SOCIAL - Journal of Anxiety Disorders. VOL 12. No. 1. pp....

Info iconThis preview shows pages 1–11. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Journal of Anxiety Disorders. VOL 12. No. 1. pp. 1-20. 1998 Copyright El 1998 Elmiu Science Lt'd P31131110“ Printed in on; USA. All rights res:ch ossmmmsswoo + .00 P11 80887-6185(97)00046-7 RESEARCH PAPERS Social Anxiety and History of Behavioral Inhibition in Young Adults MICHELE A. MICK. M.A.. AND MICHAEL J.'TELCH,1=H.D. The University of Texas a; Armin Abstract—To. evaluate the relationship between the «$131th tompcmmcnt behatioml inhibition (Eli and anxiot)‘ sympto’mology, we investigated differences inr‘wospective reports of childhood BI among undergraduates reporting one of the following: (at Saciiil anxiety (:2 = Ill). (b) gcncmlizod anxiety (I: z 13), (c) both social and generalized anxiety in = 153. undid) minimal social and generalized anxiety (n == 38). Contrary to tho hypolhcsis that BI acts as it nonspecific risk factor, for anxiety symptoms. our findings revealed that a history of childhood Ellwas associated with symptoms of social phobia but not generalized anxiety disorder. Moreover. panicipums'displaying‘synip. toms of both gcncralizcd anxiety disordcr and social phobia were no mo“: li‘k‘ély to show a childhood history of El than participants displaying social phobia symptoms alone. These data suggest that a childhood history of Bi may he more strongly associated with aduit social anxiety than some othcr types of anxiety pathology. © l998 Elsoviet Science Ltd , Kagan, Reznick, and Snidman (1988) reported identifying a laboratory-based temperamental construct, behavioral inhibition to the unfamiliar (Bi), that re— mains stable across childhood. A child is said to exhibit a BI temperament when his or her responses to novel stimuli or events are consistently characterized by excessive sympathetic arousal and behavioral withdrawal. Examples of in— hibited behavioral responses are cessation of ongoing activity and vocalization, avoidance. retreat, isolation. extended latency to interact with novel persons or objects. and clinging to caregiver. BI is estimated to be present in 10 to l5% of normal Caucasian‘z- to 3- year-olds (Kagan at al., 1988). Furthermore, re- Correspondence and reprint requests should be son! to Michacl J. Tcleh. Departmcm of Psy- chology Mazes 330. The University of Texas at Austin, Austin, TX 787l2. Email: Telch®mail. utzxassdu. searchers propose that Bl has a genetic basis (Robinson, Kagan. Rezniclt. Cerley, l99'2). is detectable early in life (as early as 9 months; Kagan 8; Snid- man. 1991), may be stable across time (predicting behavior 10 years later: Bled- erman er al.. 1993'), and, thus. influences personality development. Questions regarding the role of BI in the development of psychopathology have been raised. In particular. Rosenbaum. Biederman. Hirshfeld. Bolduc. and Chaloff (199“ and Biederrnan. Rosenbaum, Chaloi’f, and Kagan (1995) have discussed the issue of the specificity of the influence of BI. While data are lacking on the relationship between childhood BI and most forms of psycho- pathology, increased attention has been drawn to the possible linkage between childhood BI and anxiety disorders. Progress in determining whether BI is asso- ciated with anxiety disorders in general. or with specific anxiety disorders. will contribute to the understanding of the anxiety pathology and its various expres- sions. Moreover. as interventions for anxiety disorders become more disorder specific. it becomes increasingly important to determine whether a marker of increased risk for pathological anxiety is assodated with the development of particular anxiety disorders. It has been suggested that early interventions that prepare parents of behaviorally inhibited children to respond appropriately to the emergence of specific symptoms may be helpful in deterring the progression of a behaviorally inhibited temperament into an anxiety disorder (Pollock. Ro~ senbaum, Marts. Miller. & Biederman. l995; Rosenbaum, Biederman. Poi- lock, & Hirshfeld. 1994). Evidence bin link between Bland anxiety disorders has come primarily from studies of two samples of children obtained from two different populations: 3-. clinically-derived. cross—section of children whose parents received [remnant for panic disorder and agoraphobia (FDA; .Rosenbaum et 211., 1988) and a non— clinical sample followed longitudinally by Kagan and his associates (see Kagan et al. 1988). (For an excellent critique of Bi studies, see Turner. Beidel. and Wolff. 1996.) Results from both the direct diagnostic assessment of children exhibiting Bi and from family studies suggest that a childhood history of BL especially in its stable form and/or in combination with a familial history of pathological anxiety, increases risk for anxiety disorders. Studies have shown that the dren of panic disorder patients were more likely-to be behaviorally inhibited than the children of normal controls (Rosenhaum et al.. 1988). In addition. the parents of behaviorally inhibited children had a higher prevalence of any anxiety disorder, any childhood anxiety disorder. and both a child and adult anxiety diagnosis (Rosenbaum, Biederman. Hirshfeld, Bolduc. Faraone et al.. 1991:. Furthermore, behaviorally inhibited children were found to have higher rates of phobic disorders. overanxious disorder (CAD), and multiple anxiety disor— ders. (For a review. see Biedetman, 1990.) Biederman at al. (1993) described BI ‘15 a“ ' ianxiel)‘ Pmnc" diathesis that represents a risk factor for the develop mam “l” “Mie'l’ disorders in general. but not any one anxiety disorder. The proposition that childhomi Bl acts as a nonspecific risk factor for anxifly disor- l i SOCIAL ANXIETY AND BEHAVIORAL INHIBITION f, ders leads to the prediction that adults reporting either social~evaluativc anxiety associated with social phobia or anxiety symptoms associated with another anxi- ety disorder should not differ in the extent to which they exhibited Bl in child- hood. Howerer. some findings from studies of behaviorally inhibited children and their families also point to possibility that social anxiety and avoidance may be particularly associated with BI. Although social phobia was not directly assessed in the behaviorally inhibited children. Biederrnan (1990) noted that the phobic disorders found in association. with El frequently included social fears. Likewise. social concerns may contribute to a diagnosis of GAD. Further- more. :o‘oidant disorder was among those found more frequently in the behay- iorally inhibited children (Biederman. 1990). especially those who had exhibited a more Stable form of BI (Hirshfeld er al" 1992). ,At a 3-year follow-up, the significantly higher rate of multiple anxiety disorders found in the stable inhib~ ited child was attributed to higher rates of avoidant disorder. Moreover, for children without an anxiety disorder diagnosis. the parents of children with Bl were more likely to have social phobia than the parents of children without Bl (Rosenbaum et at. 1992). Using another sample of temperamentally inhibited and uninhibited children. Rickman and Davidson (1994) found that the parents of the inhibited children reported significantly less extroversion: however, they did not differ from the parents of uninhibited children with respect to neuroti~ cism. . Although Turner et al. 0996) point to a number of methodological weal‘o messes in the existing Bl studies. they concluded that children with BI and their parents. in comparison to children and parents in control groups. have more anxiety disorders. especially disorders of a social-evaluative narure. Yet, they point out that the exact nature of the relationship between BI and social anxiety is not clear. They propose that factors such as the heritable trait of introversion (in .interabtion with unfamiliar environmentsl or a familial hisrory of anxiety may underlie the observed association between BI and excessive anxiety in social—evaluative situations. Turner et a1. H996} suggested that stable BI may increase vulnerability to anxiety disorders. especially those including maladap— ti ve social anxiety, yet be neither necessary nor sufficient for their development. The. proposition that childhood Bl increases vulnerability for anxiety disor‘ ders that are social—evaluative in nature leads to the prediction that adults with Symptoms of social phobia would be more likely than those with symptoms of another anxiety disorder to report Bl during childhood. Furthermore, persons having symptoms of herb social phobia and another anxiety disorder would he expected, to report a level of Bi comparable to the level reported by persons with social phobia symptoms alone. Thus. the overall aim of the study was to compare two rival hypotheses concerning the nature of the relationship between childhood El and anxiety disorders. Specifically, we tested whether a historyof childhood Bl is associated equally with symptoms of generalized anxiety disorder or social phobia (general risk hypothesis) or whether a childhood history of BI is associated with symp- toms of social phobia but not symptoms of generalized anxiety disorder (speci- ficity hypothesis). We also examined whether controlling for depression and state anxiety altered the associations between childhood BI and social phobia or generalized anxiety disorder symptoms. The decision to compare persons reporting symptoms of social phobia with those reporting symptoms of generalized anxiety disorder (GAD) was based on several considerations. First, the diagnosis of CAD had been relatively common for the Bl children in the studies reviewed above. However. DAD has been eliminated from the fourth edition of the Diagnostic and, Statistical Manual of Mental Disorders (DSM-IV, American Psychiatric Association. 1994) with the diagnosis of GAD being recommended for use in its place. Second. it was desir- able to use an anxiety disorder known to have a relatively high rate of comorbid ity with social phobia. The high comorbidity rates of social'phobia and GAD in clinical samples (Brown & Barlow. [992) suggested that it would be possible to obtain an adequate sample (if-participants reporting symptoms of both social phobia and GAD. Finally. it was also necessary to select an anxiety disorder with an age of onset early enough to be prevalent in a population of young adults. The finding that the expected age of onset of CAD is earlier than panic disorder (Scheibe & Albus. 1992) made GAD a better candidate for a compari- son anxiety disorder. even though previous BI research involved children of parents with panic disorder. The young age and nonclinical status of participants in previous BI research influenced the selection of participants for this study. First, an undergraduate population was selected for this study in order to balance the goals of remaining focused upon a younger age group and accessing a group whose average age would exceed the expected onset for social phobia (15.5 years; Schneier. John- son, Homig, lljebowitz, & Welssmam 1992). Second. a nonclinical' population was selected because, although. significant proportions of BI children were found to have diagnosable anxiety disorders, it does not appear that most had previously engaged in treatment-seeking. Thus, our participants were recruited based upon their endorsement of anxiety symptoms and not on the basis of seeking treatment. METHOD Participants Undergraduates (N = 76) enrolled in introductory psychology at the Univer~ sity of Texas at Austin took part in the study. Selection was based on meeting criteria (see below) for one of the following four groups: (a) social anxiety (S A), ( b) generalized anxiety (GA). (c; both SA and GA (mixed), and (d) neither SA nor GA (control). Their participation fulfilled a course requirement. Demo- graphics for the total sample and each group are reported in Table 1. TABLE 1 Gnocr SIZES. GENDER AND Ernsuc COMPOSITION. AND Lieu Aces Total SA GA. Mixed Comml n 76 IO l 3 l 5 38 Gender (9%) Males .34 40 Ill 27 37 Female 66 60 69 73 63 Ethnic t “cl African American I l) 0 0 3 Asian i S 30 O 33 8 Hispanic 20 30 l 5 20 1 8 White. 6 1 30' 35 47 66 Other 4 l0 0 0 S A ge M 18.92 18.80 l8.77 19.33 [8.8-1 (SDI ( l .23) (1.23) (0.93) (1.453 (1.24) Note. SA = social'anxiety group; GA = generalized anxiety group; Mixed : group with social and generalized anxiety: Control = group without elevated social or generalized anxiety. Procedure Participant recmimzent'and screening. Participants entered the study in one or' two ways: phone solicitation or sign~up sheets. The Panic, Anxiety. and Social Phobia Quesrionnaire (PASPQ; see below), an author-constructed self—report anxiety disorders screening questionnaire was'admlnistered to approximately 2000 students in selected introductory psychology classes consisting of 100 to 500 students. Those reporting elevated social anxiety. generalized anxiety, borh social and generalized anxiety, or neither social or" generalized anxiety on the PASPQ were recruited by phone or sign-up sheet. A total of 168 students partici- pated in further tesfihg.. We administered a battery of self—report instruments (described below) to all of the 168 initial participants in groups of 5 to 30. The PASPQ was included in this battery and'responses from thisadministrationof the PASPQ were used in the participant classification procedures described below. Participants re- porting distress due to symptoms associatedwith panic disorder were excluded. Those reporting clinically significant symptoms of anxiety or depression were offered treatment referrals. Classification ofparticipants. After assessment batteries were completed, we assigned each participant to one. of the four groaps (social anxiety, generalized anxiety, both social and generalized anxiety, or neither social or generalized anxiety) based upon their scores on empirically supported instruments used in the assessment or" patients with social phobia and GAD. We adopted Turner. TABLE 2 Group CLASSIFICATION Cam-nim- AND Guour DIFFlZRENCES 0N CRITERIA w SA GA Mixed Control Classification criteria Measures assigned cut- off scores SPAi .260 <60 260 <60 PSWQ a <52 252 352 <<‘ Interference due to anxi— a ety (PASPQl Social yes — Generalized -— yes 3:5 ~ Prominent symptoms y s ‘— (PASPQl SP yes no yes no GAD no yes yes n Means (standard dcvia~ 0 dons) for classification measures Measures assigned cut- off scores 3:2,; (16.20) 28.9,0‘I (17.32) 88.73‘ (ELBO) 28.18“ El"??? 4“. or {7:63 63.31“ 7. v ' 7m Interference due to amtlo ) ( 04) 66.00 (650) 3417'“ £1591 ety item iPASPQ) Social 7 00‘ (000) 0.54” (O ‘ . . .l . .66) 2.07“ (0.26) 030° '046 Generalized 129a (o 49) '2 33» 05 . i I ’ . . . . l 2. h ‘ ' ” sympmm severity ( ) 20 (0.41) 0.6? 47.3% (PASPQ) SP 1 95‘ (0 16) 87” (O 39 . .v . . . ) 2.05‘ (0.21) 0.53‘ too: GAl) LM‘ (0.24) 2.42" (0.56) 2.18“ (0.50) 0.72“ 50.32) M" ._._.___._..___~ __ Nate. SA r—r social anxiety group; GA = generalized anxiet 'mu ‘ ' = " f and-generalized anxiety; Control = group without elevated szclal of’ggi?rea<ljizedg:::£t:ugij:lail Secret Photos and Anxiety Inventory; PSWQ = Penn Statc‘Worry Questionnaire PASP6= dean'— Anxrery, and Social Phobia Questionnaire; SP = social phobia symptoms (based lipou 4items t' m. the PASPQ); GAD = generalized anxiety disorder symptoms (based upon 4 items l’rommth: PASPQ). Means in the same row that do ' ‘ 1 . , not share so rscn ts dtffer at < ' " * honestly Significant difference comparison. PC p p ‘05 m the nu} Bcidel, Dancu. and Stanley’s (I989) recommended cut-off score of 60 on the Social Phobia and Anxiety Inventory.(SPA.l) to classify participants as hayinc significant social anxiety. Based upon Molina and Borkovec's (1994) summarv of findings for analog clinical samples, we designated a score of 52 on the Penn State Worry Questionnaire (PSWQ) as the cut-off score to claSsify participants as havmg srgniticunt generalized anxiety. The cut-off secre requirements for each group are indicated in Table 2; To better accomplish Study goals, an additional set of classification criteria (summarized in Table 2) were imposed. Because we were interested in the asso- ciation between a history of behavioral inhibition and current pathological levels of anxiety, we sought to increase the likelihood that participants in the anxiety groups were experiencing clinically significant anxiety. Thus. we eliminated participants from the anxiety groups who did not report at least moderate inter— ference due to their anxiety symptoms (mg. SA subjects had to report at least moderate interference due to social anxiety).‘ Because we were seeking to ex- amine the distinct association between behavioral inhibition and the two se- lected types of anxiety (social and generalized). we considered it desirable to improve the symptomatic homogeneity of our groups. Thus. we eliminated par- ticipants from the SA group who reported prominent symptoms of GAD. Like wise. eliminated participants fro m the GA group who reported prominent symp— toms of social phobia and eliminated'from the nonanxious control group those with prominent symptoms. of either social phobia or GAD.Z Of the 168 partici- pants originally assigned to the four groups based‘on the cut-off score classifi- cation. 63 were initially excluded for not meeting the interference criterion._and 29 were excluded for‘rcpottlng Symptoms incompatible with their initial group assignment. This resulted in a. final sample of 76 participants. MEASURES Mood Measures Stare-Trait Anxiety [uranium-State (ST/US). The. STAI-S (Spielberger. Gor- such. & Lushene, 1970) is a 20—item self-report questionnaire .for assessing tran- sient anxiety using, a 4~point Llltert response formar. Spielberger et al..reported high internal consistency (.86) for the STAI-S. We included the STAl—S to control for group differences due to the participants’ state anxiety at the time of the administration of the assessment battery. Beck Depression Inventory (EDI). The 21~item B'Dl (Beck, Ward. Mendelson, Mock. & Erbaugh, 1961) is a widely used. psychomeuically sound self—report scale for assessing depressive symptomatology. Mum—analysis of the BDI’S in— ternal consistency estimated the mean coefficient. alpha for nonpsychintric sub- jects to be .81 (Back, Steer. & Garbin, 1988). Beck ct al. (1988) reported that ‘We determined level of interference by examining particlpantt‘ responses to the item on the PASPQ that assessed level or interference due to social phobia symptoms (e.g.. none. mild, moder- ate. were. extmne) and a similar item that assessed interference due to GAD symptoms. : The elimination prowess was based upon the first’author‘s examination of participants’ responses to the PASPQ social phobia and GAD screening items. Participants assigned to the SA group based upon cut-off scores were remincd’only if they were judged‘to be relatively free of generalized anxiety symptoms and participants assigned to the GA group based upon cut-off scores. were re- tained. only if they were judged to be relatively free of social anxiety symptoms. Furthermore. those retained in the nonanxious control group were judged to be relatively free of both social and generalized anxiety symptoms. ' S n. A. MICK AND M. 1. rercn scores on the state sensitive BDI correlated significantly with clinical ratings and other self~repon measures of depression. in addition, they noted the capacity of the BDI to differentiate between outpatients with primary depressive disor- ders and those with GAD. We included the B131 to control for group differences due to the panicipants‘ depressive symptomology at the time of the administra— tion of the assessment battery. .4 Classification Measures Social Phobia and Anxierr Inventory (SPAI). The SPAl (Turner et al.. 1989) is a self-report index of social and agoraphobic anxiety with a 5~point Likert scale response format. The SPAI contains two subscales: social phobia (32 items) md'agoraphobia (13 items). Beidel. Turner, Stanley. and Dancu (1989) reported good test~rctest reliability (.86) for the total scale and high internal consistency for the social phobia subscul‘etc’oefficient alpha is .96). The SPA! has been shown to be effecrive for discriminating between those with and with- out social phobia. Penn State Worry Questionnaire (PSWQ). The PSWQ (Meyer, Miller. Metz- ger. & Borkovec. 1990) is a 16—item self-report questionnaire assessing the ten- dency to engage in excessive, generahzed. and uncontrollable worry. Respon- dents rate how typically they engage in worry behaviors using a 5-point Likert scale. High internal consistency (coefficient alpha is .93) and good test-retest reliability (.75) have been reported for the PSWQ (Molina & B'orltovec. 1994). Meyer et al. reported that the PSWQ significantly discriminated college students meeting all, some. or none of the DSM—HI—R (American Psychiatric Association, 1987) criteria for GAD. Panic. Social Phobia, and Anxiety Questionnaire (PASPQ). The PASPQ is an author—constructed self'report anxiety disorders screening questionnaire. Re— spondents indicate the presence/absence and severity (Likert format: e.g., 0 = none, 1 = mild. 2 = moderate, 3 == severe. 4 == extreme) of selected DSM—I'V symptoms of panic disorder, generalized anxiety disorder, and social phobia. Respondents are allowed to “skip out“ of the panic disorder and generalized anxiety disorder seetions if essential symptoms are not present (e.g.. concern about panic attacks. significant time spent worrying). Available psychometric data for the PASPQ is limited to that gathered during this study. We psychomenically evaluated the set of four GAD symptoms items (percentage of time spent worrying, excessivcness of worry, controllability of worry. and interference due to worry) and four social phobia symptoms items (frequency of anxiety in listed social/performance situations, frequency of avoidance of these “situations. excessiveness of social anxiety, and interference due social anxiety). Over the interval between the initial and second administra— tions of the PASPQ (M = 40 days: SD 2 23.8 days), the average test—retest SOCIAL ANXIETY AND BEHAVIORAL INHIBITION 9 reliability of the social phobia items was .62 (N = 39) with the stabilitv of items ranging from .48 for the excessiveness item to .65 and .60 for the ansietv and avoidance items. respectively. and .74 for the interference item. Because of the skip-out feature. less data for the GAD items were available (N = 261. Nonetheless. a similar average reliability was found ( .64). Reliabilities for the time worrying. excessive worry..._and interference items were .59. .62. and .59. respectively, while the reliability of the controllability of worry item was .T-l. ' Participants‘ average ratings on the PASPQ‘s four social phobia symptom items (administered with the complete assessment battery) were highly corre- lated to their SPAl scores. r (75) = .82. p < .0001. as were their responses on the single item rating interference clue to social anxiety. r (75) = .79, p < .0001. The social anxiety interference item and the set of four social phobia symptom items (which included the social anxiety interference item) were highly corre- lated. r (75) == .33. p < -0001. The participants‘ average ratings on the PASPQ’S four GAD items were highly correlated with their PSWQ scores, r (53) = .87. p < .0001, as were their responses on the single-itemrating interference due to generalized anxietv item. r63) = .3141 < .(JDOL The generalized anxiety interference item an set of four generalized satiety phobia items (including the generalized anxiezv interference item) were highly correlated. r (75) = .91. p < .OOOl. ' Behavioral Inhibition [ndices Retrospective SelfiRepm‘r of Inhibition {RSRI} Total Score and Subscales, RSRItReznick. Hegeman. Kaufman. Weeds. 8: Jacobs. 1992) is a 30-item seirl report questionnaire assessing a broad range of childhood behaviors associated with El, using a 5-point Likert scale response format. Selection of content in RSRI items was primarily influenced by interviews with children and parents of the Kagan et al. longitudinal study of B1 and the theoretical assumption that BI is a broad construct with several components (e.g., social. nonsocial. and generalized fears. as well as somatic complaints). Internal consistencies (coef- ficient alphas) ranging from .79 for undergraduate samples to .91 for a samcie that included psychiatric outpatients have been reported. Although parents- 71‘ (undergraduate: children were found. on average, to report consistently lower levels of B1 in their children than the children reported. the agreement between undergraduates and their parents was high (r = .63). 'Reznick et al.’s, anal: :55 revealed two factors: Social/School (12 items) and Fear/illness (12 items).IBe~ cause we wanted to analyze group differences on nonsocial fear and illness items separately. tve thctor analyzed the .168 participants’ responses on Reinisk et al.’s Fear/illness subscale» and forced the extraction of two rotated factczsz. items were used for interpreting the factors and retained for use on the subscales only if they loaded clearly onto only one ..factor and did so aids a loading of at least «.40. The six items loading significantly onto the ':"=t factor (items 5. 6. 7. 10. 12, and 27) were primarily related to nonsocial M. A. Mch AND M. J. TELCH fears and the five items loading significantly onto the second factor (items 1. 2. 3, 4. and 16) were primarily related to illness. Only one item (number 26) from Reznick er al.‘s Fear/Illness was eliminated. The new Illness and Nonsocial Fear factors were highly correlated with the total RSRI: Nonsocial Fear. r (168) = .62. p < .0001: Illness. r (168) .58, p < .0001. The new Illness and Nonsecialfegr factors were moderately correlated. r (168) = .43. p < .0001. Although bath new factors were significantly correlated with Reznick et al.'s original faétor Social/School. the correlations were modest. Illness: r t168) = .27. p < .001 and Nonsocinl Fear: r (168) = .24, p < .01. Reznick ct alfs original {actors and our new Illness and Nonsocial Fear factors were treated as su‘bscalcs in this study and. together with the total R’SRl score. were used as indices of El. STATISTECAL ANALYSES Differences between the SA. GA. mixed, and control groups on the demo- graphic variables were evaluated using chi-square tests. Group differences on. the set of classification measures were analyzed usinga multivariate analysis of variance (MANOV'A; alpha level of .05). Following the MANOVA. analyses of variance (ANOVAS) were performed on each measure separately followed by multiple comparisons of group differences using Tukey’s HSD test. Group differences on the indices of B1 were analyzed using an ANOVA for the RSRI total score and a iyL-XNOVA the four RSRI subscales described above (Social! School, Fear/Illness, Illness. Nonsocial Fear). The subscale MANOVA was fol- lowed by ANOVAS for each subscale. as well as multiple comparisons for each using Tnlcey’s HSD test. Group differences on the classification measures and on the behavioral inhibition indices were reanalyzed with multivariate analyses of covariate (MANCOVAS) and analyses of covariance (ANCOVAS) using state anxiety and depression as a covaria‘tes. Planned contrasts were then used to determine if original, patterns of group differences were maintained. The results of these covariatc analyses are reported only in cases where the patterns of group differences varied from original results. In our final analyses. we set aside group classifications and examined both gender and ethnic group dif— ferences in BI (using ANOVAS'); a well as the strength of the relationship be tween childhood BI and current 50cial anxiety (in regression analyses with the SPAI). R ESU LTS Demographics The four groups did not significantly differ with respect to gender, age. or ethnicity. wow. mute: r emu ntHAVJUKAL INHIBleN ll Measures Mood measures. The groups were found to differ significantly on the measures of State anxiety [STALState F(3. 7.2) = 24.64.12 < .0001] and dcpressjon [BBL H3. 72) = 30.18. p < .0001}. With respect to the STAl—S [8A2 M = 4490‘ SD = 12.20: GA: M = 44.38, SD = 11.44; Mixed: M = 54.60, SD = 943; Control M = 30.32. SD = 8.82} and BDI (SA: M = 10.80. SD = 6.61; GA: .11 = 8.62. SD = 7.52: Mixed: M = 19.80. SD = 7.20; Control M = 4.08. SD = 3.821, the SA and GA groups did not significantly differ but each scored significantly higher than thecontrol group. The SA and GA groups both scored significantly lower than the mixed group on the BDI. While the GA group scored significantly lower than them'tsed group on the STALS. the SA group did not. Classmmn‘on Measures Means. standard deviations. and group differences for the classification mea- sures are presented in Table 2. The groups differed significantly on the set of classification measures (SPAl, PSWQ- social nutrient interference, generalized anxiety interference. social phobia s3mptoms. and GAD symptoms; Wilks‘ lambda = 0.02.Approx. H18. 124.94) = 19.47. p < .0001). The groups dif— fered on thePASPQ item assessing interference due to social anxiety. F (3. 713) = 81.13. p < .0001. and the item assessing interference due to generalized anxiety, F (3. 49.) = 37.68, p < .0001. Multiple group comparisons indicated that the SA and mixed groups reported, comparable levels of interference due. to social anxiety and each reported significantly more such interference than both the GA and control groups, whose levels of interference from social anti: cry did not differ. With respect to interference due to generalized anxiety, the SA group reported significantly more interference from generalized anxiety than the control group; however. the SA "and control groups each reported signifi» cantly lower levels of such interference than both the GA and mixed groups. whose level of interference due to generalized anxiety did not differ. The only difference observed when controlling for-EDI and STAI-S scores was that the SA and control groups no longer differed in terms of interference due to general- ized anxiety. Each participant’s responses to the set of four social phobia screening items on the PASPQ were averaged and treated as an index of social phobia symp- tomology. Similarly, the set of four GAD screening items were treated as an index of generalized anxiety symptomology. Significant group differences were found on the social phobia symptom screen, H3. 71) = 86.92, p < .0001. and the GAD symptom screen, ['13. 49,) == 51.22, p < .0001. Multiple group comparisons indicated that even though the GA group reported significantly more social phobia symptomology than the control group, each of these groups reported significantly less social phobia symptomology than both the SA and l: M. A. MICK AND at. J. TELCH mixed groups. whose level of social phobia symptoms did not differ. The GA and mixed groups reported comparable levels of GAD symptomology and each reported significantly higher levels of GAD symptomology than both the SA and control groups. whose levels of GAD symptoms did not differ. Reanalysis using the EDI and STAI—T as covariates only led to two changes in these find- ings: The GA group no longer had higher levels of social phobia symptomology than the control group and the GA group’s level of generalized anxiety syrup- tomology was significantly greater than that of the mixed group, R3, 47) = 26.83. p < .0001. Cot-off scores on the SPAI and PSWQ had been used to make the initial group assignments and. as expected. the four groups differed significantly on the SPAl. F(3, 72l = 64.96. p < .0001. and the PSWQ. Fl}. 72) = 92.1l, p < .0001. Multiple comparisons among the four groups indicated that the SA and mixed groups each had higher SPAI scores, than both the GA and the control groups. The SA and the mixed groups did not differ significantly from each other on the SPAl. Likewise. the GA and control groups did not differ on the SPAI. Comparisons of PSWQ group means indicated that the GAD and the mixed groups did not differ and that each scored significantly higher than the SA group that, in turn. scored significantly higher than the control group. When analyses of the SPAI and PSWQ were repeated with the EDI and STAl-S as covatiatcs, the pattern of results, did not differ except that the SA and control groups no longer differed on the PSWQ. In the absence of diagnostic structured interviews, the diagnostic status of the participants could not be ascertained. However, these analyses of the classi— fication measures indicated that the groups differed as expected with respect to social and generalized To clarify whether participants in the various groups were reporting normal levels of social and/or generalized anxiety or levels of anxiety observed in persons with diagnosable anxiety disorders. we compared the SPAl and PSWQ scores of the four groups with previously re- ported means for anxious and nonanxious samples. The mean SPAI scores for the SA and the mixed groups were more than one standard deviation above the cutoff of 60. Their mean SPAI scores were also above the mean reported by Turner et al. (1989) for socially anxious college students (M = 72.2), of which 90¢? had been diagnosed with social phobia using the Anxiety Disorders Inter- view Schedule (ADIS: DiNardo. O‘Brien. Barlow, Weddell. & Blanchard, l933l. Furthermore. the mean SPAI scores for the SA and mixed groups were less than one standard deviation below the mean SPAI score reported by Turner ct al. for a clinical sample of social phobics (M = 94.0). On the other hand. the mean SPAl scores for the GA and control groups were more than one standard deviation below the cut—off score and at the mean reported by Turner et al. for non-socially~anxious college students (M = 32.7). Thus. the GA and control groups were reporting relatively normal levels of social anxiety while the SA and mixed groups Were reporting levels of social anxiety in the range typically observed for social phobics. ULan .‘s‘Axul i any Dunn u xvnru. I-lI-IIIALIIV ‘ TABLE 3 NIEANS on. 51161)an Devmrioss (SD) AND Gnome COMPARISONS FOR BEHAt‘tottAL lNHlBl'I'lON INDICES I SA GA Mixed Control ‘ in = l0l (n : IS) (I! = 15) (n = 38) 81 index M (SD) .11 (SD) M (SD) M (SD‘) Resnlel; et al. (1992) WWW Subscales Social/school i 93: til on) ‘-‘ 06“ ‘ u _. . . . -. $165) 3.l7‘ (0,81) 12.0lb (0.46'l FearttllneSs 2J4” (0.5% 1.78“ (0.40) 2.19‘ (0.63l 1.70“ (0,43) Mick and Telch Dirision’ of Fear/Illness Nonsocial fear illness Tomi RSRI ; (Ob-ll 1.6.35. (0.56) 142‘ (0.89) 1.89“ (0.61} " (OASl lo‘)“ (0.41) 1.92“ (0.603 1.42” (0.352 £0.43) 1.99‘ {0.37) 2.67’ (0.6M LS9" (0.3M .V’ore. Bl = behaxioral inhibition: SA = social anxiety group; GA 2 generaiired anxiety rouo' Mixed. group with social and generalized anxiety: Control = group without elevated sofinl generalized anxiety: 12512! = Retrospective Self~Report of Inhibition. "Means in the same row the do not share superscripts differ at {2 < .05 in the Tukey honestly significant difference corrtpan‘sonl The mean PSWQ scores for the GA and the mixed groups were more than one Standard deviation above the cut-off of 52 and within one standard deviation of means reported by Molina and Borkot'ec (1994) for subjects they diagnosed With GAD from their’tmalog’lll/l = 65.77. SD = 9.60) and clinical (M = 67 66 SD = 8.86) samples usingrthe ADlS-R (DiNardo &Barlow, 1988). In addition the mean PSWQ scores for the SA and the control groups were more than one standard deviation below the cutoff score. The mean of the SA croup was at the mean of the suhgecrs reported by Molina and Borkovec to be relatively untroubled by GAD symptoms but not necessarily free of other diagnosable anxrety disorders (M = 44.27, SD = 11.44). The mean of the control group was at the mean Molina and Borkovec‘reponed for subjects confirmed by the ADIS-R to have no diagnosable anxiety disorder (M = 30.98. SD = 8 13) Thus. the degree of generalized anxiety reported by the control group was. ate level typical for persons free of pathological anxiety, while the SA group‘s level or generalized anxiety was in the range expected for persons troubled by anxiety symptoms but not by generalized anxiety disorder. In contrast. those in the GA and mixed group reported levels of generalized anxietv in the range typically observed for persons with GAD. ' I Behavioral Inhibition Indices Means and standard deviations for'thc RSRI total score. the two Remick et a]. t 199%) subscales (Several/School and Fear/Illness). and the subscales we de- rived ()onsoczal Fears and Illness) are presented in Table 3. A nnivariate test hi. .\. MICK AM.) M. J. HzLLN revealed significant group difference for the t0tal RSRI. F(3. 72) = 16.63, p < .0001, and a MANOVA revealed that the groups also differed on the set of RSRI subscales. Wilks‘ lambda = 0.46. Approx. Fill 182.85) = 5.42. p < .0001. Follow-up univnriate tests indicated that the groups differed on Reznick ct al.’s Social/School subscale. F0. 72) = 19.05.12 < .0001. and Fear] Illness subscale, F( 3. 72) = 4.79. p < .01.. With respect to the newly derived subscales, the univariate te5ts were significant for both Illness, FI3, 72') = 5.34, p < .01. and Nonsocial FEM. £13. 72) = 4.82. p < .01. After controlling fol- depression and State anxiety. group differences for the Fenr/lliness and Illness subscales were no longer significant. Multiple comparisons of group means for both the toad RSRI and the Social! School subscale revealed that the GA group did not differ from the control group and each had significantly lower mean scores than both the SA and mixed groups. whose mean scores did not differ significantly. In contrast, multiple comparisons of group means for both the Fear/illness and Illness subscales revealed that only the mixed group scored significantly higher than the control group; the intermediate scores of the SA and GA groups did not differ from the scores of either the mixed or control groups. On the Nonsocial Fear subscale, none of the anxiety groups significantly differed from the control group. Inter— estingly, however. the GA group scored significantly lower than both the SA and mixed groups on Nonsocial Fear. As noted above. controlling for the ED] and STAI—S eliminated any group differences on the Fear/Illness and illness subscales. Furthermore. the mixed group's total RSRI score was no longer sig— nificantly higher than that of the control group. However. the patterns of group differences on the Social/School and Nonsocial Fear subscales remained un— changed. Behavioral Inhibitian. Social Attriery. Gender. and Ethnicity Setting our anxiety and non-anxious control groupings aside. We conducted regression analyses to determine the strength of the relationship between our behavioral inhibition indices and Our primary measure of social anxiety (SPAI). Moderately high associations were found between SPAI and both the total RSRI, r (76) = .70, pu< .0001. and the Social/School subsonic, r (76’) .75, p < .0001. Furthermore, the Nonsocial Fear and the Illness subscales were also significantly correlated with the SPAI. r (76) = .38. p < .001; r (76) z .39. p < .01; respectively. Gender differences on indices of Bl were also examined. The total RSRI scores of males and females did not differ. However. using a MANOVA to analyze gender differences on the set of RSRI subscales. we found that males and females differed significantly [Willey lambda = 0,82, Exact F(3, 72) = 3.56. p < .05]. FOIIOW'UP ANOVAS indicated that gender differences were not present for the Social/School and Illness subscales; however. females (M = 2.15, SD = .73) reported significantly more inhibition than males (M = 1.76. a\l\ me. nun-L.- e .xnw u..unnu...._ ......_. SD = .63) on the Nonsociul Fears subscule. Ftl. 74) = 5.33. p < .05. This prompted us to conduct a Group X Gender ANOVA for Nousocial Fear. The effecr of group. H3. 68) = 5.08. p < .Ol. was significant while the gender and interacrion effecrs were not. Thus. group differences on the Nonsocial Fear component of BI cannot be attributed to gender differences on this subscale. Finally. ethnic differences on indices of BI were assessed. To do so, we excluded the six participants whose ethnicity did n01 match our designated eth- nic categories and the one African American. Then we examined BI differences among our three best represented groups (11' Asians. 15 Hispanics. and 46 whites-i. The total RSRI differed among ethnic groups. HZ. 691 = 7.96, p < .001. Multiple groupcomparisons indicated that Asians (M = 2.6—1. SD e 0.63» scored significantly higher than WhitestM = 2.01. SD = 0.42). The mean RSRI score of the Hispanic group «(M = 2.22. SD = 0.53) fell between the menus for the Asian and White groups and did not differ significantly from either. A MANOVA on the RSRI'subseales (Social/School, Nonsocial Fear. Illness; revealed that the ethnic groups differed on the set of subscales {Willts’ lambda = 0.74. Approx. F (6. 13-11 = 3.64.}; < .()1].'Follow—up ANOVAs indicated that ethnic differences were not present for theNonsocial Fear and Illness subscales. However. the ethnic groups differed significantly on the Socialr’Schooll subscale. Ft}. 159) = 10.78, p < .001. with Asians (.11 = 3.21. SD = 0.73) reporting significantly more inhibition than HiSparn'cs (M = 2.44, SD 0.7l) and whites 1M = 2.15. SD = 0.66;. This led us to conduct a Group x Ethnicity ANOYA for the Social/School subscale. The effecr of group, 1412. 61) = 12.70, p < .0001. was significant. while the effect of ethnicity and of the interaction of group and ethnicity were net significant. Thus. group differences on the Sociel/ School component of Bl cerium be. attributed to ethnic differences on this sub- Scale. DISCUSSION Results from the present study fail to support the hypothesis that a childhood history of behavioral inhibition is associated with anxiety symptoms in general. This conclusion is supported by our finding that participants in the generalized anxiety group were no more likely to report childhood behmioral inhibition than were participants in the nonunxious control group. Instead. our findings provide some support forum specificity of an association between behavioral inhibition and social anxiety: Participants reporting current impairment due to Social erudety, whether alone or in combination with generalized anxiety, re ported significantly more childhood behavioral inhibition relative to participants teponing only generalized anxiety or participants reporting neither social nor generalized anxiety {nonanxious controls }. Taken together these findings point to the possibility that a childhood history of behavioral inhibition may be partic- ularly characteristic of persons reporting elevated social anxiety in adulthood. While preliminary, our results suggest that future investigations examining the 16 M. .\. MICK .txo M. J. 'l‘ELCH linkage between behavioral inhibition and specific anxiety disorders need to assess and control for social anxiety. Not surprisingly. those with current social anxiety reported significantly more childhood inhibition in social and school situations than participants with eichterted generalized anxiety or unnanxious controls. More interesting was, our- finding that groups with elevated social anxiety also scored highest on a seismic oi" behavioral inhibition unrelated to social fears (Nonsoeial Fear). This lat at finding provides some. support for the conclusion that the linkage between adult social anxiety and reports of childhood behavioral inhibition not Stm~ ply due to the overlap in social fear‘items. . Although previous Studies have shown a relationship between behavioral inhibition and multiple anxiety disorder diagnoses tfor a review. see Biederman, 199m. these studies have not examined the relative influence of the different components believed to make up’hehavioral inhibition. Note that only partici- pants in Our combined social and generalized anxiety groups reported more childhood illness and somatic concerns than nonan'xious controls. Although speculative. one interpretation of this finding is that illness and somatic concerns present in childhood may be linked to the later development of multiple anxiety problems. However. it should be noted that the Illness sttbscale score of our combined anxiety group was no longer significantly higher than that of the nonatutiou‘s controls after controlling for depression and state anxiety. The relationship observed between current social anxiety and behavioral ini hibition was not completely consistent across the features thought to character- ize behavioral inhibition in childhood. However.~our finding that all the compo- nents of behavioral inhibitioit we examined were significantly associatEd with our primary measure of social anxiety tSPAIl further supports the'link between crnidhood behavioral inhibition and current social anxiety. Were the GAD symptoms reported by those in the generalized anxiety and ii ‘ed groups a function of depression? Several factors argue against this by pottesis. First. it should be noted that the SA and GA groups did not differ significantly with respect to depression. Second. analyses of the group differ- ences on the generalized anxiety measure (PSWQ) using depression (BDI) as a :owuiate yielded a pattern of results identical to those reported above. In acdition. the observed differences in behavioral inhibition between the GA and SA eroups remained significant even after controlling for depression. Moreover. the ESDI means for the SA and GA groups were at the mean reported by Beck et '. £1988) for a sample of GAD outpatients (M = l4.46. SD = 6.10) and weir below the mean. of. a comparison sample of outpatients with depressive disorders (M: 26.37, SD = 6.94). Previous behavioral inhibition studies have not reported, gender differences because they have primarily focused upon C aucasians(Turner et al.. 1996), not reported ethnic differences. Our finding that females scored higher than maies on the Nonsocial Fear subscaie of our behavioral inhibition measure SOCIAL {NXIETY AND BEHAVIORAL INHIBITION is consistent with findings that females report more nonstxciul phobias than males (Bourdon. Boyd. Rae. Burns. Thompson. & Locke. 1988). However. fur— ther analyses revealed that these observed gender differences did not account for the observed differences among our anxiety groups on inhibition in response to nonsocial fear. Ethnic differences were observed on the Social/School sub‘ scale of our behavioral inhibition measure with Asians reporting more inhibition in this domain than Hispanics and Whites. Again. further analyses revealed that these ethnic differences did not account for the observed differences among our anxiety groups on social and school-related inhibition. Nonetheless, more thorough investigation of gender and ethnic differences in behavioral inhibition is warranted. Our findings must be considered in light of the study‘s limitations. Firsr. despite screening a large number of students. our final sample sizes were small. This may have resulted in low statistical power for detecting lbw or moderate effect size difference between groups. Another limitation is the retrospecrive nature of the study. It is possible that the-association between childhood behaw ioral inhibition and adult social anxiety may have been due to exaggerated re ports of childhood behavioral inhibition by those with current social anxiety. Likewise. persons with symptoms of GAD may not accurately report their his-to ties of nonsocial inhibition. Thus. including independent informants (e.g., par- ents l to rate participants‘ childhood behavioral inhibition would have strength- ened our conclusions. However. note that Reznick et al. (1992) rbtind a strong agreement between participants‘ and their parents‘ reports of the participants' childhood behavioral inhibition. The criteria for classifying participants deserve comment. The purpose of applying classification criteria beyond cut-off scores was to create homogeneous groups and to increase the likelihood that participants in the anxiety groups were experiencing interference in funcrioning due to their anxiety symptoms. These stringently defined groups were used in order to examine the extent to which a childhood history of behavioraiiinhibition is uniquely associated with current elevations in social or generalized anxiety. However, we acknowledge that our reliance on self~report measures is problematic and precluded a determi- nation of whether participants met diagnosnc criteria for social phobia or GAD. Replication using structured diagnostic interviews is needed for determining social phobia and GAD diagnoses. What might account for the apparent asymmetry in linkages between behav- ioral inhibition and anxiety disnrders? Our findings taken together with those from previous work suggest that behavioral inhibition may be related to panic disorder/agoraphobia {FDA} as well as social—evaluative anxiety, but not symp- toms of generalized anxiety. One possibility is that the association between behavioral inhibition and FDA may be a function of cornorbid seeial anxiety symptoms. To test this hyporhesis. r‘utttre studies examining the linkage between behavioral inhibition and FDA should control for the effects of social anxiety. til ,\I. .\. Mer AND M. J. 'rELt'n RQSenbaum et al.‘s Ijl99~lt offer an alternative explanation for the observed linkages between behavioral inhibition and both PDA and social phobia. They propose that behavioral inhibition serves 115 a diathesis that may be expressed differently across the life span. Another explanation of the asymmetry in linkages between behavioral inhibi— tion and anxiety disorders stems from the observation that PDA patients and social phobias often exhibit avoidant behavior that seems consistent with a be haviorally inhibited temperament. This tendency to avoid situations and events does not seem to be as characteristic of GAD patients. Comparing histories of behavioral inhibition for groups of PD patiean with and without agoraphobic avoidance and social phobics with high and low levels of phobic avoidance may help clarify whether the presence of an avoidant copingstyle accounts for the association, of childhood behavioral inhibition with social phobia and PDA. Turner et al. (19%) offer :1 mo =i which behavioral inhibition is viewed as one of several factors that may increase vulnerability to anxiety disorders. We hypothesize that u behaviorally inhibited temperament may contribute to avoidance while other feetors influence which specific anxiety disorder devel— ops. For example. heightened anxiety sensitivity in childhood may combine with a behaviorally inhibited temperament to increase risk for PDA. whereas a behaviorally inhibited temperance. in the presence of extreme introversion may increase risk for social phobia. Cztrrentiy, we are examining prospectively the relationship between childhood history of behavioral inhibition and anxiety sensitivity and the contribution of each to the later development of spontaneous panic attacks and elevated social-ex emotive anxiety. ln Pyer‘s (l993l excellent review or” the heritability of social anxiety, she concluded that there is no direct evidence that behavioral inhibition in childhood is related to adolescent or adult scald”: anxiety. While caution must be used in drawing inferences from retrospccnve studies. our findings suggest that a childhood history of behavioral inhibition is highly characteristic of young adults reporting current inten‘erence due to elevated social anxiety. If confirmed by prospective. studies. our finding or a continuity'betwccn child and adult social anxiety and inhibition strongly points to the need to develop cl‘fective‘childhood interventions for pathological social t ixiety and avoidance. REFS." EN’CES American Psychiatric Association. I 1987?. Disgeosnt: and statistical manual affine/Ital disorders I3rd cd.. Rem. Washington. DC"; Author American Psychiatric Association. -. 199M. («ltli ed). Washington. DC: Author. Beck. A. 'l‘.. Ward. C. H.. Mendeison. M. .’.. block. 1., & Erhnug'n. J. 0961). An inventory for measuring depression. Arr/hirer n! Gt'rier...‘ .’.';.'rhiau‘_r, 4. 561~57L Beck. A. ’l”.. Steer. R. .-\.. a: Corbin. .‘sl. (3. Dssycltotnctric properties ofthc Beck Depression Inventory: Twenty—five years or :nni nicrtl Phycizuiog)’ Review. 8. 77400. Bcidel. D. C.. Tumor. S. M. Staple" . ..‘ :cu. C. V. (1989». The Social Phobia and Anxiety inventory: Concurrent and external -. ’ieiiuriur Therapy». 20. sill—427. ..g':/)Si!i(7 and statistical manual ofmemul disorders ..‘ SOCIAL ANXKETY AND BEHAVIORAL INHIBITION l9 Bicdem 5. s {990). The diagnosis and treatment or' adolescent anxiety disordcm. Journal offlini- mi 5’._ cizium'. SlfSuppl, 5). 20—36. 1.. Rosenhnum. J. F.. Boldquurphy. E. A.. Routine. 5. VI. Chulol‘l‘. 1.. liirshi'eld. . . .i; Regan. .l. (1993l. A 3-}cttr followup of children with and without behavioral inhibi- tion. -‘nrtmai of the American .lrrra'emy of Child and Adolescent Psychiatry. 32. S l 4—32 l. ' 5.. Rosenbaum. .l. E. Chaloff. J.. 8: Regan. .l. (1995). Behavioral inhibition as a risk or .in\icty disorders. In J. S. Marsh (Edit. Anxiety (lisora’err in citildrwt and adolescents it. fact 2: :' r ' —:' Grzczr éhterences in phobias: Results of the EC.\ community survey. Juimml' qum'iery Disor- i . 2. 227—241. New York: Guilford Press. . Boyd. J. H.. Rae. D. 5.. Burns. 8. J.. Thompson. .1. w.. 5.: Locke. B. Z; (1988). . 33.. 5c Barium D. H. H993}. Comor‘oidity among anxiety disorders: Implications for . :mi DSM—Zl". Journal of Consulting and Clinical Psychology. 60. $35—$44. . .. ct Barlow. D. H. il988l. Anxiety Disorder: [men-law Schedule —— Revised {AD/S- KY: Phobia and Anxiety Disorders Clinic. State University of New York. .. Ll'Brien. (i. T.. Barlow. D. H.. WMdell. M. T.. 8: Blanchard. B. t 198.3). Reliabil- vIIIanxi'ety disordercztegories using :1 new structured intewiew. Airlift-er.tnFGmemI 41). 1070-1074. 993. Heritability of social anxiety; -\ brief rexiew. Journal of Clinical Pry/tinny. . lO—lZ. Hirshieii D. R.. Rosenbaum. I. F.. Biederman. L Bolduc. E. .—\.. Fat-none. S. V. Snidmztn. N.. 11:21:23.}. 5.. 3r. Kagan. l. (19922.5tnble behavioral inhibition and its association with anxiety discrier. Mama! qr" the American Academy of Child and Adolescent Pot/tinny. 31. 103—» ll; Remit-k. S. L. & Snidmun. .\.'. (1988l. Biological basics of childhood shyness. Science. 240. /. 3—1. .3; Stud mun. .\' . U991). Infant predictors of inhibited and uninhibited profiles. Psychologi- cm are. 2. 40—44. Meyer. 7. .. Miller. MJ... Manger. R. l... & Borkcvec. T. D. (19901. Development and validation of 1” Penn State Worry Quesuuimaire. Behaviour Research and Timmy. 28. 487—495. Moll 2;. Borkovcc. T. D. l 1994. I. We Penn State \\ on}: Questionnaire: Psychometric proper— ‘ . associated characteristics. In G. C. L. Davey & F. F Tailis IEds. l. ll’nmmgr' Perm“- :Iremy. asterimenr. and treatment (pp. 265—283). Chichester. UK: John Wiley & Sons. Pollock. Roscnhnum. .l. ii. Marts. .-\.. Miller. B. 3.. a: Biedemiun. l. r1995). Anxiety disorders ‘ cod: Implications for adult psychopathology. file Psychiatric C (info: Qf'Norrh America. 1 . ’-‘<—'es. ' ,. Hegeman. l. .\l.. Kaufman. E. R.. Woods. 5. W.. at; Jacobs. M. I l992t. Retrospeczive “’tcurreru self-report: of behavioral inhibition and their relation to adult mental health. _ men! and Psychopariiot'ogy. 4. 301—33. 31 3;: Davidson. R. .l. 1994). Personality llnd benai ior Ell parents of ternperamentally 2:; end uninhibited chil .n. Developmental Pryciwlogy 30. Fidel“. '. I... Kngan. 1., Reznicii. I. 5:. & Carley. R. {1992'}. The heritability of inhibited and tee. behavior: A nvin'stud}. Developmenmi Psyclrut‘agv. 28. lO30~1037. . 1. F... Bicderman. Bnidnc. F. A.. Hirshl‘eld. D. R.. Famone. S. V.. <9: Kngan. J. . Ccmorhidity of parental anxiety. disorder for childonset anxiety in inhibited . Amert'erm Journal anwjrt'n’nrxy. 149, 475—381 '. F.. ‘Biedennan. 1.. Gemcn. M. Hirshfeld D. R.. Memingcr. S. R.. Herman. 1., 3.. Reznick. J. 5.. & Snidman. N. '(IQESL Behavioral inhibition in children of parents isotder and agoraphobia: A controlled study. Arc/111:6: (If General Psychiam‘. 45. .. E. Bicdcrmun, 1.. llirsht‘eler. IL. Bolduc. E. .-‘\.. 3; Chalot't'. .l. {199] }. Behavioral n in children: A possible precursor to panic disorder or >0cial phobia. Jum‘mil of Clini» . rum: 52(Suppi. ll). 5—24". 20 M. A. MICK AND M. J. TELCH Ruscnbuum. J. 17.. Bicdcnuun, 1.. llirshl'cl'd. I”). R.. linlduc, E. A.. l-‘ummnc. S. V.. Kzigm. J.. thl- man. N.. & Reznlck, 1.5. (1991). Funher evidence of an association between behavioral inhlbl- tion and anxiety disorders: Results from a family smdy of children from a non-clinical sample. Journal of Psychiatric Research, 25, 49—65. Rnscnbaum. J. R, Bicdcmum, 1., Pollock, R. A.. & llirshfcld, l). R. (1994). The etiology of smzinl phobia. Jamaal (afCli/liml I’xw‘llialry. ‘55lSnppL .6], 10—45. Schcibc, (1., 8L Albus, M. (1992). Age at Onset, precipitating events, sex. distribution, and czwccun ten“: of anxiety disurders. Psyclmlmllmlvgy. 25. 11—48. Scllncicr, 1?. 12., lohnsnn. 1.. Homig, C. D..- l..ichuwilx,> M. I"... 8:. Wciwnan‘. M. M. (1992). Social phnhia: Cmnm’bidity and murhidity in nu cplilcminlngic sample. Arr/livw' 11/110qu! I’xyrhiu— try. 49. 282-2“. Spicllmrgcr, (l. 1)., (lumucll, R. l... & Lushcnc. R. E. (1971)). STA! mrmmtl jhr rim Shim-7‘an Anxiety Inwmnry. P1110 Alln. CA: Conwng Psyclmlagisls Press. 'l'urncr. S. M.. 1161ch I). (1.. Damn. C. V.. S: Smnlny. M. A. (1080). An empirically llci'ivul inventory In mcnsuw simian! I'unrs mul unxicly: 'Ihc‘xuuiul l‘lmlviu and Anxiety luvuulmy. my. rlmlugil'nl Asuwwn'm: A Jmimul If]?! 'nnmltiiug mlrl ( ‘linil‘ill I'.\l\*t'linlugy\ I. 35 .1111. Tl‘urncr. S. M.. lluidul. l). (‘.. «\‘L Wolff, I‘. L, (109(1). ls lwlmvinml inhibilion rclnlcd In llm :mx’lcly disul'llm‘s'! l' 'Iihn'nl liu'rltnlngv er‘ium "I. 157 I72. ...
View Full Document

Page1 / 11

SOCIAL - Journal of Anxiety Disorders. VOL 12. No. 1. pp....

This preview shows document pages 1 - 11. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online