Research reportTesting the tripartite model in young adolescents: Is hyperarousal specific for anxiety and not depression?
Introduction
Anxiety and depressive problems in childhood and adolescence occur frequently (Treffers, 2000, Verhulst et al., 1997), result in considerable suffering and impairment, and tend to persist (Ferdinand et al., 1999, Ferdinand and Verhulst, 1995, Pollack et al., 1996, Treffers and Öst, 2001). Therefore, it is important to investigate their etiology, and to develop an adequate taxonomy, that can serve as a cornerstone for high quality assessment.
Since anxiety and depression often co-occur, it is doubted if the two represent distinct constructs (Angold et al., 1999, Axelson and Birmaher, 2001, Essau et al., 2000, Goodwin, 2002, Kessler et al., 1999, Stein et al., 2001). In an attempt to improve the taxonomy of anxiety and depression, Clark and Watson introduced the tripartite model (Clark and Watson, 1991). According to this model, anxiety and depression share negative affect (NA) as a common factor, whereas depression is specifically characterized by low levels of positive affect (PA), and anxiety by physiological hyperarousal (PH; i.e. autonomic hyperactivity, motor tension). Since the introduction of this model, it has been frequently used and tested, both in children as well as in adults, and it has become well known in the field of anxiety and depression research.
Empirical evidence for the usefulness and fit of the tripartite model was provided by studies that employed factor analyses (Chorpita, 2002, Laurent and Ettelson, 2001). While several studies investigated the association between anxiety or depression and NA or PA (Chorpita, 2002, Joiner et al., 1996, Laurent et al., 1999), fewer have investigated whether an association with PH was specific for anxiety and not for depression (Joiner et al., 1999, Laurent et al., 2004). In the studies that have investigated the association between PH and anxiety or depression, questionnaires were used to measure PH. However, the items that were used to measure PH tended to overlap with items tapping vegetative symptoms of anxiety, which might explain the associations that were found between PH and anxiety. In our opinion, physiological measures representing arousal would be more appropriate to measure PH (see also Laurent and Ettelson, 2001). To our knowledge, investigations aimed at testing the validity of the PH component of the tripartite model against physiological measures representing arousal, are lacking.
Physiological measures that give an impression of the activity of the autonomic nervous system (ANS) are highly related to arousal. In a state of high arousal, heart rate (HR) is high. HR is influenced by two competing autonomic branches. The sympathetic branch has the function of increasing HR. Hence high arousal is associated with high sympathetic activity. The parasympathetic/vagal branch takes care of decreasing HR. High arousal is therefore associated with low vagal activity. Thus, to assess levels of arousal, it is important to measure HR. In addition, it is important to obtain other estimates of ANS functioning to give an impression of the activity of the separate branches of the ANS.
Some studies have investigated the relation between anxiety and autonomic functioning in children or adolescents (Gerra et al., 2000, Kagan et al., 1988, Mezzacappa et al., 1997). All of these studies found associations between higher HR and higher levels of anxiety. However, these studies were confined to small, nonrepresentative samples. None of these studies focused on both anxiety and depression, so firm conclusions about the specific association between anxiety and autonomic measures representing arousal, in comparison to depression, could not be drawn.
In a previous study, we investigated the association between internalizing and externalizing problems and autonomic functioning in the TRacking Adolescents' Individual Lives Survey (TRAILS) general population sample (Dietrich et al., 2007). For instance, evidence was found for higher HR in individuals with internalizing problems (affective, anxiety and somatic problems taken together). In the present study, anxiety and depressive problems were investigated separately. The TRAILS study provided the opportunity to investigate associations with HR, but also with heart rate variability (HRV). HRV reflects changes in beat-to-beat variations in HR. HRV can be analyzed by means of spectral analysis, which portrays the variance in HR as a function of frequency. The frequency range can be divided into low frequency HRV (generally between 0.04 and 0.14 Hz), and high frequency HRV (above 0.14 Hz). Low frequency HRV (HRV LF) measured in standing position, is primarily sympathetically mediated and vagal effects are inhibited, whereas in the supine posture vagally mediated effects predominate. HRV in the high frequency band is often called respiratory sinus arrythmia (RSA), and is primarily vagally mediated (Mezzacappa et al., 1997). Therefore, based on the tripartite model, we expected anxiety problems to be associated with high HRV LF in standing posture and high HR in both postures. Further, we expected anxiety problems to be related to low HRV LF in supine posture and low RSA in both postures. No signs of hyperarousal were expected to be found in depression. Since the tripartite model does not further describe arousal in depression, it was unclear if we would expect no associations between depression and arousal at all, or even low arousal levels in depression.
In addition to our previous work that only concerned parent-reports, self-reported anxiety and depressive problems were investigated. Examining self-reported problems may be an important extension, since anxiety and depression are highly subjective, and some of the more unobservable symptoms are often under-reported by parents (Comer and Kendall, 2004).
In summary, the aim of the present study was to investigate the putative associations of both parent- and self-reported anxiety problems and depressive problems with physiological measures representing arousal (HR, HRV LF, and RSA) in a large population sample of young adolescents. Based on the tripartite model (Clark and Watson, 1991), we expected specific associations between these physiological measures representing arousal and anxiety, indicating hyperarousal in anxiety, but not in depression.
Section snippets
Sample and procedure
Participants were 10- to 13-year-old young adolescents who participated in the TRacking Adolescents' Individual Lives Survey (TRAILS), a large Dutch general population study (n = 2230). In the TRAILS study, not only young adolescents, but also their parents and their teachers participated. The young adolescents filled out questionnaires at school, in the classroom, under the supervision of one or more TRAILS assistants. In addition to that, a number of physiological and neurocognitive parameters,
Descriptives
Mean scores and standard deviations of all the anxiety, depression, and physiological measures are shown in Table 1:
Parent-reports: CBCL
Table 2 shows the results of the linear regression analyses performed with the Anxiety Problems and Affective Problems scales of the CBCL as predictors, and physiological measures (HR, HRV LF, and RSA, in supine and standing posture) as dependent variables. Betas, p-values and effect sizes are presented. Betas show the direction of the association, while effect sizes give an idea
Discussion
In the present study, the tripartite model (Clark and Watson, 1991) was tested by investigating putative associations of both parent- and self-reported anxiety problems and depressive problems with physiological measures of arousal (HR, HRV LF, and RSA) in a large population sample of young adolescents. Based on the tripartite model, we expected to find specific associations between these physiological measures and anxiety problems, indicating hyperarousal in anxiety, but not in depression.
Acknowledgements
This research is part of the TRacking Adolescents' Individual Lives Survey (TRAILS). We thank all children, their parents and their teachers who willingly participated and everyone who worked on this project and made it possible. Participating centers of TRAILS include various Departments of the University of Groningen, the Erasmus Medical Center of Rotterdam, the University of Nijmegen, the University of Leiden, and the Trimbos Institute the Netherlands. TRAILS has been financially supported
References (43)
- et al.
Assessment of symptoms of DSM-IV anxiety and depression in children: a revised child anxiety and depression scale
Behav. Res. Ther.
(2000) - et al.
Psychometric properties of the Revised Child Anxiety and Depression Scale in a clinical sample
Behav. Res. Ther.
(2005) - et al.
A symptom-level examination of parent–child agreement in the diagnosis of anxious youths
J. Am. Acad. Child Adolesc. Psych.
(2004) - et al.
Frequency, comorbidity, and psychosocial impairment of anxiety disorders in German adolescents
J. Anxiety Disord.
(2000) - et al.
Associations between behavioural and emotional problems in adolescence and maladjustment in young adulthood
J. Adolesc.
(1999) The polyvagal theory: phylogenetic substrates of a social nervous system
Int. J. Psychophysiol.
(2001)- et al.
A model of neurovisceral integration in emotion regulation and dysregulation
J. Affect. Disord.
(2000) - et al.
Advances in empirically based assessment: Revised cross-informant syndromes and new DSM-oriented scales for the CBCL, YSR, and TRF: Comment on Lengua, Sadowksi, Friedrich, and Fisher (2001)
J. Consult. Clin. Psychol.
(2001) - et al.
DSM-oriented and empirically based approaches to constructing scales from the same item pools
J. Clin. Child Adolesc. Psychol.
(2003) - et al.
Comorbidity
J. Child Psychol. Psychiatry
(1999)