: a 26-year prospective study of the 1966 northern Finland birth cohort. Am J Psychiatry 154:840–5.[Abstract]
  12. Brennan PA, Mednick SA, Hodgins S. (2000) Major mental disorders and criminal violence in a Danish birth cohort. Arch Gen Psychiatry 57:494–500.[Abstract/Free Full Text]
  13. Arseneault L, Moffitt TE, Caspi A, et al. (2000) Mental disorders and violence in a total birth cohort: results from the Dunedin Study. Arch Gen Psychiatry 57:979–86.[Abstract/Free Full Text]
  14. Steadman HJ, Mulvey EP, Monahan J, et al. (1998) Violence by people discharged from acute psychiatric inpatient facilities and by others in the same neighborhoods. Arch Gen Psychiatry 55:393–401.[Abstract/Free Full Text]
  15. Singleton N, Bumpstead R, O'Brien M, et al. (2001) Psychiatric morbidity among adults living in private households, 2000. (The Stationery Office, London, United Kingdom).
  16. Shrout PE and Newman SC. (1989) Design of two-phase prevalence surveys of rare disorders. Biometrics 45:549–55.[CrossRef][Web of Science][Medline]
  17. Kish L. (1965) Survey sampling. (Wiley & Sons, London, United Kingdom).
  18. Bebbington P and Nayani T. (1994) The Psychosis Screening Questionnaire. Int J Methods Psychiatr Res 5:11–19.
  19. First MB, Gibbon M, Spitzer RL, et al. (1997) Structured clinical interviews for DSM-IV Axis-II personality disorders. (American Psychiatric Press, Washington, DC).
  20. Singleton N, Meltzer H, Gatward R, et al. (1998) Psychiatric morbidity among prisoners in England and Wales. (The Stationery Office, London, United Kingdom).
  21. Lewis G, Pelosi AJ, Araya RC, et al. (1992) Measuring psychiatric disorder in the community: a standardised assessment for lay-interviewers. Psychol Med 22:465–86.[Web of Science][Medline]
  22. Babor TF, de la Fuente JR, Saunders J, et al. (1992) AUDIT: the Alcohol Use Disorders Identification Test: guidelines for use in primary healthcare. (World Health Organization, Geneva, Switzerland).
  23. Stockwell T, Murphy D, Hodgson R. (1983) The Severity of Alcohol Dependence Questionnaire: its use, reliability and validity. Br J Addict 78:145–55.[Web of Science][Medline]
  24. Singleton N, Lee A, Meltzer H. (2002) Psychiatric morbidity among adults living in private households, 2000: technical report. (Office for National Statistics, London, United Kingdom)19–26.
  25. Rasbash J, Browne W, Goldstein H, et al. (2000) A user's guide to MlwiN. (Institute of Education, University of London, London, United Kingdom).
  26. Woodward M. (2004) Epidemiology, study design and data analysis. 2nd ed (Chapman & Hall, London, United Kingdom)146–52 283–5.
  27. Robins LN, Tipp J, Przybeck T. (1991) Antisocial personality. In Robbins LN and Regier DA (Eds.). Psychiatric disorder in America(Free Press, New York, NY) pp. 258–90.
  28. Dodd T, Nicholas S, Povey D, et al. (2004) Crime in England and Wales 2003 /2004. (Home Office, London, United Kingdom) (Home Office statistical bulletin 10/04).
  29. Langan PA and Farrington DP. (1998) Crime and justice in the United States and in England and Wales, 1981–96. (Bureau of Justice Statistics, Office of Justice Programs, US Department of Justice, Washington, DC).
  30. Moffitt TE, Caspi A, Rutter M, et al. (2001) Sex differences in antisocial behaviour: conduct disorder, delinquency, and violence in the Dunedin Longitudinal Study. (Cambridge University Press, Cambridge, United Kingdom).
  31. Wallace C, Mullen PE, Burgess P. (2004) Criminal offending in schizophrenia over a 25-year period marked by deinstitutionalization and increasing prevalence of comorbid substance use disorders. Am J Psychiatry 161:716–27.[Abstract/Free Full Text]
  32. Zimmerman M. (1994) Diagnosing personality disorders. A review of issues and research methods. Arch Gen Psychiatry 51:225–45.[Abstract/Free Full Text]
  33. Wittchen HU, Ustun B, Kessler RC. (1999) Diagnosing mental disorders in the community: a difference that matters? Psychol Med 29:1021–7.[CrossRef][Web of Science][Medline]
  34. First MB, Gibbon M, Spitzer RL, et al. (1996) User's guide for the Structured Clinical Interview for DSM-IV Axis II personality disorders (SCID-II, version 2.0). (Biometrics Research Department, New York State Psychiatric Institute, New York, NY).
  35. Taub NA, Morgan Z, Brugha TS, et al. (2005) Recalibration methods to enhance information on prevalence rates from large mental health surveys. Int J Methods Psychiatr Res 14:3–13.[CrossRef][Web of Science][Medline]
  36. Meltzer H, Gill B, Pettigrew M, et al. (1995) OPCS surveys of psychiatric morbidity in Great Britain report 1: the prevalence of psychiatric morbidity among adults living in private households. (The Stationery Office, London, United Kingdom).
  37. Winett LB. (1998) Constructing violence as a public health problem. Public Health Rep 113:498–507.[Web of Science][Medline]
  38. Stanistreet D. (1999) Violence: developing a policy agenda. J Epidemiol Community Health 53:2–3.[Web of Science][Medline]
  39. White JH. (1994) Violence: a public health epidemic. Health Prog 75:18–21.[Medline]
  40. Rosenberg ML, Powell KE, Hammond R. (1997) Applying science to violence prevention. JAMA 277:1641–2.[Abstract/Free Full Text]
  41. Hutchinson IL, Magennis P, Shepherd JP, et al. (1998) The BAOMS United Kingdom survey of facial injuries, part 1: aetiology and the association with alcohol consumption. Br J Oral Maxillofac Surg 36:3–13.[CrossRef][Web of Science][Medline]
  42. Sivarajasingam V, Shepherd J, Matthews K, et al. (2002) Trends in violence in England and Wales 1995 –2000: an accident and emergency perspective. J Public Health Med 24:219–26.[Abstract/Free Full Text]
  43. Office Home. (2002) Liquor licensing, England and Wales, July 2000–June 2001. (Home Office, London, United Kingdom).
  44. Hobbs D, Hadfield P, Lister S, et al. (2005) Violence and governance in the night-time economy. (Oxford University Press, Oxford, United Kingdom).
  45. Arsenault L, Moffitt TE, Caspi A, et al. (2002) The targets of violence committed by young offenders with alcohol dependence, marihuana dependence and schizophrenia-spectrum disorders: findings from a birth cohort. Crim Behav Ment Health 12:155–68.[CrossRef][Medline]
  46. Graham K, Leonard KE, Room R, et al. (1998) Current directions in research on understanding and preventing intoxicated aggression. Addiction 93:659–76.[CrossRef][Web of Science][Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 Br. J. PsychiatryHome page
M. Large, G. Smith, N. Swinson, J. Shaw, and O. Nielssen
Homicide due to mental disorder in England and Wales over 50 years
The British Journal of Psychiatry, August 1, 2008; 193(2): 130 - 133.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. PsychiatryHome page
J. Coid, M. Yang, A. Roberts, S. Ullrich, P. Moran, P. Bebbington, T. Brugha, R. Jenkins, M. Farrell, and N. Singleton
Authors' reply:
The British Journal of Psychiatry, February 1, 2007; 190(2): 177 - 178.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
164/12/1199    most recent
kwj339v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (7)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Coid, J.
Right arrow Articles by Singleton, N.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Coid, J.
Right arrow Articles by Singleton, N.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?