Psychosomatic Medicine Tips for Better Browsing
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zanstra, Y. J.
Right arrow Articles by Kooistra, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zanstra, Y. J.
Right arrow Articles by Kooistra, L.
Related Collections
Right arrow Therapeutic Interventions
Psychosomatic Medicine 68:583-590 (2006)
© 2006 American Psychosomatic Society

ORIGINAL ARTICLES

Vagal and Sympathetic Activity in Burnouts During a Mentally Demanding Workday

Ydwine J. Zanstra, Jan M. H. Schellekens, PhD, Cas Schaap, PhD and Libbe Kooistra, PhD

From the University of Aberdeen, College of Life Sciences and Medicine, School of Psychology, Aberdeen (Y.J.Z.); University of Groningen, Groningen, The Netherlands (J.M.H.S., C.S.); Department of Pediatrics, Behavioural Research Unit, Alberta Children’s Hospital, University of Calgary, Calgary, Alberta, Canada (L.K.).

Address correspondence and reprint requests to Jan M. H. Schellekens, Department of Psychology, University of Groningen, Grote Kruisstraat 2-1, 9712 TS Groningen, The Netherlands. E-mail: j.m.h.schellekens{at}ppsw.rug.nl


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Objective: We study differences in task performance and related sympathetic-vagal reaction patterns between burnouts and controls during a mentally demanding workday.

Method: Thirty-nine adults with burnout and 40 healthy controls performed mental tasks during a simulated workday. At pretest, just before lunch (lunch test) and at the end of the day (posttest), a Stroop color word task was administered as a probe task. Efficiency (the relation between performance and effort during the probe task), performance (reaction time and errors), and effort (self-report) were measured, as well as cardiovascular indices of sympathetic (blood pressure) and vagal (respiratory sinus arrhythmia) activity.

Results: Performance and effort investment of both burnouts and controls did not differ during pretest. As the day progressed the performance of controls improved more than the performance of burnouts. Moreover, the control group showed a decrease of blood pressure in response to mental task demands, a decrease in respiratory sinus arrhythmia activity, and no change in experienced effort. In the burnout group, no change could be demonstrated in blood pressure, suggesting a sympathetic predominance in the sympathetic-vagal balance. Burnouts experienced an increase in effort and were more tired at the end of the workday.

Conclusion: Burnouts and healthy controls differ in their pattern of sympathetic-vagal activity only after long-lasting work demands. Findings give limited support to Porges’s view that in healthy subjects, the vagal system is more responsive to challenging task situations than in chronically stressed individuals. The distinction between two phases in the burnout on the basis of behavioral and physiological characteristics is discussed.

Key Words: burnout • cardiovascular reactivity • autonomic control • blood pressure • heart rate variability • parasympathetic-vagal balance

Abbreviations: CAR = cortisol awakening rise; ECG = electrocardiogram; HF-HRV = high frequency band of heart rate variability; HPA = hypothalamic-pituitary-adrenal; HRV = heart rate variability; MBP = mean blood pressure (in mmHg); MI2 = squared modulation index; eln(MI2) = natural log of the squared modulation index; RSME = rating scale of mental effort; RT = reaction time; SAM = sympathetic adrenergic medullary system; SE = standard error; TPR = total peripheral resistance; UBOS = Utrechtse Burnout Scale; ANOVA = analysis of variance; IBI = interbeat interval.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Burnout is widely considered to be a state characterized by symptoms of mental and physical fatigue, detachment from work, and experienced incompetence (1). People with burnout typically complain about their lack of endurance in performing attention-demanding daily activities. While this has been attributed to dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis, research findings to date are inconclusive with regard to the underlying pathophysiological mechanisms (2). Accumulating evidence supports the contention that burnout serves as a risk factor for hypertension and cardiovascular disease (2–4). It is unknown which mechanisms may account for this relationship. Many studies on healthy subjects have been carried out examining the relationship between task demands and cardiovascular activity using blood pressure and heart rate as indices of invested effort (5,6). Activation of the sympathetic-adrenergic-medullary (SAM) axis can be assessed with cardiovascular measures. Positive associations have been established between job strain and chronic stress on the one hand and BP and HR on the other hand (5,6). DeVente et al. (2) provided evidence that SAM and HPA axes were disturbed among burnout patients during a brief laboratory session. The question addressed in the current study is whether individuals with a lack of endurance as a function of burnout differ in their cardiovascular activity in response to long lasting task demands. In this study, both task performance and cardiovascular activity were investigated in relation to mental task demands and time-on-task during a simulated workday. It was expected that the individuals with burnout would be more affected by the task demands per se and by the accumulated effects of the task demands over time than the healthy controls.

Changes in mental workload are typically reflected in cardiovascular adjustments of both sympathetic and parasympathetic origin. Increased blood pressure and heart rate, as well as additional release of adrenaline, are indices of the increased activation of the SAM system preparing an individual to deal with increased demands. The key determinants of arterial blood pressure are cardiac output (heart rate x stroke volume) and total peripheral resistance (TPR) (7). TPR and ventricular stroke volume are mainly determined by sympathetic activity (7); heart rate, by both sympathetic and parasympathetic activity. Because the contribution of heart rate to blood pressure is small, it is assumed that short-term blood pressure level is a good indicator of sympathetic activity. (A change of interbeat interval [IBI] from 925 to 525 ms as result of vagal blockade resulted in an increase of the mean blood pressure [MBP] of 3 mm Hg.) (8) Bunnell (9) states that cardiac output can increase only 5% to 10% under myocardial stimulation alone. In simulation studies of baroreflex regulation, these findings were confirmed (8). Vagal activity can be estimated by respiratory sinus arrhythmia (RSA) (10,11), which can be derived from heart rate variability (HRV), especially the high frequency band of heart rate variability (HF-HRV; 0.15–0.40 Hz) (10). Increased HRV is observed when subjects are relaxed and not engaged in mentally demanding tasks (12). On the other hand, decreased HRV is associated with attention demanding cognitive operations in which subjects are allocating mental effort (12). Compensatory effort can be derived from changes in the performance-effort balance (i.e., efficiency) while executing a repeatedly administered probe task (pre-/posttest paradigm) (13). Efficiency decreases if, for example, relatively more effort needs to be invested in order to produce the same quality and quantity of performance.

A variety of overlapping models has been put forward to account for the role of autonomic influences in shaping adaptive and maladaptive adjustment to effortful conditions. In the view of Berntson et al. (10), task demands may yield distinct patterns of autonomic activation in terms of sympathetic-parasympathetic balance. Both branches of the autonomic nervous system can respond from strongly related to completely unrelated (e.g., in patterns of reciprocal, coactive, or uncoupled sympathetic and parasympathetic responsivity). Berntson et al. (10) emphasize the importance of the use of a bivariate model of autonomic control because specific patterns of autonomic control can be more closely linked with behavioral states and health outcomes. Uchino et al. (14) found that the sympathetic branch of the autonomic nervous system was related to stress-induced activation of the HPA (e.g., cortisol changes) and to changes in immune response to stress. High HPA activation is generally found in individuals with a lack of control and with ineffective coping (15). According to Porges (11,16), in normal healthy subjects, a challenging stimulus produces a phasic decrease in parasympathetic (vagal) activity. In chronically stressed individuals, vagal activity is tonically suppressed and sympathetic control of the nervous system predominates (11,17). This would be a sign of poor homeostasis and increased neurophysiologic vulnerability to the detrimental effects of stress and high task demands (11,16). Low HF-HRV has been shown to be related to coronary artery disease (18) and increased mortality (19). In the current study, therefore, it is expected that people with burnout perform equally well as controls, but at the cost of increased mental effort to compensate for their lack of endurance. A phasic decrease in vagal activity is expected in individuals who perceive the task demands as a challenge and are still engaged in task-related activities. Also, it may be expected, following Porges (11), that in individuals with burnout, vagal activity is tonically suppressed or will be suppressed sooner as a consequence of increasing fatigue.

In addition to performance and physiological indices of effort, Wierwille and Eggemeier (20) recommended the use of subjective measurements of mental workload. Zijlstra (21) suggested that subjective experience affects behavior and thus performance and physiological responses. If individuals feel tired, they will likely behave as being tired and, accordingly, adopt strategies congruent with their state (e.g., changing the speed-accuracy tradeoff); also, their willingness to spend time and energy will be affected. The experience of fatigue is usually examined by means of rating scales or questionnaires (20). The current study used the Rating Scale of Mental Effort (RSME) (21) to examine the subjects’ feelings of fatigue and effort investment.

The present study examined the effectiveness and efficiency of mental task performance in adults with burnout versus healthy controls during a simulated workday (22). It was predicted that over the course of the day, the state necessary for optimal task performance would become less optimal, as indicated by a decline in efficiency and/or effectiveness of task performance. These effects were expected to be most pronounced in the subjects with burnout as this syndrome is characterized by exhaustion and lack of endurance (1).


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Subjects
Thirty-nine patients (21 women, 18 men; mean age = 46 years; range = 30–59 years) on sick leave and clinically diagnosed with burnout syndrome took part in this study from September to December 2003. Patients participated on a voluntary basis and were recruited at the university outpatient clinic for cognitive behavior therapy. Patients’ burnout diagnosis was based on the outcome of two clinical interviews using the DSM-IV-TR criteria for undifferentiated somatoform disorder (23) and the ICD-10 criteria for neurasthenia (24). Inclusion criteria were the presence of stress symptoms (mental and physical fatigue, loss of energy, concentration problems); the complaints had to be work related; the patients’ sick leave had to be burnout related; scores on the Utrechtse Burnout Scale (UBOS), the Dutch version of the Maslach Burnout Inventory (25,26), had to be high (>2.2) on the exhaustion scale and the distance scale and/or low (<3.51) on the competence scale. The mean item scores of the burnout patients on the exhaustion scale were 4.4 (standard error [SE] = 0.1); on the distance scale, 3.7 (SE = 0.1); and on the competence scale, 3.5 (SE = 0.1). All items were scored on a 7-point rating scale, ranging from 0 "never" to 6 "every day." Based on the outcome of the Composite International Diagnostic Interview Short Form (CIDI-SF), subjects with primary axis I disorders (e.g., major depressive episode, bipolar disorder, anxiety disorder, dysthymia) were excluded from the study. Additional exclusion criteria were use of prescription medication during the 4 weeks preceding the study (except oral contraceptives); cardiovascular complaints (e.g., hypertension); a history of immune disease, diabetic or other medical conditions associated with enduring complaints of fatigue.

The subjects with burnout syndrome were compared with a group of 40 healthy controls (27 women, 13 men; mean age = 40 years; range = 19–62 years). These subjects were all regularly employed in jobs of at least 16 hours a week. Control subjects were recruited by means of advertisements in local newspapers. Inclusion criteria for the control subjects were low (<0.81) scores on the exhaustion scale of the UBOS, low (<0.49) scores on the distance scale, and high (>5.0) scores on the competence scale. Exclusion criteria were equal to that of burnouts. The controls were paid for attending the laboratory day.

All participants read and signed an informed consent statement approved by the Medical Ethical Committee of the Faculty of Medicine (University of Groningen).

The Stroop Color Word Task
Subjects performed a computerized version of the Stroop color word task (27). The names of single colors appeared on the computer screen either in the same or in a different color code. A match between word meaning and print color required a yes response by pressing a button; a mismatch between word meaning and print color required a no response by pressing a different button. A wrong choice was defined as an error; a response given after presentation of the next stimulus was defined as an omission. The subjects responded with the index finger of their preferred hand. Task demands were manipulated by varying stimulus presentation rate. To ensure that subjects would perform at their maximum capacity, stimulus presentation rates were individualized based on the subjects’ performance obtained in a 15- to 20-minute training session, which took place at the start of the day. In this session, increasing the event rate with 10 events every 2 minutes systematically increased task difficulty. Each subject started from an event rate of 70 trials per 2 minutes and continued until the task became too difficult and a predefined error criterion of 6 errors was exceeded. The corresponding event rate was defined as the subject’s maximum (i.e., 100%) Stroop capacity was used to define the six levels of task difficulty for the remaining Stroop tasks; i.e., a 40%, 60%, 80%, 100%, 120%, and 140% task load. These levels remained unchanged throughout the day.

Mean reaction time (RT), percentage of errors, and number of omissions were calculated for the six time blocks of the Stroop, each lasting 2 minutes.

RSME
The RSME (21) is an unidimensional scale very similar to the CR-10 Ratings of Perceived Exertion scale designed by Borg (28). Levels of experienced effort are indicated by a cross on a vertical line. The scale consists of 16 numbered categories ranging from 0 to 150 and 9 verbal cues ranging from "not at all effortful"(= 0) to "extremely effortful"(=100) and "absolute maximum"(=150). The scale is scored by measuring the distance between zero and the cross on the vertical line (in mm). Reliability and validity of the scale are satisfactory and comparable to that of the CR-10 (21,28).

Physiological Variables
Task performance data and physiological data were simultaneously recorded during each Stroop task. It is important to note that the physiologic variables related to task conditions were all expressed as task-rest (blood pressure, respiration frequency) or rest-task differences (IBI, high frequency band of the HRV). In addition, the rest (baseline) values of the physiological variables will also be referred to. (For a more detailed description of physiological apparatus, see Schellekens et al. (22).)

IBI and HF-HRV
Heart rate was recorded using standard electrocardiogram (ECG) equipment. Electrode placement and data acquisition were carried out according to the procedures described by Mulder (29). R-wave occurrence times were determined by a hardware QRS detector with an accuracy of 1 ms. IBI is the interval between two successive R-peaks. Spectral analysis on the heart rate data was performed with the CARSPAN spectral analysis program (29). The spectral power index of heart rate was expressed in terms of the squared modulation index (MI2). This means that all variations were expressed as deviations relative to the mean value of the time series in the measured period. This procedure helps to minimize age and sex effects of heart rate interference on power spectra (29). The spectral power in the frequency band ranging from 0.15 to 0.40 Hz was defined as the HF-HRV. In order to correct for the skewness of the distribution, the MI2 was transformed to logarithmic values (natural log).

Blood Pressure
Blood pressure data were obtained using the FIN-A-PRES (Ohmeda 2300) (30). A finger cuff placed around the middle finger of the nonpreferred hand allowed continuous registration of blood pressure without disrupting task performance. Blood pressure was sampled at 1000 Hz. As preliminary analyses indicated that results for diastolic, systolic, and MBP were similar, only the findings for MBP (mm Hg) are presented. MBP was defined as the mean of the blood pressure values between two successive R-peaks (29).

Respiration Frequency
Respiration was measured using a respiration belt attached around the chest at the level of the xyphoid process. The resistance of this belt changed as a function of chest volume, which reflects respiration. Only the frequency of respiration (expressed in Hertz) was measured. Sample rate was 1000 Hz.

Procedure
In order to simulate the accumulating fatigue effects of an ordinary work day, individual subjects were invited to the laboratory, where they spent the whole day (8:30 AM to 5:00 PM) performing cognitively demanding activities, such as filling out questionnaires, paper-and-pencil tests, and computerized attention and memory tasks. (See Schellekens et al. (22). for more details.) It is important to note that these tasks were used to create tiredness and fatigue; performance results on these tasks were not included in the analyses. Similar tasks were used in the morning (9:15–12:30 AM) and the afternoon (1:15–4:30 PM).

Fatigue effects were examined using a standard Stroop task as a probe task. The Stroop task was administered at the start of the day (pretest), at the end of the morning period before lunch (lunch test), and at the end of the day (posttest). Repeated administration of the probe task allowed us to evaluate how efficiency and effectiveness were affected over time. The efficiency of probe task performance was based on the relationship between task performance (i.e., RT, percentage of errors and omissions) and invested effort, i.e., changes in sympathetic-parasympathetic balance indicated by blood pressure and HF-HRV, and subjective experience of fatigue and effort as measured by the RSME. The effectiveness of probe task performance was derived from RT and error rate.

Before the actual start of the "working day," subjects were asked to complete several questionnaires on sleep quality, breakfast, and recent travel experiences. This was followed by a training session on the Stroop, which lasted 15 to 20 minutes. Next, the ECG electrodes, finger cuff, and respiration belt were attached. At 9:15 AM, the pretest Stroop was administered; at 12:00 AM, the lunch test, and at 4:15 PM, the posttest. Each administration of the Stroop task consisted of a 5-minute rest period (baseline) followed by six 2-minute blocks of increasing task load. (See Stroop task.) During the administration of the Stroop, heart rate, respiration, and blood pressure were continuously recorded. After each of the three Stroop tasks, subjects rated their effort on the RSME-scale.

During the lunch hour (12:30 PM to 1:15 PM), subjects had a light meal and could relax. Halfway through the morning and afternoon sessions, there was a 15-minute break.

Statistical Analysis
Cardiovascular measures recorded during the rest preceding the Stroop tasks were used as baseline data. A 3 (time of day) x 2 (groups) repeated-measures analysis of variance (ANOVA) was used for testing the cardiovascular baseline data (rest values of MBP, IBI, and HF-HRV). Age was used as covariate in all analyses, and respiration frequency (baseline) was used as an additional covariate in the HF-HRV analyses.

Reactivity in MBP, respiration, IBI, and HF-HRV (0.15–0.40 Hz) was defined as the difference between rest and task values. A 3 (time of day) x 6 (task loads) x 2 (group) repeated-measures ANOVA was conducted on each of the abovementioned dependent variables (performance data: RT, percentage of errors, number of omissions; cardiovascular reactivity data: MBP, IBI, HF-HRV). In these analyses, time of day (pretest, lunch test, posttest) and task load (i.e., 40%, 60%, 80%, 100%, 120%, 140%) were within factors and group (burnout, control) was the between factor. The above covariates were used in these analyses. If the ANOVA showed significant effects, further contrast analyses were used to assess which effects were responsible for the overall significance (31). Significance was set at p ≤ .05. The Greenhouse-Geisser correction was applied when the sphericity assumption was violated.

The RSME data were analyzed using a 2 (groups) x 3 (time of day) repeated-measures ANOVA, with group as the between factor and time of day as the within factor; age was used as a covariate.

Age was controlled for because groups differed in mean age, and age significantly correlated with the Stroop task variables and physiological data. Since no main or interaction effects for gender were obtained, no correction for gender was applied.

When using HF-HRV as an index of vagal activity, it is important to consider the possibility of confounding influences from potential respiratory changes (32). To correct for such confounding effects, respiratory frequency was used as a covariate in the analyses. This is considered to be a conservative correction procedure because it carries a risk of affecting some of the true experimental effects.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 NOTES
 REFERENCES
 
Performance Data and Self-reported Effort
RT
Load manipulation had a significant effect on RT (Load: F(1.8,137.0) = 21.5, p < .001): each increase of load decreased the RT significantly, except the change from 120% to 140%. (See Figure 1.) RTs of the burnouts did not significantly change during the day. In contrast, however, the RTs of the controls decreased significantly in the lunch test and posttest compared with the pretest (Time of day x Group: F(2.0,152.0) = 26.8, p < .001). This decrease in RT of the controls was most pronounced in the lower-load conditions (Time of day x Load x Group: F(4.4,330.6) = 5.05, p < .001).


Figure 111
View larger version (17K):
[in this window]
[in a new window]
  		Figure 1. Mean reaction time (and SE) in the pretest (left), the lunch test (middle), and the posttest (right). (Diamonds for results of burnout group; crosses for controls.)

 

Percentage of errors
The percentage of errors increased as a function of increasing task demands (Load: F(2.5,186.5) = 17.56, p < .001). The controls performed better than the burnouts, especially in the higher task load conditions (Load x Group: F(2.5,186.5) = 2.86, p = .047). The controls had a lower percentage of errors than the burnouts only during the lunch test and posttest, not during the pretest (Time of day x Group: F(1.4,102.0) = 4.00, p = .035). (See Figure 2.)


Figure 211
View larger version (18K):
[in this window]
[in a new window]
  		Figure 2. Percentage of errors (and SE) in the pretest (left), the lunch test (middle), and the posttest (right). (Diamonds for results of burnout group; crosses for controls.)

 

Omissions
There was a significant main effect for load (Load: F(1.4,108.2) = 7.69, p = .003), indicating that participants overall made more omissions as a function of increasing load (Figure 3). Compared with the controls, the burnouts showed more omissions (Group: F (1, 76) = 7.04, p = .01), particularly in the high-load conditions (Load x Group: F(1.4,108.2) = 6.31, p = .007). Additional contrast analyses showed that the differences between burnouts and controls were most pronounced for the 100% load and the 140% load. The between-group difference in the amount of omissions was relatively small in the pretest but increased substantially in the lunch and posttest (Time of day x Group: F(1.8,135.3) = 5.73, p = .006).


Figure 311
View larger version (13K):
[in this window]
[in a new window]
  		Figure 3. Mean number (and SE) of omissions in the pretest (left), the lunch test (middle), and the posttest (right). (Diamonds for results of burnout group; crosses for controls.)

 

RSME
The overall scores on the RSME were significantly higher for burnouts (mean = 66.8, SE = 3.4) than for controls (mean = 50.1, SE = 3.4) (Group: F(1,74) = 11.5, p ≤ .001). However, in the pretest burnouts and controls did not differ (both mean = 55, SE = 4.1). After the pretest, the RSME-scores of lunch test and posttest increased in burnouts (lunch: mean = 71, SE = 4.2; posttest: mean = 74, SE = 4.1) and RSME scores decreased in controls (lunch: mean = 50, SE = 4.3; posttest: mean = 45, SE = 4.1) (Group x Time of day: F(2,148) = 11.3, p ≤ .001).

In sum, slower RTs and more errors and omissions in the burnouts indicated that their performance levels were substantially lower than those of the healthy controls. These between-group differences increased as the day progressed. This was mainly due to the improved performance of the controls in terms of shorter RTs and reduced percentage of errors. Group differences with regard to omissions were mainly due to a performance decline in the burnout group. Overall, these effects were more pronounced in the higher task loads, especially in the lunch and posttest assessments. Self-reported effort increased in the burnout group and decreased in the control group as a function of time of day.

Physiological Data
MBP
There were no significant between-group differences and within-subject differences in MBP rest values (Figure 4). However, MBP during the Stroop task decreased significantly as the day progressed (Time of day: F(2.0,154.0) = 4.38, p = .014). A main effect for task load indicated that MBP increased with increasing task load (Load: F(2.4,188.0) = 122.2, p < .001). These load and time-of-day effects were different between groups as indicated by a significant time of day x load x group interaction (F(5.1,396.5) = 2.75, p = .018). (See Figure 4.) Additional contrast analyses showed that in the control group the increase of MBP as a consequence of load manipulation was significantly smaller in both the lunch- and the posttest compared with the pretest (especially a decrease in load levels at 60% and 100% to 140%). In the burnout group, the increase in MBP as a consequence of load manipulation did not change significantly as a function of time of day.


Figure 411
View larger version (23K):
[in this window]
[in a new window]
  		Figure 4. Mean blood pressure (task-rest and rest) in the pretest (left), the lunch test (middle), and the post- test (right). (Diamonds for results of burnout group; crosses for controls.)

 

HF-HRV
No significant differences in the HF-HRV were found in the rest values of the pretest, the lunch test or the posttest (Figure 5).


Figure 511
View larger version (16K):
[in this window]
[in a new window]
  		Figure 5. High-frequency band of heart-rate variability (rest-task) in the pretest (left), the lunch test (middle), and the posttest (right). The higher the rest-task difference, the more suppression of HF-HRV (to interpret as less vagal activity) in comparison with the rest. (Diamonds for results of burnout group; crosses for controls.)

 

Reactivity was indicated by the rest-task values of HF-HRV: the higher the rest-task difference, the more suppression of HF-HRV in comparison with the rest. After controlling for the effect of respiration frequency, only an interaction between time of day and group was obtained (Time of day x Group: F(2.0,143.8) = 3.65, p = .029). Additional contrasts showed that the pretest versus posttest x group interaction contributed mostly to the time of day x group effect; the pretest versus lunch test x group was not significant (p = .067). In the control subjects, HF-HRV was more suppressed during both the lunch test (mean = 0.513 natural log of the squared modulation index (eln(MI2)]; SE = 0.082 eln(MI2)) and the posttest (mean = 0.515 eln(MI2); SE = 0.081 eln(MI2)) compared with the pretest (mean = 0.418 eln(MI2); SE = 0.081 eln(MI2)). In the burnout patients, the HF-HRV was less suppressed in the posttest (mean = 0.385 eln(MI2); SE = 0.078 eln(MI2)) than in both the pretest (mean = 0.537 eln(MI2); SE = 0.078 eln(MI2)) and the lunch test (mean = 0.498 eln(MI2); SE = 0.078 eln(MI2)). No further significant main effects or interactions were found.

IBI
Significant differences in IBI were found in the rest values of the pretest, the lunch test, and the posttest (Time of day: F(2.0,152.0) = 7.47, p < .001). Further analyses showed that this general effect is explained by an increase of the IBI in the lunch test compared with pretest and posttest. (See Figure 6.)


Figure 611
View larger version (21K):
[in this window]
[in a new window]
  		Figure 6. Interbeat interval of heart activity (rest-task and rest) in the pretest (left), the lunch test, and the posttest (right). (Diamonds for results of burnout group; crosses for controls.)

 

Reactivity was indicated by the rest-task IBI values: the higher the rest-task difference, the more heart rate is increased (shorter IBIs) during task performance in comparison with the rest. With increasing task load, the IBI became shorter (Load: F(2.2, 165.4) = 4.14, p = .015). This load effect was affected by time of day (Time of day x Load: F(7.3, 555.9) = 2.14, p = .035). From additional contrast analyses, it seemed that this load effect was significantly stronger in the lunch test than in the pre- and posttest. Furthermore, groups differed with regard to the time of day x load effect as indicated by the significant three-way interaction time of day x load x group (F(7.3, 555.9) = 2.10, p = .04). The time of day x load effect was completely due to the control group. The IBIs at the lunch test in the controls compared with the burnouts were significantly shorter in all task loads in comparison with the pretest and posttest.

Respiratory Frequency
No significant differences in the rest values of the pretest, the lunch test, and the posttest of respiratory frequency were obtained.

Reactivity was indicated by rest-task respiratory values: the higher the rest-task difference, the more respiratory frequency increased compared with the rest value. Only a group x load effect was obtained (Group x Load: F(3.7,283,6) = 3,63, p = .008). From additional contrast analyses, it seemed that the between-group differences were the highest at highest load levels (120% and 140%). No further significant effects were found. (See Table 1.)


View this table:
[in this window]
[in a new window]
  		TABLE 1. Respiratory Frequency (rest-task) in the Pretest (left), the Lunch Test, and the Posttest (right) (means and standard errors)

 

In sum, the MBP, the respiratory frequency, and the HF-HRV obtained in the rest periods did not change over time and did not differ between burnouts and controls. Rest values of IBI were higher during the lunch test than during the pre- and posttest in both groups. In the control group, the blood pressure and the IBI across task loads decreased after the pretest. In the burnout group, changes in blood pressure and IBI as a function of task load were small over time and not significant. HF-HRV was more diminished after the pretest session in the controls, whereas the suppression became less (particularly in posttest compared with the earlier test sessions) in the burnout group. No time-of-day-related effects could be detected in respiratory frequency.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
NOTES
 REFERENCES
 
At the start of the simulated working day, burnouts and controls did not differ with regard to either their performance data or their physiologic data. As the day progressed, Stroop task performance of the controls improved: RT, percentage of errors, and number of omissions all decreased. Burnouts, in contrast, did not show these improvements. Cardiovascular data showed no differences in baseline-rest values between burnouts and controls. In contrast, across task loads, controls showed a substantial decrease in blood pressure and HF-HRV as the day progressed. In the burnout patients, changes in blood pressure and heart rate across task loads as a function of time of day were small and insignificant; their HRV was significantly increased in the posttest. The experienced effort data (RMSE) showed that the Stroop task performance in both the lunch and the posttest required more effort from the burnouts than the controls.

Our finding that the burnout group showed a substantial performance decline over time is notable because in the majority of studies, such a decline, albeit in healthy normals, has not been found (12,33). Meijman (12) emphasized that performance rarely deteriorates as a consequence of fatigue, as long as individuals are willing to compensate by investing more effort (i.e., compensatory effort). Wilkinson (33) was among the first to provide empirical support for the validity of the compensatory effort concept. He showed that the effects of 32 to 56 hours of sleep deprivation were less marked in compensating subjects, as indicated by high increments in forearm electromyography during a vigilance task. Interestingly, though, the burnouts in the current study did not show signs of extra invested effort for maintaining their task performance, as indicated by blood pressure and HF-HRV. Instead, it seemed that the burnout patients demonstrated more and more "giving-up behavior" after the pretest, especially in the highest task loads. This is surprising, given the fact that our use of individualized task loads was intended to create a situation in which performance levels could be increased without pushing the individual beyond his/her processing limitations.

The reduction of blood pressure and HF-HRV during task performance indicated that the controls reduced their sympathetic reactivity after the pretest and switched to more vagal reactivity. Burnouts, in contrast, continued to respond with the initial level of sympathetic reactivity despite the decline in their task performance. This change in sympathetic-vagal balance can be demonstrated by depicting both systems as one regression line in a bivariate plane (10) in which the sympathetic system is reflected by Z-transformed MBP data (x-axis) and the vagal system by Z-transformed HF-HRV data (y-axis). In the burnout group, the slope of the regression line was 0.8 in the pretest and decreased to 0.7 in the posttest, suggesting a change toward more reactivity in the blood pressure component. In the control group, the slope changed from 0.8 in the pretest to 1.2 in the posttest, suggesting a relative decrease in the blood pressure component. This lack of vagal involvement in the burnout group would be in accordance with Porges (11,16), who suggested that burnouts compared with healthy controls have a vagal system that is less responsive to demanding conditions. Melamed et al. (34) demonstrated an association between burnout and cardiovascular disease, whereas others have suggested that low responsiveness in HRV is associated with coronary artery disease (18) and increased mortality (19). One may, therefore, speculate that in our burnout patients low responsiveness in HRV put them at risk for cardiac disease. Even though the current findings were obtained with experimentally designed stressors, the results may be generalized to daily stressors (14).

The physiologic reaction pattern of the burnout group in the current study resembles the characteristics of sympathetic overtraining (35) (36). Overtraining is defined as an increase in training volume or intensity of exercise with inadequate recovery periods between workouts, resulting in long-term performance decrements (36). Iellamo et al. (37) described that the effect of intensive training indicates a shift from a parasympathetic toward a sympathetic predominance. Similarly, Sluiter et al. (36) distinguished between sympathetic and parasympathetic overtraining. In their view, sympathetic overtraining is characterized by increased sympathetic activity, whereas parasympathetic overtraining shows decreased sympathetic activity and parasympathetic predominance (36). Importantly, according to Sluiter et al. (36), sympathetic overtraining develops before the parasympathetic syndrome. Hackney and Viru (38) found that a greater catecholamine elevation during the day, for example, as a consequence of intensive exercise, could result in lower nighttime circulating catecholamine levels, which would, in turn, reduce ACTH and thus cortisol. The CAR was especially affected.

It is generally assumed that burnout and overtraining share a common etiology and symptom course. Melamed et al. (34) proposed a two-phase process in burnout. The first phase ("tense burnout") shares characteristics with sympathetic overtraining. The second and more chronic phase ("listless burnout") is similar to parasympathetic overtraining. In the early tense stage of burnout, individuals may employ active and direct coping strategies to enhance and protect their resources. These individuals are described as overcommitted and high in need for control (39). Steptoe et al. (40) found a positive correlation between overcommitment level and blood pressure over the working day; the CAR was also affected. Hanson et al. (17) found that lower HF-HRV levels were related to high need for control. In our pretest and lunch test results, blood pressure baseline levels of burnouts were higher and HF-HRV baseline levels were lower than in controls. These results, although not significant, are in line with the results of Steptoe et al. (40) and Hanson et al. (17), suggesting a tense burnout phase. In the more advanced stage of burnout, however, when burnout appears to be coupled with listlessness and apathy, indirect and inactive coping behavior prevails, paralleled by overexcitation of the HPA-axis (41). It is, thus, very likely that the different stages of burnout are associated with different physiological characteristics in terms of activation of the SAM- and HPA-axes. The physiologic reaction pattern of the early stages of burnout (i.e., predominance of sympathetic responsiveness, reduced nighttime cortisol levels) may be associated with a typical pattern of overcommitment and need for control; the later stages of burnout (i.e., a parasympathetic predominance, elevated nighttime cortisol) may be associated more with depression. The time course of these "burnout phases" is unknown. Therefore, distinguishing more precisely the several phases in the burnout process may produce more conclusive results in the future regarding burnout and SAM- and HPA-axis activity.

To conclude, our present findings suggest that the combined effects of work intensity and work duration accumulate over the day. These effects can most clearly be detected by studying the relationship between performance and effort investment. In the current study, these effects were more pronounced in the burnout group than in the control group, not at the start in the pretest but only as function of time of day. Moreover, both groups demonstrated distinct modes of autonomic responsivity to task and time-of-day demands. Compared with the controls, the burnout group did not shift to more vagal suppression. This suggested a sympathetic predominance in the sympathetic-vagal balance. Our results are in accordance with Porges’ view about the responsiveness of the vagal system to task demands in chronically stressed individuals. Finally, results indicate that an approach in which physiologic reactivity is considered in terms of a bivariate model of autonomic control (10) may prove very helpful in identifying both the nature and the stage of burnout and may thus constitute a promising approach in evaluating treatment in burnout patients.

We thank Dr. Bonny Kaplan and Dr. David Shapiro for critically reviewing the text of this paper.


   NOTES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
REFERENCES
 

Received for publication November 10, 2004; revision received February 6, 2006.

DOI:10.1097/01.psy.0000228012.38884.49


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 

  1. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Ann Rev Psychol 2001;52:397–422.[CrossRef][Medline]
  2. DeVente W, Olff M, van Amsterdam JGC, Kamphuis JH, Emmelkamp PMG. Physiological differences between burnout patients and healthy controls: blood pressure, heart rate, and cortisol responses. Occup Environ Med 2003;60(suppl I):i54–61.[Abstract/Free Full Text]
  3. Melamed S, Ugarten U, Shirom A. Chronic burnout, somatic arousal and elevated salivary cortisol levels. J Psychosom Res 1999;46:591–7.[CrossRef][Medline]
  4. Shirom A, Westman M, Shamai O, Carel RS. Effects of work overload and burnout on cholesterol and triglycerides levels. J Occup Health Psychol 1997;2:275–88.[CrossRef][Medline]
  5. Steptoe A, Cropley M, Joekes K. Job strain, blood pressure, and responsivity to uncontrollable stress. J Hypertens 1999;17:193–200.[Medline]
  6. Tsutsumi A, Kayaba K, Tsutsumi K. Association between job strain and prevalence of hypertension. Occup Environ Med 2001;58:367–73.[Abstract/Free Full Text]
  7. Mohrman DE. Cardiovascular Physiology. 5th ed. London: McGraw-Hill; 2003.
  8. van Roon AM, Mulder LJM, Althaus M, Mulder G. Introducing a baroreflex model for studying cardiovascular effects of mental workload. Psychophysiology 2004;41:961–81.[CrossRef][Medline]
  9. Bunnell DE. Non-invasive measurement of sympathetic influences on the heart. In: Orlebeke JF, Mulder G, van Doornen LJP, eds. The Psychophysiology of Cardiovascular Control. New York: Plenum Press; 1985:221–35.
  10. Berntson GG, Cacioppo JT, Fieldstone A. Illusions, arithmetic, and the bidirectional modulation of vagal control. Biol Psychol 1996;44:1–17.[CrossRef][Medline]
  11. Porges SW. Cardiac vagal tone: a physiological index of stress. Neurosci Biobehav Rev 1995;19:225–33.[CrossRef][Medline]
  12. Meijman TF. Mental fatigue and the efficiency of information processing in relation to work times. Int J Industrial Ergonom 1997;20:31–42.
  13. Veldman JBP, Mulder LJM, van Roon AM, van der Veen FM, Mulder G. Test measurements are a powerful tool in determining cardiovascular effects of long lasting mental work. J Psychophysiol 1998;12:338–52.
  14. Uchino BN, Cacioppo JT, Malarkey W, Glaser R. Individual differences in cardiac sympathetic control predict endocrine and immune responses to acute psychological stress. J Pers Soc Psychol 1995;69:736–43.[CrossRef][Medline]
  15. Olff M, Brosschot JF, Godaert G, Benschop RJ, Ballieux RE, Heijnen CJ, de Smet MBM, Ursin H. Modulatory effects of defense and coping on stress-induced changes in endocrine and immune parameters. Int J Behav Med 1995;2:85–103.[CrossRef][Medline]
  16. Porges SW. The polyvagal theory: phylogenetic substrates of a social nervous system. Int J Psychophysiol 2001;42:123–46.[CrossRef][Medline]
  17. Hanson EKS, Godaert GLR, Maas CJM, Meijman TF. Vagal cardiac control throughout the day. Biol Psychol 2001;56:23–44.[CrossRef][Medline]
  18. Martin GJ, Magid NM, Myers G. Heart rate variability and sudden death secondary to coronary artery disease during ambulatory electrocardiographic monitoring. Am J Cardiol 1987;60:86–9.[CrossRef][Medline]
  19. Kleiger RE, Miller JP, Bigger JT. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 1987;59:256–62.[CrossRef][Medline]
  20. Wierwille WW, Eggemeier FT. Recommendation for mental workload measurement in a test and evaluation environment. Hum Factors 1993;35:263–81.[Medline]
  21. Zijlstra FRH. Efficiency in Work Behaviour. Thesis. Delft: University Press; 1993.
  22. Schellekens JMH, Sijtsma GJ, Vegter E, Meijman TF. Immediate and delayed after-effects of mentally demanding work hours. Biol Psychol 2000;53:37–56.[CrossRef][Medline]
  23. Sperry L. Handbook of Diagnosis and Treatment of DSM-IV-TR Personality Disorders. 2nd ed. New York, NY: Brunner-Routledge; 2003.
  24. World Health Organization. Multiaxial classification of child and adolescent psychiatric disorders: the ICD-10 classification of mental and behavioural disorders in children and adolescents. Cambridge University Press; 1996.
  25. Schaufeli W, van Dierendonck D. Utrechtse Burnout Schaal (Manual). Lisse: Swets & Zeitlinger BV; 2000.
  26. Schaufeli WB, Leiter MP, Maslach C, Jackson SE. The Maslach Burnout Inventory-General Survey. In: Maslach C, Jackson SE, Leiter MP, eds. Maslach Burnout Inventory Manual. 3rd ed. Palo Alto, CA: Consulting Psychologists Press; 1996:19–26.
  27. MacLeod CM. Half a century of research on the Stroop effect: an integrative review. Psychol Bull 1991;109:163–204.[CrossRef][Medline]
  28. Borg GAV. Borg’s Perceived Exertion and Pain Scales. Champaign, IL: Human Kinetics; 1998.
  29. Mulder LJM. Assessment of Cardiovascular Reactivity by Means of Spectral Analysis. Thesis. Groningen; 1988.
  30. Wesseling KH. Finapres, continuous noninvasive finger arterial pressure based on the method of Peñáz. In: Rüddel H, Curio I, eds. Non-invasive Continuous Blood Pressure Measurement. Frankfurt-am-Main: Verlag Peter Lang GmbH; 1991:9–17.
  31. Stevens JP. Applied Multivariate Statistics for the Social Sciences. 4th ed. Mahwah, NJ: Erlbaum; 2003.
  32. Grossman P, Karemaker J, Wieling W. Prediction of tonic parasympathetic cardiac control using respiratory sinus arrhythmia: the need for respiratory control. Psychophysiology 1991;28:201–16.[Medline]
  33. Wilkinson RT. Muscle tension during mental work under sleep deprivation. J Exp Psychol 1962;64:565–71.[CrossRef][Medline]
  34. Melamed S, Kushnir T, Shirom A. Burnout and risk factors for cardiovascular diseases. Behav Med 1992;18:53–60.[Medline]
  35. Lehmann MJ, Lormes W, Opitz-Gress A, Steinacker JM, Netzer N, Foster C, Gastmann U. Training and overtraining: an overview and experimental results in endurance sports. J Sports Med Phys Fitness 1997;37:7–17.[Medline]
  36. Sluiter JK, Frings-Dresen MHW, Meijman TF, van der Beek AJ. Reactivity and recovery from different types of work measured by catecholamines and cortisol: a systematic literature overview. Occup Environ Med 2000;57:298–315.[Abstract/Free Full Text]
  37. Iellamo F, Legramante JM, Pigozzi F, Spataro A, Norbiato G, Lucini D, Pagani M. Conversion from vagal to sympathetic predominance with strenuous training in high-performance world class athletes. Circulation 2002;105:2719–24.[Abstract/Free Full Text]
  38. Hackney AC, Viru A. Twenty-four-hour cortisol response to multiple daily exercise sessions of moderate and high intensity. Clin Physiol 1999;19:178–82.[CrossRef][Medline]
  39. van Vegchel N, de Jonge J, Bosma H, Schaufeli W. Reviewing the effort-reward imbalance model: drawing up the balance of 45 empirical studies. Soc Sci Med 2005;60:1117–31.[CrossRef][Medline]
  40. Steptoe A, Siegrist J, Kirschbaum C, Marmot M. Effort-reward imbalance, overcommitment, and measures of cortisol and blood pressure over the working day. Psychosom Med 2004;66:323–9.[Abstract/Free Full Text]
  41. Schulz P, Kirschbaum C, Pruessner JC, Hellhammer DH. Increased free cortisol secretion after awaking in chronically stressed individuals due to work overload. Stress Med 1998;14:91–7.[CrossRef]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zanstra, Y. J.
Right arrow Articles by Kooistra, L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zanstra, Y. J.
Right arrow Articles by Kooistra, L.
Related Collections
Right arrow Therapeutic Interventions

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS