 |
 |
|
|
|
Road rage has recently appeared as a new problem for drivers in many countries, with reports coming from Australia (1), Canada (2), the UK (3), and the US (4). Newspaper reports on road rage are increasing in Canada (5) and the US (6,7). Reported cases of road rage increased by a factor of 15 in Canadian newspapers between 1996 and 2000 (5). Similarly, in the 1990s, annual newspaper reporting of road rage incidents in the US numbered in the thousands (7). There is no generally accepted definition of road rage, although it has been defined as “a situation where a driver or passenger attempts to kill, injure, or intimidate a pedestrian or another driver or passenger or to damage their vehicle in a traffic incident” (2). Only a few studies have examined the characteristics of road rage victims and perpetrators beyond their age and sex. Several studies have shown that angry drivers are more likely to be young, to be men, and to engage in more arguments and violent confrontations with other drivers (3,5). However, the source of anger and aggression among drivers is less well understood. While some research suggests that anger and aggressiveness increase owing to traffic congestion or the poor driving of others, the results of available studies have shown inconsistent evidence (8). Epidemiological research has linked anger and aggressive driving to psychiatric problems. Many types of violence are more common among those with psychiatric problems. For example, the Epidemiological Catchments Area studies in the US found that, of those reporting violent behaviour, 30% met the criteria for some psychiatric disorder (9). Whether psychiatric distress is, in fact, associated with road rage is an important question, with very little research done so far. Exposure to violence and threat can have long-lasting psychological effects, such as posttraumatic stress disorder and depression (10–12). Being a victim of road rage in its more serious forms could thus exert effects similar to exposure to other forms of violence and abuse (11,13) and could result in increases in self-reported depression and other problems among those exposed. However, this possibility has not yet been tested. Fong and others reported on psychiatric morbidity and road rage in a sample of 131 people in general practice clinics in England (14). They found that, in small samples of victims and perpetrators, scores on the Clinical Interview Schedule were higher for both perpetrators and victims of road rage than for control subjects, but there were no differences on the Screening Test for Comorbid Personality Disorders, Alcohol Problems scores, or the Life Events Schedule. Many road rage perpetrators display exceptionally high levels of aggression for little or no apparent reason. Recent clinical research (15,16) has assessed the presence of intermittent explosive disorder (IED) in aggressive drivers. IED refers to recurrent, problematic, impulsive, and aggressive behaviour that has been linked to genetic, biological, and epidemiologic correlates (17). This research demonstrates that aggressive drivers, particularly those drivers who have been court- referred for psychiatric treatment, are more likely to meet the criteria for IED as well as anxiety disorder (17). While road rage in the current context is more broadly defined to include less innocuous forms of aggression (such as shouting and verbal threats), it is apparent that more extreme forms of road rage may have important clinical implications. Existing research has treated road rage as a more or less unitary dimension among victims and perpetrators. In a previous study, we observed a substantial but not complete overlap between being a road rage victim and perpetrator (18). This observation suggests the existence of subgroups of road rage victims and perpetrators that may differ on important dimensions including indicators of psychiatric distress. Clinically significant subgroups or clusters of individuals have been found among convicted drinking drivers and alcohol abusers, among other groups (19,20). We report a study of road rage and psychiatric distress based on a large representative sample of adults in Ontario. The General Health Questionnaire (GHQ; 21,22) is used as an indicator of current psychiatric distress. We hypothesize that psychiatric distress scores would be higher for victims than perpetrators of road rage and that these differences would be even greater for those involved in more road rage incidents and for those involved in more serious cases of road rage. We develop a clustering of road rage cases based on the type and frequency of road rage behaviour to examine differences in mental health among different groups of people affected by road rage. MethodsThe data for this paper are drawn from the Centre for Addiction and Mental Health (CAMH) Monitor, a repeated cross-sectional telephone survey of Ontario adults conducted by the CAMH and administered by the Institute for Social Research at York University, Toronto. Each cycle of the CAMH Monitor is regionally stratified and consists of 12 independent monthly samples (January to December). Respondents are selected via random-digit–dialing methods with the help of computer-assisted telephone interviewing. Monthly sample sizes were between 212 and 240 respondents, with response rates that ranged from 56% to 61%, rates similar to recent Canadian household surveys. Overall, these data are representative of Ontario adults aged 18 years and over (22). Data from July 2001 to June 2002 are employed, with a total sample of 2610. Road rage indicators are adopted from a taxonomy of road rage behaviour developed by Smart and Mann (5). Two sets of 4 indicators are employed: one set directed at experiences of road rage victimization and the second set focusing upon road rage offending. These indicators quantify the frequency of involvement in progressively more severe forms of road rage behaviour, beginning with general expressions of anger and frustration directed at other drivers (for example, waving hands, gesturing, and shouting) to physical intimidation (for example, tailgating, cutting in and out, and blocking traffic on purpose), verbal threats, physical injury, damage to other vehicles, and death. A similar set of road rage items has been developed and successfully tested in empirical studies of road rage in the US (4,6). Descriptive statistics for these items are presented in Table 1.
The psychiatric distress indicator used is the 12-item version of the GHQ (21,22). It is a widely used scale useful in detecting nonpsychotic psychiatric illness and capturing psychological distress, anxiety, and social functioning in particular. A large number of studies have established the validity (24–27) and reliability of the GHQ when used in general population samples (28–30). We use the binary scoring system for the GHQ with the standard cut point of two-thirds for classifying people as showing symptoms of psychiatric distress (22,31). Six demographic measures (sex, age, employment status, marital status, educational attainment, and geographic locale) are adopted to control for variations in road rage and mental health. Age is a continuous measure; sex, employment status, and geographic locale are dichotomous measures; and marital status and educational attainment are categorical measures. Table 1 provides a demographic profile of survey respondents. Descriptive statistics presented in Table 1 are not weighted; however, bivariate (see Table 2) and cluster analysis (see Table 3) results are weighted to account for sampling and poststratification adjustments (23).
Data analysis occurs in 3 stages. First, a statistical overview of each road rage measure and the GHQ is provided, as well as a description of how each measure varies across demographic indicators (that is, sex, age, education, marital status, and urban vs rural). Next, an agglomerative clustering procedure is employed to develop a typology of road rage behaviour and to determine whether individuals form distinct groups based on their patterns of road rage involvement as either a victim or offender. Cluster analysis is used to identify distinct groups of relatively homogenous entities (32). As Aldenderfer and Blashfield note, “A clustering method is a multivariate statistical procedure that starts with a data set containing information about a sample of entities and attempts to reorganize these entities into relatively homogenous groups” (33, p 7). Finally, after the road rage typology is created, differences in GHQ reporting and demographic indicators are explored. Our goal is to assess whether types of road rage are linked with psychiatric distress and other demographic patterns. Multivariate analysis of variance (MANOVA) procedures are employed to tests for among-group differences for all indicators in the cluster analysis. The goodness of fit of the resulting cluster solution is confirmed with F-tests and other post hoc tests specific to cluster analysis.
ResultsLevels of Road Rage Reporting Compared with the mean level of educational attainment, those with less than a high school education are significantly less involved in road rage as either a victim or offender (39% and 25%), while a greater proportion of those who have completed university have been the victims of road rage (53%). Individuals with some postsecondary education also exhibit levels of road rage offending (38.1%) that are significantly greater than the mean. Similarly, full-time employment is associated with being a victim of road rage (49.0% vs 43.6%) and a road rage offender (37.3% vs 27.5%). Marital status and residential location are also linked to road rage. A greater proportion of respondents who have never been married report involvement in road rage as both a victim and offender, compared with those individuals who are married or previously married. Those residing in an urban locale experienced greater victimization (48.2% vs 35.2%) and are more involved in offending (33.7% vs 25.2%) than rural residents. The next 2 rows in Table 2 show the percentage of road rage offenders who have experienced victimization in the past year and the percentage of road rage victims who admit to road rage offending. In both instances, there is a substantial and significant overlap. Three-quarters of road rage victims admit to involvement in road rage offending in the past year, while more than one-half of road rage offenders have experienced victimization in the past year. The last row in Table 2 describes the GHQ levels of road rage victims and offenders. At a bivariate level, a greater proportion of individuals reporting psychiatric distress have experienced road rage victimization. Fifty-five percent of individuals reporting more than 3 symptoms on the GHQ have experienced road rage victimization in the previous year. There are no significant differences in road rage offending. Levels of Symptom Reporting A Typology of Road Rage Behaviour Structure-Seeking Phase Classification Phase Validation Phase The second largest cluster (n = 612) is the verbal victims (Cluster 2). Members of this cluster experience significantly greater verbal victimization and are significantly less involved in road rage offending. Their involvement with other types of road rage hovers around the sample mean, as does their mental health. Moreover, respondents in this cluster are situated around the mean for all demographic indicators. The verbal victim–offenders cluster (Cluster 4) is the third largest, with 336 individuals. Members of this cluster have elevated levels of verbal road rage as both a victim and offender but average or lower-than-average involvement in all other types of road rage, mental health, and demographic indicators. The verbal-threat offenders (Cluster 1) is the second smallest cluster with 237 members. These individuals report significantly greater levels of verbal road rage perpetration, either verbally or through threats to hurt others, relative to the mean. However, their experiences with other types of road rage are minimal. Members of this group are also significantly more likely to be men and younger (37.5 years), although there are no other demographic differences. The smallest cluster, with 69 members, contains the most serious offenders (referred to hereafter as the hard core road rage perpetrators). Respondents in this cluster are significantly different from other clusters in several ways. They report elevated levels of involvement in all forms or road rage offending and have experienced significantly higher rates of all types of road rage victimization. In particular, it is the only cluster to be significantly involved in the 2 most serious forms of road rage: attempting to damage other person’s car and attempting to hurt others. Moreover, this is also the only group to report significantly more experience with these same types of road rage as victims. A far larger proportion (27.5%) of this cluster score 3 or more on the GHQ scale, indicating substantial levels of psychiatric distress. This rate is more than double the sample average and more than 10% higher than the next nearest cluster (the verbal victims). Respondents in this cluster are more likely to be men and have the youngest average age (34.5 years). Moreover, they are the least likely to be married or living common-law. While they report the lowest mean level of educational attainment and are the most likely to live in an urban locale, these differences are not significant from the mean. Overall, the differing levels of psychiatric distress evident across clusters, as well as differences in sex, age, and marital status, provide a degree of external validity to the cluster solution. Moreover, repeated clustering of multiple subsamples from the data (not reported here) confirms the presence of a 5-cluster solution. The cluster solution represents a significant improvement over the null model that assumes no clusters are present. Additionally, the number of clusters is small, while the relative size of each cluster is not too small. Collectively, these criteria imply a good model fit for the data. DiscussionThe results of this study show that road rage is more likely for men, those aged 30 to 39 years, those with higher educational levels, those living in urban areas, and those never married. Those who were victims of road rage were more likely to report psychiatric distress on the GHQ, but there was no similar effect for road rage offending. A cluster analysis of road rage behaviour found 5 groups: those with little or no involvement, verbal-threat offenders, verbal victims, verbal victim– offenders, and hard core road rage perpetrators. Only the hard core road rage perpetrators had significantly higher scores (3 or greater) on the GHQ, indicating greater psychiatric distress in this group. In general, the results confirm our hypotheses. Psychiatric distress scores were high for road rage victims in the preliminary analyses and highest among those with the greatest involvement in road rage incidents (the hard core road rage cluster). Those with little or no involvement in road rage had the lowest GHQ scores, as expected. There was a suggestion that the verbal victims cluster also had high psychiatric distress, but the results were not significant. We note especially the GHQ results for the hard core road rage group. Their GHQ scores were far higher than for those with little or no involvement in road rage or other cluster groups. These results show clearly that those with a heavy involvement in road rage incidents have elevated levels of psychiatric distress. This confirms and extends the finding of Fong and others (14) and is also consistent with studies of violence and mental health (8,9). Although our correlational findings are of substantial interest, we cannot show that road rage directly results from, or directly affects, the mental health of drivers. Prospective investigations would be needed to examine these issues of cause and effect and to assess the impact of potentially important covariates such as driving exposure. Additionally, clinical research on aggressive drivers suggests that road rage, as defined here, may be part of a larger psychiatric phenomenon (14–17). The more extreme forms of road rage involving physical threats and violence, a common activity of the hard core road rage group, appear to fall within the criteria for IED. Thus, the current study demonstrates a relation between serious road rage and psychiatric distress, particularly as a response to road rage victimization, and it is possible that serious road rage is associated with the clinical disorder of IED. The GHQ questions deal mainly with anxiety, stress, and depression, and there are no questions on aggression or violence. Further research would be needed to assess violent tendencies among road rage perpetrators, with a special interest in the most serious offenders. It may be that road rage represents 1 manifestation of a general tendency toward violence and aggression. Therefore, more research is needed on other types of violence among road rage perpetrators (for example, domestic violence and assault) to assess the presence of a general violence syndrome. We should also note that the study reveals a relatively small sample (n = 69) of hard core road rage perpetrators, representing approximately 2.8% (69/2440) of the full sample. While these individuals may or may not be regular drivers, they have experienced road rage as a victim or offender. When extrapolated to the current population of Ontario residents (11 895 000 × 0.028 = 333 060), we see that the potential number of hard core road rage perpetrators in Ontario is quite large. Moreover, the observation that 27% of hard core road rage perpetrators experience elevated levels of psychiatric distress further stresses the need to understand the nature of road rage and its link to mental health. Our results clearly suggest that psychiatric distress is an important factor, at least for those identified as the most heavily involved in road rage; however, more work is needed to determine whether psychiatric problems lead to road rage, whether road rage leads to psychiatric problems, or whether both processes may be occurring. Funding and SupportThis research was supported by the Centre for Addiction and Mental Health and by a grant from AUTO21, one of the Networks of Centres of Excellence. References1. Harding RW, Morgan FH, Indemann D, Ferrante AM, Blagg H. Road rage and the epidemiology of violence: something old, something new. Studies in Crime Prevention 1998;7:221–38. 2. Smart RG, Mann, RE. Is road rage a serious traffic problem? Journal of Traffic Medicine 2002;3:183–9. 3. Joint M. Road rage. Washington (DC): AAA Foundation for Traffic Safety; 1995. 4. Wells-Parker E, Ceminsky J, Hallberg V, Snow RW, Dunaway G, Guiling S, and others. An exploratory study of the relationship between road rage and crash experience in a representative sample of US drivers. Accident Analysis and Prevention 2002;34:271–8. 5. Smart RG, Mann, RE. Deaths and injuries from road rage: cases in Canadian newspapers. CMAJ 2002;167:761–2. 6. James L, Nahl H. Road rage and aggressive driving. Amherest (MA): Prometheus; 2002. 7. Fumento M. Road rage vs reality. Atlantic Monthly 1998;August:12–7. 8. Lajunen T, Parker D. Are aggressive people aggressive drivers? A study of the relationship between self-reported general aggressiveness, driver anger and aggressive driving. Accident Analysis and Prevention 2001;22:243–55. 9. Swanson JW, Holzer CE, Eanju VK, Jono RT. Violence and psychiatric disorder in the community; evidence from the Epidemiological Catchment Area Studies. Hosp Community Psychiatry 1990;106:321–31. 10. Blanchard EB, Hickling EJ, Taylor AE, Loos WR, Gerardi RJ. Psychological morbidity associated with motor vehicle accidents. Behav Res Ther 1994;32:283–90. 11. Udwin O. Children’s reaction to traumatic events. J Child Psychol Psychiatry 1993;34:115–27. 12. Breslau N. Epidemiological studies of trauma, posttraumatic stress disorder, and other psychiatric disorders. Can J Psychiatry 2002:47:923–9. 13. Garmezy N, Rutter M. Acute reactions to stress. In: Rutter M, Hersov L, editors. Child and adolescent psychiatry: modern approaches. Oxford: Blackwell; 1985. p 152–67. 14. Fong G, Frost D, Stansfeld S. Road rage: a psychiatric phenomenon? Soc Psychiatry Psychiatr Epidemiol 2001;36:277–86. 15. Coccaro EF. Intermittent explosive disorder. Curr Psychiatry Rep 2000;2:67–71. 16. Galovski T, Blanchard EB, Veazey C. Intermittent explosive disorder and other psychiatric comorbidity among court-referred and self-referred aggressive drivers. Behav Res Ther 2002;40:641–51. 17. Galovski T, Blanchard EB. The effectiveness of a brief psychological intervention on court-referred and self-referred aggressive drivers. Behav Res Ther 2002;40:1385–402. 18. Asbridge M. Smart RG, Mann RE. The ‘homogamy’ of road rage: understanding the relationship between victimization and offending among aggressive and violent motorists. Forthcoming. 19. Wells-Parker E, Cosby PJ, Landrum JW. A typology for drinking driving offenders: methods for classification and policy implications. Accident Analysis and Prevention 1986;18;443–53. 20. Miller M, Azrael D, Hemenway D, Solop FI. ‘Road rage’ in Arizona: armed and dangerous. Accident Analysis and Prevention 2002;34:807–14. 21. Goldberg DP. The detection of psychiatric illness by questionnaire. London: Oxford University Press; 1972. 22. Goldberg DP. Manual of the General Health Questionnaire. Windsor (UK): NFER Publishing Company; 1978. 23. Adlaf EM, Ialomiteanu A. CAMH Monitor 2001: technical guide. Toronto: Centre for Addiction and Mental Health; 2002. 24. D’Arcy C, Siddique CM. Psychological distress among Canadian adolescents. Psychol Med 1984;14:615–28. 25. Goldberg DP, Gater R, Sartorius N, Ustun TB, Piccinelli M, Gureje O, and others. The validity of two versions of the GHQ in the WHO study of mental illness in general health care. Psychol Med 1997;27:191–7. 26. Hardy GE, Shapiro DA, Haynes CE, Rick JE. Validation of the General Health Questionnaire-12: using a sample of employees from England’s health care services. Psychol Assess 1999;11:159–65. 27. Donath S. The validity of the 12-item General Health Questionnaire in Australia: a comparison between three scoring methods. Aust N Z J Psychiatry 2001;35:231–5. 28. Goldberg DP, Hillier VF. A scaled version of the General Health Questionnaire. Psychol Med 1979;9:139–45. 29. Banks MH. Validation of the General Health Questionnaire in a young community sample. Psychol Med 1983;13:349–53. 30. Pevalin DJ. Multiple applications of the GHQ-12 in a general population sample: an investigation of long-term retest effects. Soc Psychiatry Psychiatr Epidemiol 2001;35:508–12. 31. Reijneveld SA, Schene AH. Higher prevalence of mental disorders in socioeconomically deprived urban areas in the Netherlands: community or personal disadvantage? J Epidemiol Community Health 1998;52:2–7. 32. Everitt BS, Landau S, Leese M. Cluster analysis. 4th ed. London: Arnold; 2001. 33. Aldenderfer ME, Blashfield RK. Cluster analysis. Thousand Oaks (CA): Sage Publications Inc; 1984. 34. Bailey KD. Typologies and taxonomies: an introduction to classification techniques. Thousand Oaks (CA): Sage Publications Inc; 1994. Author(s)Manuscript received February 2003, revised, and accepted April 2003. 1. Principal and Senior Scientist, Centre for Addiction and Mental Health, Toronto, Ontario. 2. Assistant Professor, Department of Community Health and Epidemiology, Dalhousie University, Halifax, Nova Scotia. 3. Senior Scientist, Centre for Addiction and Mental Health, Toronto, Ontario; Associate Professor, Department of Public Health Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario. 4. Head, Population and Life Course Studies, Centre for Addiction and Mental Health, Toronto, Ontario; Associate Professor, Department of Public Health Sciences, Faculty of Medicine, University of Toronto, Toronto, Ontario. Address for correspondence: Dr RG Smart, Centre for Addiction and Mental Health, 33 Russell Street, Toronto, ON M5S 2S1 e-mail: Reg_Smart@camh.net
1 | 2
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||