• ulcerative colitis;
  • social support;
  • autonomic nervous system;
  • heart-rate variability


  1. Top of page
  2. Abstract


The relationship of psychological stress to relapse in ulcerative colitis (UC) is inconsistent. This may be due to a failure to identify patient characteristics, such as social support, which moderate the transduction of stress from the central nervous system to the immune system. In this study we tested the hypothesis that social support enhances parasympathetic modulation of heart rate in UC.


An indirect measure of autonomic function (heart rate variability; HRV) was measured in 108 patients with UC in remission during a standard protocol involving periods of stress, paced breathing, and relaxation. Social support was measured with the Social Support Questionnaire.


After controlling for age, which is strongly related to HRV, both satisfaction with social support (F = 5.7, significance = 0.002) and its interaction with age (F = 7.8, significance <0.001) were associated with high-frequency HRV, which measures parasympathetic modulation of heart rate. Social support was associated with higher levels of high-frequency HRV at almost all points in the stress protocol. Neither age nor social support was associated with differences in the LF/HF ratio, which measures sympathetic modulation of heart rate.


Social support is related to parasympathetic activity in UC. Given previous evidence of an antiinflammatory role for the parasympathetic nervous system, this suggests that autonomic function could serve as a mediating link between social support and reduced inflammatory activity.

Prospective studies of psychological stress as a provocative factor in ulcerative colitis (UC) suggest a relationship that is significant but of inconsistent strength.1–6 Studies that mix UC and Crohn's disease patients have often found no relationship at all between stress and disease outcomes,7–9 suggesting that specificity is important. The inconsistency of these findings and the modest effect sizes may be due to a failure to identify patient characteristics which moderate the physiological transduction of stress. This report concerns one such characteristic, social support, and its relationship to a plausible mediator of the impact of psychological stress on inflammation, autonomic nervous system (ANS) activity.

The acute physiological stress response occurs primarily in the hypothalamic–pituitary–adrenal axis and the ANS. Both of these systems have a role in modulating immune function, and thus may serve to mediate links between the experience of stress and immune function. In the autonomic system, a proinflammatory role for sympathetic activation and an antiinflammatory role for parasympathetic (vagal) activity10–14 are consistent with experimental models of colitis.15–18 Measuring heart rate variability (HRV) is a noninvasive method of estimating the activity of the parasympathetic and sympathetic systems.19 Compared to healthy persons, HRV indices reflecting parasympathetic function are reduced in inflammatory bowel disease (IBD) patients in remission compared to healthy controls,20, 21 which suggests that abnormal autonomic regulation could have a role in the pathophysiology of UC. Furthermore, an anomalous HRV response to a standardized stress (a pattern consisting of a slow HRV response to acute stress followed by slow recovery to baseline) at one point in time predicted UC disease activity at a follow-up visit 7–37 months later.22

Since there is substantial individual variation in both the impact of psychological stress on the course of UC, and in autonomic responses to standardized stress among UC patients,22 individual characteristics that moderate autonomic activity are likely to be relevant. Social support is known to moderate stress reactivity in general.23, 24 With respect to HRV in particular, in non-IBD populations social support moderates or buffers the relationship between psychological variables (stress or distress) and HRV25–27 and variables that are related to social support (social isolation, marital status, attachment style) are also related to HRV in the expected direction.28–31 These studies are inconsistent as to whether social support is primarily related to changes in HRV in response to stress or to tonic differences in HRV (i.e., irrespective of fluctuations related to stress and recovery).

While individual differences in social support could account for inconsistencies in stress reactivity among UC patients, social support has typically has been assessed with respect to its influence on coping and quality of life32–34 rather than on physiology. Two possibilities require investigation. The first is that social support may moderate the severity of acute changes in autonomic reactivity and recovery following exposure to stress, and that differences in this dynamic response predict disease activity over subsequent months.22 The second possibility is that social support is related to individual differences in tonic autonomic function. The purpose of this study was to test the hypothesis that social support enhances parasympathetic modulation of heart rate as measured by spectral analysis of HRV during a controlled sequence of stress events and recovery.


  1. Top of page
  2. Abstract

Patient Selection

In all, 361 patients with UC confirmed by endoscopy or biopsy, age >18, were identified by chart review and contacted by telephone. Exclusion criteria were colectomy and factors that interfere with interpretation of HRV: diagnosis of valvular heart disease or cardiomyopathy, congestive heart failure of NYHA class IV, unstable angina, grade 2–3 atrioventricular block, revascularization procedures, atrial flutter, atrial fibrillation, frequent ventricular or atrial ectopy, or sick sinus syndrome. Cardiac exclusion criteria were determined by subject's report and cross-checked against current reported prescription drugs. Forty-one percent (149 patients) consented and participated in a study that has been previously reported.35 All participants were examined endoscopically by a gastroenterologist with the observed mucosa rated as no inflammation, friable, or bleeding spontaneously. The current study excluded the 41 patients in the latter two endoscopic categories and thus reports on 108 patients with UC in remission.

Stress Protocol and HRV

Spectral analysis of HRV was assessed during successive 5-minute tasks, each of which was recorded in two 2.5-minute epochs. The stress protocol consisted of eight tasks: 1) baseline, 2) breathing paced at 12 per minute, 3) paced breathing plus self-directed relaxation, 4) recall of a stressful event, 5) recovery from stress event recall, and 6) the stress-inducing Paced Auditory Serial Addition Task (PASAT) procedure,36, 37 7) recovery from the PASAT, and 8) final recovery with paced breathing and self-directed relaxation. In each case the recording from the second 2.5-minute interval was used for analysis. The stress event recall task directed subjects to identify a salient stressful situation that occurred within the past month. Each event was rated on a Likert-type scale that ranged from 0 (not at all stressful) to 10 (extremely stressful). The stressful event was selected if it was rated by the subject as >7.5. Otherwise, subjects were instructed to identify an event meeting this criterion that occurred more than 1 month previously. Brief instruction was then provided to direct subjects to speak about their stressful event using descriptive language that focused on subjectively stressful aspects of their event. Respiration rate was not less than 12 per minute throughout each task. During each task of the stress protocol subjective stress was measured on a visual analog scale from −10 (relaxed) to +10 (stressed).

Lead II ECG signals were obtained through a Spacelabs 413 monitor. Analog-to-digital conversion was performed at 500 samples per second. Raw ECG data were analyzed offline using an M-file written in MatLab (MathWorks, Natick, MA). The time series of R–R intervals was edited prior to analysis in order to identify deviations in interbeat intervals (IBI) greater than 30%, which are attributable to ectopic beats, movement artifact, or signal noise. Power spectral computations were performed on successive intervals of 256 sample points obtained over 128-second epochs. HRV was assessed according to recommended standards for short-term recordings of successive 5-minute epochs which mirror each segment of the stress reactivity assessment protocol.

HRV indices are reported in absolute units (ms2/Hz). The spectral power of high frequency (HF, 0.15–0.40 Hz) and low frequency (LF, 0.04–0.15 Hz) components of HRV was calculated with the fast Fourier transform. The HF spectral bandwidth is a reliable marker of vagal modulation of R–R variability in a cohort of subjects without heart disease.38 The LF spectral bandwidth is a marker of a more complex set of regulatory factors including sympathetic and parasympathetic contributions. The LF/HF ratio was also calculated as an index of sympathetic function (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology).19

Social Support

Social support was measured with the six-item Social Support Questionnaire (SSQ-6).39, 40 This questionnaire measures primarily emotional (as opposed to instrumental) support as the respondent reports whom s/he can “really count on” to 1) be dependable when you need help, 2) help you feel more relaxed when you are under pressure or tense, 3) care about you, regardless of what is happening to you, 4) help you feel better when you are feeling generally down in the dumps, 5) console you when you are very upset, and 6) who accepts you totally, including both your worst and your best points. For each item, this scale measures both the number of people providing support and satisfaction with the support received. The distinction is important because previous research suggests that different aspects of social support differentially affect health outcomes. Specifically, whereas size of social network may directly protect health,41 satisfaction with social support may buffer the effects of stress.42

Statistical Analysis

The relationship between satisfaction with social support and heart rate variability was tested by repeated measures analysis of variance, including age as a covariate. This analysis included measures in each of the eight tasks of the stress protocol, and was repeated for log HF-HRV, log LF-HRV, and LF/HF. The direction of significant relationships and nature of interactions was explored graphically post-hoc by grouping participants into four groups by age (≤40, >40 years) and satisfaction with social support (median split). The analysis was repeated substituting number of supportive persons for satisfaction with social support.

In order to optimize comparability with the previously reported relationship between typical and atypical stress reactivity (defined as an increase or decrease in HF-HRV from stress recall to recovery respectively) and prognosis in UC, we performed a secondary analysis in which we grouped participants into typical and atypical HRV response groups and tested the difference between groups in age and satisfaction with social support using Student's t-test. Statistical analysis was performed using SPSS 17.0 for Windows (Chicago, IL) using two-tailed tests with P < 0.05 considered significant.

Ethical Considerations

This study was approved by the Research Ethics Boards of Mount Sinai Hospital and the University Health Network, Toronto, Canada.


  1. Top of page
  2. Abstract

Of the 108 UC patients in remission who were studied, 66 (61%) were men. Mean age was 43.4 ± standard deviation 3.0 years. Seventy-six participants (70%) were married or living common-law. Although 64 participants (59%) had required corticosteroid treatment at some point in their illness, only eight (7%) were taking corticosteroids at the time of the study. Overall, satisfaction with social support was high. The median score was 33 (possible scores range from 6 to 36; the actual range in this sample was 12–36). The median number of persons providing support was 3 (range 1–9).

As detailed in Table 1, there was a main effect of age on both low- and high-frequency HRV and their ratio over the course of the stress protocol. In addition to the effect of age, social support (satisfaction) also contributed significantly to HF-HRV. An interaction of age and social support also contributed significantly to LF-HRV, HF-HRV, and their ratio. Post-hoc testing shows the pattern of these relationships (Fig. 1). Figure 1 indicates that social support and youth are each associated with higher levels of HF HRV and LF-HRV at almost all points in the stress and relaxation protocol. Neither age nor social support appeared to be associated with consistent differences in the LF/HF ratio. When these analyses were repeated, testing the number of supportive persons reported rather than satisfaction, there was no significant contribution of support to low-frequency HRV, high-frequency HRV, or LF/HF.

Table 1. Between-Subjects Effects of Age and Satisfaction with Social Support on Heart Rate Variability Indices During an 8-Stage Stress Protocol
 Satisfaction with SupportAgeInteraction (Age*Support)
 F (df = 14)Sig.F (df = 41)Sig.F (df = 29)Sig.
  1. LF-HRV, low-frequency heart rate variability; HF-HRV, high-frequency heart rate variability.

Log LF-HRV2.
Log HF-HRV5.70.0026.40.0017.8<0.001
thumbnail image

Figure 1. Comparison of HRV indices during an 8-stage stress protocol by age and satisfaction with social support (estimated marginal means of HRV indices). For illustration, satisfaction with social support is dichotomized with a median split and age is dichotomized with a cutoff of 40 years. Sixty-four subjects (59.3%) were 40 years old or older.

Download figure to PowerPoint

With respect to reactivity of HRV indices to stress and recovery tasks, we defined two groups: persons with a typical pattern (an increase in HF-HRV from stress event recall to recovery 5 minutes later) and persons with an atypical pattern (failure of HF-HRV to increase during this transition). There was no difference in social support reported by those in the typical (mean social support satisfaction 32.3 ± standard deviation 4.4) versus the atypical reactivity groups (32.2 ± 3.7, P = 0.9). Nor was there a difference between typical and atypical groups in age (44.9 ± 13.2 years vs. 42.3 ± 13.1 years, P = 0.3).


  1. Top of page
  2. Abstract

It has previously been reported that the impact of stressful life events on health outcomes is less among people who report a high level of satisfaction with social support,43 which suggests that social support moderates the stress response by acting as a buffer. The current study indicates a potential mechanism for this effect in UC, in that perceived social support was associated with HF-HRV and, by extension, with enhanced vagal tone. The finding that there was no main effect of social support on LF/HF ratio, a more specific index of sympathetic function than LF-HRV, supports the interpretation that social support is specifically linked to parasympathetic function rather than the overall spectral power of HRV.

There were further complexities in the findings of this study which require discussion and cautious interpretation. First, there is a distinction between the relationship of social support to tonic levels of HF-HRV as opposed to dynamic reactivity of HF-HRV to stress and recovery. The finding that higher social support is associated with higher levels of HF-HRV irrespective of stressful tasks and effort to relax is compatible with a previous finding in healthy persons that attachment avoidance (a relatively stable trait, which is associated with interpersonal distance and lower social support) is associated with lower HF-HRV at all points of a stress protocol.44 This suggests that greater satisfaction with social support may be associated with consistently higher vagal tone—and not just with buffering the autonomic response to stress. Given the generally antiinflammatory neuroimmune effects of parasympathetic function13 and the evidence from animal models for vagal inhibition of the inflammation of UC,18 this suggests a possible physiological mechanism for a protective role for social support in IBD.

On the other hand, social support was not associated with typical or atypical reactivity of HRV to stress. In a previous study, an atypical response of HF-HRV to stress (i.e., failure of HF-HRV to increase during recovery from stress event recall) was associated with an overall “sluggish” pattern of HRV response—slow to change in response to stress and slow to recover afterwards.22 Over a period of several months following the stress protocol, the atypical pattern was associated with a reduced likelihood of active disease over several months. The current findings suggest that while social support is associated with greater vagal tone in general, this relationship is distinct from reactivity to stress as defined in the typical/atypical pattern distinction. It is noteworthy that the cohorts participating in these two studies overlap but they were not identical. The previous study included participants with active disease, who were excluded from the current study, and excluded subjects for whom a second assessment time was not available, who were included in the current study. Thus, the current finding cannot be easily reconciled with the previously reported relationships between atypical stress response and disease course. It is not known if the differences in HRV that are associated with social support in the current study are related to subsequent disease activity and, unfortunately, the available sample size of subjects with relevant data at a second timepoint is too small to analyze. Further research is required.

The result that social support has a direct relationship with vagal tone is consistent with Porges' “polyvagal theory,” which hypothesizes a bidirectional link between parasympathetic function and social interaction.45 It is thought that social competence (and therefore the capacity to develop and maintain supportive relationships) and the capacity for effective parasympathetic regulation are linked to one another because they have common developmental antecedents and common neurobiological substrates. In particular, optimal development of the neurological substrates of both autonomic regulation and social competence in close, supportive relationships occurs in the context of responsive face-to-face interactions between infants and their primary caregivers.46 Furthermore, in adults Porges proposed that vagal inhibition of arousal “enables the individual to rapidly engage and disengage with objects and other individuals and to promote self-soothing behaviors and calm behavioral states”45 (p. 509), which suggests an interaction by which autonomic function and social interaction might continue to be mutually reinforcing throughout life and not only during early development. The idea that the seemingly unrelated constructs of autonomic regulation and effective social interaction are linked is bolstered by evidence that insecure patterns of attachment are associated with reduced high-frequency HRV in healthy adults.29

The primary limitation on conclusions from the current study is that it is cross-sectional and cannot provide evidence for causal relationships. Further limitations are that this is a secondary analysis of data that was collected in the context of a separate investigation35 and that the sample reported here represents only 30% of the subjects contacted for participation in that study, which may introduce sampling bias. With respect to the excluded subjects, psychological indices for persons with active disease are expected to differ from those in remission (disease typically being associated with greater distress and higher support), which is one reason for excluding these individuals. More problematically, it is not known how persons in remission who declined to participate in the study differ, if at all, from the group who were studied. Because of these limitations, this analysis must be considered to provide initial support for a novel hypothesis which warrants further research.

The link between vagal tone and social support which was hypothesized and supported in this study is not expected to be specific to UC and is broadly consistent with studies of social support (and related indices) and HRV in non-IBD populations.23–31 Nonetheless, a common psychophysiological relationship may have particular importance in an inflammatory disease, because of the evidence that parasympathetic function downregulates inflammation.13, 18, 47 Preliminary observations from our group indicate that fluctuations in HRV markers of parasympathetic function over time occur in concert with inverse fluctuations in an inflammatory marker, C-reactive protein, measured at the same intervals, which further supports the hypothesis. Thus, the association reported here introduces the possibility that social support may contribute to the maintenance of remission of UC and other inflammatory conditions through its influence on autonomic regulation of immune mediators.


  1. Top of page
  2. Abstract
  • 1
    Maunder RG. Evidence that stress contributes to inflammatory bowel disease: evaluation, synthesis, and future directions. Inflamm Bowel Dis. 2005; 11: 600608.
  • 2
    Maunder RG, Levenstein S. The role of stress in the development and clinical course of inflammatory bowel disease: epidemiological evidence. Curr Mol Med. 2008; 8: 247252.
  • 3
    Rampton D. Does stress influence inflammatory bowel disease? The clinical data. Dig Dis. 2009; 27( suppl 1): 7679.
  • 4
    Bernstein CN, Singh S, Graff LA, et al. A prospective population-based study of triggers of symptomatic flares in IBD. Am J Gastroenterol. 2010; 105: 19942002.
  • 5
    Levenstein S, Prantera C, Varvo V, et al. Stress and exacerbation in ulcerative colitis: a prospective study of patients enrolled in remission. Am J Gastroenterol. 2000; 95: 12131220.
    Direct Link:
  • 6
    Bitton A, Sewitch MJ, Peppercorn MA, et al. Psychosocial determinants of relapse in ulcerative colitis: a longitudinal study. Am J Gastroenterol. 2003; 98: 22032208.
    Direct Link:
  • 7
    Vidal A, Gomez-Gil E, Sans M, et al. Life events and inflammatory bowel disease relapse: a prospective study of patients enrolled in remission. Am J Gastroenterol. 2006; 101: 775781.
  • 8
    North CS, Alpers DH, Helzer JE, et al. Do life events or depression exacerbate inflammatory bowel disease? Ann Int Med. 1991; 114: 381386.
  • 9
    Riley SA, Mani V, Goodman MJ, et al. Why do patients with ulcerative colitis relapse? Gut. 1989; 31: 179183.
  • 10
    Elenkov IJ, Wilder RL, Chrousos GP, et al. The sympathetic nerve—an integrative interface between two supersystems: the brain and the immune system. Pharmacol Rev. 2000; 52: 595638.
  • 11
    Matsunaga K, Klein TW, Friedman H, et al. Involvement of nicotinic acetylcholine receptors in suppression of antimicrobial activity and cytokine responses of alveolar macrophages to Legionella pneumophila infection by nicotine. J Immunol. 2001; 167: 65186524.
  • 12
    Bernik TR, Friedman SG, Ochani M, et al. Pharmacological stimulation of the cholinergic antiinflammatory pathway. J Exp Med. 2002; 195: 781788.
  • 13
    Tracey KJ. The inflammatory reflex. Nature. 2002; 420: 853859.
  • 14
    Borovikova LV, Ivanova S, Zhang M, et al. Vagus nerve stimulation attenuates the systemic inflammatory response to endotoxin. Nature. 2000; 405: 458462.
  • 15
    McCafferty DM, Wallace JL, Sharkey KA. Effects of chemical sympathectomy and sensory nerve ablation on experimental colitis in the rat. Am J Physiol. 1997; 272( 2 Pt 1): G272G280.
  • 16
    Furlan R, Ardizzone S, Palazzolo L, et al. Sympathetic overactivity in active ulcerative colitis: effects of clonidine. Am J Physiol Regul Integr Comp Physiol. 2006; 290: R224R232.
  • 17
    Eliakim R, Karmeli F, Rachmilewitz D, et al. Effect of chronic nicotine administration on trinitrobenzene sulphonic acid-induced colitis. Eur J Gastroenterol Hepatol. 1998; 10: 10131019.
  • 18
    Ghia JE, Blennerhassett P, Collins SM. Impaired parasympathetic function increases susceptibility to inflammatory bowel disease in a mouse model of depression. J Clin Invest. 2008; 118: 22092218.
  • 19
    Heart rate variability: standards of measurementphysiological interpretation and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Circulation. 1996; 93: 10431065.
  • 20
    Sharma P, Makharia GK, Ahuja V, et al. Autonomic dysfunctions in patients with inflammatory bowel disease in clinical remission. Dig Dis Sci. 2009; 54: 853861.
  • 21
    Coruzzi P, Castiglioni P, Parati G, et al. Autonomic cardiovascular regulation in quiescent ulcerative colitis and Crohn's disease. Eur J Clin Invest. 2007; 37: 964970.
  • 22
    Maunder RG, Greenberg GR, Nolan RP, et al. Autonomic response to standardized stress predicts subsequent disease activity in ulcerative colitis. Eur J Gastroenterol Hepatol. 2006; 18: 413420.
  • 23
    Ellis BJ, Essex MJ, Boyce WT. Biological sensitivity to context: II. Empirical explorations of an evolutionary-developmental theory. Dev Psychopathol. 2005; 17: 303328.
  • 24
    Uchino BN, Cacioppo JT, Kielcolt-Glaser JK. The relationship between social support and physiological processes: a review with emphasis on underlying mechanisms and implications for health. Psychol Bull. 1996; 119: 488531.
  • 25
    Martin LA, Doster JA, Critelli JW, et al. The ‘distressed’ personality, coping and cardiovascular risk. Stress Health. 2011; 27: 6472.
  • 26
    Cosley BJ, McCoy SK, Saslow LR, et al. Is compassion for others stress buffering? Consequences of compassion and social support for physiologica. J Exp Soc Psychol. 2010; 46: 816823.
  • 27
    Schwerdtfeger A, Friedrich-Mai P. Social interaction moderates the relationship between depressive mood and heart rate variability: evidence from an ambulatory monitoring study. Health Psychol. 2009; 28: 501509.
  • 28
    Randall G, Bhattacharyya MR, Steptoe A. Marital status and heart rate variability in patients with suspected coronary artery disease. Ann Behav Med. 2009; 38: 115123.
  • 29
    Maunder RG, Lancee WJ, Nolan RP, et al. The relationship of attachment insecurity to subjective stress and autonomic function during standardized acute stress in healthy adults. J Psychosom Res. 2006; 60: 283290.
  • 30
    Diamond LM, Hicks AM. Attachment style, current relationship security, and negative emotions: the mediating role of physiological regulation. J Soc Personal Relat. 2005; 22: 499518.
  • 31
    Horsten M, Ericson M, Perski A, et al. Psychosocial factors and heart rate variability in healthy women. Psychosom Med. 1999; 61: 4957.
  • 32
    Rini C, Jandorf L, Valdimarsdottir H, et al. Distress among inflammatory bowel disease patients at high risk for colorectal cancer: a preliminary investigation of the effects of family history of cancer, disease duration, and perceived social support. Psychooncology. 2008; 17: 354362.
  • 33
    Maunder RG, Greenberg GR, Lancee WJ, et al. The impact of ulcerative colitis is greater in unmarried and young patients. Can J Gastroenterol. 2007; 21: 715720.
  • 34
    Oliveira S, Zaltman C, Elia C, et al. Quality-of-life measurement in patients with inflammatory bowel disease receiving social support. Inflamm Bowel Dis. 2007; 13: 470474.
  • 35
    Maunder RG, Greenberg GR, Hunter JJ, et al. Psychobiological subtypes of ulcerative colitis: pANCA status moderates the relationship between disease activity and psychological distress. Am J Gastroenterol. 2006; 101: 25462551.
    Direct Link:
  • 36
    Diehr MC, Cherner M, Wolfson TJ, et al. The 50 and 100-item short forms of the Paced Auditory Serial Addition Task (PASAT): demographically corrected norms and comparisons with the full PASAT in normal and clinical samples. J Clin Exp Neuropsychol. 2003; 25: 571585.
  • 37
    Mathias CW, Stanford MS, Houston RJ. The physiological experience of the Paced Auditory Serial Addition Task (PASAT): does the PASAT induce autonomic arousal? Arch Clin Neuropsychol. 2004; 19: 543554.
  • 38
    Tsuji H, Venditti FJ Jr, Mander ES, et al. Determinants of heart rate variability. J Am Coll Cardiol. 1996; 28: 15391546.
  • 39
    Sarason IG, Levine H, Basham RB, et al. Assessing social support: the social support questionnaire. J Pers Soc Psychol. 1983; 44: 127139.
  • 40
    Sarason IG, Sarason BR, Shearin EN, et al. A brief measure of social support: Practical and theoretical implications. J Soc Personal Relat. 1987; 4: 497510.
  • 41
    Cohen S, Doyle WJ, Skoner DP, et al. Social ties and susceptibility to the common cold. JAMA. 1997; 277: 19401944.
  • 42
    Rosengren A, Orth-Gomer K, Wedel H, et al. Stressful life events, social support, and mortality in men born in 1933. BMJ. 1993; 307: 11021105.
  • 43
    Orth-Gomér K, Rosengren A, Wilhelmsen L. Lack of social support and incidence of coronary heart disease in middle-aged Swedish men. Psychosom Med. 1993; 55: 3743.
  • 44
    Mikulincer M, Shaver PR. Attachment in Adulthood: Structure, Dynamics, and Change. New York: Guilford, 2007.
  • 45
    Porges SW. The Polyvagal Theory: phylogenetic contributions to social behavior. Physiol Behav. 2003; 79: 503513.
  • 46
    Schore AN. Affect Regulation and the Origin of the Self. Hillsdale, NJ: Lawrence Erlbaum Associates, 1994.
  • 47
    Eskandari F, Webster JI, Sternberg EM. Neural immune pathways and their connection to inflammatory diseases. Arthritis Res Ther. 2003; 5: 251265.