Personality disorders and nonfatal unintentional injuries among US adults
- 1Center for Injury Research and Policy, Research Institute at Nationwide Children’s Hospital, College of Medicine, Ohio State University, Ohio, USA
- 2Center for Innovation in Pediatric Practice, Research Institute at Nationwide Children’s Hospital, College of Medicine, Ohio State University, Ohio, USA
- 3Center for Biobehavioral Health, Research Institute at Nationwide Children’s Hospital, College of Medicine, Ohio State University, Ohio, USA
- Dr H Xiang, Center for Injury Research and Policy, Research Institute at Nationwide Children’s Hospital, 700 Children’s Drive, Columbus, OH 43205, USA;
- Accepted 20 January 2008
Objective: To investigate the association between personality disorders and nonfatal unintentional injuries in a representative sample of US adults.
Methods: Data on self-reported nonfatal unintentional injuries during the 12 months before the interview were obtained from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) were analyzed; 43 093 adults ⩾18 years participated in the NESARC wave I survey in 2001–02. Personality disorders were determined using the NIAAA Alcohol Use Disorders and Associated Disabilities Interview Schedule-DSM-IV.
Results: Individuals with at least one personality disorder had a significantly higher 12-month incidence of injuries than people without any personality disorder (p<0.001). After accounting for sociodemographic characteristics or other mental disorders, OR was 1.54 (95% CI 1.39 to 1.71) for individuals with one personality disorder and 1.80 (95% CI 1.58 to 2.05) for individuals with two or more personality disorders compared with people with no personality disorder.
Conclusion: Personality disorders were associated with a significantly increased risk of unintentional injuries. This information has important implications for the treatment of patients with these disorders.
Previous research has shown that a person’s psychological profile is associated with his or her risk of injury. The most well-known such risk is the association between depression or certain personality disorders (e.g. borderline personality disorder) with suicidal behavior.1–3 However, some data suggest that psychiatric diagnoses are also associated with increased risks for non-suicidal injuries. For example, depression is associated with increased rates of unintentional fractures in older women.45 In addition, previous studies have consistently reported alcohol and substance use disorders as major risk factors for injuries such as violence, vehicle crashes, injuries from bicycling and boating events, and fires.46–9
Many personality traits, such as external locus of control,10 antisocial behaviors,11 and type A behavior pattern12 are associated with increased risks of unintentional injuries. However, to our knowledge, there have been no epidemiological studies of the association between personality disorder diagnoses and rates of nonintentional injuries. There are several reasons why individuals with personality disorders might be at increased risk for injuries. Certain personality disorders may be associated with poor judgment and impulsivity, characteristics that could increase injuries. Personality disorders may also occur more commonly among individuals in high-risk environments. To date, no study has investigated the association of personality disorders and nonfatal unintentional injuries using a large population-based survey data in which a comprehensive list of personality orders was measured by standardized assessment methods. We report here our findings on associations of personality disorders and nonfatal unintentional injuries in a large representative sample of US adults who participated in the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC).
Subjects were 43 093 non-institutionalized, US civilians aged 18 years or older who participated in the NESARC, a nationally representative face-to-face survey conducted in 2001 and 2002 by the National Institute on Alcohol Abuse and Alcoholism (NIAAA). The details of sampling methods and data collection for the study are provided elsewhere.13
The NESARC data were weighted to reflect the probability of selection of primary sampling units (PSUs) within strata and for the selection of housing units within the sample PSUs. The weighted data were then adjusted to be representative of the US civilian, non-institutionalized population for socioeconomic variables including region, age, sex, and race/ethnicity using the 2000 Decennial Census of Population and Housing.13
Diagnostic assessment of psychiatric disorders
Diagnosis of psychiatric disorders was made using the NIAAA Alcohol Use Disorders and Associated Disabilities Interview Schedule-DSM-IV version (AUDADIS-IV).14 Diagnoses examined in the NESARC included personality disorders, mood disorders, alcohol abuse/dependence, nicotine dependence, and drug abuse/dependence. Borderline, schizotypal, and narcissistic personality disorders were not included in the wave I data of the NESARC and therefore were not examined in this study. The reliability and validity of the AUDIS-IV for personality disorders, mood disorders, alcohol abuse/dependence, nicotine dependence, and drug abuse/dependence were well documented in clinical and general populations.15–18
Unintentional injury definition
Our definition of unintentional injury was as follows: (1) an injury that resulted in medical help-seeking; and/or (2) an injury that resulted in a reduction in usual activities for more than half a day. We did not include injuries experienced as a result of suicide, self-injury, victimization, or injuries during criminal activity. These types of injuries are considered to have different underlying causes and intervention strategies are usually different.
Weighted injury incidence in the past 12 months was estimated using methods that accounted for the complex design effects of NESARC survey by socio-demographic factors, personality disorders, and axis I psychiatrics disorders. Odds ratios (OR) from univariate and multivariate logistic regression models were used to examine the associations between personality disorders and self-reported unintentional injuries. Annual incidence of nonfatal unintentional injuries, ORs and associated 95% CI were estimated using SAS and SUDAAN statistical software.19 In all analyses, any comparison with a two-sided p-value <0.05 was considered statistically significant.
Unintentional injury incidence
The overall annual injury incidence for persons aged 18 years or more was 16.9%. Table 1 reports injury incidence stratified by sociodemographic characteristics. The incidence of injuries differed significantly by gender, age, race/ethnicity, education, family income, and marital status (p<0.001).
Tables 2 and 3 present injury percentage by personality disorders and axis I psychiatric disorders. Overall, individuals with personality disorder had a significantly higher injury incidence than those without a personality disorder (table 2). Among different personality disorders examined in this study, individuals with a dependent disorder (a condition characterized by an over-reliance on others that leads to submissive and clinging behavior and fears of separation, diagnosed as a personality disorder when these behaviors become persistent and very disabling or distressing) had the highest injury incidence (37.5%, 95% CI 29.3 to 46.5), followed by individuals with histrionic disorders (31.9%, 95% CI 28.0 to 36.1) and individuals with antisocial disorders (31.5%, 95% CI 28.5 to 34.6). Individuals with an axis I psychiatric disorder had a statistically significantly higher injury incidence (p<0.001) than individuals without one (table 3).
Logistic regression results
Logistic regression models (table 4) were fitted to examine the associations between personality disorders and unintentional injuries while controlling for potential confounding effects. Among the different personality disorders, individuals with dependent personality disorders had the highest OR of injuries (OR 2.97, 95% CI 2.05 to 4.31). After controlling for gender, age, race/ethnicity, family income, education, marital status, and axis I psychiatric disorders, the OR of injuries dropped slightly. However, all the personality disorders remained significantly associated with nonfatal unintentional injuries in the multivariate models.
Individuals with more personality disorders were more likely to be injured than individuals with fewer or no disorders in the logistic regression models (table 4). However, after controlling for sociodemographics and axis I psychiatric disorders, the difference in OR of injuries was not significant between individuals with one personality disorder and those with two or more personality disorders (OR 1.54, 95% CI 1.39 to 1.71 vs OR 1.80, 95% CI 1.58 to 2.05; model III).
Our results indicated that individuals with personality disorders were significantly more likely to be injured during the previous 12 months than individuals without personality disorders. This association remained statistically significant even after controlling for sociodemographics and axis I psychiatric disorders.
In these data, both axis I psychiatric disorders and personality disorders were associated with unintentional injuries. Individuals with personality disorders were more likely to have alcohol abuse/dependence, drug abuse/dependence, and nicotine dependence.15–182021 Previous research has shown that alcohol abuse/dependence,422–25 drug abuse/dependence,2627 nicotine dependence,9 and mood disorders228 were associated with a significantly increased risk of nonfatal unintentional injuries. However, after controlling for alcohol and substance dependence/abuse, mood disorders, and major sociodemographics in our analysis, personality disorders were still significantly associated with injuries. Thus, our results suggest that personality disorders contribute independently to the risk status beyond these other axis I disorders, substance dependence/abuse, and major sociodemographics.
To determine the underlying mechanism for the association of personality disorders and increased risk of unintentional injuries is a challenging task. Increased injury risk among individuals with personality disorders may be caused by errors in a person’s judgment in injury risk environments or other unknown factors. Injuries as the causative agent for personality disorders seem unlikely in this particular study because the symptoms of personality disorder were longstanding traits, while the injuries were all noted during the 12 months prior to the interview. Because data from the NESARC were cross-sectional, caution should be exercised in making causal interpretations. When wave II data of the NESARC are released in the future, a longitudinal cohort study using NESARC data could be designed to examine further the association between personality disorders and injuries.
Several limitations of this study should be considered. First, the respondents themselves reported unintentional injuries in the NESARC survey. It is possible that recall bias caused fewer injuries to be identified because injuries were self-reported; this may have been most likely when injuries were minor or occurred further in the past. The incidence of nonfatal unintentional injuries might have been underestimated in this study, but in all likelihood, should not alter the core results. Second, the increased injury incidence among those with personality disorders we observed could have been caused simply by overreporting of injuries among individuals with personality disorders. Finally, we examined the association between personality disorders and unintentional injuries, not the causal effects of personality disorders on injury risk because the NESARC wave I data were cross-sectional.
Our results also point to important topics for future research. Although a wide range of psychiatric disorders were assessed in the NESARC, respondents were not asked about injury characteristics and context (e.g. cause of injuries, locale of injuries). Persons with personality characteristics may not only have more injuries than others, they may also have injuries from different causes and in different situations. Patterns of injuries may have important implications for injury prevention and medical care of injuries among those individuals with personality disorders.
Implications for injury prevention
Our study provides new evidence that personality disorders are associated with nonfatal unintentional injuries among US adults. Many individuals in the USA with personality disorders may face an increased risk of unintentional injuries. Health professionals who interact with persons with personality disorders should be aware of the increased risk of injuries and provide advice to help individuals with personality disorders reduce the risk of unintentional injuries.
Individuals with personality disorders are significantly more likely to be injured than individuals without personality disorders.
Individuals with two or more personality disorders have a higher injury prevalence than persons with one personality disorder.
This association remains significant after accounting for sociodemographic characteristics or other mental disorders.
Health professionals who interact with persons with personality disorders should provide injury prevention information to help individuals with personality disorders reduce the risk of unintentional injuries.
Funding: GC, HX and SS were funded in part by a grant from the Centers for Disease Control and Prevention (HX, grant number R49/CE000241-01). The contents of this study are solely the responsibility of the authors and do not necessarily reflect the official views of the Centers for Disease Control.
Competing interests: None.