Combined effect of depressive symptoms and hostility on autonomic nervous system function

Abstract

Depression and hostility have been separately related to indicators of sympathetic hyperactivation and parasympathetic hypoactivation. We examined the associations of depressive symptoms, hostility, and their interaction with pre-ejection period (PEP) and high frequency heart rate variability (HRV), specific indices of sympathetic and parasympathetic cardiac control, respectively. Healthy, young adults (N = 120) completed questionnaires assessing depressive symptoms and hostility and underwent autonomic testing. Although main effects were not observed, a depressive symptoms × hostility interaction was detected for PEP (β = .25, p = .01). Simple slope analyses revealed that hostility was negatively related to PEP among individuals with low depressive symptoms but was not associated with PEP among those with mild-to-moderate depressive symptoms. No interaction effect was detected for high frequency HRV. Our findings suggest that depressive symptoms may moderate the link between hostility and sympathetic activation such that hostility is accompanied by sympathetic hyperactivation only when depressive symptoms are minimal.

Introduction

Accumulating evidence suggests that negative affective traits, such as depression and hostility, are independent risk factors for coronary artery disease (CAD). For instance, recent meta-analytic results indicate that persons with elevated depressive symptoms and those with increased hostility levels have a 64% and a 19% greater risk of developing CAD, respectively, even after adjustment for traditional risk factors (Chida and Steptoe, 2009, Rugulies, 2002, Wulsin and Singal, 2003). Although several candidate mediators have been proposed – including poor health behaviors, hypothalamic–pituitary–adrenal axis dysregulation, augmented systemic inflammation, and altered platelet function (Kop, 1999, Suls and Bunde, 2005) – the mechanisms underlying these associations have remained elusive. Another potential mechanism through which depression and hostility may confer an increased risk of CAD is autonomic nervous system (ANS) dysfunction (Kop, 1999, Suls and Bunde, 2005), which can manifest as heightened sympathetic nervous system (SNS) activity and/or diminished parasympathetic nervous system (PNS) activity. Sympathetic hyperactivation, parasympathetic hypoactivation, or both can have a deleterious influence on cardiovascular health, as various measures of ANS dysfunction have been found to predict CAD events and cardiovascular mortality (Curtis and O'Keefe, 2002, Dekker et al., 2000, La Rovere et al., 1998, Liao et al., 1997, Tsuji et al., 1996). It has been proposed that sympathetic (β-adrenergic) hyperactivation may contribute to the pathophysiology of CAD by promoting endothelial dysfunction (Curtis and O'Keefe, 2002), stimulating arterial inflammation (Black and Garbutt, 2002), and increasing sheer stress-mediated platelet activation (Markovitz and Matthews, 1991). Furthermore, parasympathetic hypoactivation may promote CAD by diminishing the anti-inflammatory actions of the PNS (Sajadieh et al., 2004, Sloan et al., 2007), increasing blood pressure variability (Sloan et al., 1999), and reducing the electrical stability of the heart (Huikuri, 1997).

Findings from largely separate literatures suggest that depression and hostility may be accompanied by ANS dysfunction. With respect to the SNS literature, individuals with major depressive disorder or elevated depressive symptoms display several signs of heightened sympathetic tone, including higher levels of plasma and urinary catecholamines (especially norepinephrine) (de Villiers et al., 1987, Hughes et al., 2004, Lake et al., 1982, Roy et al., 1985, Wyatt et al., 1971), increased norepinephrine release from sympathetic nerves (Veith et al., 1994), greater systemic peripheral resistance (Matthews et al., 2005), and shorter pre-ejection periods (PEPs) (Light et al., 1998). Similarly, there is evidence of an association between hostility and sympathetic hyperactivation. Hostile persons have been found to exhibit elevated levels of plasma and urinary norepinephrine (Gerra et al., 1996, Sherwood et al., 2004, Zhang et al., 2006), as well as reduced β-adrenergic receptor density and blunted responsiveness (Hughes et al., 2003, Sherwood et al., 2004, Suarez et al., 1998, Suarez et al., 1997).

Regarding the PNS literature, depression has been linked to multiple indicators of diminished parasympathetic tone. For instance, Rottenberg's recent meta-analysis of 13 cross-sectional studies (Rottenberg, 2007) revealed that an inverse relationship exists between major depression and indices of parasympathetic cardiac control, such as high frequency heart rate variability (HRV). Of note, the overall effect size fell in the small-to-medium range (d = 0.33). HRV measures reflecting parasympathetic activation appear to be attenuated among individuals with subclinical depressive symptoms as well (Bleil et al., 2008, Kim et al., 2005, Light et al., 1998). Furthermore, it has been reported that depressed patients exhibit decreased baroreflex sensitivity (Broadley et al., 2005, Davydov et al., 2007, Koschke et al., 2009), another marker of reduced parasympathetic activation (Reyes del Paso et al., 1996). Although not as extensively studied as depression, the available evidence does suggest that hostility may also be associated with parasympathetic hypoactivation, as indicated by decreased high frequency HRV (Demaree and Everhart, 2004, Sloan et al., 1994). It is worth noting, however, that the existing findings are not uniformly positive, as several investigations have not detected a relationship between either depression or hostility and markers of ANS dysfunction (Brinkmann et al., 2009, Grossman and Potter, 1999, Virtanen et al., 2003, Yeragani et al., 1991).

To date, no investigation has examined whether depression and hostility, which tend to occur together within individuals (Brummett et al., 2000, Felsten, 1996, Smith and Frohm, 1985), may act together as interacting factors to influence ANS dysfunction. The lack of research addressing this issue is somewhat surprising, given that depression × hostility interactions have been detected for another candidate mediator of the depression–CAD and hostility–CAD relationships: augmented inflammation. In two of these studies (Stewart et al., 2008, Suarez, 2003), depression and hostility exerted a synergistic effect on inflammatory markers predictive of CAD (Pearson et al., 2003, Ridker et al., 2000) such that associations of hostility with interleukin-6 and C-reactive protein were strongest among persons with elevated depressive symptoms. In a third study, however, the opposite pattern of results was observed (Miller et al., 2003). Depression had a blunting effect, as the relationships between hostility and inflammatory marker levels were weakest among those reporting severe depressive symptoms. Because evidence indicates that sympathetic activation promotes inflammation (Gornikiewicz et al., 2000, Liao et al., 1995, Mohamed-Ali et al., 2001, Path et al., 2001) and that parasympathetic activation curbs inflammation (Sajadieh et al., 2004, Sloan et al., 2007, Tracey, 2002, von Kanel et al., 2008), the three aforementioned studies provide the impetus for investigations evaluating whether there is also an interactive effect of depression and hostility on ANS dysfunction.

Accordingly, the objective of the present study was to examine whether depressive symptoms, hostility, and their interaction are associated with two noninvasive indicators of tonic ANS activation, resting PEP and resting high frequency HRV. PEP, a systolic time interval measuring the period between the onset of electromechanical systole and the onset of left ventricular ejection (Sherwood et al., 1990), is considered to be a specific index of sympathetic (β-adrenergic) cardiac control. Pharmacologic studies have demonstrated that PEP increases following SNS blockade but is generally unaffected by PNS blockade (Cacioppo et al., 1994). Consequently, lower values of PEP are indicative of heightened tonic β-adrenergic activation (Newlin and Levenson, 1979). High frequency HRV refers to periodic fluctuations in heart rate occurring in the respiratory (.15–.40 Hz) frequency band (Berntson et al., 1997). High frequency HRV is regarded as a specific index of parasympathetic cardiac control because PNS blockade markedly decreases high frequency HRV, whereas SNS blockade has only a minimal effect (Akselrod et al., 1981, Cacioppo et al., 1994, Pomeranz et al., 1985). Thus, lower high frequency HRV values reflect diminished parasympathetic activation. The conceptualization of resting PEP and high frequency HRV as indicators of tonic sympathetic and parasympathetic activation, respectively, is consistent with the approach used by Berntson et al. (2008).