| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Department of Public Health Sciences (M.H., A.P., S.P.W., K.O.-G.), Division of Preventive Medicine, Karolinska Institute, Stockholm, Sweden; Department of Industrial Economics and Management (M.E.), Royal Institute of Technology, Stockholm, Sweden; Department of Cardiology (K.S.-G.), Karolinska Hospital, Stockholm, Sweden.
Address reprint requests to: Myriam Horsten, MA, Department of Public Health Sciences, Division of Preventive Medicine, Karolinska Institutet, Novum, Plan 7, S-141 57 Huddinge, Sweden. E:mail: Myriam.Horsten{at}phs.ki.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: The study group consisted of 300 healthy women (median age 57.5 years) who were representative of women living in the greater Stockholm area. For the measurement of social isolation, a condensed version of the Interpersonal Support Evaluation List was used and household size assessed. Anger was measured by the anger scales previously used in the Framingham study and depressive symptoms by a questionnaire derived from Pearlin. Health behaviors were measured by means of standard questionnaires. From 24-hour ambulatory electrocardiographic monitoring, both time and frequency domain measures were obtained: SDNN index (mean of the SDs of all normal to normal intervals for all 5-minute segments of the entire recording), VLF power (very low frequency power), LF power (low frequency power), HF power (high frequency power), and the LF/HF ratio (low frequency by high frequency ratio) were computed.
RESULTS: Social isolation and inability to relieve anger by talking to others were associated with decreased heart rate variability. Depressive symptoms were related only to the LF/HF ratio. Adjusting for age, menopausal status, exercise and smoking habits, history of hypertension, and BMI did not substantially change the results.
CONCLUSIONS: These findings suggest heart rate variability to be a mediating mechanism that could explain at least part of the reported associations between social isolation, suppressed anger, and health outcomes.
Key Words: psychosocial factors, • social support, • heart rate variability, • women.
Abbreviations: BMI = body mass index (kg/m2);; SBP = systolic blood pressure (mm Hg);; DBP = diastolic blood pressure (mm Hg);; HRT = hormone replacement therapy;; CHD = coronary heart disease;; JMP = John’s Macintosh program;; HRV = heart rate variability;; SDNN = SD of the normal to normal intervals (ms);; SDNN index = average of the SDs of all normal to normal intervals for each 5-minute interval of the entire recording (ms);; LF power = low frequency power (ms2);; HF power = high frequency power (ms2);; VLF power = very low frequency power (ms2);; LF/HF ratio = ratio of low-frequency to high-frequency power (none).
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
The mechanisms through which psychosocial factors exert their effects are unclear. Associations between psychosocial factors and standard risk factors for atherosclerosis, including smoking, lack of exercise, and unhealthy dietary habits, have been proposed as causal pathways. Furthermore, neuroendocrine and psychophysiological mechanisms have been suggested to mediate the effects between psychosocial factors and CHD. Among these mechanisms, a decreased HRV or disturbed autonomic balance, operating through altered autonomic nervous system function, seems to be one of the most plausible pathways (11). HRV is widely recognized as an important index of autonomic function in normal individuals and in those recovering from acute myocardial infarction (12, 13). Decreased HRV measured over a 24-hour period has been shown repeatedly to be a strong predictor of fatal outcome after a myocardial infarction (14–17). Furthermore, decreased HRV has been associated with increased risk of all-cause mortality in community studies of male and female populations such as the Framingham study (18) and the recent Zutphen study from Holland by Dekker et al. (19). In addition, results of the Atherosclerosis Risk in Communities (ARIC) study recently showed that a decreased HRV is associated with the risk of developing nonfatal and fatal CHD (20). These results extend the previous literature to newly occurring CHD and to the general population.
The concept of allostasis (21) emphasizes that the physiological systems within the body fluctuate to meet demands from external forces. This ability can be diminished by chronic activation of the sympathetic branch or chronic suppression of the parasympathetic branch of the autonomic nervous system caused by chronic stress or harmful social conditions. It is possible, that it is this phenomenon that is captured by the measurement of HRV. Decreased HRV could then be conceptualized as a lack of ability to respond by physiological variability and complexity, making the individual physiologically rigid and, therefore, more vulnerable. This is consistent with Porges’ proposal that a chronically depressed vagal tone would reflect poor homeostasis and a neurophysiological vulnerability to the deleterious effects of stress (22).
Decreased HRV has been demonstrated in both depressed psychiatric patients and in depressed cardiac patients (23–27). Among other psychosocial factors that have been associated with HRV are anger (28), Type A behavior pattern (29), and mental stress (30, 31). We have found no studies directly examining social isolation or support in relation to HRV. However, an association can be hypothesized on the basis of earlier studies showing associations between low social support and increased catecholamine levels (32, 33), and low social support and sustained increased heart rates (34), which in turn are related to low HRV and autonomic dysfunction (14). Furthermore, poor emotional and social support has been related frequently to depressive symptoms (35, 36). Therefore, the present study examines associations between social isolation, depressive symptoms, anger, and HRV.
Besides the paucity of studies investigating relations between psychosocial risk factors and HRV, even less attention has been given to female subjects. When women were included in such studies, the number was usually too small to analyze the data by gender. In addition, as described above, both decreased HRV and psychosocial strain (eg, depression, social isolation) have been related to CHD incidence and all-cause mortality justifying the hypothesis that psychosocial strain may be associated with a decreased HRV in a healthy population.
Therefore, the aim of the present study was to investigate associations between psychosocial risk factors and HRV indices in healthy women.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
A letter was sent to prospective subjects that explained the objectives and the focus of the study and invited them to participate. Those who did not call the clinic spontaneously were contacted by phone from the research clinic. When a subject refused to participate, a new subject was selected to obtain a final number of 300 healthy women. Of those eligible, 17% declined to participate, mainly because of difficulties in arranging time off from work for the study.
Procedure
The first day of the study included a physical examination and placement of a 24-hour Holter EKG monitor. During the second day, anthropometric measures were obtained, blood pressure was measured, and the Holter EKG monitor was detached. Questionnaires on psychosocial and life style factors, including depressive symptoms (38), social support (39), and anger (40), were sent to the subjects 1 week before their visit to the research clinic. Subjects completed the questionnaires at home and brought them to the clinic. The research nurse reviewed the questionnaires with the subjects and completed any answers that were missing. Internal nonresponse was less than 10%.
Psychosocial Factors
Depressive symptoms were measured by means of a 10-item questionnaire derived from Pearlin et al. (38). One question, about sexual activity, was excluded in an effort to avoid potentially threatening items. The present version thus included 9 questions with yes/no alternatives. The yes answers were summed, with a low score indicating a low degree of depressive feelings. The scale includes questions on mood, sleeping problems, appetite, energy, interest in normal activities, crying, and feelings about the future. Some examples of questions are: Do you feel bored or do you have little interest in doing things? Do you feel downhearted or blue? Do you feel hopeless about the future? The scale had an adequate internal consistency (Cronbach’s 
= .85) and was significantly correlated (r = .71) with the Beck Depression Inventory in a subsample of the study population (N = 30).
For the measurement of social support, a condensed version of the Interpersonal Support Evaluation List (ISEL) was used (39). Social support functions included belongingness (5 items), tangible support (3 items), and appraisal support (5 items). Examples of items were: There are several different people with whom I enjoy spending time; When I feel lonely, there are several people I could call and talk to; I regularly meet or talk with members of my family or friends (belongingness); When I need suggestions for how to deal with a personal problem, I know there is someone I can turn to (appraisal support); If I needed some help in moving to a new home, I would have a hard time finding someone to help me (tangible support). Items were scored on a scale from 1 (is very true) to 4 (is not at all true). The total scale thus comprised 13 items with scores ranging from 13 to 52, a high score indicating more social support.
Anger was measured by a 12-item scale previously used in the Framingham study (40). The scale consisted of four subscales: anger-symptoms (5 items), anger-in (3 items), anger-out (2 items), and anger-discuss (2 items). Examples of items included: How do you react when you are really angry or annoyed? Get tense or worried (anger-symptoms); Keep it to yourself (anger-in); Take it out on others (anger-out); Talk to a friend or relative (anger-discuss). The items were scored on a scale from 0 to 1 (1 indicating presence of the trait characteristic). Responses to each item were summed and the sum was divided by the number of items per subscale.
Demographic and Lifestyle Factors
Age at examination was obtained from the date of birth given in the person number of the census register. The number of persons in the household, marital status, educational level, history of hypertension, physical exercise, and smoking behavior were assessed by means of questionnaires and interview procedures. Educational level was classified as follows: (I) mandatory only (corresponding to 9 years of school education); (II) high school or college. Physical exercise was assessed according to the World Health Organization criteria and graded into: (I) reading, watching TV, or other sedentary leisure activities; (II) walking, cycling, or other forms of physical activity; (III) exercises to keep fit, heavy gardening, etc., for at least 4 hours a week; and (IV) hard training or participation in competitive sports several times a week. In the analyses, physical exercise in leisure time was dichotomized into; sedentary lifestyle (I) and nonsedentary lifestyle (II-IV). Blood pressure, weight, and height were measured by the research nurse during examination. BMI was calculated as weight (kg)/height (m2). In univariate analyses, obesity was defined as BMI > 28.6 kg/m2 (41). In the analyses, a cutoff for high systolic blood pressure was taken at > 140 mm Hg and for high diastolic blood pressure at > 90 mm Hg, defining hypertension as systolic blood pressure >140 mm Hg or diastolic blood pressure > 90 mm Hg.
Menopausal status was assessed in a gynecological interview by the research nurse. Postmenopausal status was defined as having had no menses for at least 6 months. A history of gynecological surgery was obtained. A patient was classified as having had surgical menopause if she had undergone a bilateral oophorectomy. A complete history regarding HRT was also obtained. Women who had begun HRT before menopause were considered menopausal if they were more than 50 years of age.
Ambulatory EKG Monitoring and Analysis of Heart Rate Variability
A two-channel EKG recording was performed for 24-hour periods during normal daily activities. A Spacelabs model 90205 recorder (Spacelab Inc., Redmond, WA) was used to obtain the EKG data. Electrodes were attached at the positions of CM-V5 (left anterior axillary line sixth rib) and CS-V1 (fourth rib at the sternal border). Subjects were encouraged to continue their normal daily activities during the period of monitoring. There were 264 complete 24-hour Holter EKG recordings available. Recordings were excluded if they showed more than 10% nonsinus rhythm (N = 4), extreme outliers on total power (N = 2), medication for hypertension (N = 3), or less than 50% original material (N = 6), resulting in 249 recordings available for the analyses.
The 24-hour tape recordings were digitized and QRS-labeled using a commercially available PC-based system (Aspect Holter System, Daltek, Borlänge, Sweden). An automatic analysis of arrhythmias was first made and the QRS complexes classified. The consecutive R-R intervals were expressed in centiseconds and their corresponding classification codes were exported to an ASCII text file. The text file was additionally analyzed using our own custom-made software on a personal computer. Additional analysis was made in units of 5-minute epochs. To be accepted for additional analysis, we required an interval to have at least 96% of the QRS-complexes classified as normal by the Daltek system (42).
Five-minute (300 s) epochs of data were analyzed by custom-made software. The time series of R-R intervals were resampled at a sampling frequency of two samples per second. Gaps in the time series due to nonnormal R-R intervals (QRS-labeled by the Aspect System classification as noise or ectopic beats) were filled with values calculated by linear interpolation between adjacent normal R-R intervals. The computer also automatically checked for misclassified dropped beat deviating more than 3.0 SD from the normal R-R interval of each epoch.
After editing was computed, the time domain measures mean of normal RR intervals and mean of the SDs of all normal to normal intervals for all 5-minute segments of the entire recording (SDNN index, in msec) (43) were obtained from the time series of normal R-R intervals. SDNN index has also been referred to as SDNNIDX in the literature (44).
The frequency domain of the time series of R-R intervals was analyzed with an autoregressive method described by Kay and Marple (45). The model order and number of coefficients in the polynomial describing the time series was constantly set to 18. The mean R-R interval of each time series was subtracted and then detrended by applying linear regression. Frequency domain measures of R-R variability were computed by integrating over their frequency intervals in three frequency bands as described by Bigger et al. (46) and later recommended by the Task Force (43) 0.003 to < 0.04 Hz (VLF, in msec2), 0.04 to < 0.15 Hz (LF, in msec2), and 0.15 to 0.4 Hz (HF, in msec2). In addition, the LF/HF ratio was calculated.
Vagal activity is the major contributor to the HF component. Disagreement exists in respect to the LF component, which is considered by some as a marker of sympathetic modulation (especially when expressed in normalized units) and by others as a parameter that includes both sympathetic and vagal influences. Consequently, the LF/HF ratio is considered by some investigators to mirror sympathovagal balance. In results from 24-hour periods, the interpretation of the LF/HF ratio is more complicated because physiological mechanisms of heart period modulations responsible for LF and HF power components cannot be considered stationary during the 24-hour period. The physiological interpretation of VLF power is yet unclear (although the major component of the spectral density is located there) (43).
Statistical Analyses
We used JMP Statistics for the Apple Macintosh Version 3.1 (47) and STATA 3.1 (48) for the statistical analyses. All HRV variables were nonnormally distributed and, therefore, log transformed in the analyses. All correlation’s reported are Spearman 
correlations. Multivariate linear regression analyses were performed to test associations between the risk factors and the HRV variables, the HRV variables always being the dependent variable. The anti-logs of the least squares mean (LSM) were reported after controlling for other risk factors. For variables with more than two categories, the linear trend was assessed by computing the p value for trend (49). When the linear trend was nonsignificant but the LSM suggested a difference between one of the groups versus the rest, this was tested and reported when significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
|
|
|
|
As expected, a statistically significant correlation existed between marital status and household size (r = .68, p < .0001). The correlation between social support and household size was not statistically significant (r = .10, p = .13). The effects of household size were also tested while controlling for social support. Although the associations between household size and HRV parameters were weakened, they remained statistically significant (all p values > .01 and < =.05).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
The strongest and most consistent results were found for small household size, a proxy measure for social isolation, and for lack of social support. The associations between household size and HRV remained statistically significant after controlling for educational level, indicating that they were not due to a confounding effect of low socioeconomic status, which has been found to be related both to smaller households and to cardiac mortality (9). When the effects of household size on HRV were tested while controlling for social support, the associations were weakened, but remained statistically significant. A tentative explanation for this could be that household size is an indicator of support from close family (spouse, children) that is not totally covered by the social support measure.
The social support scale describes three functions that are usually provided by the more extended social network. These include appraisal support, which provides help to orient oneself in life, to adequately perceive and estimate stressors, and to find ways to cope with them. The second function, belongingness, describes the basic human need for a sense of belonging to groups of people and to share interests and values with them. The third function, tangible support, assesses the availability of concrete, practical help (39, 50).
As far as we know, no other study has investigated relations between social support, household size, and HRV in women. However, Undén et al. (34) reported persons with low social support at work to have sustained increased heart rates, using 24-hour ambulatory EKG monitoring. Sustained increased heart rates have been associated with lower HRV (14), thereby supporting the results of the present study. Both measures have been viewed as measures of cardiac autonomic dysfunction.
Depressive symptoms were only statistically significantly associated with the LF/HF ratio. Women with depressive symptoms had a lower LF/HF ratio than women without depressive symptoms. This finding may suggest less sympathetic modulation in the women with depressive symptoms. However, interpretation of the LF/HF ratio in spectral analyses performed on the entire 24-hour period is difficult, inasmuch as averages of the modulations attributable to the LF and HF components are provided. Such averages obscure the detailed information about autonomic modulation of R-R intervals, that is available in shorter recordings (43).
In the present study, depressive symptoms were not statistically significantly associated with the other HRV measures, including the SDNN index, VLF power, LF power, and HF power. In previous studies, decreased HRV in both depressed psychiatric patients and in depressed cardiac patients has been reported (23–27). In these studies, significant associations with various time domain HRV measures were reported. The one study that reported frequency domain measures of HRV in relation to depression did not base these on 24-hour EKG recordings (26), complicating comparison with our results. Furthermore, they all used patient study groups. It is possible that associations between depressive symptoms and HRV measures in patient populations are different from those in normal populations (as used in the present study).
Interestingly, of the anger scales only the scale "anger-discuss" was statistically significantly associated with HRV. Not being able to relieve anger by talking to a friend or relative was associated with lower HRV. An association with HRV has not been previously reported, but in the Framingham study the same scale was an independent predictor of the incidence of CHD and angina pectoris in women under 65 years of age (6).
As discussed earlier, the physiological meaning of the HRV measures, especially of the VLF power component, is not very well established. We know that decreased HRV is related to both cardiac and all-cause mortality (14–19) and that vagal activity is the major contributor to the HF component (43). In the present study, the time domain measure SDNN index was used. The rationale for using the SDNN index was that only the normal R-R intervals included in the 5-minute epochs used for spectral analysis should be included in the time domain measures, so that time and frequency domain measures are based on the same 5-minute epochs. Consequently, normal beats adjacent to periods of, eg, ectopic activity, were not included in the analysis. VLF power and LF power have been suggested as the frequency domain equivalents for the SDNN index because they are highly correlated (r 
.90) (44, 51). Furthermore, in a study by Bigger et al. (51), the correlation between the SDNN and the SDNN index was .81, and both measures had an equally strong association with death after myocardial infarction.
We would like to see decreased HRV in the light of the statement by Porges (22) that "clinically, stasis or lack of endogenous variability in neurally mediated peripheral systems, such as gastric motility and heart rate, is a sign of severe physiological compromise." The decreased HRV in the socially isolated women and those not being able to relieve their anger by talking to others, could then be conceptualized as a lack of ability to respond by physiological variation and complexity, making them physiologically rigid and, therefore, more vulnerable.
In conclusion, women who lived alone, who reported lack of social support, and who reported not to relieve their anger by talking to others showed evidence of lower HRV during daily life compared with women with larger households, higher social support, and women that could discuss their anger. Of all HRV measures, low SDNN, its frequency domain equivalent low ultra LF power (44, 51) and low VLF power have been most consistently and strongly related to mortality after myocardial infarction and to all-cause mortality (14, 15, 19, 51). In addition, social isolation has also been related to all-cause mortality, to CHD incidence, and to prognosis after myocardial infarction (1, 2, 5, 8–10, 52). Furthermore, suppressed anger has been related to all-cause mortality (3), to the progression of atherosclerosis (53), and in the Framingham study—using the same scale as the present study—to CHD incidence (6). Our results suggest that HRV may be one of the possible mechanisms that explain these reported associations.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication December 1, 1997.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  V. Vaccarino, R. Lampert, J. D. Bremner, F. Lee, S. Su, C. Maisano, N. V. Murrah, L. Jones, F. Jawed, N. Afzal, et al. Depressive Symptoms and Heart Rate Variability: Evidence for a Shared Genetic Substrate in a Study of Twins Psychosom Med, July 1, 2008; 70(6): 628 - 636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. A. Wellenius, K. J. Mukamal, A. Kulshreshtha, S. Asonganyi, and M. A. Mittleman Depressive Symptoms and the Risk of Atherosclerotic Progression Among Patients With Coronary Artery Bypass Grafts Circulation, May 6, 2008; 117(18): 2313 - 2319. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Bleil, P. J. Gianaros, J. R. Jennings, J. D. Flory, and S. B. Manuck Trait Negative Affect: Toward an Integrated Model of Understanding Psychological Risk for Impairment in Cardiac Autonomic Function Psychosom Med, April 1, 2008; 70(3): 328 - 337. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Ohira, A. V. D. Roux, R. J. Prineas, M. A. Kizilbash, M. R. Carnethon, and A. R. Folsom Associations of Psychosocial Factors With Heart Rate and Its Short-Term Variability: Multi-Ethnic Study of Atherosclerosis Psychosom Med, February 1, 2008; 70(2): 141 - 146. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-J. Lee, Y. Chae, H.-J. Park, D.-H. Hahm, K. An, and H. Lee Turo (Qi Dance) Training Attenuates Psychological Symptoms and Sympathetic Activation Induced by Mental Stress in Healthy Women Evid. Based Complement. Altern. Med., November 1, 2007; (2007) nem120v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Ritz and B. Dahme Implementation and Interpretation of Respiratory Sinus Arrhythmia Measures in Psychosomatic Medicine: Practice Against Better Evidence? Psychosom Med, July 1, 2006; 68(4): 617 - 627. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. T. Lewis, S. A. Everson-Rose, L. H. Powell, K. A. Matthews, C. Brown, K. Karavolos, K. Sutton-Tyrrell, E. Jacobs, and D. Wesley Chronic Exposure to Everyday Discrimination and Coronary Artery Calcification in African-American Women: The SWAN Heart Study Psychosom Med, May 1, 2006; 68(3): 362 - 368. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. Diamond and A. M. Hicks Attachment style, current relationship security, and negative emotions: The mediating role of physiological regulation Journal of Social and Personal Relationships, August 1, 2005; 22(4): 499 - 518. [Abstract] [PDF]
|
|
|
|
|
|
H. Hemingway, M. Shipley, E. Brunner, A. Britton, M. Malik, and M. Marmot Does Autonomic Function Link Social Position to Coronary Risk?: The Whitehall II Study Circulation, June 14, 2005; 111(23): 3071 - 3077. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Sher Type D personality: the heart, stress, and cortisol QJM, May 1, 2005; 98(5): 323 - 329. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Denollet DS14: Standard Assessment of Negative Affectivity, Social Inhibition, and Type D Personality Psychosom Med, January 1, 2005; 67(1): 89 - 97. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Rennie, H. Hemingway, M. Kumari, E. Brunner, M. Malik, and M. Marmot Effects of Moderate and Vigorous Physical Activity on Heart Rate Variability in a British Study of Civil Servants Am. J. Epidemiol., July 15, 2003; 158(2): 135 - 143. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Jones, J. T. Bromberger, K. Sutton-Tyrrell, and K. A. Matthews Lifetime History of Depression and Carotid Atherosclerosis in Middle-aged Women Arch Gen Psychiatry, February 1, 2003; 60(2): 153 - 160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Steptoe and M. Marmot The role of psychobiological pathways in socio-economic inequalities in cardiovascular disease risk Eur. Heart J., January 1, 2002; 23(1): 13 - 25. [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. Diamond Contributions of Psychophysiology to Research on Adult Attachment: Review and Recommendations Personality and Social Psychology Review, November 1, 2001; 5(4): 276 - 295. [Abstract] [PDF]
|
|
|
|
 |
|
 V. Vaccarino, S. V. Kasl, J. Abramson, and H. M. Krumholz Depressive symptoms and risk of functional decline and death in patients with heart failure J. Am. Coll. Cardiol., July 1, 2001; 38(1): 199 - 205. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Hemingway, M. Malik, and M. Marmot Social and psychosocial influences on sudden cardiac death, ventricular arrhythmia and cardiac autonomic function Eur. Heart J., July 1, 2001; 22(13): 1082 - 1101. [PDF]
|
|
|
|
 |
|
 K. Orth-Gomer, S. P. Wamala, M. Horsten, K. Schenck-Gustafsson, N. Schneiderman, and M. A. Mittleman Marital Stress Worsens Prognosis in Women With Coronary Heart Disease: The Stockholm Female Coronary Risk Study JAMA, December 20, 2000; 284(23): 3008 - 3014. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M Horsten, M.A Mittleman, S.P Wamala, K Schenck-Gustafsson, and K Orth-Gomer Depressive symptoms and lack of social integration in relation to prognosis of CHD in middle-aged women. The Stockholm Female Coronary Risk Study Eur. Heart J., July 1, 2000; 21(13): 1072 - 1080. [Abstract] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
