اختلالات شخصیت آشکار با بزرگسالی و خطر برای اختلالات اضطرابی در بزرگسالی میانه

Abstract Data from the Children in the Community Study, a prospective longitudinal investigation, were used to investigate the association of personality disorder (PD) traits, evident by early adulthood, with risk for development of anxiety disorders by middle adulthood. Individuals without a history of anxiety disorders who met diagnostic criteria for ≥1 PD by early adulthood were at markedly elevated risk for agoraphobia, generalized anxiety disorder, obsessive-compulsive disorder, and panic disorder by middle adulthood. Antisocial, avoidant, borderline, dependent, depressive, histrionic, obsessive-compulsive, passive-aggressive, and schizotypal PD traits, evident by early adulthood, were associated with elevated risk for ≥1 anxiety disorder during middle adulthood. These associations remained significant after a history of anxiety disorder and co-occurring Axis I psychiatric disorder was controlled statistically. Findings of this study suggest that some types of PD traits that become evident by early adulthood may contribute to increased risk for the development of anxiety disorders by middle adulthood.

. Introduction The association of personality disorder (PD) with risk for anxiety disorders is of considerable interest to researchers, and it has important clinical and theoretical implications. Clinicians are interested in this association because they recognize the importance of being well informed about the outcomes that may be associated with PD and other mental disorders. A variety of conceptual and theoretical models has been advanced regarding the associations of specific types of PDs with anxiety disorders and other Axis I disorders (e.g., Lyons, Tyrer, Gunderson, & Tohen, 1997). Some PDs may contribute to increased vulnerability for anxiety disorders (Faravelli et al., 2000). Common etiological factors, such as parental overprotection, have been hypothesized to underlie the development of certain types of PDs and anxiety disorders (Latas, Starcevic, Trajkovic, & Bogojevic, 2000). Another hypothesis is that some PDs and anxiety disorders (e.g., avoidant PD and social anxiety disorder; obsessive-compulsive PD and obsessive-compulsive disorder) may occupy different points along a common spectrum (Bejerot, Ekselius, & von Knorring, 1998; Dyck et al., 2001; Tillfors, Furmark, Ekselius, & Fredrikson, 2001). Some associations between PDs and Axis I disorders may also be attributable, in part, to overlapping diagnostic criteria (Widiger & Shea, 1991). Although research has demonstrated that PDs often co-occur with anxiety disorders and other Axis I disorders (e.g., Dyck et al., 2001, McGlashan et al., 2000 and Oldham et al., 1995), there are significant gaps in the scientific literature. Much of the information that is currently available on the association between PD and anxiety disorders has been obtained from cross-sectional studies of Axis I–Axis II comorbidity. These studies have yielded findings suggesting that some types of PDs may be particularly associated with specific anxiety disorders (e.g., Bejerot et al., 1998; Comtois, Cowley, Dunner, & Roy-Byrne, 1999; Dyck et al., 2001; Hoffart, Thornes, & Hedley, 1995; Noyes, Woodman, Holt, Reich, & Zimmerman, 1995; Skodol et al., 1995 and Zanarini et al., 1998; Zimmerman & Mattia, 1999). However, it is not yet clear whether the associations between various types of PDs and anxiety disorders are characterized by high, low, or moderate specificity. Another concern has been that cross-sectional data do not permit inferences regarding the directionality of the associations between PDs and anxiety disorders. Studies conducted with samples of patients with anxiety disorders have investigated PD sequelae, such as the association of PD with treatment outcomes. These investigations have indicated that patients who have anxiety disorders and PDs tend to have poor outcomes (Nurnberg et al., 1989 and Perry, 1993; Reich & Green, 1991; Steketee, Eisen, Dyck, Warshaw, & Rasmussen, 1999; Turner, 1987; Yonkers, Dyck, Warshaw, & Keller, 2000). However, most of these studies have used a sample of modest size, a relatively brief follow-up interval (typically one year or less), or have focused on a limited range of PDs. Moreover, findings of studies that have investigated the sequelae of PDs among patients in clinical settings may not apply to the general population. Patients with PDs differ from individuals with PDs in the remainder of the population, insofar as their symptoms tend to be more severe, and treatment may tend to have a systematic and indeterminate impact on the course and outcomes of the disorder. Prospective longitudinal studies of sizable, representative, community-based samples are, thus, needed. To date, few community-based studies have investigated the long-term mental health consequences of PDs. Studies of non-patient samples that have investigated the associations of PD with subsequent anxiety problems have focused on overall PD symptoms (Johnson & Bornstein, 1991; Johnson et al., 1996), or have not examined specific associations between PDs and a wide range of anxiety disorders (Faravelli et al., 2000 and Johnson et al., 1999). To date, there have not been any population-based investigations of associations between PDs, evident by early adulthood, and risk for the development of specific types of anxiety disorders. It is important to fill this gap in the scientific literature by investigating the long-term sequelae of PDs, which may not become evident until early adulthood, according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR; APA, 2000). We report findings from the Children in the Community Study, a community-based prospective longitudinal study, regarding the association of PD by early adulthood with risk for anxiety disorders by middle adulthood. Several hypotheses were tested. First, insofar as avoidant PD and social anxiety disorder share a similar symptom spectrum, and because avoidant PD may be a prodromal form or precursor of social anxiety disorder, individuals with avoidant PD by early adulthood were hypothesized to be at elevated risk for social anxiety disorder by middle adulthood. Second, insofar as obsessive-compulsive PD and obsessive-compulsive disorder (OCD) share a similar symptom spectrum, and because obsessive-compulsive PD may be a prodromal form or precursor of OCD, we examined the hypothesis that individuals with obsessive-compulsive PD by early adulthood are at elevated risk for OCD by middle adulthood. Third, individuals with Cluster C PDs (avoidant, dependent, and obsessive-compulsive PDs) were hypothesized to be at elevated overall risk for the development of one or more anxiety disorders by middle adulthood. This hypothesis was based on the DSM-IV-TR observation that individuals with Cluster C PDs tend to be anxious (APA, 2000). Insofar as Cluster C PDs are differentially characterized by anxiety, and because Cluster C PDs may be prodromal forms or precursors of Axis I anxiety disorders, it was hypothesized that individuals with Cluster C PDs might be at elevated overall risk for the development of anxiety disorders.

Results 3.1. Descriptive statistics Over the three assessments that were conducted between mean ages 14 and 22, 153 individuals (23.3%) were identified, cumulatively, as having met the diagnostic criteria for anxiety disorders, and 149 (22.6%) were identified, cumulatively, as having met diagnostic criteria for PD. The point prevalence of these disorders was substantially lower (e.g., prevalence of anxiety disorders ranged from 10.1% to 11.5% across the three assessments). Seventy individuals (10.6%) met the diagnostic criteria for agoraphobia (4.3%), GAD (2.7%), OCD (4.1%), panic disorder (4.6%), or social anxiety disorder (3.6%) at mean age 33. Aggregate mean PD trait levels between mean ages 14 and 22 are reported in Table 1. Table 1. Mean personality disorder (PD) trait level at mean ages 14–22 and risk for agoraphobia, GAD, OCD, panic disorder, or social anxiety disorder at mean age 33 (N = 658) Personality disorder (PD) Aggregate PD trait level (ages 14–22), mean (S.D.) Adjusted odds ratio and 95% Confidence interval DSM-IV Cluster A PD traits 3.83 (2.14) 1.26c 1.12–1.42 Paranoid PD traits 1.23 (0.90) 1.27 0.97–1.66 Schizoid PD traits 1.01 (0.77) 1.34 0.98–1.84 Schizotypal PD traits 1.82 (0.98) 1.99c 1.53–2.59 DSM-IV Cluster B PD traits 4.70 (2.91) 1.21c 1.10–1.33 Antisocial PD traits 1.33 (0.97) 1.51c 1.15–1.97 Borderline PD traits 1.30 (1.10) 1.71c 1.36–2.14 Histrionic PD traits 1.67 (1.08) 1.46c 1.14–1.86 Narcissistic PD traits 1.56 (1.20) 1.22 0.98–1.52 DSM-IV Cluster C PD traits 3.09 (2.01) 1.28c 1.12–1.46 Avoidant PD traits 0.97 (0.87) 1.44d 1.08–1.92 Dependent PD traits 1.26 (1.01) 1.54c 1.20–1.97 Obsessive-compulsive PD traits 0.87 (0.66) 1.56d 1.08–2.24 DSM-IV Appendix B PD traits 1.75 (1.37) 1.40c 1.18–1.66 Depressive PD Traits 0.79 (0.84) 1.45c 1.13–1.87 Passive-aggressive PD traits 0.96 (0.95) 1.49c 1.17–1.90 Total PD traits 13.20 (7.07) 1.10c 1.05–1.14 a Controlling for age, sex, and evidence of any anxiety disorder, any Axis I disorder, and any co-occurring PD between mean ages 14 and 22. b Odds ratio indicates the increase in risk for an anxiety disorder that would be associated with an elevation of 0.33 PD traits. Thus, for example, individuals with two borderline PD traits were 213% more likely than those with one borderline PD trait to develop an anxiety disorder. c P < .01. d P < .05. Table options 3.2. Personality disorder traits associated with overall risk for anxiety disorders by mean age 33 Antisocial, borderline, dependent, depressive, histrionic, passive-aggressive, and schizotypal PD traits, present between ages 14 and 22, were associated with elevated risk for anxiety disorders by mean age 33 after a history of anxiety disorders and other psychiatric disorders was controlled statistically (Table 1). DSM-IV Cluster A–C trait levels, DSM-IV Appendix B trait levels, and overall PD trait levels by mean age 22, were all significantly associated with risk for anxiety disorders by mean age 33. Individuals who met the DSM-IV diagnostic criteria for ≥1 PD by mean age 22 were at substantially elevated risk for any anxiety disorder (agoraphobia, GAD, OCD, panic disorder, or social anxiety disorder) at mean age 33. This association was significant when age, sex, and anxiety disorders and other Axis I disorders evident by mean age 22 were controlled (adjusted odds ratio (OR) OR = 2.07; 95% confidence interval (CI): 1.17–3.65). PD was not significantly associated with risk for ≥1 anxiety disorders among the 153 individuals who were identified as having anxiety disorders by mean age 22. However, of the 505 persons without a history of anxiety disorders, those (n = 82) who met the DSM-IV diagnostic criteria for ≥1 PD by mean age 22 were at significantly elevated risk for anxiety disorders at mean age 33 ( Fig. 2). This association remained significant when age, sex, and co-occurring Axis I disorders were controlled statistically (adjusted OR = 4.55; CI: 1.99–10.41). Association of personality disorder (PD), evident by mean age 22, with risk for ... Fig. 2. Association of personality disorder (PD), evident by mean age 22, with risk for anxiety disorders by mean age 33 (N = 658). Figure options 3.3. Personality disorder traits evident by mean age 22 and risk for agoraphobia at mean age 33 Antisocial, borderline, dependent, depressive, narcissistic, obsessive-compulsive, passive-aggressive, and schizotypal PD traits, evident by mean age 22, were associated with significantly increased risk for agoraphobia by mean age 33 after the covariates were controlled (Table 2). DSM-IV Cluster A–C trait levels, DSM-IV Appendix B trait levels, and overall PD trait levels by mean age 22, were all significantly associated with risk for agoraphobia by mean age 33. Individuals with PD by mean age 22 were at significantly elevated risk for agoraphobia after the covariates were controlled (adjusted OR = 2.39; CI: 1.01–5.70) (Fig. 3). PD was not associated with risk for agoraphobia among the 153 individuals who were identified as having anxiety disorders by mean age 22. However, of the 505 persons who did not have anxiety disorders by mean age 22, those who met the diagnostic criteria for ≥1 PD were at substantially elevated risk for agoraphobia at mean age 33 (OR = 3.26; CI: 1.15–9.24; P = .03 (Fisher's Exact Test)). This association remained significant when age, sex, and co-occurring Axis I disorders were controlled statistically (adjusted OR = 3.85; CI: 1.17–12.62). Table 2. Association of mean personality disorder (PD) trait level at mean ages 14–22 with risk for agoraphobia at mean age 33 (N = 658) Personality disorder (PD) Adjusted odds ratio and 95% Confidence interval DSM-IV Cluster A PD traits 1.30c 1.09–1.55 Paranoid PD traits 1.23 0.82–1.85 Schizoid PD traits 1.53 0.97–2.43 Schizotypal PD traits 2.17c 1.48–3.17 DSM-IV Cluster B PD traits 1.23c 1.08–1.40 Antisocial PD traits 1.68c 1.14–2.47 Borderline PD traits 1.70c 1.24–2.32 Histrionic PD traits 1.35 0.94–1.95 Narcissistic PD traits 1.39d 1.01–1.92 DSM-IV Cluster C PD traits 1.31c 1.09–1.56 Avoidant PD traits 1.41 0.94–2.12 Dependent PD traits 1.69c 1.19–2.41 Obsessive-compulsive PD traits 1.40d 1.02–2.94 DSM-IV Appendix B PD traits 1.58c 1.26–1.98 Depressive PD traits 1.60c 1.15–2.21 Passive-aggressive PD traits 1.85c 1.31–2.61 Total PD traits 1.11c 1.05–1.17 a Controlling for age, sex, and evidence of any anxiety disorder, any Axis I disorder, and any co-occurring PD between mean ages 14 and 22. b Odds ratio indicates the increase in risk for agoraphobia that would be associated with an elevation of 0.33 PD traits. Thus, for example, individuals with two Borderline PD traits were 210% more likely than those with one borderline PD trait to develop subsequent agoraphobia. c P < .01. d P < .05. Table options Association of personality disorder (PD), evident by mean age 22, with risk for ... Fig. 3. Association of personality disorder (PD), evident by mean age 22, with risk for agoraphobia, generalized anxiety disorder, obsessive-compulsive disorder, panic disorder, and society anxiety disorder by mean age 33 (N = 658). Figure options 3.4. Personality disorder traits evident by mean age 22 and risk for GAD at mean age 33 Borderline, histrionic, and schizotypal PD traits, evident by mean age 22, were associated with significantly increased risk for GAD by mean age 33 after the covariates were controlled (Table 3). DSM-IV Cluster A–C trait levels and overall PD trait levels by mean age 22, were significantly associated with risk for GAD by mean age 33. Individuals with PD by mean age 22 were at significantly elevated risk for GAD after the covariates were controlled (adjusted OR = 3.25; CI: 1.14–9.31). PD was not associated with risk for GAD among the 153 individuals who were identified as having anxiety disorders by mean age 22. However, of the 505 persons who did not have anxiety disorders by mean age 22, those who met the diagnostic criteria for ≥1 PD were at markedly elevated risk for GAD at mean age 33 (OR = 13.67; CI: 2.60–71.72; P = .002 (Fisher's Exact Test)). This association remained significant when age, sex, and co-occurring Axis I disorders were controlled statistically (adjusted OR = 14.41; CI: 2.26–91.72). Table 3. Association of mean personality disorder (PD) trait level at mean ages 14–22 with risk for generalized anxiety disorder (GAD) at mean age 33 (N = 658) Personality disorder (PD) Adjusted odds ratio and 95% Confidence interval DSM-IV Cluster A PD traits 1.27c 1.03–1.57 Paranoid PD traits 1.18 0.73–1.91 Schizoid PD traits 1.39 0.79–2.45 Schizotypal PD traits 2.24d 1.40–3.59 DSM-IV Cluster B PD traits 1.24c 1.07–1.45 Antisocial PD traits 1.29 0.82–2.03 Borderline PD traits 1.75d 1.23–2.50 Histrionic PD traits 1.78c 1.13–2.79 Narcissistic PD traits 1.36 0.92–2.01 DSM-IV Cluster C PD traits 1.20d 1.03–2.36 Avoidant PD traits 1.30 0.79–2.14 Dependent PD traits 1.40 0.90–2.17 Obsessive-compulsive PD traits 1.34 0.68–2.68 DSM-IV Appendix B PD traits 1.14 0.86–1.52 Depressive PD traits 1.14 0.72–1.78 Passive-aggressive PD traits 1.22 0.78–1.89 Total PD traits 1.09c 1.02–1.16 a Controlling for age, sex, and evidence of any anxiety disorder, any Axis I disorder, and any co-occurring PD between mean ages 14 and 22. b Odds ratio indicates the increase in risk for GAD that would be associated with an elevation of 0.33 PD traits. Thus, for example, individuals with two borderline PD traits were 225% more likely than those with one borderline PD trait to develop subsequent GAD. c P < .05. d P < .01. Table options 3.5. Personality disorder traits evident by mean age 22 and risk for OCD at mean age 33 Avoidant, borderline, dependent, histrionic, passive-aggressive, and schizotypal PD traits, evident by mean age 22 were associated with significantly increased risk for OCD by mean age 33 after the covariates were controlled (Table 4). DSM-IV Cluster A–C trait levels, DSM-IV Appendix B trait levels, and overall PD trait levels by mean age 22 were all significantly associated with risk for OCD by mean age 33. PD was not associated with risk for OCD among the 153 individuals who were identified as having anxiety disorders by mean age 22. However, of the 505 persons who did not have anxiety disorders by mean age 22, those who met the diagnostic criteria for ≥1 PD were at markedly elevated risk for OCD at mean age 33 (OR = 5.49; CI: 1.72–17.46; P = .006 (Fisher's Exact Test)). This association remained significant when age, sex, and co-occurring Axis I disorders were controlled statistically (adjusted OR = 4.21; CI: 1.19–14.95). Table 4. Association of mean personality disorder (PD) trait level at mean ages 14–22 with risk for obsessive-compulsive disorder (OCD) at mean age 33 (N = 658) Personality disorder (PD) Adjusted odds ratio and 95% Confidence interval DSM-IV Cluster A PD traits 1.28c 1.07–1.52 Paranoid PD traits 1.03 0.68–1.55 Schizoid PD traits 1.36 0.86–2.17 Schizotypal PD traits 2.54c 1.74–3.71 DSM-IV Cluster B PD traits 1.26c 1.11–1.43 Antisocial PD traits 1.44 0.99–2.08 Borderline PD traits 1.86c 1.38–2.52 Histrionic PD traits 1.47d 0.92–1.91 Narcissistic PD traits 1.32 0.96–1.82 DSM-IV Cluster C PD traits 1.31d 1.09–1.58 Avoidant PD traits 1.74c 1.17–2.59 Dependent PD traits 1.46d 1.01–2.12 Obsessive-compulsive PD traits 1.65 0.95–2.88 DSM-IV Appendix B PD traits 1.41c 1.13–1.75 Depressive PD traits 1.34 0.93–1.92 Passive-aggressive PD traits 1.67d 1.20–2.31 Total PD traits 1.11c 1.05–1.17 a Controlling for age, sex, and evidence of any anxiety disorder, any Axis I disorder, and any co-occurring PD between mean ages 14 and 22. b Odds ratio indicates the increase in risk for OCD that would be associated with an elevation of 0.33 PD traits. Thus, for example, individuals with two borderline PD traits were 258% more likely than those with one borderline PD trait to develop subsequent OCD. c P < .01. d P < .05. Table options 3.6. Personality disorder traits associated with risk for panic disorder at mean age 33 Borderline, dependent, schizoid, and schizotypal PD traits, evident by mean age 22, were associated with significantly increased risk for panic disorder by mean age 33 after the covariates were controlled (Table 5). DSM-IV Cluster A–C trait levels, DSM-IV Appendix B trait levels, and overall PD trait levels by mean age 22, were all significantly associated with risk for panic disorder by mean age 33. PD was not associated with risk for panic disorder among the 153 individuals who were identified as having anxiety disorders by mean age 22. However, of the 505 persons who did not have anxiety disorders by mean age 22, those who met the diagnostic criteria for ≥1 PD were at markedly elevated risk for panic disorder at mean age 33 (OR = 3.26; CI: 1.15–9.24; P = .03 (Fisher's Exact Test)). This association remained significant when age, sex, and a history of other Axis I disorder was controlled statistically (adjusted OR = 4.64; CI: 1.39–15.51). Table 5. Association of mean personality disorder (PD) trait level at mean ages 14–22 with risk for panic disorder at mean age 33 (N = 658) Personality disorder (PD) Adjusted odds ratio and 95% Confidence interval DSM-IV Cluster A PD traits 1.28c 1.08–1.52 Paranoid PD traits 1.16 0.78–1.74 Schizoid PD traits 1.68d 1.08–2.62 Schizotypal PD traits 2.06c 1.41–3.00 DSM-IV Cluster B PD traits 1.18d 1.03–1.34 Antisocial PD traits 1.24 0.84–1.83 Borderline PD traits 1.51c 1.11–2.06 Histrionic PD traits 1.30 0.92–1.86 Narcissistic PD traits 1.36 0.99–1.87 DSM-IV Cluster C PD traits 1.22c 1.01–1.46 Avoidant PD traits 1.25 0.82–1.90 Dependent PD traits 1.48d 1.05–2.09 Obsessive-compulsive PD traits 1.43 0.84–2.45 DSM-IV Appendix B PD traits 1.30d 1.04–1.64 Depressive PD traits 1.38 0.98–1.93 Passive-aggressive PD traits 1.36 0.95–1.96 Total PD traits 1.09c 1.03–1.15 a Controlling for age, sex, and evidence of any anxiety disorder, any Axis I disorder, and any co-occurring PD between mean ages 14 and 22. b Odds ratio indicates the increase in risk for panic disorder that would be associated with an elevation of 0.33 PD traits. Thus, for example, individuals with two borderline PD traits were 153% more likely than those with one borderline PD trait to develop subsequent panic disorder. c P < .01. d P < .05. Table options 3.7. Personality disorder traits associated with risk for social anxiety disorder at mean age 33 Avoidant, borderline, dependent, depressive, obsessive-compulsive, and schizotypal PD traits, evident by mean age 22 were associated with significantly increased risk for social anxiety disorder by mean age 33 after the covariates were controlled (Table 6). DSM-IV Cluster A and C trait levels, DSM-IV Appendix B trait levels, and overall PD trait levels by mean age 22 were significantly associated with risk for social anxiety disorder by mean age 33. The associations of PD with risk for social anxiety disorder at mean age 33 were not statistically significant among the respondents with (n = 153) or without (n = 505) a history of anxiety disorder. However, data from the full sample (N = 658) indicated that the individuals who were identified as having PD by mean age 22 were at significantly elevated risk for social anxiety disorder when the covariates were controlled (adjusted OR = 2.71; CI: 1.06–6.92). Table 6. Association of mean personality disorder (PD) trait level at mean ages 14–22 with risk for social anxiety disorder at mean age 33 (N = 658) Personality disorder (PD) Adjusted odds ratio and 95% Confidence interval DSM-IV Cluster A PD traits 1.21c 1.01–1.47 Paranoid PD traits 1.22 0.79–1.88 Schizoid PD traits 1.45 0.89–2.38 Schizotypal PD traits 1.64c 1.09–2.47 DSM-IV Cluster B PD traits 1.11 0.96–1.28 Antisocial PD traits 1.28 0.84–1.96 Borderline PD traits 1.63d 1.17–2.28 Histrionic PD traits 1.26 0.86–1.86 Narcissistic PD traits 0.82 0.56–1.21 DSM-IV Cluster C PD traits 1.52d 1.25–1.84 Avoidant PD traits 2.36d 1.55–3.60 Dependent PD traits 1.86d 1.29–2.69 Obsessive-compulsive PD traits 1.96c 1.13–3.41 DSM-IV Appendix B PD traits 1.32c 1.04–1.69 Depressive PD traits 1.82d 1.29–2.57 Passive-aggressive PD traits 0.96 0.61–1.51 Total PD traits 1.09c 1.03–1.16 a Controlling for age, sex, and evidence of any anxiety disorder, any Axis I disorder, and any co-occurring PD between mean ages 14 and 22. b Odds ratio indicates the increase in risk for social anxiety disorder that would be associated with an elevation of 0.33 PD traits. Thus, for example, individuals with two borderline PD traits were 189% more likely than those with one borderline PD trait to develop social anxiety disorder. c P < .05. d P < .01.