Relationships between anxiety and externalizing disorders in youth: the influences of age and gender

https://doi.org/10.1016/j.janxdis.2006.06.004Get rights and content

Abstract

Minimal information about the relationship between anxiety disorders and externalizing disorders in youth is available. This study examined relationships between different specific anxiety and externalizing disorders and examined whether these associations varied by age and gender. The Methods for the Epidemiology of Child and Adolescent Mental Disorders (MECA) data set, consisting of youth from ages 9 to 17 recruited at four sites across the United States using a probability sampling method, was used. Results indicated that all externalizing disorders (attention-deficit hyperactivity disorder, oppositional defiant disorder, and conduct disorder) were positively related to a range of anxiety disorders. The magnitude of these associations tended to be stronger for males than for females (particularly for associations between social phobia and all externalizing disorders) and at younger, compared to older, ages (particularly for the association between oppositional defiant disorder and overanxious disorder). The cross-sectional positive relationships between externalizing and anxiety disorders vary somewhat based on gender, age, and which specific pair of disorders is examined; this may help explain the discrepant findings of previous research in this area.

Introduction

The study of comorbidity of child and adolescent psychopathology increased rapidly during the 1990s (Angold, Costello, & Erkanli, 1999), and it is clear that disorders co-occur more often than would be expected based on base rates within the population (e.g., Caron & Rutter, 1991). Comorbidity can arise due to both methodological reasons (such as shared diagnostic criteria) and substantive reasons (such as one disorder causing another disorder) and has implications for the classification and etiology of mental disorders (see Lilienfeld, 2003, for a detailed discussion of the causes and implications of comorbidity). Associations between anxiety and externalizing disorders among young people have proven especially difficult to understand. Although overall it appears that anxiety and externalizing disorders are positively associated (e.g., Angold et al., 1999; Lewinsohn, Hops, Roberts, Seeley, & Andrews, 1993), some studies appear to show that anxiety may actually be protective against severe conduct problems (e.g., Walker et al., 1991).

In addition to the methodological reasons that comorbidity may be found (as discussed in Lilienfeld, 2003), there are several substantive reasons that anxiety and externalizing disorders should be positively associated. First, comorbidity is the norm, not the exception, for mental health problems (e.g., Kessler et al., 1997). Although research has demonstrated that a two-factor structure—internalizing and externalizing—best characterizes mental disorders, the two factors are significantly intercorrelated (Hinden, Compas, Howell, & Achenbach, 1997; Krueger, Caspi, Moffitt, & Silva, 1998). Second, certain risk factors for anxiety and externalizing disorders could be similar. For example, a child living in poverty could be at risk for both types of disorders due to the variety of stressors that are associated with potential correlates of poverty, such as living in dangerous neighborhoods. Third, it is possible that youth with externalizing disorders generate anxiety-provoking situations for themselves. This idea is supported by the Frick, Lilienfeld, Ellis, Loney, and Silverthorn's (1999) hypothesis that antisocial behavior leads to recurrent state anxiety; this state anxiety may then lead to anxiety that is difficult to differentiate from anxiety disorders. Fourth, certain symptoms of anxiety and externalizing disorders may be similar enough that a youth with one type of disorder would be at increased risk of also qualifying for the other type of disorder. For example, fidgeting may be characteristic of both attention-deficit hyperactivity disorder (ADHD) and anxiety. Although this is unlikely to result in a broad association between these two syndromes, it could account for some degree of association between certain disorders.

However, there are also reasons to expect that anxiety and externalizing disorders might be negatively associated. For example, anxiety is associated with high levels of inhibition (e.g., Biederman et al., 1990) and excessively timid behavior (e.g., Tellegan, 1985), while externalizing disorders are associated with low levels of constraint (Krueger, McGue, & Iacono, 2001) and impulsivity (indeed, this characteristic is even built into one subtype of ADHD [the hyperactive/impulsive subtype]). These opposing tendencies would seem to be mutually exclusive.

Although there may be an overall association between anxiety and externalizing behavior, associations between different specific pathologies of each type may differ. These differing associations may fall along diagnostic lines (e.g., the relationship between social phobia [SOP] and conduct disorder [CD] may differ from the relationship between generalized anxiety disorder [GAD] and CD), or they may fall along more broad syndrome lines. For example, a recent study (Lahey et al., 2004) indicated that some anxiety symptoms (generalized worries and maybe social anxiety) appear to be part of the same syndrome as depression, while others (separation anxiety, fears, and compulsions) may be part of a separate anxiety dimension; perhaps these two dimensions of anxiety symptoms would have differing relationships with externalizing behaviors. Thus, it is critical to examine relationships between different pairs of disorders separately.

Importantly, there may also be both gender and developmental differences in these relationships. Anxiety disorders are more common among girls, while externalizing disorders are more common among boys (e.g., Simonoff et al., 1997); however, it is unclear whether these two classes of disorders should be differentially associated for girls and boys. It is possible that boys who are anxious may feel pressure to hide their anxiety and may do so by engaging in acting-out behavior, while girls may not feel a similar pressure and therefore be less likely to act-out as a consequence of anxiety. If this were the case, these two sets of disorders would be more closely linked in boys, compared to girls. These two classes of disorders may be more strongly associated among younger children, compared with older children, as would be suggested by the notion that psychopathology in young children may be relatively amorphous, while older children and adults may tend to exhibit more clearly differentiated forms of psychopathology (e.g., Nottelmann & Jensen, 1995). A direct examination of age and gender differences in these relationships may help resolve some of the discrepancies in the literature in this area, as discussed below.

The purpose of the present study was to comprehensively describe associations between specific anxiety and externalizing disorders in a population-based sample of 9–17-year-olds. Due to the potential for gender and developmental differences in these relationships, an important question to be addressed was whether or not these associations differed by gender and/or age. Diagnoses both requiring and not requiring impairment were examined in order to examine any differences in these relationships that may have arisen based on clinical impairment (as suggested by Simonoff et al., 1997).

Hypotheses were developed based on prior research in the area. ADHD and anxiety disorders overall co-occur at rates greater than chance (as reviewed in Biederman, Newcorn, & Sprich, 1991; a more recent meta-analysis of this relationship was reported on in Angold et al., 1999), and there may be a specific association between ADHD and simple phobia (SIP) when clinically significant impairment is present (Simonoff et al., 1997). However, results vary across studies and sometimes do not indicate a significant association between these two types of disorders at all; this may be due to some studies’ inclusion of all anxiety disorders and other studies’ examination of specific anxiety disorders or to the varying age and gender compositions of the different studies. This latter notion is partially supported by the fact that some studies that do not find a positive relationship between anxiety disorders and ADHD used a relatively old, all-male sample (ages 16–23; Gittelman, Mannuzza, Shenker, & Bonagura, 1985; Mannuzza, Gittelman-Klein, Horowitz-Konig, & Giampino, 1989).

When CD and related problems (e.g., aggression) are considered, the evidence to date is more conflicting. Cross-sectional studies tend to report a positive association between anxiety and CD (e.g., Angold et al., 1999, Kashani, Deuser, & Reid, 1991) this relationship may be particularly strong for SOP and post-traumatic stress disorder (Goodwin & Hamilton, 2003) or for overanxious disorder (OAD; whether or not impairment was required for a diagnosis) and SIP (if impairment was required for a diagnosis; Simonoff et al., 1997). However, longitudinal studies frequently find that anxiety attenuates antisocial behavior (e.g., Mason et al., 2004; Pine, Cohen, Cohen, & Brook, 2000; Raine, Venables, & Williams, 1995; Walker et al., 1991), though some find the opposite (Ialongo, Edelsohn, Werthamer-Larsson, Crockett, & Kellam, 1996; Roza, Hofstra, van der Ende, & Verhulst, 2003). With the exception of the Walker et al. study, all of these studies suggesting that anxiety attenuates antisocial behavior have included mainly older participants (i.e., ages 15–29 in Raine et al., ages 10/11 to 21 in Mason et al., and mean ages of 14–22 in Pine et al.); therefore, it is possible that there are developmental differences in the relationship between CD and anxiety.

In the one study to specifically examine associations between oppositional defiant disorder (ODD) and anxiety disorders (ODD behaviors are commonly considered together with CD), ODD was found to be positively associated with SAD and SIP, whether or not impairment was required for the diagnoses; in addition, it was associated with SOP if impairment was required for the diagnoses and SAD if impairment was not required for the diagnoses (Simonoff et al., 1997).

Thus, it was expected that ADHD, ODD, and CD would be positively associated with anxiety disorders overall, although it was considered likely that the exact patterns of association between each externalizing disorder and different anxiety disorders would differ. Overall, generalized anxiety (generalized anxiety disorder [GAD] and OAD) was expected to be less-strongly associated with severe externalizing behavior (such as CD) than was focused anxiety (e.g., SIP). Other relationships were expected to be positive and generally significant, though due to the highly specific nature of SIP, agoraphobia (AGOR), and separation anxiety disorder (SAD), it was less clear whether these would be consistently associated with externalizing disorders. Predictions related to age and gender differences were also made. Due to the possibility that psychopathology becomes more differentiated with age and the limited evidence (reviewed above) regarding potential age differences in the relationship between CD and anxiety, it was expected that relationships between anxiety disorders and externalizing disorders would generally be stronger at younger ages, compared to older ages. It was also tentatively expected that relationships would be stronger for boys than girls.

Section snippets

Participants

Participants in the Methods for the Epidemiology of Child and Adolescent Mental Disorders (MECA) study were included in this study. The public access version of this database, which became available in 2004, was used. A complete description of the sample and selection procedures for the MECA study is provided in Lahey et al. (1996). Briefly, participants were selected from among youths aged 9 through 17 years living in specific regions of Connecticut, Georgia, New York, and Puerto Rico. A

Overall associations between anxiety disorders and externalizing disorders

Results of chi-square analyses examining associations between each anxiety disorder and each externalizing disorder and requiring impairment for a diagnosis are presented in Table 1. ADHD was positively associated with SOP, AVD, GAD, OAD, and SAD. ODD was positively associated with all the anxiety disorders that were examined. CD was positively associated with SIP, SOP, GAD, and OAD. Odds ratios for significant results ranged from approximately 2 (2.27) to approximately 7 (7.02). All odds

Discussion

The results of this study support the notion that anxiety and externalizing disorders are positively associated, and the strength of the associations between these two sets of disorders differs based on which specific pair of disorders is examined. In addition, the relationships differ somewhat based on gender and age group. Some of the discrepant findings of past research may be accounted for by differing pairs of disorders examined, as well as different age and gender compositions of samples.

Acknowledgments

The author thanks the investigators of the Methods for the Epidemiology of Child and Adolescent Mental Disorders (MECA) study, and specifically Benjamin Lahey and Karen Bourdon for their work in making the public use version of this data set available. The MECA Program was an epidemiological methodology study which began in 1989 and was performed by four independent research teams in collaboration with staff of the National Institute of Mental Health, Rockville, MD. The NIMH Principal

References (35)

  • A. Angold et al.

    Comorbidity

    Journal of Child Psychology and Psychiatry

    (1999)
  • J. Biederman et al.

    Comorbidity of attention deficit hyperactivity disorder with conduct, depressive, anxiety, and other disorders

    American Journal of Psychiatry

    (1991)
  • J. Biederman et al.

    Psychiatric correlates of behavioral inhibition in young children of parents with and without psychiatric disorders

    Archives of General Psychiatry

    (1990)
  • C. Caron et al.

    Comorbidity in child psychopathology: concepts, issues and research strategies

    Journal of Child Psychology and Psychiatry

    (1991)
  • S.L. Decker et al.

    Comorbidity among individuals with attention disorders

    International Journal of Neuroscience

    (2001)
  • R. Eme

    Selective female affliction in the developmental disorders: a literature review

    Journal of Clinical Child Psychology

    (1992)
  • P.J. Frick et al.

    The association between anxiety and psychopathy dimensions in children

    Journal of Abnormal Child Psychology

    (1999)
  • Cited by (0)

    View full text