The other side of the coin: Using intervention research in child anxiety disorders to inform...

Development and Psychopathology, 14 (2002), 819–841Copyright 2002 Cambridge University PressPrinted in the United States of AmericaDOI: 10.1017.S095457940200408X

The other side of the coin:Using intervention research inchild anxiety disorders to informdevelopmental psychopathology

JENNIFER L. HUDSON,a PHILIP C. KENDALL,b MEREDITH E. COLES,b

JOANNA A. ROBIN,bAND ALICIA WEBBb

aMacquarie University; and bTemple University

AbstractWith advancements in the technology of prevention and treatment of childhood anxiety disorders, informationregarding our understanding of normal and abnormal child development can be enriched. Typically, research hasfocused on developing efficacious and effective interventions with less attention devoted to the impact thisinformation may have on the field of developmental psychopathology. By reviewing the results of both treatmentand prevention studies, several potential contributions of intervention research to the field can be explored. Resultsfrom wait-list or monitoring control groups will be reviewed, providing valuable information regarding the normaltrajectory of the anxious child. Outcomes of children receiving the intervention prove that this pathway can bealtered and is not impermeable. Furthermore, a review of long-term follow-up studies addresses the question ofwhether intervention can change the long-term trajectory of an anxiety-disordered child and prevent disorders inlater life. Contributions to the etiological understanding of the anxiety disorders will also be reviewed: changes invariables considered important in the etiology and maintenance of disorder can be examined in synchronicity withchanges in symptomatology following intervention. An examination of potential developmental predictors oftreatment outcome will also contribute to this review, with a focus on the limitations of the current research ingaining a complete understanding of the relationship between developmental level and outcome. Directionsregarding future research in the study of interventions for child and adolescent anxiety disorders will be discussedwith the aim of promoting further communication between intervention research and the field of developmentalpsychopathology.

The existence of empirically supported treat- functioning have been observed and docu-mented following cognitive–behavioral ther-ments for childhood disorders in general (see

Kazdin & Weisz, 1998; Ollendick & King, apy (CBT) interventions (e.g., Barrett,Dadds, & Rapee, 1996; Kendall, 1994; Ken-2000) and anxiety disorders specifically (see

Ollendick & King, 1998) has consolidated the dall et al., 1997; Mendlowitz et al., 1999; Sil-verman et al., 1999; Spence et al., 2000). Re-notion that, to some degree, child develop-

ment is flexible and malleable and not fixed cently, prevention studies and long-termfollow-up studies have provided additionalor impermeable. Significant and meaningful

changes in children’s symptoms and general evidence that the trajectory of the anxiouschild’s development can be altered (Barrett,Duffy, Dadds, & Rapee, 2001; Dadds, Hol-land, Laurens, Mullins, Barrett, & Spence,

Address correspondence and reprint requests to: Jennifer L.1999; Dadds, Spence, Holland, Barrett, &Hudson, Department of Psychology, Macquarie University,Lauren, 1997; Kendall & Southam–Gerow,Sydney, NSW 2109, Australia; E-mail: jhudson@psy.

mq.edu.au. 1996). These results demonstrate the ability

819

J. L. Hudson et al.820

of interventions (treatment and prevention) to derscored the importance of perceived con-trol, hypothesizing that early experience of re-not only produce change but also inform de-

velopmental theory. Change produced by inter- duced control over one’s environment willincrease a child’s vulnerability to anxiety dis-vention provides evidence that maladaptive

paths can be modified. order. They further proposed that an environ-ment of reduced control leads to the develop-Typically and unfortunately, intervention

research has remained somewhat disparate ment of a cognitive bias whereby the personhas an increased probability of interpreting fu-from developmental psychology (Cicchetti &

Toth, 1992; Shirk & Russell, 1996). This sep- ture situations as uncontrollable. Low percep-tions of control are said to increase the indi-aration is apparent when one considers the

way in which treatments for childhood disor- vidual’s underlying inhibition (Gray, 1987;Kagan, Reznick, & Snidman, 1987). Theders are often developed. Many interventions

for childhood depression and anxiety are model argues that in early development an in-dividual’s cognitive vulnerability acts in adownward extensions of adult treatments and

not necessarily or sufficiently based on theory mediational role between the occurrence ofuncontrollable events and the development ofand data from child development. Not surpris-

ingly, this pattern is also evident in the diag- anxiety. In later development this cognitivevulnerability or “template” is said to act as anostic categories (DSM-IV; American Psychi-

atric Association, 1994; Jensen & Hoagwood, moderator to amplify the expression of anx-iety.1997) applied to childhood presentations of

emotional distress (Schneiring, Hudson, & Hudson and Rapee (in press) proposed amultiple pathway model that outlines the im-Rapee, 2000; Silk et al., 2000) and further

highlights the gap between clinical psychol- portance of the child’s temperament in select-ing specific environments that may furtherogy and developmental psychology (Cic-

chetti & Rogosch, 2002; Shirk, 1998). With maintain the child’s vulnerability to anxiety.This model offers multiple and fluid pathwaysdevelopmental psychopathology as a growing

endeavor, the division between clinical and toward disorder, suggesting that a child whoinherits a genetic predisposition toward anxi-developmental psychology will narrow. De-

velopmental psychopathology offers the com- ety is likely to exhibit increased sensitivity oremotionality, a more avoidant coping style,bined understanding of normal and abnormal

development in an attempt to further elucidate and greater physiological arousal to threat.Greater emotionality and greater physiologi-the causal processes involved in the etiology

of disorder and can, with a shift to interven- cal arousal can give rise to an increased ten-dency to interpret situations as threatening.tion, allow developmental theory to inform in-

tervention research. The benefits of the inter- An inhibited child who avoids novel stimuliis prevented from habituating to potentiallyplay have emerged and will continue to do so

(e.g., Kendall, 1984; Kendall, Lerner, & fearful stimuli and instead the child’s restric-tion of exposure to the world increases theCraighead, 1984; Masten & Braswell, 1991;

Ollendick & Vasey, 1999; Toth & Cicchetti, child’s sensitization to novel stimuli. Themodel also proposes that the child’s tempera-1999).

On the other side of the coin, however, ment allows maladaptive patterns of behaviorto develop from others in his or her environ-past results of intervention research have of-

fered little to advance theories of development (parent, siblings, teachers, peers). Forinstance, the environment provides supportmental psychopathology. In the field of the

anxiety disorders, several etiological models for the child’s avoidant style (parents whopermit or facilitate avoidance), thereby furtherhave recently emerged that attempt to de-

scribe potential pathways to disorder (Chor- increasing the child’s perception of threat, re-ducing the child’s perceived control overpita & Barlow, 1998; Manassis & Bradley,

1994; Rapee, 2001; Vasey & Dadds, 2001). threat, and ultimately increasing the child’savoidance of threat and anxiety.For example, Chorpita and Barlow (1998) un-

Intervention research to inform developmental psychopathology 821

Verbal instructions and/or modeling of these variables may play in the cause or main-tenance of the disorder.anxious behavior are also important in the

model described by Hudson and Rapee (in This paper reviews the results of treatmentand prevention studies of youth with anxietypress). A child may observe and copy the anx-

ious behavior of other significant persons in disorders and highlights the ways in which theresults inform theories of the developmentalthe environment (parents, peers, siblings) as

well as learn how to behave in the face of psychopathology of anxiety. Access to datafrom carefully controlled randomized treat-fearful or ambiguous stimuli. Finally, the on-

set of disorder may be shaped or triggered by ment and prevention trials for childhood dis-orders (and hence more rigorous designs)the child’s social, cognitive, and emotional

development; traumatic experiences; and cul- allows a worthwhile examination of the po-tential bridges between the subdisciplines.tural factors. These factors influence the fluid

transition between normality and “disorder.” Studies that show the superiority of an activetreatment condition to a wait-list control con-In considering the development of anxiety

disorders, no single common pathway can be dition provide evidence of both the plasticityand the stability of the course of the disorder:defined. Rather, the paths are many and var-

ied (Cicchetti, 1993; Cicchetti & Rogosch, without intervention the child’s disorder re-mains stable, and with intervention there are1996; Sroufe & Jacobvitz, 1989). Develop-

mental psychopathologists have argued that observed changes. Additional questions canbe addressed when intervention studies followboth multifinality and equifinality are key

concepts in understanding the complexity and treatment recipients for periods longer thanthe typical 6–12 months. For example, areintricacy of causal processes (Cicchetti & Ro-

gosch, 1996), and they apply here. That is, changes made in childhood maintained acrossthe child’s development and into maturity?there are multiple pathways to one disorder

(equifinality) and one pathway may have mul- Does successful treatment alter the trajectoryfor the sequelae of the disorder?tiple outcomes (multifinality). With regard to

anxiety, the presence of a parent who encour- Also central to this discussion is the degreeto which developmental variables (e.g., age,ages the child’s avoidance may or may not

lead to an anxiety disorder in the child. Alter- level of cognitive or emotional development)may be predictive of a positive treatment out-nately, the presence of an anxiety disorder

does not necessitate the presence of an envi- come. If treatment shows comparative effi-cacy regardless of the child’s initial develop-ronment (e.g., parent) that was encouraging of

avoidance. ment, then one might conclude thatdevelopmental variations (within those in-The present paper will address the ways in

which treatment and prevention of anxiety cluded in the study) are not important in treat-ment effectiveness. Alternately, if certain de-disorders in youth can inform theories of nor-

mal and abnormal development. To do so, one velopmental variables do moderate outcomes,then critical periods in the course of the disor-must first recognize and acknowledge the lim-

itations of intervention research to inform de- der and the timing of treatment are revealed.Important, too, is knowledge of normal de-velopmental theory and etiological theory. We

must, for example, be cautioned not to assume velopment. An understanding of the etiologyof anxiety disorders benefits from an under-that just because a variable changes due to an

intervention that the same variable was there- standing of normal development. In terms ofintervention research, this understanding isfore involved in the etiology or maintenance

of the disorder. Variables that show change crucial in determining the point at which anintervention has been effective. Children mayfollowing intervention may be consistent with

a certain theory of abnormal development but improve over the course of treatment, but howmeaningful are these changes? How do thedo not offer proof of the theory. The docu-

mentation of intervention-produced change treated children, at posttreatment, compare tochildren with normal levels of anxiety? Hasprovides fuel for investigation of the roles that


treatment returned once extreme cases to sures of anxiety and measures of coping anddepressive symptoms. Children on the wait-within normal limits (Kendall, Marrs–Garcia,

et al., 1999)? The usefulness of the informa- list generally showed no change (or verymodest change) over the 8-week duration. Outtion provided by normative comparisons in

intervention research will be discussed. of 20 wait-list children, only 1 child (5%)failed to meet criteria for their primary diag-Finally, we discuss the impact of develop-

mental psychopathology on treatment imple- nosis following the 8-week wait-list. Anxietydisorders appear relatively stable in 9- to 13-mentation with anxiety-disordered youth. De-

spite concerns that intervention research is year-olds over a 2-month period. Consistentdata were reported from a second randomizedatheoretical, the increased communication be-

tween clinical and development psychology clinical trial (Kendall et al., 1997) where,again, CBT-treated youth showed signifi-has enabled intervention research to be guided

by theories of development. cantly greater improvement than wait-listchildren and the children on the wait-list ex-hibited only a very modest change over the 8-

Treatment and Prevention Effortsweek period. Only 6% of wait-listed children

Compared to Wait-Listfailed to meet criteria for their primary diag-

or Monitoring Conditionsnosis compared to 53% of the treated chil-dren.Mostly within the last decade, randomized

clinical trials have been conducted to examine Results from other laboratories are fairlyconsistent: there is little improvement overthe efficacy of CBT for childhood anxiety dis-

orders (Barrett, 2001; Kendall, 1993; Ollen- wait-list periods for anxious children. For in-stance, Silverman and colleagues (1999) re-dick & King, 1998). The majority of large

randomized clinical trials to date have in- ported that following an 8–10 week wait-list,only 13% of the wait-list children no longercluded samples of children with one of three

primary diagnoses: Generalized Anxiety dis- met criteria for their primary diagnosis (com-pared to 64% of the CBT-treated children).order (GAD, formerly Overanxious Disorder),

Separation Anxiety Disorder (SAD), or Social The only somewhat discrepant data set (Bar-rett et al., 1996) had a slightly longer waitPhobia (SP, earlier labeled Avoidant Disor-

der). Generally, these studies have shown that period (12 weeks) and indicated that 26% ofwait-listed cases showed some improvementCBT reduces anxiety for school-aged children

suffering from diagnosable levels of these dis- (compared to 70% of CBT-treated children).The rates of change were higher in this studyorders (e.g., Barrett, Dadds, & Rapee, 1996;

Kendall, 1994; Silverman et al., 1999). Of for both the wait-list and the treated cases.Our review of the evidence from random-special interest to the current paper are the

data from the untreated control conditions. ized clinical trials suggests that only a smallpercentage of anxious children “lose” their di-Specifically, many of these studies used wait-

list controls (to control for the effects of matu- agnosis of an anxiety disorder without treat-ment: childhood anxiety disorders did not typ-ration, naturally occurring events, regression

to the mean, the effects of repeated assess- ically remit naturally after reaching a clinicallevel. This conclusion is consistent with otherments, and other threats to internal validity).

The data from wait-list controls are useful, as research demonstrating the chronicity of anxi-ety disorders (e.g., Kovacs & Devlin, 1998)they provide an indication of the natural

course of children after they have developed and the fact that anxious adults retrospectivelyreport higher levels of anxiety symptoms orclinically significant levels of anxiety.

The first randomized clinical trial for anxi- anxiety-related problems as children com-pared to nonanxious adults (Lipsitz et al.,ety in youth (ages 9–13) compared 16 weeks

of CBT to an 8-week wait-list (Kendall, 1994).In addition to considering treatment out-1994). Children who had received CBT (n =

27) showed a steeper rate of improvement come for youth with anxiety disorders, it isalso worthwhile to consider prevention ef-than children on the wait-list on multiple mea-


forts. The development of prevention pro- tion developed a full anxiety disorder over a6-month follow-up period. This revealed thatgrams aimed at the anxiety disorders is a rela-

tively new endeavor, and hence the number of 54% of the children in the monitoring groupdeveloped a full anxiety disorder compared totreatment studies far outweighs the number of

prevention trials. Nevertheless, several studies only 16% of the children in the interventiongroup, suggesting the superiority of the CBhave begun to emerge in the literature (e.g.

Barrett & Turner, 2001; Dadds et al., 1997; intervention in preventing the onset of fullanxiety disorders.LaFreniere & Capuano, 1997). Prevention

programs have the potential to limit the devel- Rapee and colleagues (Rapee & Jacobs,2002; also cited in Lyneham & Rapee, inopment or progression of anxiety to clinical

levels, and they may be more cost effective press) have been examining the efficacy of aselective prevention program for behaviorallyin reducing the overall incidence of childhood

disorders and their cost to society (Dadds et inhibited preschoolers (3.5–4.5 years). Be-havioral inhibition has been identified as aal., 1997). Recently, Barrett and Turner

(2001) conducted a universal prevention pro- temperamental risk factor for the later devel-opment of anxiety disorders (Biederman etgram designed specifically to prevent anxiety

disorders in sixth grade (10–12 years) chil- al., 1993). In this study, 50% of the behavior-ally inhibited children received a CB interven-dren in a number of schools. The study com-

pared a CB intervention (either teacher led or tion while the other half received no interven-tion. The intervention program was providedpsychologist led) to a monitoring only condi-

tion and found significantly greater changes exclusively to the parents and involved a six-session educational program addressing suchon self-reported anxiety in both teacher-led

and psychologist-led intervention conditions issues as parental overinvolvement, modelingof anxiety, parent anxiety reduction, and ex-from pre- to postintervention compared to the

monitoring condition. Interestingly, the study posure hierarchies for the child. The prelimi-nary results suggest that children whose par-showed a main effect for time, indicating that

all three groups significantly improved on the ents received the intervention were lower ininhibition and anxiety “problems” than chil-self-report measures of anxiety (Spence Chil-

dren’s Anxiety Scale). The study also exam- dren in the monitoring group. Interestingly,these results seem to indicate that the childrenined high-risk children (i.e., children who

scored in the clinical range in the self- in the monitoring group show an increase inanxiety while anxiety in children in the inter-reported anxiety measures). Although the

sample size was small, the evidence points to vention group decreases. These promising re-sults suggest that intervention at even aspromising news that high-risk children in the

interventions groups were more likely to young as 3 years of age can be beneficial inaltering the trajectory of the anxious child.move into the healthy range than high-risk

children in the monitoring condition. Similar In contrast, LaFreniere and Capuano(1997) randomly assigned 42 anxious–with-results were found in a purely teacher-led in-

tervention for fifth to seventh graders (Lowry, drawn preschoolers (ages 31–70 months) toeither an intervention that targeted parent–Barrett, & Dadds, 2001).

In an indicated prevention trial, Dadds and child attachment or a control group. The inter-vention produced significant improvementscolleagues (1997) compared the efficacy of a

10 session school-based child- and parent- on parent–child interactions, parental stress,parental control, child motivation, and childfocused CB intervention to a monitoring

group. In this study, both groups showed re- competence; but no significant improvementsin teacher rated anxious–withdrawn behaviorductions in anxiety at posttreatment, but at the

6-month follow-up this effect was only evi- were observed. The comparison of this studyto the ongoing project conducted by Rapeedenced in the intervention group. Addressing

prevention most directly, Dadds and col- and Jacobs (2002) suggests that for preschool-ers, intervening with parents targeting theleagues examined how many children who

had been diagnosis free before the interven- child’s inhibited behavior may be more effec-


tive in reducing anxiety than targeting the at- trajectory that an anxiety disorder might oth-erwise have on the child’s life, possibly pre-tachment bond.

Together these prevention programs have venting the development of comorbid condi-tions. Later in the paper we will discuss thisprimarily shown that, in comparison to pro-

viding no treatment at all, intervention can issue further by examining longer term fol-low-up data on children and adolescents re-prevent the development of unwanted levels

of anxiety. Moreover, these studies have ceiving interventions.Although not proof of an etiological role,shown that for most children, anxiety levels

overall improve over time, even without inter- it is nonetheless informative to consider whatfactors actually change as a result of an effec-vention. This result is consistent with evi-

dence from studies examining the course and tive treatment for clinically anxious youth.Specifically, what variables change followingprevalence of fears in children, which indicate

that increasing age is associated with a de- efficacious treatment (e.g., cognitive biases,family functioning, etc.) and how might thesecrease in fearfulness (e.g., Gullone, 2000).

Conversely, these prevention studies have findings fit with developmental theories ofanxiety? For example, Barrett, Dadds, andshown that anxiety levels in children at risk

for developing anxiety disorders are likely to Rapee (1996) found that treated children(CBT and family management training forworsen over time.

Are there special benefits to intervening parents) showed significant cognitive changein terms of reductions in the number of threatduring childhood? Prompt intervention for

clinical anxiety means that the child will ex- interpretations of ambiguous situations. Theseauthors also reported decreases (compared toperience immediate relief in anxiety and re-

lated negative states. Skills learned in child- pretreatment) in the avoidant solutions fol-lowing family discussions. In other words,hood may be used, reused, adapted, and used

again. When CBT is effective with children, family discussions no longer lead to avoidantsolutions. Treadwell and Kendall (1996) dem-they become armed with coping skills that can

be carried throughout a lifetime (Kendall, onstrated the role of processing biases as me-diators of treatment outcome. They examined1994). The continued application of coping

skills may lead to further benefits. For exam- anxious children’s self-talk and found thatnegative, but not positive, self-talk mediatedple, Silverman et al. (1999) reported that chil-

dren who had received CBT continued to therapeutic improvement. These findings,taken together, suggest that changes in inter-show meaningful symptomatic improvements

over the posttreatment follow-up period. Fur- pretation biases, environmental support ofavoidance, and negative self-talk influencether, numerous self-report and parent-report

measures also showed gradual, but continual, gains linked to intervention.How do these results mesh with develop-improvements after posttreatment assess-

ments. Similarly, Barrett, Dadds, and Rapee mental theories of anxiety? Consider themodel proposed in Hudson and Rapee (in(1996) reported continued improvements after

therapy was discontinued. These findings sug- press; see also Kendall, 2000) in which oneof the variables defining an anxious vulnera-gest that gains can continue after the interven-

tion and buttress the belief that early interven- bility is a cognitive processing bias. Changesin negative self-talk (Treadwell & Kendall,tion is to be valued. A final potential benefit

of intervention lies in the impact on the devel- 1996) and changes in interpretations of am-biguous situations (Barrett, Rapee et al.,opment of other disorders. Some data suggest

that anxiety may lead to other conditions such 1996) following treatment are consistent withthe hypothesized role of a processing bias inas depression or substance abuse (Brady &

Kendall, 1992; Dobson, 1985; Kendall, the maintenance and/or development of anxi-ety. Another component was environmentalBrady, & Verduin, 2001; Strauss, Last,

Hersen, & Kazdin, 1988; see also Kessler & support, and the results from the Barrett,Rapee, et al. (1996) study are consistent withPrice, 1993). Early intervention may alter the


the benefits of reducing the environmental outcome studies (e.g., Kendall, 1994; Kendallet al., 1997). The total sample was 155 chil-support for avoidance. Taken together the

findings are consistent with current theories dren at posttreatment and 107 children at1-year follow-up. Using diagnostic data fromof childhood anxiety, but they are not direct

tests of etiology. Even when anxiety symp- these points, the children were placed into twogroups: poor response (an anxiety disorder di-toms have been alleviated and a change is ob-

served in family functioning or negative self- agnosis) and good treatment response (noanxiety disorder diagnosis). Although age didtalk, the conclusion cannot be drawn that

these variables were necessarily involved in not predict severity at pretreatment, olderchildren were significantly more likely to beetiology. Future studies should develop re-

search designs to directly test the existing in the poor response group than younger chil-dren at posttreatment, but not at 1-year fol-models (e.g., prospective longitudinal re-

search in so-called high-risk children). low-up. Perhaps older youth had a more per-sistent form of anxiety or were lessdevelopmentally “nonnormative” and, as a re-

Developmental Predictors of Outcomessult, their symptoms may have interfered

for Anxious Youthmore with the developmental challenges thatthese youth face. It is also worth recalling that

Agethe child’s developmental level plays an inte-gral role in the development of responses toPolicymakers (and health care companies)

may look to the data to identify critical peri- adult authority (Kendall et al., 1984): childrenmay become less cooperative with their adultods for intervention (cost-effective plans) for

anxious children. At what developmental age therapist as they grow older, and interventionsgeared for middle childhood may need further(stage) should one implement various strate-

gies to maximize intervention gains and en- adjustment for older youth (Albano, 2000).The onset of adolescence coincides with ansure the maintenance of their benefits? Evalu-

ations of CBT for anxious youth have increase in reports of anxiety associated withnegative evaluation (Schniering et al., 2000),reported efficacy, yet age as a predictor of

treatment outcome has only recently received and anxious adolescents may experiencegreater difficulty building a therapeutic alli-attention (e.g., Berman, Weems, Silverman, &

Kurtines, 2000; Kendall et al., 1997; Sou- ance as a result of their increased social con-cerns. It has also been suggested that thetham–Gerow, Kendall, & Weersing, 2001).

Treatment and prevention studies involving many CBT interventions are geared more to-ward children and less toward adolescentsanxious youth have focused on ranges in age

from 7 to 17 years, but not all of the studies (Southam–Gerow et al., 2001).In another study that provides further in-have reported analyses of age as a predictor

of outcome (e.g., Dadds et al., 1997; Mend- formation regarding age and treatment out-come, Last, Hansen, and Franco (1998) ran-lowitz et al., 1999; Spence, Donovan, &

Brechman–Toussaint, 2000). In those studies domly assigned 56 children (aged 6–17) withschool phobia to a CBT or an attention-pla-that do examine age, although the results are

mixed, it seems that younger children may cebo control condition. At posttreatment,younger children were more likely than olderbenefit slightly more from CBT and more

from parental involvement in treatment than children to achieve 95% school attendance.Although the authors were not able to exam-do older children.

Southam–Gerow, Kendall, and Weersing ine whether age differentially affected the out-comes in the separate conditions (because of(2001) examined predictors of treatment re-

sponse in a combined sample of anxious small sample size), results are consistent withthe findings reported by Southam–Gerow andyouth (ages 7–15). All of these youth had re-

ceived individual CBT (Coping Cat program), colleagues (2001). Perhaps breaking the dys-functional patterns of anxious cognition andand many were a part of previously reported


avoidance may be more difficult for adoles- CBT (based on Kendall et al., 1990), childCBT plus family anxiety management train-cents because their coping styles are more es-

tablished than younger children, the result of ing (CBT+FAM), and a wait-list control con-dition. Children aged 7–14 with a primary di-having had more time to practice maladaptive

coping. agnosis of overanxious disorder, SAD, or SPwere randomly assigned to one of the threeThese results are somewhat surprising be-

cause, when compared to studies examining 12-week conditions. Although both treatmentswere effective, within the younger age groupchildren with both internalizing and external-

izing difficulties, there are some data suggest- (aged 7–10), children were more likely tobe diagnosis free at posttreatment in theing that older children benefit more from CBT

than younger children. For example, Durlak, CBT+FAM condition (100%) than the CBTcondition (56%). In contrast, for the olderFuhrman, and Lampman (1991) conducted a

meta-analysis on CBT for a wide range of group (aged 11–14), there were nonsignifi-cant posttreatment differences between thechildhood psychopathologies based on cogni-

tive developmental stage. A larger effect size CBT (60%) and CBT+FAM (60%) condi-tions. At 1-year follow-up the same age ef-was observed for the group aged 11–13 years

(.92) than for the two younger age groups, fects were observed. Parent training may bemore beneficial for younger than for olderaged 5–7 years (.57) and 7–11 years (.55).

The authors asserted that a certain level of children.Considerations of normal developmentalcognitive development may be essential for

cognitive therapy to be most effective. As changes can inform our understanding of therole of age and parental involvement in treat-children cognitively mature and become capa-

ble of abstract thinking and hypothetic deduc- ment on outcomes. As Daleiden, Vasey, andBrown (1999) suggested, younger childrentive reasoning, they may become more com-

petent at the skills involved in cognitive might have responded more favorably to fam-ily involvement due to the powerful role thattherapy (Weisz & Weersing, 1999; Weisz,

Weiss, Han, Granger, & Morton, 1995). If this parents play in the lives of younger children.Parents of younger children play a larger roleis the case, then what explanation can be pro-

vided for the success of younger anxious chil- in the level of their child’s exposure to out-side social influences than parents of olderdren? Perhaps CBT for anxious youth (a treat-

ment that is often applied and evaluated for children who may have less impact on theirchild’s social exposure. Improving parentingyouth who are at least 8 years of age) matches

well with the developmental level of children skills may be important for younger children,but for older children individual cognitive ex-this age.

Although further exploration is required, ercises and exposure to feared stimuli may besufficient to produce change in anxiety levels.the results suggest that age relations for anx-

ious youth are somewhat different from those Older children are in the process of increasingtheir autonomy and reducing their contactfor youth suffering from other types of psy-

chopathology. Perhaps younger children suf- with the family environment, which maymake parent involvement in treatment morefering from anxiety disorders experience more

distress over their symptoms than do younger difficult.Another factor that has been shown to im-children with externalizing disorders, al-

though the latter group’s symptoms are more pact the influence of parental involvement intreatment on outcome is the presence of pa-distressing to individuals in their environ-

ment. Consequently, anxious children may, in rental psychopathology. This influence ap-pears to depend on the child’s developmentalgeneral, be more motivated to participate in

treatment. level. Berman and colleagues (2000) investi-gated specific predictors of treatment outcomeThe child’s age can be important when

determining whether family involvement in an exposure-based CBT with 106 anxietydisordered children and adolescents (aged 6–in treatment is preferred. Barrett, Dadds,

and Rapee (1996) compared a child-only 17). Although, the authors did not find that


age predicted outcome, the presence of paren- velopmental level and the presence of parentalpsychopathology in the treatment of anxioustal psychopathology appeared to be more

troublesome when treating younger children youth requires further examination. The re-sults with respect to age, parental involve-as compared to older children. The finding is

consistent with the notion of transfer of con- ment, and parental psychopathology seem tobe inconsistent. These results suggest thattrol (Ginsburg, Silverman, & Kurtines, 1995;

Silverman & Kurtines, 1996). In this model, younger children tend to do better when par-ents are involved in the child’s treatment, sug-control is gradually transferred from the thera-

pist to the parent and then to the child. The gesting that parental involvement is not neces-sary for older children. However, oldertherapist uses a reinforcement system to man-

age the child’s behavior and then teaches the children do not improve as rapidly if the par-ents are anxious and the parents receive noparent to use this system. The parent then uses

the system to aid the child in developing self- treatment. In future randomized clinical trials,the issues of age, parental psychopathology,control over his or her own symptoms. The

presence of parental psychopathology may and parental involvement need to be furtherdelineated. Larger sample sizes in randomizedhinder this process and limit the efficacy of

treatment for younger children. clinical trials will allow for a closer examina-tion of what is occurring.In a study by Cobham, Dadds, and Spence

(1998), parental psychopathology again ap- It is also important to understand age ef-fects in prevention outcome research. Al-peared as an influence. The study compared

CBT alone and CBT plus Parental Anxiety though it is clear that there are a number ofrisk factors associated with the developmentManagement (PAM) for two groups of anx-

ious children, those with anxious parents and of anxiety (e.g. parental anxiety, traumatic,stressful, and negative life events), how thesethose without anxious parents. It is worth not-

ing that the parental anxiety management con- risk factors affect children may be influencedby developmental level. Not surprisingly,dition here differed from the FAM condition

reported in Barrett, Dadds, and Rapee (1996), Donovan and Spence (2000) suggest that pre-ventive efforts should take the child’s devel-in that the focus of PAM was on the parent’s

anxiety and not on the child. Cobham and col- opmental level into account when designingpreventive interventions.leagues found an effect at posttreatment (but

not 1-year follow-up) that held for those in At what age should preventive programsfor anxiety be implemented? Currently, thethe CBT alone condition: anxious youth with-

out an anxious parent were more likely to be results on developmental level as a predictorof outcome are inconclusive. For example,diagnosis free than anxious youth who had

one or more anxious parents, but only for the Barrett and Turner (2001) and Lowry–Web-ster and colleagues (2001) conducted studiesolder group (11–14 years). Note also that

there was no effect of age for participants in examining universal prevention programs foranxiety. Although children aged 10 –13 yearsthe CBT+PAM condition (neither at posttreat-

ment nor at 1-year follow-up). These results can benefit from universal prevention pro-grams, information regarding age as a predic-suggest that, when the parent does not receive

treatment for their own anxiety, the older tor of outcome was not reported. Using datacollected from their previous study (Barrett &child does not improve as rapidly. Perhaps

consistent with developmental theory, it is im- Turner, 2001), Barrett, Johnson, and Turner(in press) examined the role of age in preven-portant for the older child to remain autono-

mous and receive individual rather than fam- tive interventions of 692 children at low,moderate, and high risk for anxiety in grade 6ily therapy (therapy with the parents).

However, when parental anxiety exists, these (ages 9 and 10) and grade 9 (ages 14–16).Results indicate that at children in grade 6 re-issues may need to be addressed for the child

to improve at the same rate as older children ported more anxiety at preintervention andsignificantly greater reductions in levels ofwithout an anxious parent.

It is clear that the relationship between de- anxiety at postintervention compared to chil-


dren in grade 9. However, the authors stated tion (89%) compared to females who partici-pated in CBT alone (20%). Furthermore, fe-that the appropriate timing of preventive inter-

ventions remains unclear due to the differen- males who participated in CBT alone weremore likely to be anxiety free if their parenttial intervention effects found at postinterven-

tion and at 1-year follow-up. did not have an anxiety disorder (78%), com-pared to females with an anxious parent(20%). Both of these findings did not occur

Genderfor males, and no gender effects were foundat 1-year follow-up. One can speculate thatDevelopmental theory considers the role of

gender in the negotiation of the tasks of nor- parental psychopathology may not affect fe-males and males in the same manner and thatmal development. Consideration of gender is

warranted, too, in the study of abnormal de- girls may be more sensitive to changes intheir parents’ modeling of anxious behaviorvelopment (Cicchetti & Sroufe, 2000). Cic-

chetti and Sroufe (2000) argued that being than boys.An explanation of these findings may bemale may be a risk factor for certain types of

psychopathology (i.e., conduct disorder) consistent with previous research reportinggender differences in the way in which par-while serving as a protective factor for other

types of psychopathology (i.e., anorexia). ents interact with anxious children. Krohneand Hock (1991) reported that mothers ofGender has been identified as an important

variable in the literature on depression, but it high-anxious girls (versus low-anxious girls)were more likely to intervene competitively inhas not been a consistent factor in anxiety

treatment outcome, with few studies reporting a problem-solving task on which the childwas working independently; the effects weregender effects in outcome for child anxiety

(e.g., Kendall, 1994; Kendall et al., 1997; the opposite for the boys. Perhaps the familytreatment was most applicable to mothers ofTreadwell et al., 1995). Whereas some studies

have reported that gender did not significantly anxious girls and thus produced greaterchanges among girls than boys (e.g., by re-predict outcome for child anxiety (e.g., Sou-

tham–Gerow et al., 2001), some studies have ducing overcontrolling behavior and increas-ing granting of psychological autonomy).reported significant gender effects (e.g., Bar-

rett, Dadds, & Rapee, 1996; Mendlowitz et In studies aimed at preventing anxiety dis-orders in children, the role of gender in pre-al., 1999). Some interesting findings have

emerged. For example, gender was associated dicting outcome is not clear. Lowry–Websterand colleagues (2001) randomly assigned 594with outcome in a study by Mendlowitz and

colleagues (1999). Regardless of treatment children (ages 10–13) to a family group CBTprogram or to a monitoring group. In thiscondition, females reported both less anxiety

at posttreatment and greater use of the coping study, gender was not a predictor of preven-tion outcome. However, Dadds and col-strategies than did males. In another study,

Barrett, Dadds and Rapee (1996) reported that leagues (1999) reported that gender was a pre-dictor of posttreatment outcome in a studygender also played a role in treatment out-

come. Females, who participated in the that compared a school-based intervention toa monitoring group. At postintervention, girlsCBT+FAM condition, were more likely to

have no anxiety diagnosis (83%) compared to were more likely to meet the criteria for anxi-ety disorder than boys were, regardless offemales who participated in the CBT-alone

condition (37%). Findings reported by Cob- whether they received the intervention. Simi-lar results were found in Barrett and Turner’sham et al. (1998) are consistent with Barrett,

Dadds, and Rapee (1996), in that both stud- (2001) universal prevention program. Girls re-ported higher levels of anxiety than boys pre-ies suggest that the child’s gender may affect

the benefits of parental involvement in treat- and postintervention. These findings are con-sistent with community prevalence studiesment. Females with an anxious parent were

more likely to be anxiety diagnosis free if that show a higher prevalence rate of anxietydisorders in females than males. Further re-they had participated in the CBT+PAM condi-


search should be conducted to clarify the role that treatment could have on the sequelae ofanxiety disorders: depressive symptomatologyof gender as a predictor of outcome of preven-

tative interventions. and substance use problems. Eighty-fiveyouth (90% of the original 94) who had beentreated in a randomized clinical trial (Kendall

Longitudinal Researchet al., 1997) an average of 7.4 years prior,

in an Intervention Contextwere evaluated through structured diagnosticinterviews and multiple self-report measuresWithin intervention research, longitudinal

data can be used to assess the maintenance of to assess anxiety levels, depressive symptoms,substance use, and other comorbid disorders,intervention effects over longer periods of

time and the degree to which intervention for both currently and since treatment.Analyses indicate that the treatment gainsthe target disorder may have a desirable im-

pact on the disorder and other comorbid con- were maintained from posttreatment to long-term follow-up on measures of anxiety, de-ditions or the sequelae of the target disorder.

Carefully designed longitudinal studies can pression, coping skills, and internalizing andexternalizing symptoms. Moreover, some sig-also potentially elucidate developmental vari-

ables (e.g., age of onset, gender, duration of nificant improvements were made in copingskills, internalizing symptoms, and externaliz-disorder, and timing of risk and protective

factors) as they relate to therapeutic change ing symptoms in the time from posttreatmentto the 7-year follow-up assessment. Theseand to the maintenance of change over time.

Recently, the results from several long-term findings hint of a developmental maturity, a“sleeper effect” of treatment, or a combina-follow up intervention studies have begun to

emerge in the literature, allowing these question of the two factors. In any case, the resultsbuttress the advantage of intervening at agestions to begin to be addressed. Unfortunately,

longer term follow-up of childhood anxiety 9–13 in the modification of both the develop-mental trajectory of anxiety and the develop-prevention efforts has yet to be carried out.

The follow-up studies that have been con- ment of depression and other sequelae ofchildhood anxiety (Brady & Kendall, 1992).ducted have supported the long-term mainte-

nance of gains to up to 6 years posttreatment Immediate gains were achieved, which thenseem to have equipped the youth to navigate(e.g., Barrett et al., 2001; Kendall & Sou-

tham–Gerow, 1996). In Kendall and Sou- the developmental tasks of adolescence in amore normative and age-appropriate manner.tham–Gerow’s 3-year follow-up study, cli-

ents’ maintenance of gains were evident on Treatment may have facilitated developmentalmaturity, which then led to the improvedthe children’s diagnostic status and self- and

parent-report measures of anxiety, as well as treatment outcomes or vice versa. Another ad-mittedly optimistic explanation could attributeself-reported anxious self-talk and depression.

Barrett and colleagues (2001) reported that at the larger improvements at longer term fol-low-up to a sleeper effect by which initial im-6 years posttreatment, 85.7% of children no

longer met criteria for any anxiety disorder. provements are seen, but more sizable im-provements lie dormant, to be seen at a laterMoreover, they found the individual and fam-

ily-based CBT to be comparably effective at developmental period (at long-term follow-up). Because wait-listed clients eventually re-long-term follow-up. Although both studies

support the long-term clinical utility of CBT ceived treatment, untreated cases are not fol-lowed for extended periods and alternative ex-in the treatment of childhood anxiety disor-

ders, they do not report on developmental planations for the results cannot be ruled out.The lack of anxiety-related problems at thevariables (such as age or gender) as predictors

of treatment response. longer term follow-up may have resultedfrom, for example, maturity or co-occurringIn a very recent project, Kendall, Safford,

Flannery–Schroeder, and Webb (2002) evalu- outside events and not the longer term effectsof CBT. However, the majority of longitudi-ated, not only the long-term maintenance of

treatment gains but also the potential impact nal epidemiological data suggest that anxiety


does not remit without treatment and anxiety- psychological) from drug use than youthtreated successfully. Furthermore, they weredisordered children tend to become anxiety-

disordered adults (e.g., Gittelman, 1986; Last, more likely to have experienced drug with-drawal and overdose.1988).

Preliminary analyses from Kendall and In terms of predictors of substance abuseand dependence, age (at intake) predicted to-colleagues (2002) revealed that the main pre-

dictors of the presence of the primary anxiety bacco, alcohol, and marijuana use: olderyouth were more likely to have tried thesediagnoses (SAD, GAD, and SP) at 7.4-year

follow-up were poor coping skills at pretreat- substances, intuitively they have had moreyears in which to initiate use. In addition, itment (as measured on both child and parent

versions of the Coping Questionnaire) and the is interesting that boys in the sample were lesslikely to have tried alcohol and youth withchild (now adolescent) having experienced

negative life events. Other variables were ex- married parents at intake were more likely tohave tried tobacco. Other analyses revealedamined as well. Self- and parent reports of

anxiety (e.g., symptoms) and the number of that youth with better school performance atintake were less likely to have had substanceself- and parent-reported diagnoses were also

predicted by the occurrence of negative life use problems at long-term follow-up. Exter-nalizing symptomatology at posttreatment andevents after completing the anxiety treatment

program and by impoverished family func- the experience of negative life events werepredictive of hard drug use and substance usetioning in areas such as communication, af-

fective involvement, and behavior control. problems. Further analyses are being con-ducted to more closely examine the relation-Further analyses examined the effects of

anxiety treatment on the developmental tra- ship between treated childhood anxiety andlater substance use, abuse, and dependencejectory of other disorders, such as depression

and substance abuse. Importantly, individuals (see Kendall et al., 2002).In addition to long-term changes associ-who continued to have one or more anxiety

diagnoses (SAD, GAD, SP) at the time of ated with child anxiety treatment, the durabil-ity of prevention efforts warrant investigation.long-term follow-up were significantly more

likely to have depression than children with- Note that longer term follow-up of childhoodanxiety prevention efforts has only recentlyout an anxiety disorder. This finding confirms

previous reports suggesting that anxiety may been undertaken, with a 2-year follow-up byDadds and colleagues (1999). They conductedlead to depressive disorders (Brady & Ken-

dall, 1992). Also, it appears that individuals 1- and 2-year follow-ups of the 128 childrenparticipating in their early intervention andwho met criteria for depression at some point

between posttreatment and long-term follow- prevention group program (Dadds et al.,1997). Although both the treatment and moni-up were significantly more likely to have re-

ceived additional treatment after leaving our toring groups showed improvements (i.e., re-duction in rates of existent anxious disordersprogram. Note that risk factors commonly

found to be associated with the development and prevention of new anxiety) immediatelyand 6 months postintervention, the groups didof depression (gender, negative life events,

previous depression) were not found to be not significantly differ in improvements at 12months postintervention. However, the superi-predictive in the current study.

Similar analyses were conducted to exam- ority of the intervention group was again evi-dent at the 2-year follow-up. In fact, at thatine whether successful treatment of child

anxiety had a preventative effect on the ado- time, the intervention group showed the low-est diagnosis rate at any point in the study.lescents’ use and abuse of substances. Prelim-

inary analyses found that youth treated unsuc- Thus, the importance of longer term follow-up is again underlined to demonstrate the truecessfully for anxiety reported more days of

drinking, were more likely to have smoked effects of prevention efforts, which were notseen at the 1-year follow-up. As with Kendallmarijuana, and reported more consequences

(e.g., social or interpersonal and physical or et al.’s (2002) 7.5-year follow-up, the notion


of the sleeper effect and/or the role of devel- of statistical significance describe whether ornot the treated participants are functioningopmental maturity is suggested, whereby siz-

able changes lie dormant to be manifested within normal limits at the end of treatment.Kendall and colleagues describe how norma-later in the participant’s experience. Dadds

and colleagues also examined age and gender, tive comparisons can be used to evaluate theclinical significance of therapeutic interven-among other potential predictors of chronicity

at long-term follow-up. Although gender was tions (Kendall & Grove, 1988; Kendall,Marrs–Garcia, et al., 1999; Kendall & Shel-predictive of outcome at postintervention (as

discussed earlier), in that girls tended to im- drick, 2000).Normative comparisons allow researchersprove less than boys, this effect disappeared

by 24 months. Pretreatment severity was the to investigate the effectiveness of an interven-tion against a criterion that is independent ofonly predictor of chronicity at 2 years post-

treatment. the initially distressed or diagnosed group. Byusing a predetermined standard, one can po-Results regarding developmental factors

and other possible predictors of long-term tentially identify which of two effective treat-ments might be more beneficial. Moreover,outcome are informative about the degree to

which an anxious developmental trajectory normative comparisons may give insight intothe nature of psychopathology by providingcan be modified and the sequelae of anxiety

changed. The results from long-term treat- information on disorders that rarely reach nor-mative levels, despite otherwise therapeuticment follow-up (Kendall et al., 2002) indicate

that it is possible for successful treatment to intervention.A central issue that occurs when using nor-prevent associated problems such as depres-

sion and substance use. Recently Kendall and mative comparisons is deciding what is con-sidered “normal.” From a medical model, nor-Kessler (in press) made a call to researchers to

shift their thinking toward the consideration mality is viewed as “health,” in that thepresence of psychopathology is abnormal,of child and adolescent treatment not only as

treatment for a target disorder per se but also whereas its absence is viewed as normal. Re-searchers with this viewpoint would be likelyas prevention. The longer term follow-up re-

sults provide testament to this. Future research to use exclusionary criteria when conductingnormative comparisons (Sabshin, 1989). Forin both treatment and prevention should exam-

ine the role of developmental factors and the example, a researcher would exclude subjectsfrom the normative sample if they displayedtiming of risk and protective factors that influ-

ence long-term intervention-produced gains. the characteristics of disorder that are similarto those being addressed in the intervention.Using exclusion criteria runs the risk of creat-

Normative Comparisonsing a nonrepresentative, “supernormal” sam-ple, and comparing the treated sample withNormative data are crucial to an understand-

ing of child and adolescent psychopathology this nonrepresentative sample would be set-ting up an overly stringent criterion.and also central to interventionists’ research.

Normative development is the backdrop From a developmental perspective, nor-mality could be viewed as “average” and rep-against which to judge not only the presence

of psychopathology but also the convincing- resented by a bell-shaped curve showing acontinuous distribution of individual scoresness of treatment-produced outcomes. In ran-

domized clinical trials, efficacy is demon- on a given characteristic. Although this per-spective has its merits, Kendall, Marrs–Gar-strated by testing the statistical significance of

outcomes: changes exceed what would be ex- cia, et al. (1999) caution that researcherscannot assume that scores are normallypected by chance alone. These results do not

identify the “clinical significance” or mean- distributed or that the average is normal (i.e.,report of hallucinations). Moreover, it is notingfulness of the amount of change associ-

ated with the intervention (Kendall, Flan- always clear which population should be usedas the basis for normative comparison (i.e. lo-nery–Schroeder, & Ford, 1999), nor do tests


cal vs. national data). Normative comparisons can inform developmental theory and devel-opmental psychopathology. Our discussionare undeniably associated with definitions of

normal development, and researchers are ulti- now flips the coin to address the ways inwhich developmental models inform treat-mately responsible for determining how to de-

fine normative behavior and choosing norma- ment. In this section, several issues will bediscussed. We first talk more broadly abouttive samples for purposes of the comparison.

Kendall, Marrs–Garcia, et al. (1999) pro- the influence of developmental theory on thetreatment of child anxiety disorders such asvide an example of how to use equivalency

testing for normative comparisons in the eval- the differences in treating anxious adults andanxious children and in defining abnormaluation of treatment outcome. As an illustra-

tion, they used normative comparisons to anxiety. Then we offer a more focused discus-sion on developmental sensitivity in the treat-evaluate the clinical significance of the find-

ings reported by Barrett, Dadds, and Rapee ment of anxiety in early childhood and ado-lescents.(1996). They compared the 6-month posttreat-

ment Child Behavior Checklist (CBCL) Inter-nalizing T scores (Achenbach, 1991) of chil-

Adult versus child interventiondren who had participated in CBT plus familyanxiety management (CBT+FAM) with the Early in the history of child therapy, ap-

proaches to the treatment of childhood anxietydata on normative (nondisordered) children.Prior to treatment, Barrett and colleagues re- were based on the then-current adult theories,

often using the same elements and processes.ported that children in the CBT+FAM groupscored above the clinical cutoff on the CBCL Although adult theory and data may serve as

a starting point for the development of child-(M = 66.3, SD = 7.3; T-score mean of 50;Achenbach, 1991). Six months after treat- focused interventions, this approach, without

other considerations, has potentially seriousment, treated children were within the norma-tive range (M = 45.8, SD = 7.6). Analysis us- pitfalls. When developing child-focused inter-

ventions, it is important to recognize thating the normative comparisons procedure (seeKendall, Marrs–Garcia, et al., 1999) indicated “children” are not a uniform group (Kendall,

1984) and that they are not “little adults”that treated children were indistinguishablefrom a normative sample, whereas the mean (Barrett, 2000). A single treatment approach,

like one used with adults, is unlikely to meetof the control children remained outside thenormative range. Thus, children who once the needs of children at various develop-

mental levels. The cognitive, emotional, andshowed deviant levels of symptoms were ef-fectively treated and progressed to a level that social needs and abilities of clients in early

childhood are quite different than those of ad-falls within the normative range.The normative comparison method allows olescents. Further, any intervention is likely

to have a different impact and meaning forintervention researchers to use knowledge ofnormal development not only to assist in the children at different developmental stages and

with different developmental histories. Forpretreatment determination (classification) ofthe presence of psychopathology but also to example, a therapist might confront and dis-

pute the irrationality of the beliefs of an adultserve as a benchmark to judge the clinical sig-nificance of the efficacy of the intervention. client, but such direct disputation might go

over (not into) the head of a child and evenSuch efforts require the collaboration of de-velopmental psychopathologists and interven- be misperceived by the child as a scolding

(see Toth & Cicchetti, 1999, on the role oftionists alike.causal reasoning in interventions).

Once a child is identified as needing treat-Using Developmental Models

ment, many aspects of his or her developmentto Inform Intervention

need to be considered in determining the bestintervention. General “child” treatments areWe have thus far discussed the ways in which

interventions (both treatment and prevention) not best: often they fail to consider develop-


mental aspects. The “developmental unifor- gosch, 2002; Toth & Cicchetti, 1999) pro-posed that points of natural developmentalmity myth” (Kendall et al., 1984) states that

it is misguided to describe and employ treat- transition, such as adolescence, presentunique opportunities for reorganization. How-ments presumed to be uniform for “children.”

Children are very heterogeneous, varying in ever, Kendall et al. (1984) noted that multiplesimultaneous transitions can be problematiccognitive abilities, behavioral histories, ge-

netic makeup, and social and emotional capa- for maintenance. As Kendall et al. (1984) pro-posed, research is needed to assess the relativebilities. Any application of general strategies

to “children” ignores development and fails to merits of intervening during times of develop-mental transition compared to times of devel-benefit from an understanding of the child’s

normal developmental changes. opmental stability.Toth and Cicchetti (1999) assert that con-Developmental knowledge can direct an

intervention. When determining what type of sidering the interface between pathology andnormality is particularly important whenclinical intervention will be most appropriate,

the child’s emotional, physical, cognitive, be- working with children, as characteristics orbehaviors that are considered normal at onehavioral, and social development should all be

considered (see Kendall et al., 1984). Further, point of development, may be seen as abnor-mal at another point of development. Thisdevelopmental changes in areas such as mem-

ory, language, conditional thinking, categori- point is clearly applicable to work with anx-ious children because anxiety is a normal partzation abilities, and perceptions of rules may

also all influence the effectiveness of treat- of childhood (Gullone & King, 1993). Knowl-edge of normal childhood anxieties helps usment interventions (see Kendall et al., 1984).

A thorough consideration of these important formulate theories of pathological anxiety andcan help define the separation of normal anddevelopmental variables will allow clinicians

to identify both potential limitations and op- abnormal anxiety. Accordingly, the assign-ment of an anxiety disorder diagnosis is bestportunities. Note that although cognitive be-

havioral treatments for anxious youth are made within the context of the child’s currentpoint of development. For example, behaviormanualized and checks of reliability to the

manual document adherence, the treatment is in response to separation from one’s motheris quite different in meaning for children agesapplied with flexibility, allowing the therapist

to adapt the treatment to fit the developmental 7 and 17. Anxiety is only considered to beof a clinical level when it is in exaggeratedneeds of the individual child (Kendall, Chu,

Gifford, Hayes, & Nauta, 1998). The thera- proportion, given the situation and child’spoint of development, and when it interferespist’s flexibility to adapt the treatment manual

to the child’s needs is considered of great im- with the child’s activities and development.Consideration of the boundary betweenportance (see Hudson, Krain, & Kendall,

2001; Kendall et al., 1998). Therapists work- normality and pathology also serves to high-light the adaptive role of anxiety in develop-ing with younger children, who have diffi-

culty with the concept of cognitive restructur- ment. The numerous forms of anxiety that areexperienced in childhood are adaptive and caning, may spend less time on this aspect of the

treatment and more on exposure-based inter- serve to facilitate normal development. Dur-ing infancy, fears of loud noises, strangers,ventions. Spence et al. (2000) reported that

therapists in their treatment study reduced the separation, and physical injury are common(Gullone, 1996; Spence, Rapee, McDonald, &focus of cognitive challenges with younger

children (ages 7–9) because they noted that Ingram, 2001). For school-aged children,however, fears shift toward evaluative and so-younger children had a difficult time grasping

cognitive restructuring exercises. cial concerns and then again to more abstractfears during adolescence (Gullone, 1996). Ex-One interesting question is whether treat-

ment is best enacted during periods of devel- perience with naturally occurring anxiety-pro-voking situations allows children to practiceopmental change or those of relative stability.

Cicchetti and colleagues (Cicchetti & Ro- discriminating between threatening and non-


threatening situations. Children develop the dall, & Steinberg, 1996). Treatment interven-tions are informed by the theory and findingsability to distinguish between situations that

must be endured and threatening situations on the behavior of parents of anxious youth;this means helping parents to break the pat-that can be avoided. Enduring mild stress in sit-

uations allows the child to try different coping terns of avoidance that they are directly andindirectly teaching their children. Treatmentsstrategies and to develop success experiences,

which will contribute to a sense of control. that involve the family have focused on teach-ing parents child management skills, skills toIn addition to utilizing appropriate lan-

guage and presenting concepts that are devel- manage their own anxiety, communicationand social skills, ways to increase the grantingopmentally appropriate for the child’s age and

cognitive, emotional, and social development, of autonomy to their children, and strategiesfor problem solving (e.g., Barrett, Dadds, etit is also important that therapeutic ap-

proaches recognize the network of relation- al., 1996; Cobham et al., 1998; Rapee, Wig-nall, Hudson, & Schniering, 2000; Silvermanships that support the child. Children are intri-

cately linked with their parents, siblings, et al., 1999). Positive results from these typesof programs were reviewed earlier.teachers, and peers, making truly individual

therapy quite different than working with Although there is uniform and widespreadacceptance of the need to consider the child’smore autonomous adults. It is important to

note the differences in how children and current developmental level, the majority ofexisting treatments are implemented withoutadults seek treatment. Parents usually seek

treatment for their child because they are the consideration of how the child’s psychopath-ology developed. Typically, children withones who believe there is a problem within

the child, whereas adults are typically respon- similar presenting problems (the same diagno-sis) are treated with a similar theoretical (andsible for identifying their own psychological

(emotional) distress. Children usually do not practical) approach. It is a practice that is con-sistent with the notion that there are numerousthink their symptoms are as problematic as

their parents report. As a result, Howard pathways to the same disorder. Indeed, theconcept of equifinality (Cicchetti & Rogosch,(1995) suggested that a distinction be made

between the patient (i.e., the child) and the 1996) specifies that a common outcome (e.g.,anxiety disorder) will develop over time fromclient (i.e., the parent or school personnel) to

differentiate between the persons invested in different starting points, indicating that di-verse processes are involved in attaining thetreatment. Further, the primary social network

of the child may vary according to the age shared outcome. Numerous pathways to anxi-ety are currently considered viable. Givenand developmental level. Young children are

primarily linked with their parents and family, also that the factors maintaining a disordermay not be the same as those that led to itswhereas peer relationships and independence

take on much more importance in adoles- development, it is crucial to incorporate con-sideration of maintaining variables into treat-cence. The challenges of treating anxious

youth are compounded by the fact that work- ment. It is possible that an understanding ofthese pathways to disorder will lead us to re-ing with children requires working with their

parents, who themselves may suffer from anx- fined treatment and prevention. Currently thetreatment programs have shown that approxi-iety.

As indicated earlier in this paper, research mately 60% of children improve followingCB treatment. Could it be that modifying thefindings point to the fact that parents play an

important role in the maintenance of anxious treatment programs based on the pathway thatled the child to disorder would bring aboutbehavior in their children by modeling and re-

inforcing anxious and avoidant behavior in- increased treatment efficacy?Most existing interventions for children arestead of encouraging autonomous, coping be-

havior (e.g., Barrett, Rapee, et al., 1996; targeted toward middle childhood (approxi-mately ages 8–14; Barrett, Dadds, et al.,Hudson & Rapee, 2001; Manassis et al., 1994;

Messer & Beidel, 1994; Siqueland, Ken- 1996; Kendall, 1994; Kendall et al., 1990).


Therefore, we will briefly consider how treat- haviorally inhibited, Rapee and Jacobs (2002)exclusively see the parents. In that programment and prevention efforts for young chil-

dren and adolescents could be informed by parents are given psychoeducation about anxi-ety on both its adaptive functions and possibledevelopmental psychopathology.interfering patterns. Parents can also beshown ways in which their behavior may

Early childhood interventionmaintain their child’s anxiety (e.g., letting achild play video games during the day afterRecent efforts to understand the presence of

anxiety symptoms in young children (Spence refusing to go to school). Teaching parentscontingency management procedures and anx-et al., 2001) have indicated that specific sub-

types of anxiety may be less discrete in pre- iety coping skills can be worthwhile. Suchtechniques will enable them to deal more ef-school than in school-aged children. Given

this, treatment of children in early childhood fectively with their child’s anxiety and betterhelp their child (and themselves) cope withis likely to present unique challenges and re-

wards. These children are unlikely to be able feared situations. Considering models of anxi-ety that emphasize the role of attachmentto grasp some of the examples or concepts

that are used with older children. For exam- (Manassis & Bradley, 1994), methods thatseek to increase the child’s sense of securityple, young children are limited in their cogni-

tive development and verbal skills, and they may also be a worthwhile area of attention inthe treatment of anxiety disorders in this agemay not possess the cognitive skills for ab-

stract processes such as meta-cognition or group. Indeed, recent application of parent–child interaction therapy, an intervention thatthinking about their thoughts. For example,

young children may be aware of things that seeks to strengthen the parent–child relation-ship, has produced promising results for thefrighten them (e.g., going to school) but un-

able to articulate the source of these fears or treatment of SAD in young children (Pincus,Choate, & Barlow, 2001).feared outcomes (e.g., fears of ill conse-

quences befalling their caretaker during sepa-ration). Indeed, various theorists have identi-

Intervention during adolescencefied the ages of 5–7 as crucial for theemergence of mediational thinking (Ken- It should not be surprising that, even if the

treatment strategies are similar, some of thedler & Kendler, 1962), and such a change hasimplications for approaches to treatment. many features of interventions that target ado-

lescents will be different from those that focusGiven the limited cognitive capacity of earlychildhood, the anxiety of young children is of- on young children. For disorders in general,

the findings indicate that intervention with ad-ten manifest in behavior, for example, cryingor running away in response to feared situa- olescents produces less positive outcomes,

suggesting that we need to modify the treat-tions. It is also worth noting that the sourceof anxiety is likely to differ for children in ment for adolescents. For instance, the Cop-

ing Cat program for 7- to 13-year-olds hasearly childhood. These children are mostlikely to exhibit fears related to fears of physi- been modified (CAT project; Kendall, Choud-

hury, Hudson, & Webb, 2002) for adolescentscal injury (Spence et al., 2001). Therapeuticinterventions such as questioning one’s anx- (14–18 years) by emphasizing developmen-

tally appropriate content. It is important thatious self-talk are likely to be more difficultand less effective with young children com- adolescents possess more complex cognitive

skills such as abstract thinking and metacog-pared to their older counterparts. Interventionsthat are primarily behavioral in nature may nition. These skills are useful for therapeutic

interventions that require reasoning throughhold the most promise. Further, caretaker par-ticipation in treatment efforts may also be par- hypothetical situations (Weisz & Hawley,

2002) and in identifying anxious self-talk andticularly important for this age group. In fact,in a prevention program described earlier developing coping self-talk. However, the

ability to consider other’s perspectives alsoaimed at 3- to 4-year-old children who are be-


leads to greater concern with other’s opinions relatively stable. Also, prevention studieshave indicated that without intervention chil-of oneself. Therefore, older children may have

a better understanding of the nature of their dren at risk for anxiety are more likely to de-velop clinically significant levels of anxietyfears (e.g., fears of going to school due to

concerns regarding negative evaluation) but over time than children receiving the interven-tion. The findings from adult retrospectivefears of negative evaluation may limit the ad-

olescent’s expression of such anxieties. In- studies suggest that adults with anxiety arelikely to have had disorders as children, indi-deed, common interview and therapy proce-

dures ask the client to describe personal cating that the impact of anxiety is long term.Can intervention change this trajectory and re-feelings and fears to a stranger, a commonly

feared situation for anxious adolescents. As- duce the distress the child experiences acrosshis or her lifetime and, as stressed by Kendallsessment and therapy procedures could be

modified based on such concerns. Normaliz- and Kessler (in press), prevent disorders inlater life? The findings from long-term fol-ing the adolescent’s social concerns and high-

lighting the appropriate developmental level low-up studies have suggested that interven-tion can indeed alter the development of anof his or her concerns may improve the alli-

ance and increase the adolescent’s willingness anxious child. Future research should con-tinue to examine the long-term benefits of in-to disclose social fears.

Adolescents are also likely to have a grow- terventions. In the case of prevention researchlongitudinal follow-up is required of not onlying desire for autonomy. Adolescence is typi-

cally accompanied with a shift in which fam- the cohorts who received the intervention butalso cohorts who were simply monitored.ily interactions become less prominent than

interactions with peers. This shift is consistent Analysis of the monitoring-only cohorts pro-vides valuable information regarding long-with the normal course of development, but

it may not be handled in an optimal fashion. term outcomes when no intervention takesplace. Treatment studies typically provide in-Adolescents vary greatly in the degree to

which they have met earlier developmental tervention for those children in wait-list con-trol conditions, hence, longitudinal data onchallenges, and therefore, the manner and de-

gree to which they strive for autonomy (Cic- these children is not possible or ethical.Often assumed in the field is the notionchetti & Rogosch, 2002). As highlighted by

Cicchetti and Rogosch (2002), interventions that there is applicability of an interventionfor both boys and girls of a range of ages.should optimize strivings for psychological

autonomy. Parental involvement may be used The extant literature has provided inconsistentdata regarding age and gender as predictorsto renegotiate the balance of parental author-

ity. Finally, building on adolescents’ strengths of outcome. Contrary to some research, olderanxiety-disordered children tend to haveand promoting competence will likely be ben-

eficial in both prevention and treatment inter- poorer outcomes than younger anxiety-disor-dered children. Also, there has been some evi-ventions (Cicchetti & Rogosch, 2002).dence that anxious girls may be more sensi-tive to parental involvement in treatment thananxious boys. The results are inconsistent, andSummary and Future Directionsit seems clear that future research in the treat-ment of anxiety-disordered children needs toAdvancements in the field of child and ado-

lescent anxiety have led to confidence in the use larger sample sizes to produce more com-pelling analyses examining age and gender.interventions offered to children and their

families. The growing body of intervention A different approach would be to design,implement, and evaluate age- or gender-focusedresearch has also contributed to the expansion

of knowledge in developmental psychopathol- programs. A prevention or treatment programfor 8-year-old girls may be different from aogy of the anxiety disorders. Wait-list con-

trolled treatment studies inform us that with- program for 12-year-old boys, even if someof the guiding theory and intervention proce-out intervention children’s anxiety remains


dures are similar. In such an approach there is one of a small number of interventions tohave been rigorously and favorably evaluated.not likely to be acceptance of a uniformity

myth about all children or an acceptance of However, we need to move forward and takethe next step to the conceptual level. What arean adult program simply applied to youth. In

this paper, we have offered suggested modifi- some of the psychological forces and conceptsthat contribute to beneficial gains? One areacations of the intervention based on the

child’s age. Further research examining the that may assist in understanding the treatmentmore intimately is the examination of thedevelopmental appropriateness and effective-

ness of the overall treatment package as well treatment processes in child treatment (e.g.,therapeutic alliance, therapist flexibility; seeas the individual treatment components for

children of varying age and gender groups is Russell & Shirk, 1998) and their role in bene-ficial treatment outcomes. The systematicrequired.

Although age is frequently used as a study of therapy moderators and mediatorsexplores specific mechanisms that may pro-marker for developmental change, it may be

a poor proxy for more influential forces. Fur- duce unique psychological change (Kazdin,1995, 2001; Kendall, Flannery–Schroeder, etther, age is also a poor proxy for the diversity

of changes that occur at different rates. Age, al., 1999; March & Curry, 1998). Unfortu-nately, however, the analysis of client andat best, is a simple marker for the various so-

cial, cognitive, physical, and emotional fac- therapist variables in the therapy process withchildren is a relatively neglected field (Mook,tors that reflect psychological maturation. In-

stead, measures specific to the important 1982a, 1982b; Kazdin & Weisz, 1998;Shirk & Russell, 1996; Shirk & Saiz, 1992).developmental forces are needed at pretreat-

ment and posttreatment. The inclusion of Process research, done on treatments knownto be effective, will contribute meaningfullyother more meaningful measures of develop-

mental level at pretreatment will provide the to our conceptual understanding of treatmentoutcomes, and perhaps be better able to pointintervention evaluator with an opportunity to

test for potential developmental moderators of to specific developmental variables in predict-ing change. Earlier in this paper we specu-outcome and determine whether an interven-

tion is differentially effective for youth at dif- lated that older children might not do as wellin treatment because the anxious adolescentferent points in their social, cognitive, physi-

cal, or emotional development. For example, experiences more difficulties in building analliance with the therapist. Alternately, oneCB treatment relies on children being able to

participate in problem solving and cognitive might also speculate that an adolescent, inpart due to the need for autonomy, would dorestructuring; therefore, limitations in the

child’s cognitive development may be predic- better in individual treatment as compared tofamily therapy. Although developmental the-tive of a poor outcome. Not only may the

child’s cognitive development be predictive ory and data would suggest that such hypothe-ses have merit, such notions need much moreof outcome but other indicators of develop-

ment may also prove to be important. The application and evaluation within the field ofprevention and intervention evaluation. Anal-availability of data at pre- and posttreatment

would permit focused analyses of changes on yses of the process of therapy with adolescentclients, for example, can begin to addressspecific developmental factors and the associ-

ation of these changes with treatment-pro- these and related developmentally informedspeculations. We know so little about what isduced gains. The potential to identify media-

tors of therapeutic gain, as well as moderators effective about our treatments, and the explo-ration of the psychotherapy process is the keyof change, more than justifies the inclusion of

developmental assessments. to the advancement of knowledge in this area.As mentioned at the outset, successfulAs we have discussed, CBT for anxiety-

disordered youth has been described as a treatments cannot assume to provide conclu-sive information about how a disorder devel-“probably efficacious” treatment (e.g., Kaz-

din & Weisz, 1998; Ollendick & King, 2000), oped. Nevertheless, some variables (e.g., fam-


ily enhancement of avoidant responding suggested here and elsewhere (Kendall et al.,1999) that it is important to consider the de-[FEAR] effect, cognitive bias), identified by

intervention researchers, change with success- gree to which an intervention can return a pre-viously extremely anxious youth to within aful treatment. It is possible that these vari-

ables represent the mechanism by which normative range of anxiety. Such normativecomparisons (e.g., Kendall et al., 1999) pro-change in anxiety occurs: treatment that pro-

duces change in the FEAR effect and cogni- vide both a metric for evaluating change andan illustration of yet another valuable use fortive biases will produce change in anxiety.

Given that etiological theories of anxiety normative developmental information.In conclusion, much can be learned aboutplace emphasis on the role of cognition and

the role of parents in the etiology and/or developmental psychopathology from inter-vention research given the presence of ade-maintenance of anxiety, this explanation is

plausible. Future intervention research in quately designed studies. To further the com-munication between the subdisciplines, futurechild and adolescent anxiety disorders needs

to more adequately examine the mechanisms intervention research needs to focus on (a) in-cluding measures that assess developmentalthat produce change in childhood anxiety

symptomatology to be able to investigate variables in addition to age and gender, (b)collecting larger sample sizes to allow for ex-more substantially the knowledge of etiology

and maintenance of anxiety. Etiology should amination of possible the differential effectsof the intervention for children with varyingbe more adequately assessed by carefully de-

signed longitudinal studies. developmental levels, and (c) examining theeffective components of treatment and theDevelopmental psychology offers theory

and research that can inform and guide the mechanisms that bring about change. Inter-vention research has also much to gain fromgrowth of intervention. In this paper we also

discussed the importance of using data from developmental psychology in the advance-ment of fine-tuning developmentally appro-studies of normal development to help iden-

tify areas where an anxiety-disordered child priate interventions. Developmental psychol-ogy offers theory, and theories can assist inor adolescent has gone astray. Normative data

offer information about the processes that un- the development of an intervention. Clinicalchild and adolescent psychology contributefold as anxiety is diminished, as well as about

the forces that maintain unwanted anxious the methods of randomized clinical trials andthe search for empirically supported treat-distress. When the efficacy and effectiveness

of an intervention is examined, normative de- ments. Together, the chances for advancementare multiplied.velopment must also be considered. We have

References

Achenbach, T. M. (1991). Manual for the Child Behavior Family treatment of childhood anxiety: A controlledtrial. Journal of Consulting and Clinical Psychology,Checklist/4-18 and 1991 profile. Burlington, VT:

University of Vermont. 64, 333–342.Barrett, P. M., Duffy, A. L, Dadds, M. R., & Rapee,Albano, A. M. (2000). Treatment of social phobia in ado-

lescents: Cognitive behavioral progress as focused on R. M. (2001). Cognitive–behavioral treatment of anx-iety disorders in children: Long-term (6 year) follow-interventions and prevention. Journal of Cognitive

Psychotherapy, 14, 67–76. up. Journal of Consulting and Clinical Psychology,69, 1–7.American Psychiatric Association. (1994). Diagnostic

and Statistical Manual of Mental Disorders (4th ed., Barrett, P. M., Johnson, S., & Turner, C. (in press). Ex-amination of developmental differences in schoolRev.). Washington, DC: Author.

Barrett, P. M. (2000). Treatment of childhood anxiety: based prevention of anxiety and depressive symptomsand disorders. Clinical Child Psychology and Psychi-Developmental aspects. Clinical Psychology Review,

20, 479–494. atry.Barrett, P. M., Rapee, R. M., Dadds, M. R., & Ryan,Barrett, P. M. (2001). Current issues in the treatment of

childhood anxiety. In M. W. Vasey & M. R. Dadds S. (1996). Family enhancement of cognitive style inanxious and aggressive children. Journal of Abnormal(Eds.), The developmental psychopathology of anxi-

ety. New York: Oxford University Press. Child Psychology, 24, 187–203.Barrett, P., & Turner, C. (2001). Prevention of anxietyBarrett, P. M., Dadds, M. R., & Rapee, R. M. (1996).


symptoms in primary school children: Preliminary re- Gray, J. A. (1987). The psychology of fear and stress(2nd ed.). Cambridge: Cambridge University Press.sults from universal school-based trial. British Jour-

nal of Clinical Psychology, 40, 399–410. Gullone, E. (1996). Developmental psychopathology andnormal fear. Behaviour Change, 13, 143–155.Berman, S. L., Weems, C. F., Silverman, W. K., & Kur-

tines, W. M. (2000). Predictors of outcome in expo- Gullone, E. (2000). The development and normal fear: Acentury of research. Clinical Psychology Review, 20,sure-based cognitive and behavioral treatments for

phobic and anxiety disorders in children. Behavior 429–451.Gullone, E., & King, N. J. (1993). The fears of youth inTherapy, 31, 713–731.

Biederman, J., Rosenabum, J. F., Bolduc–Murphy, E. A., the 1990s: Contemporary normative data. Journal ofGenetic Psychology, 154, 137–153.Faraone, S. V., Chaloff, J., Hirshfeld, D. R., & Kagan,

J. (1993). A 3-year follow-up of children with and Hibbs, E. D., & Jensen, P. S. (Eds.). (in press). Psychoso-cial treatments for child and adolescent disorders:without behavioral inhibition. Journal of the Ameri-

can Academy of Child and Adolescent Psychiatry, 32, Empirically based strategies for private practice.Washington, DC: American Psychological Associa-814–821.

Brady, E. U., & Kendall, P. C. (1992). Comorbidity of tion.Howard, K. (1995). Discussant’s comments. Annualanxiety and depression in children and adolescents.

Psychological Bulletin, 111, 244–255. Meeting of the Society for Psychotherapy Research,Vancouver, Canada.Chorpita, B. F., & Barlow, D. H. (1998). The develop-

ment of anxiety: The role of control in the early envi- Hudson, J. L., Krain, A., & Kendall, P. C. (2001). Ex-panding horizons: Adapting manual-based treatmentsronment. Psychological Bulletin, 124, 3–21.

Cicchetti, D. (1993). Developmental psychopathology: for anxious children with comorbid diagnoses. Cogni-tive and Behavioral and Practice, 8, 338–346.Reactions, reflections, projections. Developmental

Review, 13, 471–502. Hudson, J. L., & Rapee, R. M. (2001). Parent–child inter-actions and the anxiety disorders: An observationalCicchetti, D., & Rogosch, F. A. (1996). Equifinality and

multifinality in developmental psychopathology. De- analysis. Behaviour Research and Therapy, 39,1411–1427.velopment and Psychopathology, 8, 597–600.

Cicchetti, D., & Rogosch, F. A. (2002). A developmental Hudson, J. L., & Rapee, R. M. (in press). From tempera-ment to disorder: An etiological model of generalizedpsychopathology perspective on adolescence. Journal

of Consulting and Clinical Psychology, 70, 6–20. anxiety disorder.Jensen, P. S., & Hoagwood, K. (1997). The book ofCicchetti, D., & Sroufe, L. A. (2000). Editorial: The past

as prologue to the future: The times, they’ve been a- names: DSM-IV in context. Development and Psycho-pathology, 9, 231–249.changin. Development and Psychopathology, 12,

255–264. Kagan, J., Reznick, J. S., & Snidman, N. (1987). Thephysiology and psychology of behavioral inhibition inCicchetti, D., & Toth, S. L., (1992). The role of develop-

mental theory in prevention and intervention. Devel- children. Child Development, 58, 1459–1473.Kazdin, A. (1995). Bridging child, adolescent, and adultopment and Psychopathology, 4, 489–494.

Cobham, V. E., Dadds, M. R., & Spence, S. H. (1998). psychotherapy: Directions for research. Psychother-apy Research, 5, 258–277.The role of parental anxiety in the treatment of child-

hood anxiety. Journal of Consulting and Clinical Psy- Kazdin, A. (2001). Progression of therapy research andclinical application of treatment require better under-chology, 66, 893–905.

Dadds, M. R., Holland, D. E., Laurens, K. R., Mullins, standing of the change process. Clinical Psychology:Science and Practice, 8, 143–151.K. R., Barrett, P. M., & Spence, S. H. (1999). Early

intervention and prevention of anxiety disorders in Kazdin, A. E., & Weisz, J. R. (1998). Identifying anddeveloping empirically supported child and adoles-children: Results at 2-year follow-up. Journal of Con-

sulting and Clinical Psychology, 67, 145–150. cent treatments. Journal of Consulting and ClinicalPsychology, 66, 19–36.Dadds, M. R., Spence, S. H., Holland, D. E., Barrett,

P. M., & Laurens, K. R. (1997). Prevention and early Kendall, P. C. (1984). Social cognition and problem solv-ing: A developmental and child–clinical interface. Inintervention for anxiety disorders: A controlled trial.

Journal of Consulting and Clinical Psychology, 65, B. Gholson & T. Rosenthal (Eds.), Applications ofcognitive–developmental theory (pp. 115–148). New627–635.

Daleiden, E. L., Vasey, M. W., &, Brown, L. M. (1999). York: Academic Press.Kendall, P. C. (1993). Cognitive–behavioral therapiesInternalizing disorders. In W. K. Silverman & T. H.

Ollendick (Eds.), Developmental issues in the clinical with youth: Guiding theory, current status, andemerging developments. Journal of Consulting andtreatment of children (pp. 261–278). Needham

Heights, MA: Allyn & Bacon. Clinical Psychology, 61, 235–247.Kendall, P. C. (1994). Treating anxiety disorders in chil-Dobson, K. S. (1985). The relationship between anxiety

and depression. Clinical Psychology Review, 5, 307– dren: Results of a randomized clinical trial. Journalof Consulting and Clinical Psychology, 62, 100–110.324.

Durlak, J. A., Fuhrman, T., & Lampman, C. (1991). Ef- Kendall, P. C. (2000). Guiding theory for therapy withchildren and adolescents. In P. C. Kendall (Ed.),fectiveness of cognitive–behavioral therapy for mala-

dapting children: A meta-analysis. Psychological Bul- Child and adolescent therapy: Cognitive–behavioralprocedures (2nd ed., pp. 3–27). New York: Guilfordletin, 110, 204–214.

Ginsburg, G. S., Silverman, W. K., & Kurtines, W. K. Press.Kendall, P. C., Brady, E. U., & Verduin, T. L. (2001).(1995). Family involvement in treating children with

phobic and anxiety disorders: A look ahead. Clinical Comorbidity in childhood anxiety disorders and treat-ment outcome. Journal of the American Academy ofPsychology Review, 15, 457–473.

Gittelman, R. (Ed.). (1986). Anxiety disorders of child- Child and Adolescent Psychiatry, 40, 787–794.Kendall, P. C., Choudhury, M., Hudson, J. L., & Webb,hood. New York: Guilford Press.

https://www.researchgate.net/publication/279653312_Developmental_Psychopathology_and_Normal_Fear?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz

https://www.researchgate.net/publication/279653312_Developmental_Psychopathology_and_Normal_Fear?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz


A. (2002). The C.A.T. Project. Ardmore, PA: Work- Last, C. G., Hansen, C., & Franco, N. (1998). Cognitive–behavioral treatment of school phobia. Journal of thebook Publishing.

Kendall, P. C., Chu, B., Gifford, A., Hayes, C., & Nauta, American Academy of Child and Adolescent Psychia-try, 37, 404–411.M. (1998). Breathing life into a manual: Flexibility

and creativity with manual-based treatments. Cogni- Lipsitz, J. D., Martin, L. Y., Mannuzza, S., Chapman,T. F., Liebowitz, M. R., Klein, D. F., & Fyer, A. J.tive and Behavioral Practice, 5, 177–198.

Kendall, P. C., Flannery–Schroeder, E. C., & Ford, J. D. (1994). Childhood separation anxiety disorder in pa-tients with adult anxiety disorders. American Journal(1999). Therapy outcome research methods. In P. C.

Kendall, J. N., Butcher, & J. N. Holmbeck (Eds.), of Psychiatry, 151, 927–929.Lowry–Webster, H. M., Barrett, P. M., & Dadds, M. R.Handbook of research methods in clinical psychology

(2nd ed., pp. 330–363). New York: Wiley. (2001). A universal prevention trial of anxiety and de-pressive symptomatology in childhood: PreliminaryKendall, P. C., Flannery–Schroeder, E., Panichelli–Min-

del, S., Southam–Gerow, M. Henin, A., & Warman, data from an Australian study. Behaviour Change, 18,36–50.M. (1997). Therapy for youth with anxiety disorders:

A second randomized clinical trial. Journal of Con- Lyneham, H. J., & Rapee, R. M. (in press). Generalizedanxiety disorder in children and adolescents. In S.sulting and Clinical Psychology, 65, 366–380.

Kendall, P. C., & Grove, W. M. (1988). Normative com- Schneider (Ed.), Anxiety disorders in children and ad-olescents. Heidelberg, Germany: Springer–Verlag.parisons in therapy outcome. Behavioral Assessment,

10, 147–158. Manassis, K., & Bradley, S. J. (1994). The developmentof childhood anxiety disorders: Toward an integratedKendall, P. C., Kane, M., Howard, B., & Siqueland, L.

(1990). Cognitive–behavioral treatment of anxious model. Journal of Applied Developmental Psychol-ogy, 15, 345–366.children: Treatment manual. Available from the first

author, Department of Psychology, Temple Univer- Manassis, K., Bradley, S., Goldberg, S., Hood, J., &Swinson, R. P. (1994). Attachment in mothers withsity, Philadelphia, PA 19122.

Kendall, P. C., & Kessler, R. C. (in press). The impact anxiety disorders and their children. Journal of theAmerican Academy of Child and Adolescent Psychia-of childhood psychopathology interventions on subse-

quent substance abuse: Comments and recommenda- try, 33, 1106–1113.March, J. S., & Curry, J. F. (1998). Predicting the out-tions. Journal of Consulting and Clinical Psy-

chology. come of treatment. Journal of Abnormal Child Psy-chology, 26, 39–51.Kendall, P. C., Lerner, R. M., & Craighead, W. E. (1984).

Human development and intervention in childhood Masten, A. S., & Braswell, L. (1991). Developmentalpsychopathology: An integrative framework. In P. R.psychopathology. Child Development, 55, 71–82.

Kendall, P. C., Marrs–Garcia, A., Nath, S. R., & Shel- Martin (Ed.), Handbook of behavior therapy and psy-chological science: An integrative approach (pp. 35–drick, R. C. (1999). Normative comparisons for the

evaluation of clinical significance. Journal of Con- 56). New York: Pergamon Press.Mendlowitz, S. L., Manassis, K., Bradley, S., Scapillato,sulting and Clinical Psychology, 67, 285–299.

Kendall, P. C., Safford, S., Flannery–Schroeder, E., & D., Miezitis, S., & Shaw, B. F. (1999). Cognitive–behavioral group treatments in childhood anxiety dis-Webb, A. (2002). Long-term outcomes of children

treated for anxiety disorders. Manuscript in prepara- orders: The role of parental involvement. Journal ofthe American Academy of Child and Adolescent Psy-tion.

Kendall, P. C., & Sheldrick, R. C. (2000). Normative data chiatry, 38, 1223–1229.Messer, S. C., & Beidel, D. C. (1994). Psychosocial cor-for normative comparisons. Journal of Consulting

and Clinical Psychology, 68, 767–773. relates of childhood anxiety disorders. Journal of theAmerican Academy of Child and Adolescent Psychia-Kendall, P. C., & Southam–Gerow, M. (1996). Long-

term follow-up of treatment for anxiety disordered try, 33, 975–983.Mook, B. (1982a). Analyses of therapist variables in ayouth. Journal of Consulting and Clinical Psychol-

ogy, 65, 883–888. series of psychotherapy sessions with two child cli-ents. Journal of Clinical Psychology, 38, 63–76.Kendler, H. H., & Kendler, T. S. (1962). Vertical and

horizontal processes in problem solving. Psychologi- Mook, B. (1982b). Analyses of client variables in a seriesof psychotherapy sessions with two child clients.cal Review, 69, 1–16.

Kessler, R. C., & Price, R. H. (1993). Primary prevention Journal of Clinical Psychology, 38, 263–274.Ollendick, T. H., & King, N. J. (1998). Empirically sup-of secondary disorders: A proposal and agenda.

American Journal of Community Psychology, 21, ported treatments for children with phobic and anxi-ety disorders: Current status. Journal of Clinical607–633.

Kovacs, M., & Devlin, B. (1998). Internalizing disorders Child Psychology, 27, 156–167.Ollendick, T. H., & King, N. J. (2000). Empirically sup-in childhood. Journal of Child Psychology and Psy-

chiatry, 39, 47–63. ported treatments for children and adolescents. InP. C. Kendall (Ed.), Child and adolescent therapy:Krohne, H. W., & Hock, M. (1991). Relationships be-

tween restrictive mother–child interactions and anxi- Cognitive and behavioral procedures (pp. 386–425).New York: Guilford Press.ety of the child. Anxiety Research, 4, 109–124.

LaFreniere, P. J., & Capuano, F. (1997). Prevention inter- Ollendick, T. H., & Vasey, M. W. (1999). Developmentaltheory and the practice of clinical child psychology.vention as means of clarifying direction of effects in

socialization: Anxious–withdrawn pre-schoolers case. Journal of Clinical Child Psychology, 28, 457–466.Pincus, D. B., Choate, M. L., & Barlow, D. H. (2001,Development and Psychopathology, 9, 551–564.

Last, C. G. (1988). Anxiety disorder in childhood and July). Treatment of young children with separationanxiety disorder using parent–child interaction ther-adolescence. In C. Last & M. Hersen (Eds.), Hand-

book of anxiety disorders (pp. 531–541). New York: apy. In J. L. Hudson & D. Pincus (Chairs), Parentsand their role in the etiology, assessment, and treat-Plenum Press.

https://www.researchgate.net/publication/227874721_Internalizing_Disorders_in_Childhood?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz

https://www.researchgate.net/publication/227874721_Internalizing_Disorders_in_Childhood?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz

https://www.researchgate.net/publication/16141922_Analyses_of_therapist_variables_in_a_series_of_psychotherapy_sessions_with_two_child_clients?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz






ment of anxiety disorders in children. Symposium at (2001). Examining outcome variability: Correlates oftreatment response in a child and adolescent anxietythe Annual Meeting of the World Congress of Behav-

ioural and Cognitive Therapies, Vancouver, Canada. clinic. Journal of Clinical Child Psychology, 30,422–436.Rapee, R. M. (2001). The development of generalized

anxiety. In M. W. Vasey & M. R. Dadds (Eds.), The Spence, S. H., Donovan, C., & Brechman–Toussaint, M.(2000). The treatment of childhood social phobia: Thedevelopmental psychopathology of anxiety (pp. 481–

503). New York: Oxford University Press. effectiveness of a social skills training-based, cogni-tive–behavioural intervention, with and without pa-Rapee, R. M., & Jacobs, D. (2002). The reduction of tem-

peramental risk for anxiety in withdrawn preschool- rental involvement. Journal of Child Psychology andPsychiatry, 41, 713–726.ers: A pilot study. Journal of Behavioural and Cogni-

tive Psychotherapy, 30, 211–216. Spence, S. H., Rapee, R. M., McDonald, C., & Ingram,M. (2001). The structure of anxiety symptoms amongRapee, R. M., Wignall, A., Hudson, J. L., & Schniering,

C. A. (2000). Evidence-based treatment of child and preschoolers. Behavior Research and Therapy, 39,1293–1316.adolescent anxiety disorders. Oakland, CA: New Har-

binger Publications. Sroufe, L. A., & Jacobvitz, D. (1989). Diverging path-ways, developmental transformations, multiple etiolo-Russell, R. L., & Shirk, S. R. (1998). Child psychother-

apy process research. Advances in Clinical Child Psy- gies and the problem of continuity in development.Human Development, 32, 196–203.chology, 20, 93–124.

Sabshin, M. (1989). Normality and the boundaries of psy- Strauss, C. C., Last, C. G., Hersen, M., & Kazdin, A. E.(1988). Association between anxiety and depressionchopathology. Journal of Personality Disorders, 3,

259–273. in children and adolescents with anxiety disorders.Journal of Abnormal Child Psychology, 16, 57–68.Schniering, C., Hudson, J. L., & Rapee, R. M. (2000).

Issues in the assessment and diagnosis of anxiety dis- Toth, S. L., & Cicchetti, D. (1999). Developmental psy-chopathology and child psychotherapy. In S. W.orders in children and adolescents. Clinical Psychol-

ogy Review, 20, 453–478. Russ & T. H. Ollendick (Eds.), Handbook of psycho-therapies with children and families (pp. 15–44).Shirk, S. R. (1998). Cognitive development and child psy-

chotherapy. New York: Plenum Press. New York: Kluwer Academic/Plenum Press.Treadwell, K. H., Flannery–Schroeder, E. C., & Kendall,Shirk, S. R., & Russell, R. L. (1996). Change processes

in child psychotherapy: Revitalizing treatment and re- P. C. (1995). Ethnicity and gender in relation to adap-tive functioning, diagnostic status, and treatment out-search. New York: Guilford Press.

Shirk, S. R., & Saiz, C. C. (1992). Clinical, empirical, come in children from an anxiety clinic. Journal ofAnxiety Disorders, 9, 373–384.and developmental perspectives on the therapeutic re-

lationship in child psychotherapy. Development and Treadwell, K. H., & Kendall, P. C. (1996). Self-talk inanxiety-disordered youth: States-of-mind, contentPsychopathology, 4, 713–728.

Silk, J. S., Nath, S. R., Siegel, L. R., & Kendall, P. C. specificity, and treatment outcome. Journal of Con-sulting and Clinical Psychology, 64, 941–950.(2000). Conceptualizing mental disorders in children:

Where have we been and where are we going? Devel- Vasey, M. W., & Dadds, M. R. (Eds.) (2001). The devel-opmental psychopathology of anxiety. New York: Ox-opment and Psychopathology, 12, 713–735.

Silverman, W. K., & Kurtines, W. M. (1996). Anxiety ford University Press.Weisz, J. R., & Hawley, K. M. (2002). Developmentaland phobic disorders: A pragmatic approach. New

York: Plenum Press. factors in the treatment of adolescents. Journal ofConsulting and Clinical Psychology, 70, 21–43.Silverman, W. K., Kurtines, W., Ginsburg, G. S., Weems,

C. F., Rabian, B., & Serafini, L. (1999). Contingency Weisz, J. R., & Weersing, V. R. (1999). Developmentaloutcome research. In W. K. Silverman & T. H. Ollen-management, self-control, and education support in

the treatment of childhood phobic disorders: A ran- dick (Eds.), Developmental issues in the clinicaltreatment of children (pp. 457–469). Needhamdomized clinical trial. Journal of Consulting and

Clinical Psychology, 67, 675–687. Heights, MA: Allyn & Bacon.Weisz, J. R., Weiss, B., Han, S. S., Granger, D. A., &Siqueland, L., Kendall, P. C., & Steinberg, L. (1996).

Anxiety in children: Perceived family environments Morton, T. (1995). Effects of psychotherapy withchildren and adolescents revisited: A meta-analysis ofand observed family interaction style. Journal of

Clinical Child Psychology, 25, 225–237. treatment outcome studies. Psychological Bulletin,117, 450–458.Southam–Gerow, M., Kendall, P. C., & Weersing, V. R.

https://www.researchgate.net/publication/19795348_Association_Between_Anxiety_and_Depression_in_Children_and_Adolescents_with_Anxiety_Disorders?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz






https://www.researchgate.net/publication/11844162_Examining_Outcome_Variability_Correlates_of_Treatment_Response_in_a_Child_and_Adolescent_Anxiety_Clinic?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz







https://www.researchgate.net/publication/12765877_Contingency_management_self-control_and_education_support_in_the_treatment_of_childhood_phobic_disorders_A_randomized_clinical_trial?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz










https://www.researchgate.net/publication/231840507_The_reduction_of_temperamental_risk_for_anxiety_in_withdrawn_preschoolers_A_pilot_study?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz






https://www.researchgate.net/publication/232446030_Cognitive_development_and_child_psychotherapy_Perspectives_in_developmental_psychology?el=1_x_8&enrichId=rgreq-5bb0491e5a516ba396dde6671c5d81f7-XXX&enrichSource=Y292ZXJQYWdlOzEwOTMxNDkyO0FTOjE5NTkwNzg1NDMxMTQzMUAxNDIzNzE5NDgwNzUz



Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

The other side of the coin: Using intervention research in child anxiety disorders to inform...

Documents

Transcript of The other side of the coin: Using intervention research in child anxiety disorders to inform...