Cytokines and immune activation in systolic heart failure: The role of type D personality

Tilburg University

Cytokines and immune activation in systolic heart failure

Denollet, J.K.L.; Conraads, V.; Brutsaert, D.L.; de Clerck, L.S.; Stevens, W.J.; Vrints, C.J.

Published in:

Brain, Behavior, and Immunity: An international journal

Publication date:

2003

Document Version

Peer reviewed version

Link to publication in Tilburg University Research Portal

Citation for published version (APA):

Denollet, J. K. L., Conraads, V., Brutsaert, D. L., de Clerck, L. S., Stevens, W. J., & Vrints, C. J. (2003). Cytokines and immune activation in systolic heart failure: The role of type D personality. Brain, Behavior, and Immunity: An international journal, 17(4), 304-309.

General rights

Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights. • Users may download and print one copy of any publication from the public portal for the purpose of private study or research. • You may not further distribute the material or use it for any profit-making activity or commercial gain

• You may freely distribute the URL identifying the publication in the public portal

Take down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

CONFIDENTIAL

Cytokines and Immune Activation in Systolic Heart Failure:

The Role of Type D Personality

Johan Denollet*, PhD; Viviane M. Conraads†, MD; Dirk L. Brutsaert†, MD;

Luc S. De Clerck#, MD; Wim J. Stevens#, MD; Christiaan J. Vrints†, MD * Depart ment of Psychology and Healt h, Tilburg Universit y, The Net herlands

†

Depart ment of Cardiology, # Depart ment of Immunology, Universit y Hospit al Ant werp, Belgium

Brain, Behavior, and Immunit y 2003; 17:000-000, in press.

Short t it l e:

Type D, TNF-α and Heart Failure

Correspondence t o:

Johan Denollet , PhD Medical Psychology

Tilburg Universit y, Warandelaan, 2 P.O. Box 90153

5000 LE Tilburg THE NETHERLANDS

Tel: +- 31-13-466 2390

Fax: +- 31-13-466 2370

ABSTRACT

The proinflammat ory cyt okine t umor necrosis fact or-α (TNF-α) and it s soluble recept ors 1 (sTNFR1) and 2 (sTNFR2) are predict ors of mort alit y in chronic heart failure (CHF) but t he det erminant s of t hese increased levels of disease-promot ing cyt okines are largely unknown. Type D personalit y refers t o t he combinat ion of t he t endency t o experience negat ive emot ions (negat ive affect ivit y) and t he t endency t o inhibit t he expression of emot ions in social int eract ion (social inhibit ion). Type D is an independent predict or of cardiac event s in coronary pat ient s who are at risk for CHF. The present st udy examined t he effect of Type D personalit y on TNF-α, sTNFR1 and sTNFR2 in 42 men wit h CHF(mean age= 57.9 ± 10.5 years). There was a significant mult ivariat el effect of Type D on TNF-α measures (p=.006); i.e., circulat ing levels of TNF-α (4.8 ± 0.9 versus 2.5 ± 0.2 pg/ ml, p=.003), sTNFR1 (1814 ± 314 versus 1134 ± 78 pg/ ml, p=.014) and sTNFR2 (2465 ± 243 versus 1874 ± 118 pg/ ml, p=.019) were significant ly higher in Type D pat ient s as compared t o non-Type D pat ient s. The effect size (ES) of Type D personalit y ranged from rat her large (sTNFR1, ES=0.77; sTNFR2, ES= 0.73) t o large (TNF-α, ES=0.90). Aft er

cont rolling for ischemic et iology and severit y of heart failure, Type D personalit y emerged as an independent predict or of increased circulat ing levels of bot h TNF-α (OR= 9.5, 95% CI 2.1-43.8,

INTRODUCTION

“ The impact of t he immune syst em would be an int erest ing subj ect f or f ut ure research in cardiovascular behavioral medicine,

which might be of relevance f or t he onset of acut e coronary syndromes…”

- Willem Kop,1994 -

This opening quot at ion is in fact t he very last sent ence of t he closing chapt er of Dr. Willem Kop’ s innovat ive work published in 1994. His conclusion was t he harbinger for t he int egrat ion of psychoneuroimmunology wit h psychosomat ic cardiology. A number of reviews in primary j ournals (Kop, 1999; Rozanski, Blument hal & Kaplan, 1999; Ziegelst ein, 2001) have provided abundant evidence suggest ing t hat psychological fact ors may impact on t he development and course of acut e coronary syndromes. The underlying mechanisms explaining t his associat ion include indirect mechanisms such as poor adherence t o t reat ment (Ziegelst ein, Bush, Fauerbach, 1998), and more direct physiological mechanisms such as impaired plat elet funct ion, decreased heart rat e variabilit y, and t riggering of myocardial ischemia (Krant z, Kop, Sant iago & Got t diener, 1996). Lat ely, immune act ivat ion has been proposed as a novel mechanism t hat may explain t he link bet ween emot ional dist ress and acut e coronary event s (Appels, J. Bär, C. Bär, Bruggeman & de Baet s, 2000; Ishihara, Nohara, Makit a, Imai, Kubo & Hashimot o, 1999; Kop & Cohen, 2001).

Chronicheart failure(CHF)hasemergedasanepidemic as a result of anagingpopulat ionand

improved survival aft er myocardial infarct ion (Goldberg & Konst am, 1999). Each year,

approximat ely 550,000 pat ient s develop heart failure in t he Unit ed St at es (Hellermann et al., 2002). CHF is a condit ion t hat carries a high mort alit y risk (Hellermann et al., 2002) and t hat calls for t he furt her ident ificat ion of risk fact ors for a poor prognosis (Faris, Purcell, Henein & Coat s, 2002). In t he past t wo decades, t here has been a shift in t he et iology of CHF from hypert ension or valvular disease t o coronary heart disease (Gheorghiade & Bonow, 1998). There is also growing evidence t hat elevat ed concent rat ions of proinflammat ory cyt okines play an import ant role in t he pat hogenis/ progression of CHF (Deswal, Pet ersen , Feldman, Young , Whit e & Mann, 2001).

plaque. Binding of membrane-bound TNF-α recept ors wit h t heir ligand, TNF-α, result s in shedding of t he ext racellular domain, referred t o as soluble TNF-α recept or 1 and recept or 2 (sTNFR1 and sTNFR2). Plasma levels of sTNFR1 and sTNFR2 purport edly reflect exposure of t he organism t o TNF-α over longer periods of t ime. TNF-α, sTNFR1 and sTNFR2 have consist ent ly emerged as predict ors of mort alit y in pat ient s wit h CHF (Deswal et al., 2001; Ferrari, Bachet t i, Confort ini, et al., 1995; Rauchhaus, Doehner, Francis, et al., 2000; Torre-Amione et al., 1996). The det erminant s of t hese increased levels of TNF-α in CHF pat ient s are not well underst ood.

Negat ive emot ions are associat ed wit h increased product ion of proinflammat ory cyt okines including TNF-α (Kiecolt -Glaser, McGuire, Robles & Glaser, 2002). Episodic psychological risk fact ors such as depression have been associat ed wit h a poor prognosis in CHF (Jiang, Alexander, Christ opher, et al., 2001; Vaccarino, Kasl, Abramson , Krumholz , 2001; Faris, Purcell, Henein & Coat s, 2002). Chronic psychological risk fact ors may also affect clinical manifest at ions of heart disease (Kop, 1999), but individual difference variables like personalit y t rait s have received lit t le at t ent ion t o dat e in behavioral immunology research (Miller, Cohen, Rabin, Skoner & Doyle, 1999).

Type D personalit y is an individual difference variable t hat may play a role in t he prognosis of CHF. Type-D refers t o t he combinat ion of t he t endency t o experience negat ive emot ions (i.e., negat ive affect ivit y) and t he t endency t o inhibit self-expression (i.e., social inhibit ion). This personalit y predisposes t o chronic emot ional st ress in coronary pat ient s. In a 6-10 year follow-up st udy, Type D was associat ed wit h a 4-fold increased risk for mort alit y (Denollet et al., 1996). A 5-year prospect ive follow-up st udy confirmed t hat Type D pat ient s were at an increased risk of cardiac event s (Denollet et al. 2000). Type D is also predict ive of t he clinical course of coronary pat ient s who were at risk for CHF (Denollet & Brut saert , 1998). Therefore, t he purpose of t his st udy was t o examine associat ions bet ween Type D personalit y and TNF-α in pat ient s wit h CHF.

METHODS Subjects

if pat ient s were free of hospit alizat ion and st able wit h regard t o sympt oms and medical t herapy (ACE-inhibit ors 90% of pat ient s, diuret ics - spironolact one •EHWD-blockers 54%; digoxin 39%,

aspirin - amiodarone 24%) for at least one mont h. Pat ient s wit h act ive infect ion, allergy, rheumat oid disease, cancer, or t aking ant i-inflammat ory medicat ions were excluded. The st udy was approved by t he Local Et hical Commit t ee and all pat ient s gave writ t en informed consent .

Type D personality

Personalit y was assessed using t he Type D Scale-14 or DS14 (Denollet , 2002) comprising a 7-it em subscale measuring negat ive affect ivit y (t he first personalit y component of Type D) and a 7-it em subscale measuring social inhibit ion (t he second personalit y component of Type D). Cronbachs’ α of t hese subscales is 0.88 and 0.86, respect ively. A cut -off of 10 on bot h DS14 subscales was used t o classify pat ient s as Type D, result ing in 16 Type D pat ient s and 26 non-Type D pat ient s.

TNF-D, sTNFR1 and sTNFR2

An enzyme-linked immunosorbent assay (ELISA) was used t o measure circulat ing plasma levels of TNF-α, sTNFR1 and sTNFR2. Fast ing blood samples were collect ed bet ween 8 and 9 AM int o et hylenediaminet et raacet ic (EDTA) t ubes (Vacut ainer, Bect on and Dickinson, Meylan, France) and plasma was separat ed by cent rifugat ion and aliquot s were st ored at –20°C. A high sensit ivit y kit (Quant ikine HS, R&D Syst ems, sensit ivit y 0.18 pg/ ml) was used t o measure TNF-α. The manufact urer’ s specificat ions were used t o measure soluble TNF-α recept ors (Quant ikine, R&D Syst ems, sensit ivit y: 1.5 pg/ ml for sTNFR1, 1 pg/ ml for sTNFR2). All samples were run in duplicat e. The invest igat ors were blinded wit h regard t o CHF et iology and Type D st at us.

Etiology and severity of CHF

Deswal et al. (2001) showed t hat circulat ing levels of cyt okines and cyt okine recept ors were significant ly great er in CHF pat ient s wit h ischemic heart disease as compared t o pat ient s wit h idiopat hic CHF. Therefore, ischemic cause of CHF was included t o cont rol for t his maj or

cardiological det erminant of cyt okine proliferat ion. LVEF and New York Heart Associat ion [NYHA] class were included t o cont rol for disease severit y as a det erminant of TNF-α and it s recept ors.

Statistical analyses

as a whole. An ES=0.2 was considered t o indicat e a small effect , an ES=0.5 a moderat e effect and an ES=0.8 a large effect (Cohen, 1988). A median split was used t o dichot omize cyt okine levels in order t o st rat ify pat ient s in high- and low-risk groups, respect ively. Mult iple logist ic regression analyses (met hod= ent er) were used t o det ermine t he risk associat ed wit h Type D aft er cont rolling for ischemic cause of CHF and disease severit y as indicat ed by NYHA class. All analyses were conduct ed using t he SPSS soft ware package, release 11.0 (SPSS Inc., Chicago,Il).

RESULTS

- Table 1 -

Type D and characteristics of CHF

The diagnosis of Type D personalit y was not significant ly associat ed wit h age (p= .63), ischemic et iology of CHF (p= .23), or severit y of CHF as indicat ed by mean LVEF (p= .79) or proport ion of CHF pat ient s wit h a LVEF < 25% (Table 1). There was a t endency for Type D pat ient s t o be more oft en classified in NYHA class III/ IV t han non-Type D pat ient s, but t his difference was not st at ist ically significant (p= .13). Type D pat ient s did not differ from non-Type Ds in t erms of medical t reat ment , including use of ACE-inhibit ors, diuret ics, spironolact one, or bet a-adrenergic blocking agent s.

- Figure 1 -

Type D and mean levels of TNF-D

A general linear mult ivariat e model of TNF-α, sTNFR1 and sTNFR2 wit h Type D as bet ween-subj ect s fact or indicat ed a significant overall effect of Type D personalit y [main effect : F(1,40)= 8.42, p=.006]. Post -hoc analyses revealed t hat Type D pat ient s displayed significant ly increased circulat ing plasma levels of (a) TNF-α [F(1,40)= 9.66, p=.003] and (b) bot h sTNFR1 [F(1,40)= 6.64, p=.014] and sTNFR2 [F(1,40)= 5.94, p=.019], as compared t o non-Type D pat ient s (Figure 1). Ischemic et iology of heart failure was also associat ed wit h significant ly increased plasma levels of TNF-α (4.1 ± 2.9 pg/ ml versus 2.4 ± 1.1 pg/ ml, p=.03) but not sTNFR1 (p=.25) or sTNFR2 (p=.14).

ES=0.90). In comparison, t he effect of ischemic et iology –an import ant cardiological det erminant of immune act ivat ion in CHF pat ient s (Deswal et al., 2001)- ranged from rat her small (sTNFR1, ES=0.37) and moderat e (sTNFR2, ES= 0.46) t o rat her large (TNF-α, ES=0.66). These findings suggest t hat Type D was as import ant as ischemic et iology wit h reference t o immune act ivat ion.

Type D and high levels of TNF-D

A median split was used t o classify pat ient s as displaying high versus low levels of TNF-α (mean= 4.8 ± 2.7 pg/ ml versus 1.9 ± 0.7 pg/ ml). An aggregat e index was used t o cont rast pat ient s wit h high levels of bot h TNF-recept ors (mean sTNFR1= 1794 ± 930 pg/ ml and sTNFR2= 2737 ± 504 pg/ ml) wit h pat ient s scoring below t he median on at least one of t hese t wo recept ors (mean sTNFR1= 844 ± 193 pg/ ml and sTNFR2= 1468 ± 351 pg/ ml). Type D men were significant ly more likely t o display high levels of TNF-α (75% versus 23%; OR=10.0 [95%CI 2.3-42.8], p=.002) and it s recept ors (69% versus 23%; OR=7.3 [95%CI 1.8-29.6], p=.005) as compared wit h non-Type D men.

- Table 2 -

Type D as independent predictor

Apart from Type D personalit y, ischemic et iology and NYHA class were included in a mult iple logist ic regression model in order t o cont rol for et iology and severit y of CHF. This model

ret ained Type D personalit y (p=.004) as an independent predict or of high TNF-α levels (Table 2, t op). Accordingly, Type D personalit y was also an independent predict or of high TNF-α recept or levels (p=.014) aft er cont rolling for bot h t he et iology and severit y of CHF (Table 2, bot t om).

DISCUSSION

Yet , in light of t he prognost ic power of (1) TNF-α in pat ient s wit h CHF (Deswal et al., 2001) and (2) Type D personalit y in pat ient s at risk for CHF (Denollet & Brut saert , 1998), t hey also do warrant furt her research in t he complex domain of personalit y, immune act ivat ion and CHF. There has been a shift in t he et iology of CHF from hypert ension or valvular disease t o coronary heart disease (Gheorghiade & Bonow, 1998). The et iology of coronary heart disease, in t urn, includes dynamic, pat hophysiological fact ors such as immune fact ors (Ross, 1999). TNF-α may play a role in each of t hree pat hogenic st ages at t he sit e of a vulnerable at herosclerot ic plaque: (1) plaque inst abilit y, (2) plaque rupt ure, and (3) t hrombosis (Gidron et al. 2002; Kop, 1994).

First , TNF-α st imulat es adhesion and migrat ion of leukocyt es int o t he coronary endot helium and act ivat ion of macrophages, causing migrat ion of smoot h muscle cells and subsequent plaque inst abilit y. Second, TNF-α has been t hought t o st imulat e vasoconst rict ion and elevat ed blood pressure; i.e., t wo ext ra-cellular fact ors t hat may induce plaque rupt ure. Third,

proinflammat ory cyt okines regulat e t he int eract ion bet ween t he endot helium and blood plat elet s as well as clot t ing and fibrinolyt ic fact ors, causing enhanced plat elet aggregat ion and subsequent t hrombosis at t he rupt ured plaque. In addit ion, persist ence of immune processes t hought t o be crit ical in t he et iology of t he coronary event s may result from an imbalance bet ween neuroendocrine fact ors t hat enhance and fact ors t hat inhibit immune responses.

Depression (Maes, Bosmans, Melt zer, Scharpe & Suy, 1993), acut e st ress (Maes, Song, Lin et al., 1998; St ept oe, Willemsen, Owen, Flower & Mohamed-Ali, 2001), and exhaust ion (Appels et al, 2000) have also been relat ed t o elevat ions of pro-inflammat ory cyt okines. On t he one hand, t his chain of pat hophysiological processes may endure due t o lack of neuroendocrine-t o-immune negat ive feedback st emming from ‘ cort isol resist ance’ (Chrousos, 1995). This condit ion of insufficient cort isol-induced immune suppression may account for t his persist ence of PNI processes pot ent ially leading t o an acut e coronary event . On t he ot her hand, act ivat ion of t he sympat het ic nervous syst em may also promot e pro-inflammat ory cyt okines such as TNF-α.

progression in gay men who concealed t heir homosexual ident it y as compared t o t hose who did not . Their descript ion of t hese high-risk HIV pat ient s (Cole, Kemeny & Taylor, 1997) mat ches remarkably well wit h our descript ion of coronary pat ient s wit h Type D (Denollet et al., 1996).

The present findings indicat ed t hat Type D personalit y was an independent predict ors of higher values for TNF-α in men wit h CHF. The presence of chronic inflammat ion in t hese pat ient s has been widely recognized, and circulat ing levels of sTNFR1 proved t o be bet t er predict or of mort alit y t han st andard risk fact ors such as NYHA class or impaired LVEF (Deswal et al., 2001; Rauchhaues et al. 2000). Of not e, CHF pat ient s wit h a Type D personalit y displayed significant ly higher plasma levels of t his TNF-α recept or as compared t o non-Type D pat ient s. Hence, chronic psychoimmunological dysregulat ion may be a mechanism t hat mediat es t he relat ion bet ween Type D and poor prognosis in pat ient s wit h ischemic heart disease (Denollet et al., 1996).

The co-occurrence of emot ional dist ress and coronary heart disease may lead t o higher

morbidit y and mort alit y (Rozanski et al., 1999). Depression is relat ively common in pat ient s wit h CHF (Skot zko, Kricht en, Ziet owski et al., 2000). Not only depression but ot her psychological fact ors may as well impact on morbidit y and mort alit y in CHF. The findings of t he present st udy highlight t he import ance of individual difference variables such as broad and st able personalit y t rait s t hat may modify circulat ing levels of disease-promot ing proinflammat ory cyt okines. Kiecolt -Glaser et al. (2002) have argued t hat dist ress-relat ed immune dysregulat ion may be one core mechanism behind a large and diverse set of healt h risks associat ed wit h negat ive

REFERENCES

Appels, A., Bär, F.W., Bär, J., Bruggeman, C., & de Baet s, M. (2000). Inflammat ion, depressive sympt omt ology, and coronary art ery disease. Psychosom. Med. 62, 601¯605.

Chrousos, G.P. (1995). The hypot halamic¯pit uit ary¯adrenal axis and immune-mediat ed

inflammat ion. New Engl . J. Med. 332, 1351¯1362.



Cohen, J. (1988). St at ist ical Power Anal ysis f or t he Behavioral Sciences. Hilsdale, NJ: Lawrence Erlbaum Associat es 1988.

Cole, S.W., Kemeny, M.E. & Taylor, S.E. (1997). Social ident it y and physical healt h: accelerat ed HIV progression in rej ect ion-sensit ive gay men. J. Pers. Soc. Psychol . 72, 320-335.

Cole, S.W., Kemeny, M.E., Taylor, S.E., Visscher, B.R., & Fahey, J.L. (1996). Accelerat ed course of human immunodeficiency virus infect ion in gay men who conceal t heir homosexual ident it y. Psychosom. Med. 58, 219-231.

Denollet , J. (2002). Type D personalit y and vulnerabilit y t o chronic disease, impaired qualit y of life, and depressive sympt oms (abst ract ). Psychosom. Med. 64, 101.

Denollet , J., & Brut saert , D.L. (1998). Personalit y, disease severit y, and t he risk of long-t erm cardiac event s in pat ient s wit h decreased ej ect ion fract ion aft er myocardial infarct ion.

Circul at ion. 97, 167-173.

Denollet , J., Conraads, V.M., Vrint s, C.J., De Clerck, L.S., St evens, W.J., van den Broek, K., Brut saert , D.L. (2002). Type D personalit y is associat ed wit h increased TNF-α recept or levels in pat ient s wit h heart failure (abst ract ). Psychosom. Med. 64, 152.

Denollet , J., Sys, S.U., St roobant , N., Rombout s, H., Gillebert , T.C., & Brut saert , D.L. (1996). Personalit y as independent predict or of long-t erm mort alit y in pat ient s wit h coronary heart disease. Lancet . 347, 417-421.

Denollet , J., Vaes, J., & Brut saert , D.L. (2000). Inadequat e response t o t reat ment in coronary heart disease. Adverse effect s of Type D personalit y and younger age on 5-year prognosis and qualit y of life. Circul at ion. 102, 630-635.

Deswal, A., Pet ersen, N.J., Feldman, A.M., Young, J.B., Whit e, B.G., & Mann, D.L. (2001). Cyt okines and cyt okine recept ors in advanced heart failure. An analysis of t he cyt okine dat abase from vesnarinone t rial (VEST). Circul at ion. 103, 2055-2059.

Faris, R., Purcell, H., Henein, M., & Coat s, A. (2002). Clinical depression is common and

significant ly associat ed wit h reduced survival in pat ient s wit h non-ischemic heart failure.

Eur. J. Heart Fail. 4, 541-551.

Gheorghiade, M., & Bonow, R.O. (1998). Chronic heart failure in t he Unit ed St at es: a manifest at ion of coronary art ery disease. Circul at ion. 97, 282-289.

Gidron, Y., Gilut z, H., Berger, R., & Huleihel, M. (2002). Molecular and cellular int erface bet ween behavior and acut e coronary event s. Cardiovasc. Res. 56, 15-21.

Goldberg, R.J., & Konst am, M.A. (1999). Assessing t he populat ion burden from heart failure: need for sent inel populat ion-based surveillance syst ems. Arch. Int ern. Med. 159, 15-17. Hellermann, J., Jacobsen, S., Gersh, B., Rodeheffer, R., Reeder, G., & Roger, V. (2002). Heart

failure aft er myocardial infarct ion: a review. Am. J. Med. 113, 324-330.

Ishihara, S., Nohara, R., Makit a, S., Imai, M., Kubo, S., & Hashimot o, T. (1999). Immune funct ion and psychological fact ors in pat ient s wit h coronary heart disease (I).

Jpn. Circ. J. 63, 704-709.

Jessup, M. (2003). The less familiar face of heart failure. J. Am. Col l . Cardiol . 41, 224-226. Jiang, W., Alexander, J., Christ opher, E., et al. (2001). Relat ionship of depression t o increased

risk of mort alit y and rehospit alizat ion in pat ient s wit h congest ive heart failure. Arch.

Int ern. Med. 161, 1849-1856.

Kiecolt -Glaser, J.K., McGuire, L., Robles, T.F., & Glaser, R. (2002). Emot ions, morbidit y, and mort alit y: new perspect ives from psychoneuroimmunology. Annu. Rev. Psychol. 53, 83-107.

Kop, W.J. (1994). The predict ive value of vit al exhaust ion in t he clinical course af t er coronary

angioplast y. Maast richt (The Net herlands): Dat awyse, Maast richt Universit y Press.

Kop, W.J. (1999). Chronic and acut e psychological risk fact ors for clinical manifest at ions of coronary art ery disease. Psychosom. Med. 61, 476-487.

Kop, W.J., Cohen, N. (2001). Immune syst em involvement in cardiovascular disease. In: Ader, R., Felt en, D.L., Cohen, N. (Eds.): Psychoneuroimmunology. 3rd ed. San Diego (CA): Academic Press.

Krant z, D.S., Kop, W.J., Sant iago, H.T., & Got t diener, J.S. (1996). Ment al st ress as a t rigger of myocardial ischemia and infarct ion. Cardiology Cl inics. 14, 271-287.

Kronfol, Z., & Remick, D.G. (2000). Cyt okines and t he brain: implicat ions for clinical psychiat ry.

Am. J. Psychiat ry. 157, 683¯694.

Maes, M., Bosmans, E., Melt zer, H.Y., Scharpe, S., & Suy, E. (1993). Int erleukin-1 bet a: a put at ive mediat or of HPA axis hyperact ivit y in maj or depression. Am. J. Psychiat ry.

150, 1189¯1193.

Maier, S.F., & Wat kins, L.R. (1998). Cyt okines for psychologist s: implicat ions of bidirect ional immune-t o-brain communicat ion for underst anding behavior, mood, and cognit ion.

Psychol. Rev. 105, 83-107.

Masoudi, F.A., Havranek, E.P., Smit h, G, Fish, R.H., St einer, J.F., Ordin, D.L., & Krumholz, H.M. (2003). Gender, age, and heart failure wit h preserved left vent ricular syst olic funct ion. J. Am. Col l . Cardiol . 41, 217-223.

Miller, G.E., Cohen, S., Rabin, B.S, Skoner, D.P., & Doyle, W.J. (1999). Personalit y and t onic cardiovascular, neuroendocrine, and immune paramet ers. Brain Behav. Immun. 13, 109-123.

Rauchhaus, M., Doehner, W., Francis, D.P., Davos, C., Kemp, M., Liebent hal, C., Niebauer, J., Hooper, J., Volk, H.D., Coat s, A.J., & Anker SD. (2000). Plasma cyt okine paramet ers and mort alit y in pat ient s wit h chronic heart failure. Circulat ion. 102, 3060-3067.

Ross, R. (1999). At herosclerosis: an inflammat ory disease. New Engl . J. Med. 340, 115¯126.

Rozanski, A., Blument hal, J.A., & Kaplan, J. (1999). Impact of psychological fact ors on t he pat hogenesis of cardiovascular disease and implicat ions for t herapy. Circul at ion. 99, 2192-2217.

Skot zko, C.E., Kricht en, C., Ziet owski, G., Alves, L., Freudenberger, R., Robinson, S., Fisher, M., & Got t lieb, S.S. (2000). Depression is common and precludes accurat e assessment of funct ional st at us in elderly pat ient s wit h congest ive heart failure. J. Cardiac Fail. 6, 300¯305.

St ept oe, A., Willemsen, G., Owen, N., Flower, L., & Mohamed-Ali, V. (2001). Acut e ment al st ress elicit s delayed increases in circulat ing inflammat ory cyt okine levels. Cl in. Sci.

101, 185¯192.

Torre-Amione, G., Kapadia, S., Lee, J., Durand, J.B., Bies, R.D., Young, J.B., & Mann, D.L. (1996). Tumor necrosis fact or-α and t umor necrosis fact or recept ors in t he failing human heart . Circul at ion. 93, 704-711.

Vaccarino, V., Kasl, S.V., Abramson, J., & Krumholz, H.M. (2001). Depressive sympt oms and risk of funct ional decline and deat h in pat ient s wit h heart failure. J Am Col l Cardiol. 38, 199-205.

Ziegelst ein, R.C. (2001). Depression in pat ient s recovering from a myocardial infarct ion. JAMA.

286, 621-1627.

Ziegelst ein, R.C., Bush, D.E., & Fauerbach, J.A. (1998). Depression, adherence behavior, and coronary disease out comes. Arch Int ern Med.158, 808-809.

Table 1 Baseline characteristics as a function of Type D personality. _________________________________________________________________________ Personalit y Type ___________________________________________________________ Non-Type D Type D (n = 26) (n = 16) _________________________________________________________________________ Age 58.5 (10.7) 56.9 (10.5) p = .63

Severity & Etiology of CHF

LVEF (%) 25.0 (7.8) 25.7 (6.9) p = .79

LVEF < 25% 10 (39%) 7 (44%) p = .74

NYHA class III/ IV 10 (39%) 10 (62%) p = .13

Ischemic et iology 13 (50%) 11 (69%) p = .23 Medical treatment ACE-inhibit ors 24 (92%) 14 (87%) p = .62 Diuret ics 24 (92%) 13 (81%) p = .34 Spironolact one 22 (85%) 13 (81%) p = .69 Bet a-blockers 15 (58%) 8 (50%) p = .53 Digoxin 12 (46%) 4 (25%) p = .32 Aspirin 8 (31%) 2 (13%) p = .28 Amiodarone 8 (31%) 2 (13%) p = .28 _________________________________________________________________________

Table 2. Predictors of Increased Levels of TNF-D and Its Receptors

__________________________________________________________________________ Logist ic regression model

______________________________________________ Variable Odds ratio 95% CI p-value __________________________________________________________________________

TNF-D

Type D personalit y * 9.5 2.1 t o 43.8 .004

Ischemic et iology 2.7 0.6 t o 12.4 .20 NYHA class III/ IV 0.9 0.2 t o 4.1 NS

Receptors

Type D personalit y * 6.1 1.4 t o 25.8 .014

Ischemic et iology 1.8 0.4 t o 8.2 NS NYHA class III/ IV 2.3 0.5 t o 9.7 NS

__________________________________________________________________________

* Independent predict or of enhanced immune act ivat ion (p <.05); increased

FIGURE LEGEND

Figure 1. Plasma Levels of TNF-D and Its Receptors, Stratified by Type D Personality