Journal of Affective Disorders 110 (2008) 156–160www.elsevier.com/locate/jad

Brief report

β-adrenoceptor affinity as a biological predictor of treatmentresponse to paroxetine in patients with acute panic disorder☆

In-Soo Lee a, Kyung-Jeong Kim b, Eun-Ho Kang c, Bum-Hee Yu c,⁎

a Department of Psychiatry, Semin Mental Hospital, Sungkyunkwan University School of Medicine, Seoul, South Koreab Samsung Biomedical Research Institute, Sungkyunkwan University School of Medicine, Seoul, South Korea

c Department of Psychiatry, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea

Received 14 September 2007; received in revised form 11 December 2007; accepted 11 December 2007Available online 15 January 2008

Abstract

Background: Few studies have reported on the functional differences of the β-adrenoceptor between treatment responders and non-responders in panic disorder (PD). The aim of this study was to compare the nature of the β-adrenoceptor function and clinicalvariables between treatment responders and non-responders to paroxetine treatment in acute PD patients.Method: Paroxetine was administered to all of the panic patients for 12 weeks. The lymphocyte β-adrenoceptor density (Bmax),affinity (1/Kd), and sensitivity (cAMP ratio) were measured in 22 untreated outpatients with acute PD and 22 age, sex and BMImatched control subjects. Psychological assessments were conducted using the HAM-A, and HAM-D, STAI-S and STAI-T,Anxiety sensitivity index (ASI), and Acute panic inventory (API).Results: A significantly higher Kd was observed in the panic patients before treatment as compared with the control subjects, butthere was no significant difference in Kd between the panic patients and control subjects after the treatment. Among the 22 patients,the 11 treatment responders (50%) showed a significantly higher Kd and lower mean scores of HAM-D, STAI-S, STAI-T, and ASIat baseline, compared with the non-responders. Logistic regression revealed that the pretreatment Kd and HAM-D weresignificantly reliable predictors for treatment response (pb0.05).Conclusion: The β-adrenoceptor affinity (1/Kd) was decreased and adaptively normalized after treatment with paroxetine in theacute panic patients. In addition, a low pretreatment β-adrenoceptor affinity (1/Kd) was found to predict the treatment response andcan be suggested as a biological predictor of treatment response in acute PD.© 2007 Elsevier B.V. All rights reserved.

Keywords: β-adrenoceptor affinity; Panic disorder; Paroxetine; Treatment response

☆ Part of this article was presented at the Annual Meeting of theKorean Neuropsychiatric Association, which was held on October20th, 2006 at the 63rd building, Seoul, South Korea.⁎ Corresponding author. Department of Psychiatry, Samsung

Medical Center, Sungkyunkwan University School of Medicine,Ilwon-Dong 50, Gangnam-Gu, Seoul, South Korea 135-710. Tel.: +822 3410 3583; fax: +82 2 3410 0050.

E-mail address: bhyu@skku.edu (B.-H. Yu).

0165-0327/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.jad.2007.12.007

1. Introduction

Recent studies have found evidence of abnormalfunctioning of peripheral β-adrenoceptor system in panicdisorder(PD) ( Aronson et al., 1989; Brown et al., 1988;Maddock et al., 1993a; Nesse et al., 1984). Moreover,Maddock et al. (1993a) found a relationship betweena lower β-adrenoceptor density and better treatment

157I.-S. Lee et al. / Journal of Affective Disorders 110 (2008) 156–160

response to adinazolam, and proposed the pretreatmentβ-adrenoceptor density as a possible predictor of treat-ment response (Maddock et al., 1993a). Park and Yu(2005) also reported that the pretreatment β-adreno-ceptor affinity (1/Kd) and density could be useful bio-logical markers that predict the treatment response toparoxetine (Park and Yu, 2005). However, these resultswere inconsistent and few studies directly compared thefunctional differences of β-adrenoceptor between treat-ment responders and treatment non-responders in PD.

This study was designed to compare the nature of theperipheral β-adrenoceptor function and clinical vari-ables between treatment responders and non-respondersin acute PD patients before and after 12 weeks ofparoxetine treatment. The lymphocyte β2-adrenoceptoris a useful model for measuring peripheral β-adreno-ceptor function in man ( Mills et al., 1990), so theauthors hypothesized that the lymphocyte β-adrenocep-tor density (Bmax), affinity (1/Kd), and sensitivity(cAMP ratio) of patients with PD are lower than those ofnormal control subjects and that they are different be-tween treatment responders and non-responders beforeand after paroxetine treatment.

2. Materials and methods

2.1. Subjects

Forty-six patients with acute PD, ranging in age from20 to 50 years, were recruited from the outpatient clinic oftheDepartment of Psychiatry in SamsungMedical Center.The diagnosis was made using the Anxiety DisorderInterview Schedule for DSM-IV (ADIS-IV) (First et al.,1997). Only those patients whose duration of PD did notexceed 1 year were enrolled in this study. Patients withmedical illnesses or with other psychiatric illnesses wereexcluded from the study, because such illnesses mightaffect the adrenergic receptor responsiveness (Dao et al.,1998; Johnson, 1998; Johnson and Lydiard, 1995; Joyceet al., 1992; Mills et al., 1995a; Pandey et al., 1987).Thirty-nine patients were assigned to 12 weeks ofparoxetine treatment. However, 17 of these patientsdid not complete the 12-week treatment. Thus, thefinal study population comprised 22 patients (male: 16,female: 6). Twenty two control subjects who werematched for age, sex, and BMI and had no majorpsychiatric or medical disorders were recruited (male: 17,female: 5). The Institutional Review Board of SamsungMedical Center approved the study and written informedconsent was obtained from all participants.

All of the participants were required to be free of anymedication for at least 2 weeks prior to their participa-

tion in the study. The BMI (kg/m2) was required to bebetween 18 and 30 in all subjects. All of the womenparticipants were required not to be pregnant and not totake oral contraceptives during the study period.

2.2. Treatment

Paroxetine was administered to all of the panic patientsfor 12 weeks. The starting dosage was 10–20 mg/day andthemaximumdosage during the 12weekswas 40mg/day.No other psychotropic medications were allowed exceptfor alprazolam.

2.3. Biochemical measure

All of the subjects were instructed to refrain fromdrinking caffeinated beverages or alcohol and from smok-ing for the 24-h period prior to the study. The study beganat 8 a.m., whereupon the subjects were seated and a 20-gauge catheter was inserted into a forearm vein permittingblood collection. After instrumentation, the subjectsrested quietly for 30 min. Then, the physiological con-ditions of the subjects were examined and blood sampleswere collected. The same procedure was followed in all ofthe subjects after the 12-week treatment with paroxetine.

2.4. Beta-adrenergic receptor binding data analysis

The lymphocyte isolation and receptor assay wasconducted to determine the level of the β2-adrenoceptoraccording to previously published methods (Yu et al.,1999). The binding data were transformed mathemati-cally and analyzed using the Inplot program (GraphPad,Sandiego, CA). The lymphocyte β-adrenoceptor density(Bmax) and affinity (1/Kd) were determined by using agamma counter to measure the radioactivity from thelymphocyte binding after incubation with radioligand[125I]-iodopindolol at six concentrations ranging from 10to 320 pM for 1 h at 37 °C. The specific binding wasdetermined using 10−6 M 1-propranolol. The lymphocyteβ-adrenoceptor sensitivity was determined in whole cellsby quantifying the cyclic AMP (cAMP) accumulationfollowing incubationwith 10 μM isoproterenol. The basalnon-stimulated, intracellular cAMP was determined andthe βAR sensitivity defined as the ratio of the stimulatedto basal cAMP (the cAMP ratio).

2.5. Psychological assessment

Clinical variables were assessed with the Hamiltonrating scale for anxiety (HAM-A) (Hamilton, 1959),the Hamilton rating scale for depression (HAM-D)

Table 1Comparison of pretreatment demographic data, psychological statesandβ-adrenoceptor function between panic patients and normal controlsubjects

Panic(Mean±SD)(N=22)

Control(Mean±SD)(N=22)

p

Gender Male 16 (73%) 17 (77%)Female 6 (27%) 5 (23%) 0.731

Age 37.32±6.60 35.36±7.10 0.350BMI 22.34±2.65 23.42±2.35 0.161Exercise 2.55±1.34 2.68±1.29 0.818HAMA⁎⁎ 17.36±8.08 3.75±3.02 b0.001HAMD⁎⁎ 13.32±6.07 2.67±2.50 b0.001STAI-S⁎⁎ 44.91±9.59 5.10±5.44 b0.001STAI-T⁎⁎ 48.73±8.50 34.57±10.53 b0.001ASI⁎⁎ 22.64±14.07 37.41±10.36 b0.001Bmax 8.69±6.22 9.54±5.01 0.467Kd⁎ 76.38±28.17 56.12±24.97 0.015cAMP ratio 6.72±3.48 5.48±2.52 0.185

BMI: body mass index, Exercise: score 1–6 (number of exercise perweek), HAM-A: Hamilton rating scale for anxiety, HAM-D: Hamiltonrating scale for depression, STAI-S: Spielberger state-trait anxietyinventory-state, STAI-T: Spielberger state-trait anxiety inventory-trait.ASI: Anxiety sensitivity index, API: Acute panic inventory, Bmax:maximum binding capacity, cAMP ratio: isoproterenol stimulatedcAMP to basal cAMP, Kd: dissociation constant. One-way ANOVA orKruskal–Wallis analysis of variance for all comparisons, except chisquare test for gender. ⁎: pb0.05,⁎⁎: pb0.01.

158 I.-S. Lee et al. / Journal of Affective Disorders 110 (2008) 156–160

(Hamilton, 1960), the Spielberger state-trait anxietyinventory-State and Trait (STAIS and STAIT) (Spiel-berger et al., 1983), Anxiety sensitivity index (ASI)(Reiss et al., 1986), and Acute panic inventory (API)(Liebowitz et al., 1984). The psychological evaluationof the control subjects was also done by means of theHAM-A, HAM-D, STAIS, STAIT, and ASIassessment.

Treatment responders were defined as patients whoshowed a greater than 50% increase on their API scoresafter 12 weeks of paroxetine treatment. The API inquiresinto 23 symptoms that the subjects rate as absent (0),mild (1), moderate (2), or severe (3) and generates a totalscore (possible range, 0–69).

Table 2Comparison of β-adrenoceptor density, affinity and sensitivity between treparoxetine treatment

Baseline p

Responder (N=11) Non-responder (N=11)

Bmax 8.46±5.89 8.915±6.81 0.9Kd 88.50±29.64 64.27±21.53 0.0cAMP ratio 5.65±3.33 7.79±3.44 0.1

Bmax: maximum binding capacity, cAMP ratio: isoproterenol stimulated cAKruskal–Wallis analysis of variance for all comparisons. ⁎: pb0.05.

2.6. Statistics

To analyze the differences at baseline between thepanic patients and the normal control subjects and be-tween the responders and non-responders to paroxetinetreatment, the χ2-test and one-way ANOVA or Kruskal–Wallis analysis of variance were performed using threereceptor variables (Bmax, Kd, cAMP ratio) and variousclinical variables. To analyze the difference at the end-point between the responders and non-responders toparoxetine treatment, the t-test or Mann–Whitney U testwas performed. For the prediction of treatment response, astepwise logistic regression analysiswas carried out for allsignificant baseline variables. All of the statistical testswere two-tailed and were performed at a level of sig-nificance of 5%. All of the statistical analyses wereperformed using the SPSS 13.0 statistical software.

3. Results

In the comparison of the demographic data andpsychological states between the panic patients and con-trol subjects, there were no statistically significantdifferences in sex, age, BMI, or exercise level. How-ever, all of the psychological variables were signifi-cantly higher in the panic patients at baseline (Table 1).One-way ANOVA revealed a significantly higher Kd(p=0.015) in the patients compared with the controlsubjects before treatment. However, there were no sig-nificant differences in the means of Bmax and the cAMPratio between the patients and control subjects beforetreatment (Table 1).

Among the 22 patients, 11 (50%) were treatment re-sponders. There were no significant differences in themeans of age, sex, duration of illness, or fitness levelbetween the responder group and non-responder groupbefore treatment. However, the mean scores of the HAM-D (p=0.011), STAI-S (p=0.047), STAI-T (p=0.020) andASI (p=0.020) were significantly higher and the meanBMI (p=0.014) was significantly lower in the non-re-sponder group. After 12 weeks of treatment, the mean

atment responders and non-responders before and after 12 weeks of

Endpoint p

Responder (N=11) Non-responder (N=11)

22 8.63±4.61 11.28±8.46 0.37340⁎ 70.84±21.20 59.91±32.00 0.35655 9.80±6.86 8.98±6.31 0.773

MP to basal cAMP, Kd: dissociation constant. One-way ANOVA or

159I.-S. Lee et al. / Journal of Affective Disorders 110 (2008) 156–160

paroxetine dosage was not significantly different betweenthe responder group (23.98±5.86 mg/day) and non-re-sponder group (26.70±6.50 mg/day) (p=0.316), but themean scores of the HAM-A (p=0.008), API (p=0.001),STAI-S (p=0.031), and STAI-T (p=0.014) and the meandosage of alprazolam during the study were significantlylower in the responder group.

In addition, the β-adrenoceptor Kd was significantlyhigher (p=0.040) in the treatment responders (Kd=88.50±29.67) compared with the non-responders (Kd=64.27±21.53) at baseline. However, after 12 weeks oftreatment, there were no statistically significant differ-ences in any of the β-adrenoceptor measures betweenthe responders and non-responders (Table 2).

For the prediction of treatment response to 12 weeksof paroxetine treatment, a logistic regression analysiswas performed. Pretreatment Kd (p=0.050) and HAMD(p=0.021) were found to predict treatment response.

4. Discussion

In this study, the pretreatment Kd was significantlyhigher in the panic patients and was normalized afterparoxetine treatment. These results demonstrated that theβ-adrenoceptor affinity (1/Kd) was decreased in the acutepanic patients, which is consistent with previous studieswhich reported a decreased lymphocyte β-adrenoceptorbinding in PD (Brown et al., 1988;Maddock et al., 1993a;Park and Yu, 2005). The alteration of Kd may represent adifference in the ratio of the high and low affinity receptorbinding components rather than a change in the affinityfor the radioligand (Perry and U'Prichard, 1984). There-fore these results suggest the existence of an association ofPD with the conformational change of the postsynapticlymphocyte β-adrenoceptor, due to the increase in thepresynaptic noradrenergic activity in response to the panicanxiety state. This assumption can be tested by displacing125I-iodopindolol with an agonist (e.g. isoproterenol) inthe future study.

After 12 weeks of successful treatment, the Kdtended to be normalized and there was no significantdifference between the panic patients and normal con-trol subjects. Paroxetine has a secondary effect ofdecreasing the noradrenergic activity (Gorman et al.,2000). By this mechanism of action, after successfultreatment with paroxetine, the presynaptic noradren-ergic activity decreased and resulted in an increase of theβ-adrenoceptor affinity (1/Kd). Although small doses ofalprazolam were allowed in some panic patients, aprevious study reported that benzodiazepine had littleeffect on the β-adrenoceptor receptors (Lafaille et al.,1991). Thus, it seems reasonable to suppose that the

change of the β-adrenoceptor function observed in thisstudy was caused by the paroxetine treatment.

Contrary to our initial hypothesis, the pretreatmentβ-adrenergic receptor density (Bmax) and sensitivity(cAMP ratio) in the panic patients showed no significantdifference compared with those of the control subjectsand changes in Kd did not reflect changes in cAMP ratio.The decrease in the β-adrenoceptor affinity (1/Kd) mayaffect the efficiency with which agonist binding trans-lates into cAMP production. This process determines thedecreased sensitivity (desensitization) (Mills and Dims-dale, 1988). In a previous study, the β-adrenoceptorresponsiveness decreased in the female patients, but notin the male patients, in comparison with the normalcontrols ( Kim et al., 2004). Consequently, the unevengender ratio in this study might have affected this re-sult. Additionally, the existence of a racial difference inthe β-adrenoceptor sensitivity and density was demon-strated in a previous study ( Mills et al., 1995b), whichmight also affect the result of this study. The true lossof receptor numbers (down-regulation) can occur aftermore chronic exposure to high concentrations of anagonist (De Blasi et al., 1985; Motulsky et al., 1986),and a time-dependent sequential decrease of the centralβ-adrenoceptor affinity (1/Kd) was demonstrated in aprevious study (Flugge et al., 1997). In this study, themean duration of illness in the panic patients (17.19±18.20 weeks) was much shorter than that of a previousstudy (110±90 months) (Maddock et al., 1993b). In thisregard, decreased β-adrenoceptor affinity (1/Kd) couldbe a time-dependent characteristic in acute PD patients.

Our study is the first report on the functional dif-ferences in the β-adrenoceptor measures between treat-ment responders and non-responders to short-termpharmacotherapy with paroxetine in acute PD. Theresponders to pharmacotherapy with paroxetine in theacute panic patients had a lower β-adrenoceptor affinity(1/Kd) before treatment compared with the non-respon-ders. Furthermore, decreased pretreatment β-adrenocep-tor affinity (1/Kd) reliably predicted treatment response.This result is consistent with previous studies whichreported that decreasedβ-adrenoceptor function predictedgood treatment response (Maddock et al., 1993b; Park andYu, 2005) and support the hypothesis that the moreadaptive the response of the β-adrenoceptors to panicanxiety, the better the treatment response that could beobtained (Maddock et al., 1993b).

Our study has some limitations. First, the unevengender ratio of the panic patients might have affected ourresults. Second, the size of our study population was toosmall to generalize the results concerning panic disorder.Third, we did not consider the severity of agoraphobia

160 I.-S. Lee et al. / Journal of Affective Disorders 110 (2008) 156–160

for the analysis, although the presence of agoraphobiadid not differ between the treatment responders and non-responders at baseline.

In conclusion, only the β-adrenoceptor affinity (1/kd)was decreased in the acute panic patients and was adap-tively normalized after treatment. We suggest that lowpretreatment β-adrenoceptor affinity (1/kd) could be abiological marker to predict the treatment response inacute PD.

Role of funding sourceFunding for this study was provided by SBRI Grants CA00141 and

C-A6-418-2 from the Samsung Biomedical Research Institute. TheSamsung Biomedical Research Institute had no further role in studydesign; in the collection, analysis and interpretation of data; in thewriting of the report; and in the decision to submit the paper forpublication.

Conflict of interestNone.

References

Aronson, T.A., Carasiti, I., McBane, D., Whitaker-Azmitia, P., 1989.Biological correlates of lactate sensitivity in panic disorder. Biol.Psychiatry 26, 463–477.

Brown, S.L., Charney, D.S., Woods, S.W., Heninger, G.R., Tallman, J.,1988. Lymphocyte beta-adrenergic receptor binding in panicdisorder. Psychopharmacology (Berl.) 94, 24–28.

Dao, T.T.,Kailasam,M.T., Parmer, R.J., Le,H.V., LeVerge, R., Kennedy,B.P., Ziegler, G., Insel, P.A., Wright, F.A., O'Connor, D.T., 1998.Expression of altered alpha2-adrenergic phenotypic traits innormotensive humans at genetic risk of hereditary (essential) hyper-tension. J. Hypertens. 16, 779–792.

De Blasi, A., Lipartiti, M., Motulsky, H.J., Insel, P.A., Fratelli, M.,1985. Agonist-induced redistribution of beta-adrenergic receptorson intact human mononuclear leukocytes: redistributed receptorsare nonfunctional. J. Clin. Endocrinol. Metab. 61, 1081–1088.

First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 1997. Struc-tured Clinical Interview for DSM IVAxis I Disorders—Clinicianversion(SCID-CV). American Psychiatric Press, Washington, D.C.

Flugge, G., Ahrens, O., Fuchs, E., 1997. Beta-adrenoceptors in the treeshrew brain. II. Time-dependent effects of chronic psychosocialstress on [125I]iodocyanopindolol bindings sites. Cell. Mol.Neurobiol. 17, 417–432.

Gorman, J.M., Kent, J.M., Sullivan, G.M., Coplan, J.D., 2000.Neuroanatomical hypothesis of panic disorder, revised. Am. J.Psychiatry 157, 493–505.

Hamilton, M., 1959. The assessment of anxiety states by rating. Br. J.Med. Psychol. 32, 50–55.

Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg.Psychiatry 23, 56–62.

Johnson, M., 1998. The beta-adrenoceptor. Am. J. Respir. Crit. CareMed. 158, S146–S153.

Johnson, M.R., Lydiard, R.B., 1995. The neurobiology of anxietydisorders. Psychiatr. Clin. North Am. 18, 681–725.

Joyce, J.N., Lexow, N., Kim, S.J., Artymyshyn, R., Senzon, S.,Lawrence, D., Cassanova, M.F., Kleinman, J.E., Bird, E.D.,

Winokur, A., 1992. Distribution of beta-adrenergic receptor subtypesin human post-mortem brain: alterations in limbic regions ofschizophrenics. Synapse 10, 228–246.

Kim, Y.R., Min, S.K., Yu, B.H., 2004. Differences in beta-adrenergicreceptor sensitivity between women and men with panic disorder.Eur. Neuropsychopharmacol. 14, 515–520.

Lafaille, F., Welner, S.A., Suranyi-Cadotte, B.E., 1991. Regulation ofserotonin type 2 (5-HT2) and beta-adrenergic receptors in ratcerebral cortex following novel and classical antidepressanttreatment. J. Psychiatry Neurosci. 16, 209–214.

Liebowitz, M.R., Fyer, A.J., Gorman, J.M., Dillon, D., Appleby, I.L.,Levy, G., Anderson, S., Levitt, M., Palij, M., Davies, S.O., et al.,1984. Lactate provocation of panic attacks. I. Clinical andbehavioral findings. Arch. Gen. Psychiatry 41, 764–770.

Maddock, R.J., Carter, C.S., Magliozzi, J.R., Gietzen, D.W., 1993a.Evidence that decreased function of lymphocyte beta adrenor-eceptors reflects regulatory and adaptive processes in panicdisorder with agoraphobia. Am. J. Psychiatry 150, 1219–1225.

Maddock, R.J., Gietzen, D.W., Goodman, T.A., 1993b. Decreased lym-phocyte beta-adrenoreceptor function correlates with less agorapho-bia and better outcome in panic disorder. J. Affect. Disord. 29, 27–32.

Mills, P.J., Dimsdale, J.E., 1988. The promise of receptor studies inpsychophysiologic research. Psychosom. Med. 50, 555–566.

Mills, P.J., Dimsdale, J.E., Ziegler, M.G., Berry, C.C., Bain, R.D.,1990. Beta-adrenergic receptors predict heart rate reactivity to apsychosocial stressor. Psychosom. Med. 52, 621–623.

Mills, P.J., Dimsdale, J.E., Coy, T.V., Ancoli-Israel, S., Clausen, J.L.,Nelesen, R.A., 1995a. Beta 2-adrenergic receptor characteristics insleep apnea patients. Sleep 18, 39–42.

Mills, P.J., Dimsdale, J.E., Ziegler, M.G., Nelesen, R.A., 1995b. Racialdifferences in epinephrine and beta 2-adrenergic receptors. Hyper-tension 25, 88–91.

Motulsky, H., Cunningham, E.M.S., Deblasi, A., Insel, P.A., 1986.Agonists promote rapid desensitization and redistribution of betaadrenergic receptors on intact human molecular leukocytes. Am. J.Physiol. 250, 583–590.

Nesse, R.M., Cameron, O.G., Curtis, G.C., McCann, D.S., Huber-Smith, M.J., 1984. Adrenergic function in patients with panicanxiety. Arch. Gen. Psychiatry 41, 771–776.

Pandey, G.N., Janicak, P.G., Davis, J.M., 1987. Decreased beta-adrenergic receptors in the leukocytes of depressed patients.Psychiatry Res. 22, 265–273.

Park, J.E., Yu, B.H., 2005. Biological markers that predict thetreatment response of paroxetine in panic disorder. J. KoreanNeuropsychiatr. Assoc. 44, 238–246.

Perry, B., U'Prichard, D.C., 1984. Alpha adrenergic receptors in neuraltissues: methods and applications of radioligand binding assays.In: Marangos, P.J., Campbell, I.C., Cohen, R.M. (Eds.), BrainReceptor Methodologies. Academic press, Florida, pp. 255–283.

Reiss, S., Peterson, R.A., Gursky, D.M., McNally, R.J., 1986. Anxietysensitivity, anxiety frequency and the prediction of fearfulness.Behav. Res. Ther. 24, 1–8.

Spielberger, C., Gorsuch, R., Lushene, R., Vagg, P.R., Jacobs, G.A.,1983.Manual for the State–Trait Anxiety Inventory, STAI (FormY).Consulting Psychologists Press, palo Alto, CA.

Yu, B.H., Dimsdale, J.E., Mills, P.J., 1999. Psychological states andlymphocyte beta-adrenergic receptor responsiveness. Neuropsy-chopharmacology 21, 147–152.