AUTNEU-01762; No of Pages 5 Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Autonomic Neuroscience: Basic and Clinical journal homepage: www.elsevier.com/locate/autneu
Exploring the autonomic correlates of personality Daniel Shepherd a,⁎, Joseph Mulgrew a, Michael J. Hautus b a b
Department of Psychology, Auckland University of Technology, Auckland, New Zealand School of Psychology, University of Auckland, Auckland, New Zealand
a r t i c l e
i n f o
Article history: Received 22 April 2014 Received in revised form 12 May 2015 Accepted 16 May 2015 Available online xxxx Keywords: Heart rate variability Big Five personality model Neurovisceral integration
a b s t r a c t The aim of this study was to investigate the relationship between personality and resting heart rate variability (HRV) indices. Healthy volunteers (n = 106) completed a 240-item Big Five personality inventory, the state/ Trait Anxiety inventory, and a ten minute electrocardiographic recording. Time and frequency domain estimates of HRV were derived from the cardiac time series and related to the Big Five dimensions of personality, to personality types extracted from a cluster analysis, and to Trait Anxiety. Frequency domain measures of HRV (HRV-HF, LF/HF) were associated with specific dimensions of personality, but significance was not noted for the time domain measure (STD-RR). Furthermore, distressed personality types exhibited significantly greater autonomic imbalance (LF/HF) than other personality types. However, significance was not noted for the time domain measure (STD-RR). These results can be explained with reference to a contemporary model of neurovisceral integration. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Investigating the association between autonomic processes and personality is appropriate given that maladaptive autonomic responses have been linked to inflexible executive function, counterproductive coping strategies (e.g., avoidance), and high levels of arousal (Eysenck, 1967; Koelsch et al., 2012). Sympathetic over-arousal has been connected with various psychological disorders including anxiety (Melzig et al., 2009), depression (Rottenberg, 2007), attention deficit hyperactivity disorder (Lackschewitz et al., 2008), and posttraumatic stress disorder (Hauschildt et al., 2011). Thus there is an implied link between autonomic factors and personality, as classification systems categorizing psychological disorders typically describe symptoms occupying the extreme poles of personality dimensions. Thayer and colleagues (e.g., Thayer and Brosschot, 2005; Saus et al., 2012) have proposed a neurovisceral integrative model of dynamic autonomic regulation that provides a fresh way of understanding personality in terms of autonomic function. Their model consists of a variety of sparsely distributed cortical and subcortical structures (the so-called central autonomic network (CAN)) that have roles in affective, social, attentional, executive, and motivational behavior (Thayer and Lane, 2009; Riganello et al., 2012). According to this model, the amygdala is under constant inhibition via GABAergic projections from the prefrontal cortex, giving fast, flexible, and appropriate responses to both novel and familiar stimuli. During heighted stress the prefrontal cortex becomes ⁎ Corresponding author at: Department of Psychology, Auckland University of Technology, Private Bag 92006, Auckland 1142, New Zealand. Tel.: + 64 9 921 9999; fax: +64 9 921 9780. E-mail address: [email protected] (D. Shepherd).
hypoactive and the amygdala disinhibited, resulting in increased cardiac arousal by a rapid release of the heart from vagal inhibition, and then by sympathetic processes. These moment-by-moment autonomic influences on the heart can be indexed by Heart Rate Variability (HRV), with higher HRV viewed as markers of flexible responding and the ability to self-regulate (Ode et al., 2010). A study investigating the electrophysiological correlates of psychopathy in prison inmates demonstrated that effective inhibitory processes (as measured by cognitive tasks) were related to both high HRV and high levels of interpersonal skills such as superficial charm, manipulation, and pathological lying (Hansen et al., 2007). The authors argued that their findings consistently linked components of the CAN (i.e., HRV and cognitive inhibition) to personality traits. In this scheme, individuals possessing high HRV respond to his/her environment with appropriate levels of arousal, avoiding unnecessary sympathetic-mediated cycles of inflexible over-arousal (a characteristic of Neuroticism). In relation to the CAN it can be deduced that HRV will be positively related to those personality traits associated with flexible responding and broadening response alternatives to situational input (Extroversion, Openness, Agreeableness, Conscientiousness), but negatively associated to the maladaptive trait of Neuroticism, which is characterized by the inability to self-regulate. Few studies have directly examined the relationship between HRV and personality, and findings are inconsistent. Bleil et al. (2008) and Miu et al. (2009) both reported negative relationships between HRV and Trait Anxiety, the latter a component of Neuroticism. Ode et al. (2010) reported that HRV was not associated with Neuroticism, though only used a brief ten-item scale to represent the more multifactorial construct of Neuroticism. Carpeggiani et al. (2005) reported only one positive correlation between eight HRV measures and Cattell's 16
http://dx.doi.org/10.1016/j.autneu.2015.05.004 1566-0702/© 2015 Elsevier B.V. All rights reserved.
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
2
D. Shepherd et al. / Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
personality factors (insecurity and tension). Using the Tridimensional Personality Questionnaire, Huang et al. (2013) found few significant relationships between five HRV indices and 15 personality subscales, even when males and females were examined in isolation, with only a ‘harm avoidance’ subscale returning a significant negative correlation. Additionally, they found a moderate negative relationship between depression (also a facet of Neuroticism) and HRV, but only for males. Schweiger et al. (1998) related aspects of personality to a single measure of HRV and found no significant associations, while Cukić and Bates (2014) reported negative associations between baseline HRV and Openness. In this study we investigated the relationship between baseline HRV and personality. Existing studies present inconsistent findings, with personality often haphazardly measured, personality traits but not personality types investigated, small sample sizes risking Type II errors, exclusion of personality dimensions other than Neuroticism, single measures of HRV, and without theoretical guidance. The current study uses a 240 item version of the Big Five personality model, takes both personality traits and types into account, utilizes a relatively large sample, and employs both time and frequency domain measures of HRV. Specifically, it is hypothesized that higher HRV, indicative of either greater self-regulation or flexibility, will be associated with lower Neuroticism scores and higher scores across the remaining four Big-Five factors.
Pre-processing of the ECG, removal of ectopic beats, identification of QRS complexes and the determination of interbeat intervals were performed using ECGLab (Carvalho et al., 2003). Time domain and frequency domain HRV metrics, calculated using Kubios HRV (v.2) analysis software (Tarvainen and Niskanen, 2008), were selected as an indirect measure of autonomic influences on the heart. As the parasymatic system is responsible for most of the variation in the interbeat interval (Porges, 1997), greater HRV is thought to reflect greater parasympathetic dominance, and so HRV is considered a non-invasive index of vagal tone, autonomic balance, and autonomic flexibility (Rajendra et al., 2006; Task Force of the European Society of Cardiology and the North American Society of Pacing Electrophysiology, 1996). In the time domain, the standard deviation of the RR intervals (STD-RR), considered a reliable estimate of overall HRV (Berntson et al., 2008), was calculated. In the frequency domain, power spectrum density estimates were obtained for two frequency bands of interest: 0.04 to 0.15 Hz (Low Frequency: LF), and 0.15 to 0.40 Hz (High Frequency: HRV-HF). It is thought that HRV-HF is an estimate of parasympathetic influences on HRV, while LF is thought to reflect both parasympathetic and sympathetic influences (Dishman et al., 2000). The ratio LF/HF is thought to provide a meaningful index of overall autonomic balance (Montano et al., 2009).
2. Materials and methods
Personality was measured using the 240 item NEO Personality Inventory-Revised (NEO-PI-R) (Costa and McCrae, 1992), which assesses all five dimensions of the Big-Five: Neuroticism (N), Extraversion (E), Openness (O), Agreeableness (A), and Contentiousness (C). Each dimension is comprised of six facets of eight items each, which are summed to yield a total score for each dimension. Personality can also be conceptualized at the ‘type or prototype’ (nominal) level of description. Structural analyses of the Big Five dimensions have been performed to extract personality prototypes, labeled ‘Resilient’, ‘Average’, and ‘Non-desirable’ (e.g., Rammstedt et al., 2004). Resilient types are characterized by lower than average Neuroticism and higher than average scores across the other four dimensions of the Big Five, with an emphasis on Extroversion. Average types have average-to-moderate Neuroticism and mediocre scores across the other four dimensions. Non-desirable, or ‘Distressed’,1 types have higher than average Neuroticism scores, and lower than average scores on all other factors, with the emphasis on Introversion. State and Trait Anxiety was measured using the State-Trait Anxiety Inventory (STAI: Spielberger, 1983), which requires respondents to read 40 statements and respond to each by circling the number, on a four-point Likert scale, that best reflects the statements' perceived relevance, either in “this moment” (State Anxiety) or “generally” (Trait Anxiety).
2.1. Participants Data were collected from 106 postgraduate students or members of staff at the Auckland University of Technology's Faculty of Health. A total of 39 male and 67 female non-smoking participants provided data, with a mean age of 34.74 (±13.34) years. Participation was voluntary, and a shopping voucher was given to each participant on completion of the research. Participants were all non-smokers and reported excellent health. The university ethics committee approved all measures and procedures in accordance with the Declaration of Helsinki for human studies. 2.2. Procedure Participants were instructed not to consume caffeine nor engage in strenuous physical activity in the two hours prior to arriving at the laboratory. Upon presentation at the laboratory on testing day, participants were briefed on the nature of the research and, if willing to continue, were asked to sign an informed consent form. The research began with the administration of a number of psychometric scales, all completed in isolation. During the recording of electrocardiograms (ECG) the participants were seated in a comfortable arm-chair located in a quiet laboratory, and were asked to remain as still as possible for the duration of the recording. The ECG was recorded for 10 min during conditions of uncontrolled respiration. 2.3. Physiological measurements Cardiac signals were measured continuously using a 24-bit Nexus 10 (v.2) unit configured to sample at 2048 Hz, twice the recommended rate (Task Force of the European Society of Cardiology and the North American Society of Pacing Electrophysiology, 1996). Following thorough cleaning of the target areas with isopropyl alcohol skin cleansing swabs, three disposable silver–silver chloride (Ag/AgCl) electrodes filled with Redux electrolyte were positioned in a triangular chest configuration (i.e., standard Lead II placement). All recordings were undertaken between 11:00 AM and 2:00 PM. No corrections for respiratory factors were implemented as baseline ECGs collected at rest are not vulnerable to respiratory artifacts (Grossman and Kollai, 1993).
2.4. Psychometric scales
2.5. Statistical analyses Scale reliability was assessed using Cronbach's alpha (αc). For the trait analyses, the association between continuous personality measures (NEO-PI(R) and Trait Anxiety) and baseline HRV (logarithmically transformed) indices were investigated using partial correlation coefficients controlling for age and mean heart rate, both of which have been shown to be related to measures of HRV and are considered confounding variables (Kupari et al., 1993). High collinearity was observed between the two STAI scales (r = 0.854, p b 0.001). Trait Anxiety is a more temporally stable construct relative to State Anxiety, and in this way is more similar to the Big Five dimensions. For this reason, Trait 1 The term ‘non-desirable’ is inappropriate, as it is overtly judgmental. Better would be to incorporate Denollet's (2005) Type-D personality with the prototype names suggested by Rammstedt et al (2004), which would result in: Resilient, Distressed, and Average Personalities.
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
D. Shepherd et al. / Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
Anxiety was chosen as the anxiety measure to retain in analyses, while the inclusion of mean heart rate in the analyses would be expected to act as a more valid control of state anxiety than pen-and-paper scales. Following Hansen et al. (2007), these correlations were further explored with forward stepwise regression using each of the three HRV measures as dependent variables and the Big Five dimensions and Trait Anxiety as predictors, again while controlling for age and mean heart rate. The predictor variables were dichotomized into high and low, based on a median split. Preliminarily analyses showed that gender had no effect on HRV indices, in line with previous studies (e.g., Miu et al., 2009; Huang et al., 2013), so this variable was excluded from the correlation and regression analyses, but retained in further analyses as a confirmatory check. For the personality prototype analysis, group membership was determined using the two-step hierarchical cluster analysis method described by Asendorpf et al. (2001). Subsequently, a 2 (female vs. male) by 3 (Resilient vs. Distressed vs. Average) design was employed to test the null hypothesis of no difference in HRV indices across the three prototype groups. In these three analyses (STD-RR, HRV-HF, and LF/HF), age and mean heart rate were included in the model as covariates. 3. Results 3.1. Trait analysis Cronbach's alpha was calculated for each dimension of the Big Five, with satisfactory coefficients being obtained for Neuroticism (αc = 0.94), Extroversion (αc = 0.90), Openness (αc = 0.85), Agreeableness (αc = 0.88), and Conscientiousness (αc = 0.92). For the Trait Anxiety scale, Cronbach's alpha was 0.93. Table 1 presents partial correlation coefficients, controlling for age (left columns) or both age and mean heart rate (right columns), between the Big Five personality dimensions, Trait Anxiety, and the three HRV measures. For the single time-domain metric, the standard deviation of the RR intervals (STD-RR), no significant correlations are noted, with the strongest being Openness (r = − 0.147, p = 0.140). For the two frequency domain measures, HRV-HF and LF/HF, significant associations were found with Neuroticism and Openness. Additionally, Trait Anxiety correlated negatively with HRV-HF and positively with LF/HF. In line with the correlation result, regression analyses failed to uncover significant personality predictors of STD-RR. For HRV-HF, the Neuroticism dimension was the only personality predictor to be selected, predicting a negative relationship with HRV-HF (β = −0.231, p = 0.019, R2 = 0.096). For the measure of autonomic balance, LF/ HF, an equivalent finding was obtained, with β = 0.239 (p = 0.016, R2 = 0.12). 3.2. Prototype analysis To assign participants to discrete personality groups a two-stage cluster analysis was performed using the Big Five personality data. First, a hierarchical cluster analysis produced a dendrogram confirming Table 1 Partial correlation coefficients, controlling for age (left) and both age and mean heart rate (right), between dimensions of the Big Five, Trait Anxiety (TA), and autonomic measures. Partial correlations (age)
N E O A C TA
Partial correlations (age, heart rate)
STD-RR
HRV-HF
LF/HF
STD-RR
HRV-HF
LF/HF
−.033 −.069 .06 .093 −.099 .015
−.231* .134 −.235* 0.047 .077 −.244*
.231* −.113 .229* −.156 −.123 .170*
0.006 −0.049 −0.147 0.032 −0.104 0. 024
−.221* 0.147 −0.216* 0.023 0.082 −0.226*
0.220* −0.128 0.206* −0.130 −0.133 0.172*
Note: *p b 0.05 (two-tailed).
3
that a three-group solution was the best fit for the data. Second, a Kmeans cluster analysis, specifying a three-group solution, was performed to assign each participant to one of three groups. With reference to the literature (e.g., Rammstedt et al., 2004), the labels given to each grouping were Resilient (n = 31), Distressed (n = 27), and Average (n = 46) personality prototypes (see Fig. 1), with no gender bias noted in regards group membership (χ2 (2, N = 104) = 0.028, p = 0.986). An effect of age was observed, (F(2, 103) = 609.3, p = 0.031), and this variable was included in subsequent analyses as a covariate. The groups appear to be consistent with those reported by (Rammstedt et al., 2004; Asendorpf et al., 2001), especially with respect to the contrast between N and E scores observed in the Resistant and the Distressed prototypes. Fig. 1 presents mean z-scores for the Big Five dimensions and Trait Anxiety across the three personality profiles. A factorial ANCOVA failed to show a main effect of Prototype with STD-RR as the dependent variable (F(2104) = 1.531, p = 0.222). For HRV-HF, however, a main effect of Prototype emerged (F(2104) = 4.213, p = 0.018) but not Gender (F(1104) = 2.714, p = 0.103), nor the interaction term Prototype × Gender (F(2104) = 4.04, p = 0.669). Subsequent post hoc testing revealed that the Distressed group had significantly lower mean HRV-HF than the Average group (t(71) = -2.549, p = 0.007). No other pairwise comparisons reached significance. For LF/HF, there was again a main effect of Prototype (F(2104) = 4.145, p = 0.019) but not Gender (F(1104) = 1.818, p = 0.181), nor the interaction term (F(2104) = 0.033, p = 0.967). Post hoc testing revealed that the Distressed group had significantly lower means than the Average (t(71) = − 2.997, p b 0.001) and Resilient (t(71) = − 3.021, p b 0.004) groups.
4. Discussion This study investigated the relationship between resting autonomic state and personality. We report significant associations between highfrequency HRV and three aspects of personality: Neuroticism, Trait Anxiety and Openness, providing further evidence for a link between HRV and personality (Hansen et al., 2007). A link between psychological traits, including personality and autonomic processes, has been consistently noted in the medical literature, notably for anxiety and cardiovascular disease (Goodwin et al., 2009). Our findings are of interest as it is thought that personality may moderate emotional stress, which in turn interferes with the autonomic processes implicated in ischemia, myocardial infarction, and elevated blood pressure (Carpeggiani et al., 2005). Considering personality at the trait level of description, significant associations between frequency domain, but not time domain, representations of HRV were noted. For both Neuroticism and Trait Anxiety, significant negative coefficients were noted with HRV-HF, as-well-as positive correlations with LF/HF. The Trait Anxiety finding is consistent with others (Bleil et al., 2008; Miu et al., 2009), and expected if higher baseline HRV is associated with enhanced cognitive control and selfregulation and lower baseline HRV is associated with impaired selfregulation or emotional dysregulation (Ode et al., 2010), the latter being hallmarks of anxiety (Clark, 2005). The Neuroticism finding is inconsistent with Ode et al. (2010), who develop considered hypotheses stating that a negative correlation between HRV and Neuroticism would be expected if HRV indicates self-regulation, and no association if it indicates flexible responding. Thus while they support an explanation of HRV based on flexible responding, our data supports the selfregulation approach. The differences between the two studies are difficult to reconcile, given that both use an equivalent measure of high Frequency HRV and had sufficient sample sizes (n N 80). Where the two studies differ is the measurement of Neuroticism, with Ode et al. (2010) using Goldberg's (1999) ten-item inventory in contrast to our use of the 48-item NEO-PI-R scale. Thus the greater refinement in the
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
4
D. Shepherd et al. / Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
1.5 Distressed Average Resilient
1.0
Mean z score
0.5
0.0
-0.5
-1.0
-1.5 N
E
O
A
C
TA
Personality Dimension Fig. 1. Mean z-scores for each of the Big-Five dimensions and Trait Anxiety across the three personality prototypes: Distressed, Average, and Resilient.
current study's measure of Neuroticism may have served to remove noise from the analyses and be less exposed to a Type-II error. The NEO-PI-R Neuroticism dimension itself contains a facet representing anxiety, and so the relationship between itself and HRV may potentially reflect the contribution of the anxiety facet. The regression analyses suggested that, for both HRV-HF and LF/HF, the Neuroticism dimension was the most influential of the Big Five and Trait Anxiety. That the association between Trait Anxiety and HRV was not upheld in the regression analyses may be due to the anxiety facet of the Neuroticism scale accounting for variability that would otherwise be absorbed by the Trait Anxiety measure. To test for this, we repeated the regression analysis, except with the anxiety facet removed from the Neuroticism scale. The reduced Neuroticism scale was still the only significant predictor of HRV, so this possibility does not account for the lack of significance for Trait Anxiety. In addition, it suggests that the relationship between Neuroticism and HRV is not mediated by anxiety exclusively. The significant negative correlation between Openness and HRV-HF contradicts the findings of Cukić and Bates (2014), who reported a positive association between Openness and low-frequency HRV, albeit erroneously operationalizing low frequency HRV as a biomarker of sympathetic activation. McCrae and Costa (1997) conceptualize Openness as a need to expand and assess experience, and that individual differences in Openness reflect differences in neural circuits underlying reward and motivation. Thus Openness may reflect aroused (i.e., sympathetic) states related to autonomic processes promoting prolonged attention to stimuli (Cukić and Bates, 2014), though interestingly is not related to situational awareness (Saus et al., 2012). Ode et al. (2010) expended substantial effort towards differentiating selfregulation from flexible responding, and to explaining their respective contributions to HRV. It maybe, however, that self-regulation is a prerequisite of flexible responding, mediated by personality domains such as Openness. As such, adaptive responding and goal attainment may entail more complex relationships between the two autonomic branches, and future research examining reciprocal, coactive, or independent covariability between the two is needed. At the prototype level of personality description it was noted that the Distressed group had significantly lower autonomic balance (i.e., LF/HF) than either the Resilient and Average groups, and a lower mean HRV-HF score than the Average group. In regards the latter, Martin et al. (2010) reported a link between Type-D personality types, the conceptual equivalent of our Distressed Group, and lower HRV-HF,
though only while imagining a stressful event. The Distressed group in this study also scored significantly higher on the Neuroticism dimension than the Resilient and Average groups. Neuroticism manifests a vulnerability to stress and over-reactivity, and neurotic individuals are more likely to appraise environments as threatening (Saus et al., 2012). Such characteristics suggest rigidity in the face of environmental challenges, a propensity to cope using avoidance, and would be expected to covary negatively with HRV measures representing adaptive functioning via flexible executive functions. Type-D personality is a known risk factor of post-coronary mortality (O'Dell et al., 2011), though the relationship between this personality type and cardiac disease is poorly understood (Martin et al., 2010), and here we confirm that HRV may be useful approach in elucidating the relationship between the two. These results can be reflected on neurovisceral integration models of autonomic balance, and HRV. Dynamic autonomic balance, resulting from appropriate suppression of the amygdala by the prefrontal cortex, would be expected to correlate with personality (Hansen et al., 2007). First, because HRV has been argued to represent self-regulation (Thayer and Lane, 2000), defined as the process of filtering out taskirrelevant or task-impeding cognitions, then traits like Openness and creativity that are associated with more imaginative and lateral thinking styles may impair self-regulation. This is supported by our data, where a negative correlation between Openness and HRV-HF was observed. Note that a flexible responding approach (Ode et al., 2010) would predict the opposite, and so we report indirect evidence that HRV at least in part represents self-regulation. Second, tonic parasympathetic dominance results in a greater range of arousal, especially at the lower end of the spectrum (e.g., HR below 100 BPM), which would result in the prefrontal cortex remaining ‘online’ through a greater range of arousal, thus preserving executive functions including those relating to personality (e.g., impulse control, goal-directed behavior, and reasoning). However, a negative correlation between Extroversion and either HRV and HRV-HF was not found. Last, parasympathetic dominance, reflected in high HRV, would reflect the inhibition of amygdala processes, decreasing negative bias and a focus on negative stimuli that are associated with the anxiety, depression, and negative affect factors common to Neuroticism. The current study had several limitations that may have influenced the results. First, statistically speaking, the sample size was modest due to the time demands typically experienced when collecting physiological data. However, the current study had a sample size of over 100 participants, which compares favorably to other similar studies. A
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
D. Shepherd et al. / Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
further influence is that the sample was drawn from a university population, which in turn impacts external validity. Had a clinical population been sampled, then bigger effect sizes might be expected. A challenge in interpreting the findings arose with a lack of significance in the time domain. While we have no explanation for this we can only speculate that while the time domain measures reflect both branches of the ANS, the frequency domain indices manage to disentangle them to a greater (e.g., HRV-HF) or lesser (e.g., LF/HF) degree. Finally, while the results were able to highlight several interesting relationships, causality could not be assessed. However, the findings of the current study were related to the NIS, and a strong argument can be made to suggest that some relationships observed were causal. In summary, the current study investigated the link between autonomic measures and personality, with predictions guided by the NIS. Results suggest that HRV is related to personality at both the trait and prototype level. The results, however, also provided only partial support for the propositions of the NIS, and further research is still required to fully understand the relationship between the ANS and personality, especially with respect to the personality and time domain measures of HRV. Findings in this area are of importance, as further research may identify cardiac metrics of individuals at risk for cardiovascular disease, and conversely, cardiac measures can likely be refined and used to assess personality types (Ode et al., 2010). Acknowledgments This research was funded by grants from the Auckland University of Technology (CGH 21/09) and the University of Auckland (Faculty Research Development Fund: 3624433/9853). References Asendorpf, J.B., Borkenau, P., Ostendorf, F., Van Aken, M.A.G., 2001. Carving personality description at its joints: confirmation of three replicable personality prototypes for both children and adults. Eur. J. Personal. 15 (3), 169–198. Berntson, G.G., Norman, G.J., Hawkley, L.C., Cacioppo, J.T., 2008. Cardiac autonomic balance versus cardiac regulatory capacity. Psychophysiology 45 (4), 643–652. Bleil, M.E., Gianaros, P.J., Jennings, J.R., Flory, J.D., Manuck, S.B., 2008. Trait negative affect: toward an integrated model of understanding psychological risk for impairment in cardiac autonomic function. Psychosom. Med. 70 (3), 328–337. Carpeggiani, C., Emdin, M., Bonaguidi, F., Landi, P., Michelassi, C., Trivella, M.G., Macerata, A., L'Abbate, A., 2005. Personality traits and heart rate variability predict longterm cardiac mortality after myocardial infarction. Eur. Heart J. 26, 1612–1617. Carvalho, J.L.A., Rocha, A.F., Junqueira, L.F., Souza Neto Jr., J., Santos, I., Nascimento, F.A.O., 2003. A tool for time-frequency analysis of heart rate variability. 25th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 2574–2577. Clark, D.A., 2005. Intrusive Thoughts in Clinical Disorders: Theory, Research, and Treatment. Guilford Press, New York. Costa, P.T., McCrae, R.R., 1992. Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI). Psychological Assessment Resources, Odessa, FL. Cukić, I., Bates, T.C., 2014. Openness to experience and aesthetic chills: links to heart rate sympathetic activity. Personal. Individ. Differ. 64, 152–156. Denollet, J., 2005. DS14: standard assessment of negative affectivity, social inhibition, and type D personality. Psychosom. Med. 67 (1), 89–97. Dishman, R.K., Nakamura, Y., Garcia, M.E., Thompson, R.W., Dunn, A.L., Blair, S.N., 2000. Heart rate variability, trait anxiety, and perceived stress among physically fit men and women. Int. J. Psychophysiol. 37 (2), 121–133. Eysenck, H.J., 1967. The Biological Basis of Personality. Thomas, Springfield, Il. Goldberg, L.R., 1999. A broad-bandwidth, public domain, personality inventory measuring the lower-level facets of several five-factor models. In: Mervielde, I., Deary, I., De
5
Fruyt, F., Ostendorf, F. (Eds.), Personality psychology in Europe vol. 7. Tilburg University Press, Tilburg, The Netherlands, pp. 7–28. Goodwin, R.D., Davidson, K.W., Keyes, K., 2009. Mental disorders and cardiovascular disease among adults in the United States. J. Psychiatr. Res. 43 (3), 239–246. Grossman, P., kollai, M., 1993. Respiratory sinus arrhythmia, cardiac vagal tone, and respiration: within- and between-individual relations. Psychophysiology 30 (5), 486–495. Hansen, A.L., Johnsen, B.H., Thornton, D., Waage, L., Thayer, J.F., 2007. Facets of psychopathy. Heart rate variability and cognitive function. J. Personal. Disord. 21 (5), 568–582. Huang, W., Chang, L., Kuo, B.T.B.J., Lin, Y., Chen, Y., Yang, C.C.H., 2013. Gender differences in personality and heart-rate variability. Psychiatry Res. 209 (3), 652–657. Hauschildt, M., Peters, M.J.V., Moritz, S., Jelinek, L., 2011. Heart rate variability in response to affective scenes in posttraumatic stress disorder. Biol. Psychol. 88 (2–3), 215–222. Koelsch, S., Enge, J., Jentschke, S., 2012. Cardiac signatures of personality. Plos One 72 (2), e31441. Kupari, M., Virolainen, J., Koskinen, P., Tikkane, M.J., 1993. Short term heart rate variability and factors modifying the risk of coronary artery disease in a population sample. Am. J. Cardiol. 72, 897–903. Lackschewitz, H., Hüther, G., Kröner-Herwig, B., 2008. Physiological and psychological stress responses in adults with attention-deficit/hyperactivity disorder (ADHD). Psychoneuroendocrinology 33 (5), 612–624. Martin, L.A., Doster, J.A., Critelli, J.W., Lambert, P.L., Purdum, M., Powers, C., Prazak, M., 2010. Ethnicity and Type D personality as predictors of heart rate variability. Int. J. Psychophysiol. 76, 118–121. McCrae, R.R., Costa, P.T. Jr, 1997. Conceptions and correlates of openness to experience. In: Hogan, R., Johnson, J.A., et al. (Eds.), Handbook of personality psychology. Academic Press Inc., San Diego, CA, USA, pp. 825–847. Melzig, C.A., Weike, A.I., Hamm, A.O., Thayer, J.F., 2009. Individual differences in fearpotentiated startle as a function of resting heart rate variability: implications for panic disorder. Int. J. Psychophysiol. 71 (2), 109–117. Miu, A.C., Heilman, R.M., Miclea, M., 2009. Reduced heart rate variability and vagal tone in anxiety: trait versus state, and the effects of autogenic training. Auton. Neurosci. 145, 99–103. Montano, N., Porta, A., Cogliati, C., Costantino, G., Tobaldini, E., Casali, K.R., Iellamo, F., 2009. Heart rate variability explored in the frequency domain: a tool to investigate the link between heart and behavior. Neurosci. Biobehav. Rev. 33 (2), 71–80. O'Dell, K.R., Masters, K.S., Spielmans, G.I., Maisto, S.A., 2011. Does type-D personality predict outcomes among patients with cardiovascular disease? A meta-analytic review. J. Psychosom. Res. 71 (4), 199–206. Ode, S., Hilmert, C.J., Zielke, D.J., Robinson, M.D., 2010. Neuroticism's importance in understanding the daily life correlates of heart rate variability. Emotion 10 (4), 536–543. Porges, S.W., 1997. Emotion: an evolutionary by-product of the neural regulation of the autonomic nervous system. Ann. N. Y. Acad. Sci. 807, 62–77. Rajendra, A.U., Paul, J.K., Kannathal, N., Lim, C.M., Suri, J.S., 2006. Heart rate variability: a review. Med. Biol. Eng. Comput. 44 (12), 1031–1051. Rammstedt, B., Riemann, R., Angleitner, A., Borkenau, P., 2004. Resilients, overcontrollers, and undercontrollers: the replicability of the three personality prototypes across informants. Eur. J. Personal. 18 (1), 1–14. Riganello, F., Garbarino, S., Sannita, W.G., 2012. Heart rate variability, homeostasis, and brain function. J. Psychophysiol. 26 (4), 178–203. Rottenberg, J., 2007. Cardiac vagal control in depression: a critical analysis. Biol. Psychol. 74 (2), 200–211. Saus, E.-R., Johnsen, B.H., Eid, J., Thayer, J.F., 2012. Who benefits from simulator training: personality and heart rate variability in relation to situation awareness during navigation training. Comput. Hum. Behav. 28 (4), 1262. Schweiger, E., Wittling, W., Genzel, S., Block, A., 1998. Relationship between sympathovagal tone and personality traits. Personal. Individ. Differ. 25, 327–337. Spielberger, C.D., 1983. State-Trait Anxiety Inventory (Form Y) Manual. Mindgarden, Redwood City, CA. Tarvainen, M.P., Niskanen, J.P., 2008. Kubios HRV Version 2.0 Users Guide. Biosignal Analysis and Medical Imaging Group (BSAMIG). Department of Physics, University of Kuopio, Finland. Task Force of the European Society of Cardiology, The North American Society of Pacing Electrophysiology, 1996. Heart rate variability: standards of measurement, physiological interpretation, and clinical use. Circulation 93 (5), 1043–1065. Thayer, J.F., Brosschot, J.F., 2005. Psychosomatics and psychopathology: looking up and down from the brain. Psychoneuroendocrinology 30 (10), 1050–1058. Thayer, J.F., Lane, R.D., 2000. A model of neurovisceral integration in emotion regulation and dysregulation. J. Affect. Disord. 61 (3), 201–216. Thayer, J.F., Lane, R.D., 2009. Claude Bernard and the heart–brain connection: further elaboration of a model of neurovisceral integration. Neurosci. Biobehav. Rev. 33 (2), 81–88.
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
Contents lists available at ScienceDirect
Autonomic Neuroscience: Basic and Clinical journal homepage: www.elsevier.com/locate/autneu
Exploring the autonomic correlates of personality Daniel Shepherd a,⁎, Joseph Mulgrew a, Michael J. Hautus b a b
Department of Psychology, Auckland University of Technology, Auckland, New Zealand School of Psychology, University of Auckland, Auckland, New Zealand
a r t i c l e
i n f o
Article history: Received 22 April 2014 Received in revised form 12 May 2015 Accepted 16 May 2015 Available online xxxx Keywords: Heart rate variability Big Five personality model Neurovisceral integration
a b s t r a c t The aim of this study was to investigate the relationship between personality and resting heart rate variability (HRV) indices. Healthy volunteers (n = 106) completed a 240-item Big Five personality inventory, the state/ Trait Anxiety inventory, and a ten minute electrocardiographic recording. Time and frequency domain estimates of HRV were derived from the cardiac time series and related to the Big Five dimensions of personality, to personality types extracted from a cluster analysis, and to Trait Anxiety. Frequency domain measures of HRV (HRV-HF, LF/HF) were associated with specific dimensions of personality, but significance was not noted for the time domain measure (STD-RR). Furthermore, distressed personality types exhibited significantly greater autonomic imbalance (LF/HF) than other personality types. However, significance was not noted for the time domain measure (STD-RR). These results can be explained with reference to a contemporary model of neurovisceral integration. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Investigating the association between autonomic processes and personality is appropriate given that maladaptive autonomic responses have been linked to inflexible executive function, counterproductive coping strategies (e.g., avoidance), and high levels of arousal (Eysenck, 1967; Koelsch et al., 2012). Sympathetic over-arousal has been connected with various psychological disorders including anxiety (Melzig et al., 2009), depression (Rottenberg, 2007), attention deficit hyperactivity disorder (Lackschewitz et al., 2008), and posttraumatic stress disorder (Hauschildt et al., 2011). Thus there is an implied link between autonomic factors and personality, as classification systems categorizing psychological disorders typically describe symptoms occupying the extreme poles of personality dimensions. Thayer and colleagues (e.g., Thayer and Brosschot, 2005; Saus et al., 2012) have proposed a neurovisceral integrative model of dynamic autonomic regulation that provides a fresh way of understanding personality in terms of autonomic function. Their model consists of a variety of sparsely distributed cortical and subcortical structures (the so-called central autonomic network (CAN)) that have roles in affective, social, attentional, executive, and motivational behavior (Thayer and Lane, 2009; Riganello et al., 2012). According to this model, the amygdala is under constant inhibition via GABAergic projections from the prefrontal cortex, giving fast, flexible, and appropriate responses to both novel and familiar stimuli. During heighted stress the prefrontal cortex becomes ⁎ Corresponding author at: Department of Psychology, Auckland University of Technology, Private Bag 92006, Auckland 1142, New Zealand. Tel.: + 64 9 921 9999; fax: +64 9 921 9780. E-mail address: [email protected] (D. Shepherd).
hypoactive and the amygdala disinhibited, resulting in increased cardiac arousal by a rapid release of the heart from vagal inhibition, and then by sympathetic processes. These moment-by-moment autonomic influences on the heart can be indexed by Heart Rate Variability (HRV), with higher HRV viewed as markers of flexible responding and the ability to self-regulate (Ode et al., 2010). A study investigating the electrophysiological correlates of psychopathy in prison inmates demonstrated that effective inhibitory processes (as measured by cognitive tasks) were related to both high HRV and high levels of interpersonal skills such as superficial charm, manipulation, and pathological lying (Hansen et al., 2007). The authors argued that their findings consistently linked components of the CAN (i.e., HRV and cognitive inhibition) to personality traits. In this scheme, individuals possessing high HRV respond to his/her environment with appropriate levels of arousal, avoiding unnecessary sympathetic-mediated cycles of inflexible over-arousal (a characteristic of Neuroticism). In relation to the CAN it can be deduced that HRV will be positively related to those personality traits associated with flexible responding and broadening response alternatives to situational input (Extroversion, Openness, Agreeableness, Conscientiousness), but negatively associated to the maladaptive trait of Neuroticism, which is characterized by the inability to self-regulate. Few studies have directly examined the relationship between HRV and personality, and findings are inconsistent. Bleil et al. (2008) and Miu et al. (2009) both reported negative relationships between HRV and Trait Anxiety, the latter a component of Neuroticism. Ode et al. (2010) reported that HRV was not associated with Neuroticism, though only used a brief ten-item scale to represent the more multifactorial construct of Neuroticism. Carpeggiani et al. (2005) reported only one positive correlation between eight HRV measures and Cattell's 16
http://dx.doi.org/10.1016/j.autneu.2015.05.004 1566-0702/© 2015 Elsevier B.V. All rights reserved.
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
2
D. Shepherd et al. / Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
personality factors (insecurity and tension). Using the Tridimensional Personality Questionnaire, Huang et al. (2013) found few significant relationships between five HRV indices and 15 personality subscales, even when males and females were examined in isolation, with only a ‘harm avoidance’ subscale returning a significant negative correlation. Additionally, they found a moderate negative relationship between depression (also a facet of Neuroticism) and HRV, but only for males. Schweiger et al. (1998) related aspects of personality to a single measure of HRV and found no significant associations, while Cukić and Bates (2014) reported negative associations between baseline HRV and Openness. In this study we investigated the relationship between baseline HRV and personality. Existing studies present inconsistent findings, with personality often haphazardly measured, personality traits but not personality types investigated, small sample sizes risking Type II errors, exclusion of personality dimensions other than Neuroticism, single measures of HRV, and without theoretical guidance. The current study uses a 240 item version of the Big Five personality model, takes both personality traits and types into account, utilizes a relatively large sample, and employs both time and frequency domain measures of HRV. Specifically, it is hypothesized that higher HRV, indicative of either greater self-regulation or flexibility, will be associated with lower Neuroticism scores and higher scores across the remaining four Big-Five factors.
Pre-processing of the ECG, removal of ectopic beats, identification of QRS complexes and the determination of interbeat intervals were performed using ECGLab (Carvalho et al., 2003). Time domain and frequency domain HRV metrics, calculated using Kubios HRV (v.2) analysis software (Tarvainen and Niskanen, 2008), were selected as an indirect measure of autonomic influences on the heart. As the parasymatic system is responsible for most of the variation in the interbeat interval (Porges, 1997), greater HRV is thought to reflect greater parasympathetic dominance, and so HRV is considered a non-invasive index of vagal tone, autonomic balance, and autonomic flexibility (Rajendra et al., 2006; Task Force of the European Society of Cardiology and the North American Society of Pacing Electrophysiology, 1996). In the time domain, the standard deviation of the RR intervals (STD-RR), considered a reliable estimate of overall HRV (Berntson et al., 2008), was calculated. In the frequency domain, power spectrum density estimates were obtained for two frequency bands of interest: 0.04 to 0.15 Hz (Low Frequency: LF), and 0.15 to 0.40 Hz (High Frequency: HRV-HF). It is thought that HRV-HF is an estimate of parasympathetic influences on HRV, while LF is thought to reflect both parasympathetic and sympathetic influences (Dishman et al., 2000). The ratio LF/HF is thought to provide a meaningful index of overall autonomic balance (Montano et al., 2009).
2. Materials and methods
Personality was measured using the 240 item NEO Personality Inventory-Revised (NEO-PI-R) (Costa and McCrae, 1992), which assesses all five dimensions of the Big-Five: Neuroticism (N), Extraversion (E), Openness (O), Agreeableness (A), and Contentiousness (C). Each dimension is comprised of six facets of eight items each, which are summed to yield a total score for each dimension. Personality can also be conceptualized at the ‘type or prototype’ (nominal) level of description. Structural analyses of the Big Five dimensions have been performed to extract personality prototypes, labeled ‘Resilient’, ‘Average’, and ‘Non-desirable’ (e.g., Rammstedt et al., 2004). Resilient types are characterized by lower than average Neuroticism and higher than average scores across the other four dimensions of the Big Five, with an emphasis on Extroversion. Average types have average-to-moderate Neuroticism and mediocre scores across the other four dimensions. Non-desirable, or ‘Distressed’,1 types have higher than average Neuroticism scores, and lower than average scores on all other factors, with the emphasis on Introversion. State and Trait Anxiety was measured using the State-Trait Anxiety Inventory (STAI: Spielberger, 1983), which requires respondents to read 40 statements and respond to each by circling the number, on a four-point Likert scale, that best reflects the statements' perceived relevance, either in “this moment” (State Anxiety) or “generally” (Trait Anxiety).
2.1. Participants Data were collected from 106 postgraduate students or members of staff at the Auckland University of Technology's Faculty of Health. A total of 39 male and 67 female non-smoking participants provided data, with a mean age of 34.74 (±13.34) years. Participation was voluntary, and a shopping voucher was given to each participant on completion of the research. Participants were all non-smokers and reported excellent health. The university ethics committee approved all measures and procedures in accordance with the Declaration of Helsinki for human studies. 2.2. Procedure Participants were instructed not to consume caffeine nor engage in strenuous physical activity in the two hours prior to arriving at the laboratory. Upon presentation at the laboratory on testing day, participants were briefed on the nature of the research and, if willing to continue, were asked to sign an informed consent form. The research began with the administration of a number of psychometric scales, all completed in isolation. During the recording of electrocardiograms (ECG) the participants were seated in a comfortable arm-chair located in a quiet laboratory, and were asked to remain as still as possible for the duration of the recording. The ECG was recorded for 10 min during conditions of uncontrolled respiration. 2.3. Physiological measurements Cardiac signals were measured continuously using a 24-bit Nexus 10 (v.2) unit configured to sample at 2048 Hz, twice the recommended rate (Task Force of the European Society of Cardiology and the North American Society of Pacing Electrophysiology, 1996). Following thorough cleaning of the target areas with isopropyl alcohol skin cleansing swabs, three disposable silver–silver chloride (Ag/AgCl) electrodes filled with Redux electrolyte were positioned in a triangular chest configuration (i.e., standard Lead II placement). All recordings were undertaken between 11:00 AM and 2:00 PM. No corrections for respiratory factors were implemented as baseline ECGs collected at rest are not vulnerable to respiratory artifacts (Grossman and Kollai, 1993).
2.4. Psychometric scales
2.5. Statistical analyses Scale reliability was assessed using Cronbach's alpha (αc). For the trait analyses, the association between continuous personality measures (NEO-PI(R) and Trait Anxiety) and baseline HRV (logarithmically transformed) indices were investigated using partial correlation coefficients controlling for age and mean heart rate, both of which have been shown to be related to measures of HRV and are considered confounding variables (Kupari et al., 1993). High collinearity was observed between the two STAI scales (r = 0.854, p b 0.001). Trait Anxiety is a more temporally stable construct relative to State Anxiety, and in this way is more similar to the Big Five dimensions. For this reason, Trait 1 The term ‘non-desirable’ is inappropriate, as it is overtly judgmental. Better would be to incorporate Denollet's (2005) Type-D personality with the prototype names suggested by Rammstedt et al (2004), which would result in: Resilient, Distressed, and Average Personalities.
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
D. Shepherd et al. / Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
Anxiety was chosen as the anxiety measure to retain in analyses, while the inclusion of mean heart rate in the analyses would be expected to act as a more valid control of state anxiety than pen-and-paper scales. Following Hansen et al. (2007), these correlations were further explored with forward stepwise regression using each of the three HRV measures as dependent variables and the Big Five dimensions and Trait Anxiety as predictors, again while controlling for age and mean heart rate. The predictor variables were dichotomized into high and low, based on a median split. Preliminarily analyses showed that gender had no effect on HRV indices, in line with previous studies (e.g., Miu et al., 2009; Huang et al., 2013), so this variable was excluded from the correlation and regression analyses, but retained in further analyses as a confirmatory check. For the personality prototype analysis, group membership was determined using the two-step hierarchical cluster analysis method described by Asendorpf et al. (2001). Subsequently, a 2 (female vs. male) by 3 (Resilient vs. Distressed vs. Average) design was employed to test the null hypothesis of no difference in HRV indices across the three prototype groups. In these three analyses (STD-RR, HRV-HF, and LF/HF), age and mean heart rate were included in the model as covariates. 3. Results 3.1. Trait analysis Cronbach's alpha was calculated for each dimension of the Big Five, with satisfactory coefficients being obtained for Neuroticism (αc = 0.94), Extroversion (αc = 0.90), Openness (αc = 0.85), Agreeableness (αc = 0.88), and Conscientiousness (αc = 0.92). For the Trait Anxiety scale, Cronbach's alpha was 0.93. Table 1 presents partial correlation coefficients, controlling for age (left columns) or both age and mean heart rate (right columns), between the Big Five personality dimensions, Trait Anxiety, and the three HRV measures. For the single time-domain metric, the standard deviation of the RR intervals (STD-RR), no significant correlations are noted, with the strongest being Openness (r = − 0.147, p = 0.140). For the two frequency domain measures, HRV-HF and LF/HF, significant associations were found with Neuroticism and Openness. Additionally, Trait Anxiety correlated negatively with HRV-HF and positively with LF/HF. In line with the correlation result, regression analyses failed to uncover significant personality predictors of STD-RR. For HRV-HF, the Neuroticism dimension was the only personality predictor to be selected, predicting a negative relationship with HRV-HF (β = −0.231, p = 0.019, R2 = 0.096). For the measure of autonomic balance, LF/ HF, an equivalent finding was obtained, with β = 0.239 (p = 0.016, R2 = 0.12). 3.2. Prototype analysis To assign participants to discrete personality groups a two-stage cluster analysis was performed using the Big Five personality data. First, a hierarchical cluster analysis produced a dendrogram confirming Table 1 Partial correlation coefficients, controlling for age (left) and both age and mean heart rate (right), between dimensions of the Big Five, Trait Anxiety (TA), and autonomic measures. Partial correlations (age)
N E O A C TA
Partial correlations (age, heart rate)
STD-RR
HRV-HF
LF/HF
STD-RR
HRV-HF
LF/HF
−.033 −.069 .06 .093 −.099 .015
−.231* .134 −.235* 0.047 .077 −.244*
.231* −.113 .229* −.156 −.123 .170*
0.006 −0.049 −0.147 0.032 −0.104 0. 024
−.221* 0.147 −0.216* 0.023 0.082 −0.226*
0.220* −0.128 0.206* −0.130 −0.133 0.172*
Note: *p b 0.05 (two-tailed).
3
that a three-group solution was the best fit for the data. Second, a Kmeans cluster analysis, specifying a three-group solution, was performed to assign each participant to one of three groups. With reference to the literature (e.g., Rammstedt et al., 2004), the labels given to each grouping were Resilient (n = 31), Distressed (n = 27), and Average (n = 46) personality prototypes (see Fig. 1), with no gender bias noted in regards group membership (χ2 (2, N = 104) = 0.028, p = 0.986). An effect of age was observed, (F(2, 103) = 609.3, p = 0.031), and this variable was included in subsequent analyses as a covariate. The groups appear to be consistent with those reported by (Rammstedt et al., 2004; Asendorpf et al., 2001), especially with respect to the contrast between N and E scores observed in the Resistant and the Distressed prototypes. Fig. 1 presents mean z-scores for the Big Five dimensions and Trait Anxiety across the three personality profiles. A factorial ANCOVA failed to show a main effect of Prototype with STD-RR as the dependent variable (F(2104) = 1.531, p = 0.222). For HRV-HF, however, a main effect of Prototype emerged (F(2104) = 4.213, p = 0.018) but not Gender (F(1104) = 2.714, p = 0.103), nor the interaction term Prototype × Gender (F(2104) = 4.04, p = 0.669). Subsequent post hoc testing revealed that the Distressed group had significantly lower mean HRV-HF than the Average group (t(71) = -2.549, p = 0.007). No other pairwise comparisons reached significance. For LF/HF, there was again a main effect of Prototype (F(2104) = 4.145, p = 0.019) but not Gender (F(1104) = 1.818, p = 0.181), nor the interaction term (F(2104) = 0.033, p = 0.967). Post hoc testing revealed that the Distressed group had significantly lower means than the Average (t(71) = − 2.997, p b 0.001) and Resilient (t(71) = − 3.021, p b 0.004) groups.
4. Discussion This study investigated the relationship between resting autonomic state and personality. We report significant associations between highfrequency HRV and three aspects of personality: Neuroticism, Trait Anxiety and Openness, providing further evidence for a link between HRV and personality (Hansen et al., 2007). A link between psychological traits, including personality and autonomic processes, has been consistently noted in the medical literature, notably for anxiety and cardiovascular disease (Goodwin et al., 2009). Our findings are of interest as it is thought that personality may moderate emotional stress, which in turn interferes with the autonomic processes implicated in ischemia, myocardial infarction, and elevated blood pressure (Carpeggiani et al., 2005). Considering personality at the trait level of description, significant associations between frequency domain, but not time domain, representations of HRV were noted. For both Neuroticism and Trait Anxiety, significant negative coefficients were noted with HRV-HF, as-well-as positive correlations with LF/HF. The Trait Anxiety finding is consistent with others (Bleil et al., 2008; Miu et al., 2009), and expected if higher baseline HRV is associated with enhanced cognitive control and selfregulation and lower baseline HRV is associated with impaired selfregulation or emotional dysregulation (Ode et al., 2010), the latter being hallmarks of anxiety (Clark, 2005). The Neuroticism finding is inconsistent with Ode et al. (2010), who develop considered hypotheses stating that a negative correlation between HRV and Neuroticism would be expected if HRV indicates self-regulation, and no association if it indicates flexible responding. Thus while they support an explanation of HRV based on flexible responding, our data supports the selfregulation approach. The differences between the two studies are difficult to reconcile, given that both use an equivalent measure of high Frequency HRV and had sufficient sample sizes (n N 80). Where the two studies differ is the measurement of Neuroticism, with Ode et al. (2010) using Goldberg's (1999) ten-item inventory in contrast to our use of the 48-item NEO-PI-R scale. Thus the greater refinement in the
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
4
D. Shepherd et al. / Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
1.5 Distressed Average Resilient
1.0
Mean z score
0.5
0.0
-0.5
-1.0
-1.5 N
E
O
A
C
TA
Personality Dimension Fig. 1. Mean z-scores for each of the Big-Five dimensions and Trait Anxiety across the three personality prototypes: Distressed, Average, and Resilient.
current study's measure of Neuroticism may have served to remove noise from the analyses and be less exposed to a Type-II error. The NEO-PI-R Neuroticism dimension itself contains a facet representing anxiety, and so the relationship between itself and HRV may potentially reflect the contribution of the anxiety facet. The regression analyses suggested that, for both HRV-HF and LF/HF, the Neuroticism dimension was the most influential of the Big Five and Trait Anxiety. That the association between Trait Anxiety and HRV was not upheld in the regression analyses may be due to the anxiety facet of the Neuroticism scale accounting for variability that would otherwise be absorbed by the Trait Anxiety measure. To test for this, we repeated the regression analysis, except with the anxiety facet removed from the Neuroticism scale. The reduced Neuroticism scale was still the only significant predictor of HRV, so this possibility does not account for the lack of significance for Trait Anxiety. In addition, it suggests that the relationship between Neuroticism and HRV is not mediated by anxiety exclusively. The significant negative correlation between Openness and HRV-HF contradicts the findings of Cukić and Bates (2014), who reported a positive association between Openness and low-frequency HRV, albeit erroneously operationalizing low frequency HRV as a biomarker of sympathetic activation. McCrae and Costa (1997) conceptualize Openness as a need to expand and assess experience, and that individual differences in Openness reflect differences in neural circuits underlying reward and motivation. Thus Openness may reflect aroused (i.e., sympathetic) states related to autonomic processes promoting prolonged attention to stimuli (Cukić and Bates, 2014), though interestingly is not related to situational awareness (Saus et al., 2012). Ode et al. (2010) expended substantial effort towards differentiating selfregulation from flexible responding, and to explaining their respective contributions to HRV. It maybe, however, that self-regulation is a prerequisite of flexible responding, mediated by personality domains such as Openness. As such, adaptive responding and goal attainment may entail more complex relationships between the two autonomic branches, and future research examining reciprocal, coactive, or independent covariability between the two is needed. At the prototype level of personality description it was noted that the Distressed group had significantly lower autonomic balance (i.e., LF/HF) than either the Resilient and Average groups, and a lower mean HRV-HF score than the Average group. In regards the latter, Martin et al. (2010) reported a link between Type-D personality types, the conceptual equivalent of our Distressed Group, and lower HRV-HF,
though only while imagining a stressful event. The Distressed group in this study also scored significantly higher on the Neuroticism dimension than the Resilient and Average groups. Neuroticism manifests a vulnerability to stress and over-reactivity, and neurotic individuals are more likely to appraise environments as threatening (Saus et al., 2012). Such characteristics suggest rigidity in the face of environmental challenges, a propensity to cope using avoidance, and would be expected to covary negatively with HRV measures representing adaptive functioning via flexible executive functions. Type-D personality is a known risk factor of post-coronary mortality (O'Dell et al., 2011), though the relationship between this personality type and cardiac disease is poorly understood (Martin et al., 2010), and here we confirm that HRV may be useful approach in elucidating the relationship between the two. These results can be reflected on neurovisceral integration models of autonomic balance, and HRV. Dynamic autonomic balance, resulting from appropriate suppression of the amygdala by the prefrontal cortex, would be expected to correlate with personality (Hansen et al., 2007). First, because HRV has been argued to represent self-regulation (Thayer and Lane, 2000), defined as the process of filtering out taskirrelevant or task-impeding cognitions, then traits like Openness and creativity that are associated with more imaginative and lateral thinking styles may impair self-regulation. This is supported by our data, where a negative correlation between Openness and HRV-HF was observed. Note that a flexible responding approach (Ode et al., 2010) would predict the opposite, and so we report indirect evidence that HRV at least in part represents self-regulation. Second, tonic parasympathetic dominance results in a greater range of arousal, especially at the lower end of the spectrum (e.g., HR below 100 BPM), which would result in the prefrontal cortex remaining ‘online’ through a greater range of arousal, thus preserving executive functions including those relating to personality (e.g., impulse control, goal-directed behavior, and reasoning). However, a negative correlation between Extroversion and either HRV and HRV-HF was not found. Last, parasympathetic dominance, reflected in high HRV, would reflect the inhibition of amygdala processes, decreasing negative bias and a focus on negative stimuli that are associated with the anxiety, depression, and negative affect factors common to Neuroticism. The current study had several limitations that may have influenced the results. First, statistically speaking, the sample size was modest due to the time demands typically experienced when collecting physiological data. However, the current study had a sample size of over 100 participants, which compares favorably to other similar studies. A
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
D. Shepherd et al. / Autonomic Neuroscience: Basic and Clinical xxx (2015) xxx–xxx
further influence is that the sample was drawn from a university population, which in turn impacts external validity. Had a clinical population been sampled, then bigger effect sizes might be expected. A challenge in interpreting the findings arose with a lack of significance in the time domain. While we have no explanation for this we can only speculate that while the time domain measures reflect both branches of the ANS, the frequency domain indices manage to disentangle them to a greater (e.g., HRV-HF) or lesser (e.g., LF/HF) degree. Finally, while the results were able to highlight several interesting relationships, causality could not be assessed. However, the findings of the current study were related to the NIS, and a strong argument can be made to suggest that some relationships observed were causal. In summary, the current study investigated the link between autonomic measures and personality, with predictions guided by the NIS. Results suggest that HRV is related to personality at both the trait and prototype level. The results, however, also provided only partial support for the propositions of the NIS, and further research is still required to fully understand the relationship between the ANS and personality, especially with respect to the personality and time domain measures of HRV. Findings in this area are of importance, as further research may identify cardiac metrics of individuals at risk for cardiovascular disease, and conversely, cardiac measures can likely be refined and used to assess personality types (Ode et al., 2010). Acknowledgments This research was funded by grants from the Auckland University of Technology (CGH 21/09) and the University of Auckland (Faculty Research Development Fund: 3624433/9853). References Asendorpf, J.B., Borkenau, P., Ostendorf, F., Van Aken, M.A.G., 2001. Carving personality description at its joints: confirmation of three replicable personality prototypes for both children and adults. Eur. J. Personal. 15 (3), 169–198. Berntson, G.G., Norman, G.J., Hawkley, L.C., Cacioppo, J.T., 2008. Cardiac autonomic balance versus cardiac regulatory capacity. Psychophysiology 45 (4), 643–652. Bleil, M.E., Gianaros, P.J., Jennings, J.R., Flory, J.D., Manuck, S.B., 2008. Trait negative affect: toward an integrated model of understanding psychological risk for impairment in cardiac autonomic function. Psychosom. Med. 70 (3), 328–337. Carpeggiani, C., Emdin, M., Bonaguidi, F., Landi, P., Michelassi, C., Trivella, M.G., Macerata, A., L'Abbate, A., 2005. Personality traits and heart rate variability predict longterm cardiac mortality after myocardial infarction. Eur. Heart J. 26, 1612–1617. Carvalho, J.L.A., Rocha, A.F., Junqueira, L.F., Souza Neto Jr., J., Santos, I., Nascimento, F.A.O., 2003. A tool for time-frequency analysis of heart rate variability. 25th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 2574–2577. Clark, D.A., 2005. Intrusive Thoughts in Clinical Disorders: Theory, Research, and Treatment. Guilford Press, New York. Costa, P.T., McCrae, R.R., 1992. Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI). Psychological Assessment Resources, Odessa, FL. Cukić, I., Bates, T.C., 2014. Openness to experience and aesthetic chills: links to heart rate sympathetic activity. Personal. Individ. Differ. 64, 152–156. Denollet, J., 2005. DS14: standard assessment of negative affectivity, social inhibition, and type D personality. Psychosom. Med. 67 (1), 89–97. Dishman, R.K., Nakamura, Y., Garcia, M.E., Thompson, R.W., Dunn, A.L., Blair, S.N., 2000. Heart rate variability, trait anxiety, and perceived stress among physically fit men and women. Int. J. Psychophysiol. 37 (2), 121–133. Eysenck, H.J., 1967. The Biological Basis of Personality. Thomas, Springfield, Il. Goldberg, L.R., 1999. A broad-bandwidth, public domain, personality inventory measuring the lower-level facets of several five-factor models. In: Mervielde, I., Deary, I., De
5
Fruyt, F., Ostendorf, F. (Eds.), Personality psychology in Europe vol. 7. Tilburg University Press, Tilburg, The Netherlands, pp. 7–28. Goodwin, R.D., Davidson, K.W., Keyes, K., 2009. Mental disorders and cardiovascular disease among adults in the United States. J. Psychiatr. Res. 43 (3), 239–246. Grossman, P., kollai, M., 1993. Respiratory sinus arrhythmia, cardiac vagal tone, and respiration: within- and between-individual relations. Psychophysiology 30 (5), 486–495. Hansen, A.L., Johnsen, B.H., Thornton, D., Waage, L., Thayer, J.F., 2007. Facets of psychopathy. Heart rate variability and cognitive function. J. Personal. Disord. 21 (5), 568–582. Huang, W., Chang, L., Kuo, B.T.B.J., Lin, Y., Chen, Y., Yang, C.C.H., 2013. Gender differences in personality and heart-rate variability. Psychiatry Res. 209 (3), 652–657. Hauschildt, M., Peters, M.J.V., Moritz, S., Jelinek, L., 2011. Heart rate variability in response to affective scenes in posttraumatic stress disorder. Biol. Psychol. 88 (2–3), 215–222. Koelsch, S., Enge, J., Jentschke, S., 2012. Cardiac signatures of personality. Plos One 72 (2), e31441. Kupari, M., Virolainen, J., Koskinen, P., Tikkane, M.J., 1993. Short term heart rate variability and factors modifying the risk of coronary artery disease in a population sample. Am. J. Cardiol. 72, 897–903. Lackschewitz, H., Hüther, G., Kröner-Herwig, B., 2008. Physiological and psychological stress responses in adults with attention-deficit/hyperactivity disorder (ADHD). Psychoneuroendocrinology 33 (5), 612–624. Martin, L.A., Doster, J.A., Critelli, J.W., Lambert, P.L., Purdum, M., Powers, C., Prazak, M., 2010. Ethnicity and Type D personality as predictors of heart rate variability. Int. J. Psychophysiol. 76, 118–121. McCrae, R.R., Costa, P.T. Jr, 1997. Conceptions and correlates of openness to experience. In: Hogan, R., Johnson, J.A., et al. (Eds.), Handbook of personality psychology. Academic Press Inc., San Diego, CA, USA, pp. 825–847. Melzig, C.A., Weike, A.I., Hamm, A.O., Thayer, J.F., 2009. Individual differences in fearpotentiated startle as a function of resting heart rate variability: implications for panic disorder. Int. J. Psychophysiol. 71 (2), 109–117. Miu, A.C., Heilman, R.M., Miclea, M., 2009. Reduced heart rate variability and vagal tone in anxiety: trait versus state, and the effects of autogenic training. Auton. Neurosci. 145, 99–103. Montano, N., Porta, A., Cogliati, C., Costantino, G., Tobaldini, E., Casali, K.R., Iellamo, F., 2009. Heart rate variability explored in the frequency domain: a tool to investigate the link between heart and behavior. Neurosci. Biobehav. Rev. 33 (2), 71–80. O'Dell, K.R., Masters, K.S., Spielmans, G.I., Maisto, S.A., 2011. Does type-D personality predict outcomes among patients with cardiovascular disease? A meta-analytic review. J. Psychosom. Res. 71 (4), 199–206. Ode, S., Hilmert, C.J., Zielke, D.J., Robinson, M.D., 2010. Neuroticism's importance in understanding the daily life correlates of heart rate variability. Emotion 10 (4), 536–543. Porges, S.W., 1997. Emotion: an evolutionary by-product of the neural regulation of the autonomic nervous system. Ann. N. Y. Acad. Sci. 807, 62–77. Rajendra, A.U., Paul, J.K., Kannathal, N., Lim, C.M., Suri, J.S., 2006. Heart rate variability: a review. Med. Biol. Eng. Comput. 44 (12), 1031–1051. Rammstedt, B., Riemann, R., Angleitner, A., Borkenau, P., 2004. Resilients, overcontrollers, and undercontrollers: the replicability of the three personality prototypes across informants. Eur. J. Personal. 18 (1), 1–14. Riganello, F., Garbarino, S., Sannita, W.G., 2012. Heart rate variability, homeostasis, and brain function. J. Psychophysiol. 26 (4), 178–203. Rottenberg, J., 2007. Cardiac vagal control in depression: a critical analysis. Biol. Psychol. 74 (2), 200–211. Saus, E.-R., Johnsen, B.H., Eid, J., Thayer, J.F., 2012. Who benefits from simulator training: personality and heart rate variability in relation to situation awareness during navigation training. Comput. Hum. Behav. 28 (4), 1262. Schweiger, E., Wittling, W., Genzel, S., Block, A., 1998. Relationship between sympathovagal tone and personality traits. Personal. Individ. Differ. 25, 327–337. Spielberger, C.D., 1983. State-Trait Anxiety Inventory (Form Y) Manual. Mindgarden, Redwood City, CA. Tarvainen, M.P., Niskanen, J.P., 2008. Kubios HRV Version 2.0 Users Guide. Biosignal Analysis and Medical Imaging Group (BSAMIG). Department of Physics, University of Kuopio, Finland. Task Force of the European Society of Cardiology, The North American Society of Pacing Electrophysiology, 1996. Heart rate variability: standards of measurement, physiological interpretation, and clinical use. Circulation 93 (5), 1043–1065. Thayer, J.F., Brosschot, J.F., 2005. Psychosomatics and psychopathology: looking up and down from the brain. Psychoneuroendocrinology 30 (10), 1050–1058. Thayer, J.F., Lane, R.D., 2000. A model of neurovisceral integration in emotion regulation and dysregulation. J. Affect. Disord. 61 (3), 201–216. Thayer, J.F., Lane, R.D., 2009. Claude Bernard and the heart–brain connection: further elaboration of a model of neurovisceral integration. Neurosci. Biobehav. Rev. 33 (2), 81–88.
Please cite this article as: Shepherd, D., et al., Exploring the autonomic correlates of personality, Auton. Neurosci. (2015), http://dx.doi.org/ 10.1016/j.autneu.2015.05.004
