| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Department of Epidemiology and Public Health, University College London (A.S., S.K-E., N.O., P.J.F., M.M.), London, United Kingdom; and the Haemostasis, Thrombosis, and Vascular Medicine Unit, University Department of Medicine (A.R., G.D.O.L.), Glasgow, United Kingdom.
Address reprint requests to: Dr. Andrew Steptoe, Department of Epidemiology and Public Health, University College London, 1-19 Torrington Place, London WC1E 6BT, UK. Email: a.steptoe{at}ucl.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
METHODS: The study was conducted with 125 white men and 96 white women aged 47 to 58 years, drawn from the Whitehall II cohort. Socioeconomic status was indexed by grade of employment, with 82 high, 75 intermediate, and 64 low grade participants. Plasma fibrinogen and hematocrit were assessed at baseline, immediately after performance of color-word and mirror tracing tasks, and 45 minutes later.
RESULTS: Plasma fibrinogen increased from baseline to stress (from 2.85 ± 0.57 to 2.92 ± 0.58 g/liter), remaining elevated 45 minutes after stress (2.89 ± 0.58 g/liter, p < .001). Fibrinogen concentration was greater in the low than in the high or intermediate employment grade groups, independently of sex, age, body mass index, smoking status, and hematocrit. Fibrinogen responses to acute stress did not differ across employment grades. Women had higher fibrinogen levels than men, but this pattern was abolished in women taking hormone replacement therapy. Men experiencing low job control showed greater fibrinogen responses to acute stress than did those with high job control (p = .003). Fibrinogen levels were greater in socially isolated individuals, but social isolation did not affect responses to acute stress.
CONCLUSIONS: Socioeconomic status and acute stress had independent effects on the plasma fibrinogen level. Low job control may influence cardiovascular disease risk in men partly through provoking greater fibrinogen stress responses.
Key Words: fibrinogen • socioeconomic status • job control • stress • social isolation.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
A small study of 45 high status and 29 low status male civil servants identified fibrinogen as a possible link between social class and coronary heart disease (4). The inverse association between fibrinogen and socioeconomic status has since been confirmed in both men and women from the Whitehall II cohort (5, 6) and in a number of other studies of young adults and middle-aged men and women (7–11). Raised fibrinogen may be one of the pathways through which socioeconomic status increases coronary heart disease risk and is also related to factors such as cigarette smoking, menopausal status, and body mass index.
The literature relating fibrinogen with psychosocial factors has been less consistent (12). Several investigations have shown an association of fibrinogen with work stress (4, 13), in particular with low control at work (6, 10, 14) and effort-reward imbalance (15). However, negative findings have also been obtained (16–19). The explanation of these discrepancies is unclear, although many of the studies that showed no association were carried out with young adults (9, 18, 19), and relationships between chronic work stress and fibrinogen may not emerge until middle age. Sex differences may also be important, since in the Whitehall II study, higher fibrinogen was associated with low self-rated job control in men but not women (6). Another factor that has been identified as a correlate of raised fibrinogen is social isolation or low social support (10, 20), and associations with hostility and depression have also been described (9).
Further insight may emerge from studies that model the impact of mental stress on fibrinogen more acutely. Jern et al. (21) assessed the influence of 20 minutes of mental arithmetic and a color-word interference task on plasma fibrinogen in 10 young men. Fibrinogen concentration was elevated in blood drawn immediately after tasks, an effect that was sustained 10 minutes later. This pattern of results was subsequently replicated in young women (22), and reliable increases in plasma fibrinogen emerged in response to speech and mental arithmetic tasks in another study of young women (23). By contrast, no significant increase was recorded by Muldoon et al. (24) in blood drawn at the end of a 21-minute color-word task or in another study of patients with hyperlipidemia (25).
The studies of fibrinogen responses to acute mental stress published to date have been relatively small scale and have centered on immediate reactions. It is not known whether increases in fibrinogen are sustained for extended periods after stress or rapidly return to baseline levels. An additional issue concerns the increase in hemoconcentration that typically occurs in response to mental stress (24, 26). Greater hemoconcentration may lead to increases in the concentration of fibrinogen that are not due to higher circulating levels but to the same amount of fibrinogen being present in a smaller plasma volume. Accordingly, in the present study, we assessed plasma fibrinogen both immediately after stress and 45 minutes later and monitored hematocrit at the time of each observation.
The objectives of this study were to assess fibrinogen responses to acute stress in a large sample of men and women and to investigate associations with socioeconomic status and psychosocial factors. The study involved men and women from the Whitehall II epidemiological cohort, drawn from high, intermediate and low grades of employment. We hypothesized that plasma fibrinogen would be elevated in participants from lower grades and that lower grade individuals would show larger fibrinogen responses to acute mental stress. We also assessed the influence of self-rated job control and social isolation on fibrinogen, postulating that low control and high social isolation might be associated with enhanced fibrinogen stress responsivity.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Measures
Participants completed a series of questionnaire measures of recent health, working situation, and psychosocial factors. Job control or control at work was measured with the scale previously used in the Whitehall cohort (29, 30). This consists of nine items (eg, "I can decide when to take a break"), each of which is rated on a 4-point scale from "never/almost never" to "often." Ratings were converted into score ranging from 0 to 100, where 100 represents the maximum possible control. Cronbach’s 
for the scale in this study was 0.73.
Social isolation was measured with questions concerning contact with family, friends, and relatives, derived from the Close Persons Questionnaire (31). Respondents were given points on the social isolation measure if they lived alone, if they had no relatives outside their household, never visited or were never visited by relatives or friends, or had no relatives or friends who they saw at least once a month. Scores could range from 0 (no social isolation) to 3 (maximum isolation).
Blood pressure and heart rate were monitored continuously from the finger using a Portapres-2, a portable version of the Finapres device that shows good reproducibility and accuracy in a range of settings (32). Details of these measures are provided elsewhere (33). During the session, ratings of subjective stress were taken periodically on a 7-point scale where 1 = low and 7 = high.
Behavioral Tasks
Mental stress was induced by two behavioral tasks. The first was a computerized color-word interference task involving the successive presentation of target color words (eg, green, yellow) printed in another color (34). At the bottom of the computer screen were four names of colors printed in incongruous colors. The task was to press a computer key that corresponded to the position at the bottom of the screen of the name of the color in which the target word was printed. The rate of presentation of stimuli was adjusted to the performance of the participant to ensure sustained demands. The second task was mirror tracing, involving the tracing, with a metal stylus, of a star that could be seen only in mirror image (35). A mistake was registered each time the stylus came off the star, and a loud beep was emitted by the apparatus (Lafayette Instruments Corp., Lafayette, IN). Participants were told that the average person completed five circuits of the star in the time available, and they were asked to give accuracy priority over speed on both tasks.
Procedure
Participants were tested in the morning or afternoon in a light- and temperature-controlled laboratory. Before the day of testing they were instructed not to drink tea, coffee, or caffeinated beverages, or to smoke, for at least 2 hours before the study, and not to consume alcohol or exercise on the evening before or on the day of testing. The study was approved by the University College London/UCL Hospital Committee on the Ethics of Human Research.
After application of instruments and insertion of a venous cannula for periodic collection of blood samples, the participant rested for 30 minutes. A baseline blood sample was drawn, and blood pressure and heart rate were recorded for a 5-minute period. The two tasks were then administered in random order. Each lasted for 5 minutes, during which blood pressure and heart rate were recorded continuously. After each task, the participant rated task difficulty and task involvement on 7-point scales ranging from 1 = low to 7 = high. A second blood sample (stress measure) was drawn immediately after the task period, and a third after 45 minutes of quiet rest (recovery measure). Saliva samples were obtained at the end of the adaptation period, after tasks, and 20 and 45 minutes after tasks for the assessment of cortisol.
Fibrinogen and Hematocrit
Blood was collected in citrated tubes, centrifuged, and frozen at -70°C within 1 hour. Clottable fibrinogen was measured from frozen samples by an automated Clauss assay in a MDA-180 coagulometer (Organon Teknika, Cambridge, UK) using the manufacturer’s reagents and the international fibrinogen standard (36). Hematocrit was assessed immediately after each blood sample was drawn using a microhematocrit centrifuge and reader (Hawksley Gelman, Lancing, Sussex, UK).
Statistical Analysis
Plasma fibrinogen responses were analyzed by repeated-measures analysis of covariance with grade of employment and sex as between-subjects factors and trial (baseline, stress, recovery) as the within-subject factor. Concurrent hematocrit was included as a covariate in all analyses. Significant interactions involving trial were followed up with analysis of change scores between stress or recovery and baseline. In subsequent analyses the fibrinogen responses of women currently taking hormone replacement therapy were compared both with those of men and with women not taking therapy. Preliminary analysis indicated that fibrinogen did not vary according to whether sessions were held in the morning or afternoon, so time of day was not included as a factor. The influence of job control and social isolation was analyzed by introducing these variables in separate models as between-subjects factors. Participants were divided into high and low job control groups using the median cutoff of 65. Just over half (53.4%) of the participants had social isolation scores of 0, so they were compared with those with isolation scores of 1 to 3.
Cardiovascular and cortisol responses are not described in this report, but associations between fibrinogen responses and cardiovascular and cortisol reactivity are discussed. Cardiovascular reactivity was defined in a manner similar to that described by Muldoon et al. (24). Change scores in systolic blood pressure, diastolic blood pressure, and heart rate between task and baseline periods were standardized and then summed to produce a single index of reactivity. Cortisol stress reactivity was defined as the change between baseline and the 20-minute poststress sample. Correlations partialed for sex were computed between cardiovascular and cortisol reactivity measures and fibrinogen changes between baseline and stress, and changes between baseline and recovery.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
2 = 6.06, p = .014). The job control scores showed a significant gradient across grades of employment, with greater control in high status participants (F(2,217) = 21.2, p < .001). The proportion of socially isolated individuals did not differ across groups (p = .60).
|
Hematocrit increased from baseline to stress trials (mean = 39.2 ± 3.1% to 39.8 ± 3.1%, F(1,215) = 36.8, p < .001), subsequently falling to 39.6 ± 3.1% in the recovery trial. The value in the recovery trial remained above baseline (F(1,215) = 20.5, p < .001) but did not differ significantly from the value in the stress trial (p = .067).
Socioeconomic Status and Fibrinogen Responses
The analysis of plasma fibrinogen responses to tasks with grade and sex as between-subjects and hematocrit as covariate revealed main effects of grade of employment (F(2,214) = 3.98, p = .020), sex (F(1,214) = 16.6, p < .001), and trial (F(2,429) = 12.4, p < .001). There were no interactions between the three factors. These results are summarized in Figure 1. Plasma fibrinogen was higher in women than men at all time points. Levels increased from baseline to stress (p < .001) and remained above baseline levels after 45 minutes of recovery (p = .003). There was a decrease in fibrinogen from stress to recovery (mean = 2.92 ± 0.58 and 2.89 ± 0.58 g/liter) that approached significance (p = .059). The grade effect resulted from the fact that fibrinogen levels were greater in the low grade than in the intermediate and high grade groups. Separate analyses indicated that the intermediate and high employment grade groups did not differ at any point. The employment grade effect was evident in baseline, stress, and recovery samples; separate analyses of trials with sex, age, body mass index, smoking status, and hematocrit as covariates revealed significant grade effects for baseline (p = .016), stress (p = .031), and recovery (p = .024) trials.
|
Smoking and Hormone Replacement Treatment
Cigarette smoking was more prevalent in lower status participants, and fibrinogen is strongly related to smoking in population surveys (37). Analyses were therefore repeated with smoking status as an additional covariate. The effects of grade of employment (F(2,212) = 3.56, p = .030), sex (F(1,212) = 16.7, p < .001) and trial (F(2,436) = 7.76, p < .001) were unchanged. A further analysis was carried out on nonsmokers only. The association between fibrinogen level and grade was greater with smokers excluded (F(2,192) = 4.90, p = .008), and the main effects of sex and trial were replicated. The one difference from the analyses of the full sample was that the decrease between stress and recovery trials became statistically significant in nonsmokers (F(1,194) = 6.37, p = .012).
Hormone replacement therapy is associated with plasma fibrinogen in population surveys (38), so fibrinogen was compared in women who did and did not take hormone replacement therapy. A difference between the groups was apparent (F(1,93) = 10.7, p < .001); fibrinogen was lower in women taking hormone replacement therapy. This effect was independent of trial and grade of employment. The mean fibrinogen concentration across trials, adjusted for age, body mass index, smoking status, and hematocrit, was 3.22 ± 0.65 g/liter in women not taking hormone replacement therapy, compared with 2.76 ± 0.61 g/liter in those taking hormone replacement. Fibrinogen levels in the latter did not differ from the levels recorded in men (2.66 ± 0.46 g/liter). Thus, the higher plasma fibrinogen level in women compared with men was eliminated among those taking hormone replacement therapy.
Job Control and Fibrinogen Responses
Job control was introduced as an additional between-subjects factor in the analysis. The main effects for grade of employment, sex, and trial remained as before. There was no main effect of job control, but there was a significant interaction between job control, sex, and trial (F(2,423) = 4.02, p = .019). This interaction was due to differences in fibrinogen stress responses in men and women reporting low job control. Analyses of change scores between task and baseline trials (with change in hematocrit as covariate) revealed a significant job control-by-sex interaction (F(1,220) = 6.73, p = .01). As can be seen in Figure 2, the stress-induced increase in fibrinogen was greater in men reporting low than high job control (p = .003), whereas job control was not related to fibrinogen responses in women (p = .28). Thus, low job control was associated with increased fibrinogen stress responsivity, but only in men. The influence of job control was independent of the socioeconomic status difference in plasma fibrinogen level. The impact of job control was also retained when analyses were confined to nonsmokers, and it was not related to hormone replacement therapy in women.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Although the stress testing protocol elicited substantial increases in blood pressure and heart rate and in self-reported stress, we found no correlation between the magnitude of individual cardiovascular or subjective changes and fibrinogen response. Similarly, the changes in fibrinogen were not related to saliva free cortisol secretion over the session. This indicates that the stress-induced increase in fibrinogen may be mediated through biological processes that are not sensitively indexed by these markers of physiological reactivity. The increase in fibrinogen may be mediated by generalized activation as opposed to specific stress mechanisms.
The socioeconomic gradient in plasma fibrinogen recorded here replicates a pattern previously observed in the Whitehall study and elsewhere (6). However, the hypothesis that low socioeconomic status individuals would show larger fibrinogen responses to acute stress was not confirmed. There was no evidence for associations between socioeconomic status and increases in fibrinogen from baseline to task periods or for more sustained elevations in fibrinogen in the recovery period. The result was not due to differences across grades of employment in task engagement or involvement because task appraisals were comparable in the three groups. The effects of socioeconomic status and stress were independent, and both contributed to the absolute level of fibrinogen obtained. Thus, the highest absolute levels were recorded in low status individuals during the stress trials. There is evidence that exposure to chronic stressors in everyday life is more frequent in low status groups (27, 41). It is probable therefore that the occurrence of stress-induced increases in fibrinogen may also be more frequent in less affluent groups. Hence, differences in fibrinogen levels may be related to differential exposure to stressors rather than differential reactivity (42). It is also possible that acute stress testing is not an appropriate methodology with which to investigate socioeconomic differences in fibrinogen. Slower and more sustained biological processes may stimulate differences, or they may arise in part from variations in patterns of health behavior.
The sex differences in fibrinogen between middle-aged men and women have been documented previously and relate to hormonal status in women (6, 10). Hormone replacement therapy in postmenopausal women has been associated with substantial reductions in total and cardiovascular mortality (43), though these observational findings have yet to be confirmed in randomized trials. Although the mechanism remains controversial, part of the benefit seems to be mediated through hemostatic mechanisms, including fibrinogen (44). Decreased levels of fibrinogen in women taking hormone replacement therapy have been reported in a number of studies (38, 45, 46). The present data are consistent with this pattern because fibrinogen levels were lower among women taking hormone replacement therapy, falling to levels similar to those recorded in men.
Although grade of employment did not relate to fibrinogen responsivity, associations with job control were observed. Men reporting low job control showed larger stress-induced increases in fibrinogen than did those with high control (Figure 2). It is possible therefore that stress-induced alterations in fibrinogen may be relevant to the pathways through which lack of control at work increases cardiovascular disease in men. Low control at work in men has previously been linked with heightened acute blood pressure responsivity to uncontrollable tasks (47), elevated ambulatory blood pressure (48), and heightened cortisol early in the day (49). The influence of job control did not interact with socioeconomic status. However, the prevalence of low job control was greater in low status individuals; 66% of men in the low employment grade reported low control, compared with 31% of those in high and intermediate grades (a similar pattern was present in women, but is not relevant to fibrinogen). Thus, the impact of low job control in population terms is greater in lower socioeconomic status groups.
The explanation for the lack of a job control effect on fibrinogen responses in women is unclear. Job control has previously been found to show closer associations with psychobiological responses in men than women. For example, Light et al. (50) found that job strain (high demand/low control) was associated with elevated ambulatory blood pressure over the working day in men but not women; women’s blood pressure was influenced by other factors such as persistent active coping attempts (51). The lack of differences in fibrinogen level at baseline in women varying in job control stress is consistent with studies on younger samples (18, 19). In addition, Brunner et al. (6) showed that fibrinogen levels were positively rather than negatively associated with self-reported control at work in women after adjustment for age, ethnic origin, menopausal status, employment grade, body mass index, smoking, alcohol consumption, and physical activity. Thus, it seems that job control has distinct effects on fibrinogen in men and women.
The second psychosocial factor investigated in this study was social isolation. Higher levels of fibrinogen were recorded in participants who reported greater social isolation, a pattern that has previously been observed with other measures of social contact (10, 20). But social isolation showed no association with fibrinogen responses to acute stress. In contrast with the findings related to job control, the elevated fibrinogen of socially isolated individuals is not related to dysregulation of acute stress responsivity. It is possible that the factors determining the impact of chronic psychosocial influences in fibrinogen may be different from those affecting acute stress responses. It has been suggested that genetic polymorphisms for fibrinogen may have greater effects on stress-induced levels than on baseline values (52).
A number of limitations in this study should be noted. The study was carried out with white men and women (the majority group in the British civil service), so findings may not generalize to other ethnic groups. The response rate for participation in this study was 55%. Although many of those who refused to take part cited difficulties with incorporating ambulatory blood pressure monitoring into their working day, the possibility that selection biases were operating cannot be ruled out. The socioeconomic gradient in body mass index that is evident in the full Whitehall study was not observed (27). Similarly, although a gradient in cigarette smoking was present, the absolute levels of smoking were less than those recorded previously, particularly in the low grade employment group. The study may therefore have attracted relatively fit lower status participants, underestimating differences between socioeconomic groups.
Fibrinogen has been identified as part of the biological pathway through which socioeconomic status and other psychosocial factors influence cardiovascular disease risk (53). These results indicate that although the association with socioeconomic status is not related to heightened acute stress responsivity, low control at work does stimulate short-term disturbances in fibrinogen regulation in men. Future research will assess whether these alterations in fibrinogen response predict later coronary heart disease.
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication October 9, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. H. Wirtz, L. S. Redwine, U. Ehlert, and R. von Kanel Independent Association Between Lower Level of Social Support and Higher Coagulation Activity Before and After Acute Psychosocial Stress Psychosom Med, January 1, 2009; 71(1): 30 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. H. Wirtz, L. S. Redwine, C. Baertschi, M. Spillmann, U. Ehlert, and R. von Kanel Coagulation Activity Before and After Acute Psychosocial Stress Increases With Age Psychosom Med, May 1, 2008; 70(4): 476 - 481. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Dimsdale Psychological Stress and Cardiovascular Disease J. Am. Coll. Cardiol., April 1, 2008; 51(13): 1237 - 1246. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M Rosvall, G Engstrom, L Janzon, G Berglund, and B Hedblad The role of low grade inflammation as measured by C-reactive protein levels in the explanation of socioeconomic differences in carotid atherosclerosis Eur J Public Health, August 1, 2007; 17(4): 340 - 347. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Albert, R. J. Glynn, J. Buring, and P. M Ridker Impact of Traditional and Novel Risk Factors on the Relationship Between Socioeconomic Status and Incident Cardiovascular Events Circulation, December 12, 2006; 114(24): 2619 - 2626. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Steptoe and M. Marmot Psychosocial, Hemostatic, and Inflammatory Correlates of Delayed Poststress Blood Pressure Recovery Psychosom Med, July 1, 2006; 68(4): 531 - 537. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Banks, M. Marmot, Z. Oldfield, and J. P. Smith Disease and disadvantage in the United States and in England. JAMA, May 3, 2006; 295(17): 2037 - 2045. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. B. Loucks, L. F. Berkman, T. L. Gruenewald, and T. E. Seeman Social Integration Is Associated With Fibrinogen Concentration in Elderly Men Psychosom Med, May 1, 2005; 67(3): 353 - 358. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Steptoe and M. Marmot Burden of Psychosocial Adversity and Vulnerability in Middle Age: Associations With Biobehavioral Risk Factors and Quality of Life Psychosom Med, November 1, 2003; 65(6): 1029 - 1037. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), and low (
) grade of employment groups. Error bars are SEM.