Impact of regular relaxation training on the cardiac autonomic nervous system of hospital cleaners and bank employees.

Impact of regular relaxation training on the cardiac autonomic nervous system of hospital cleaners and bank employees. Scand J Work En viron Health 1993;19:319-25. The work-related strain of 50 female hospital cleaners and 48 female bank employees was recorded during a period of rationalization in the workplace, and the effect of daily relaxation to help the workers cope was tested. The subjects were arranged into age-matched pairs and randomly allocated into intervention and reference groups. The intervention period lasted six months. The relaxation method was brief and easily introduced as an alternative break in the work place. Each training session lasted 15 min. A microcomputer-based system was used to record heart rate variability in response to quiet breathing, the Valsalva maneuver, deep breathing, and active or thostatic tests. Cardiac reflexes indicated that occupational strain (especially of a mental nature) caused the functioning of the autonomic nervous system to deteriorate. Regular deep relaxation normalized the function and improved the ability to cope.

Occupational stress and burnout have becom e serious probl ems in today ' s society. Psycho somatic symptomatology and early retirement because of disability are common, especially in predomin antly caring and service vocation s among women.
Female auxiliary work in general, and the work of hospital cleaners in particular, is physically demanding (1). In addition, there is a real discrepancy between physical capacity and the output expected of (aging) women in physical occupations; and this situation leads to increa sing symptoms, both of a physical and mental nature (2). Workin g on the lowest level of the occupational hierarchy also creates stress on cleaners (3).
In offices time pressure, precision demands, role conflicts, ergonomic shortcomings, and relationships with custom ers can be stressing (4). In the Finnish banking sector, female workers belong to the lowest personnel level (>90%), while men predominantly make up the managing class. The possibilities for better salarie s, in-bank training, and career advancement are considerably worse for women than for their male colleagues (5).
This study was conducted during a period of workplace rationalization due to serious economic circum-stances in Finland. An economic crisis was especially the case in Finn ish bank s and hospitals, both of which suffered extensive cutbacks, such as a decrease in acti vities, a par ing of expenses, and a reduction in staffing level s. The occupational future of workers became insecure, and fellow workers could be conceived as rivals. The work load of the continuously decreasing number of employee s at each workplace was increa sing . In addition, in bank offices, customers poured out their discontent with high-level banking polic y to the clerk s.
Constant strain in the absenc e of adequate strategies for coping is known to lead to disease (6,7). Daily stress, especially if self-esteem is low , can cause behavioral changes, acute or chronic psychological reactions, physiological responses such as aches and frequent flu, depressed immune function , decreased work ability, and an increase in work-related diseases (8)(9)(10).
Because there are many reports indicating that relaxation might lessen stress reactions (11-15), we have tested the use of an easy relaxation program as a method of copin g. The method was designed to be so easy that it could be considered as a possible alternative to coffee and smoking breaks in the workplace.
The purpose of the study was to measure the workrelated strain of hospital cleaners and bank employees during a stressful period in their work situation from recordings of heart rate variability in cardi ac autonomic nervous system function tests and to test the effect of daily relaxation as, a method of coping. To determine wheth er people can maintain a daily relaxation practi ce on their own , or whether regular support is needed, we left a portion of the subjects Scand J Work Enviro n Health 1993, vol 19, no 5 alon e with their exercises, while others were supported by guided meetings throughout the six months of the interve ntion period.

Subjec ts
Fifty hospital cleaners and 48 bank employees , all women, gave their informed consent to partic ipate in the study, which was approve d by the Ethical Committee of the University of Kuopi o. The recruitment of the subjects occ urred in a normal occupational meeting in both workplaces. The workers were informed about the project and its aims, methods, and intervention, after which the poten tial volunteers fill ed out enrollment forms. All of the willing 48 bank employees were acce pted. Two hospital cleaners were excluded becau se of heart problem s, and 10 were deleted by lottery in order to restrict the number of the hospital group to 50. According to the information provided on the enrollment form s, none of the 98 subjec ts had any major health problems.
Because the age range of the parti cipants was 23-60 years and because autonomic nervous system (ANS) functions are known to be age dependent (16), the subj ects were arra nged (separately within each occupation) into age-m atched pairs, one memb er of the pair being randomly appointed to the intervention group and the other entering the reference grou p.
All the referent s cont inued in the study to the end. Two memb ers of the hospital intervention group and one member of the bank intervention group withdrew for personal reas ons.

Intervention
The intervention period lasted six months. The interventi on method ( 17) was an easily learned relaxation exercise perform ed while sitting in an ordinary chair. Each training session lasted IS min and consisted of deep breathing exercises (2 min), forced tension and relaxation of the abdominal muscles (3 min), and deep relaxatio n as the subject focused on her gradually calming breath ing. (10 min). The subj ects were advis ed to repeat the trainin g daily.
We also wanted to determin e how easy it was to maint ain this kind of daily trainin g for a longer period, with or without regular support, and without any special personal interes t, and whether the difference in the training practices was reflected in the measured results. To support the motiv ation of the hospital intervention group, the relaxation program was performed in guided joint meetings organiz ed durin g workhours in the hospital three times a week throu ghout the entire interven tion period. On other days the subjec ts were encourage d to train alone at home. The bank intervent ion group was taught the same program in four training sess ions, after which 320 they were left alone with their exercises but urged to maintain them on a regular basis.
In addition, the subjec ts of the intervention groups were encouraged to lower their shoulders and to relax themselves whene ver they remembered, especially at night just before falling asleep. They were also asked to document their daily exercises. The reliabilit y of the reports was emph asized.

Measurement of cardiac autonomic nervous f unction
In the beginnin g there were measurements designed to familiarize the subjects with the techniqu es, and directly after them measurement I was performed. Measurement II followed after three month s, and measurement III after six month s.
Both the measurement protocol and the age-related reference value s for the card iovascular ANS function tests were taken from Finnish standards ( 18,19). The test conditions were kept peaceful and pleasant. The subjec ts were instructed not to have meals, drink coffee, tea or cola drinks, or smoke durin g the 4 h preceding the measurements, and to abstain from alcohol durin g the two preceding days. The phase of the menstrual cycle (or the menopausal status) of each subject was chec ked at every measurement.
A microcomputer-ba sed syste m (20) for noninvasive monitoring of heart rate interv als (RRI ) was used. The analyses of the recordin gs were performed automatically by the attached software (with the possibility of corr ection in the case of abnormal QRS complexes) . Three surface electrodes were used. One was placed under each clavicle and the third at the point of V5-V6 of standard chest leads. Electrocardiogra ms were recorded for 2 min durin g four different tests as follows ( 18): Test I. Quiet breathing with breathin g cycl es (expiration + inspiration) of 5 s in a sitting position . Standard deviations of the RRI (RRI-SD) were calculated. Test 2. Valsalva maneu ver: blowing for 15 s through a mouthpiece attached to a manometer and maintaining a pressure of 40 mm Hg (",5.32 kPa ) while sitting. The following two ratios were counted: (i) tachycardia ratio (TR) = shortest RRI durin g blowing : mean RRI during resting before blowing and (ii) Valsalva ratio (VR) = longest RRI after blowing : shortest RRI dur ing blowing.
Test 3. Deep breathing in a sitting position, deep breathing at a rate of six breaths per minute . The deep breathin g difference (DBD) was calc ulated as the mean of four success ive 10-s breathing cycles: DBD = maximum instantaneous heart rate -minimum instantaneo us heart rate. of the quiet breathing test and the Valsalva ratio, and the maximum:minimum ratios of the active orthostatic test. Deep breathing differences (DBD) were transformed as follows: y = DBD raised to the power of 0.7 (18). The statistical comparison of the changes in the group means (reference group versus intervention group) from measurement I to measurement III was done with a multivariate analysis of variance (MANOVA) of repeated measures (23), or the group means of the reference group and intervention groups were compared with the use of the t-test at measurement III (table 1). The paired t-test was also used to test the significance of the changes in the means of eaeh group (separately) from one measurement to the next. The influence of the menstrual cycle and menopause on the results of the ANS function tests were examined by a two-way analysis of variance.

Questioning of work factors
At measurement III, when the increasing occupational demands had become obvious due to the cutbacks, a small inquiry was made to obtain a better understanding of the work situation of the subjects.
The bank subjects were interviewed at measurement III about their work. The discussions dealt with current items, such as the relationship with customers, the atmosphere of the workplace, their occupational future in the bank, likes and dislikes, overtime work, feelings of tiredness, working post and status of each subject (from "low-level managers" or "better" posts in administrative and business consulting offices to "crowded low-level service on the ground floor").
The hospital subjects completed a small questionnaire concerning current feelings about their occupational future, likes and dislikes, the relationships between the cleaners, how the other hospital staff seemed to value cleaners and their work, positive and negative feedback, mental and physical strain of the work, and their perceived personal health.

Statistical analysis
The statistical examination of the measured values was based on the age-related Finnish reference values (in five-year intervals) developed by Piha (18,19). The theoretical "expected values" between the "normal" limits (figure I), two-sided 80% (0.10 and 0.90) tolerance limits from Piha's data, were the values to which the measured results were compared (18,19).
Increasing age attenuates the cardiac ANS function and narrows the intervals of reference values for most parameters (16,18,21,22). The same tendency was also seen in this study (figure I). This was the reason why the measured values of the subjects of different ages could not be counted together into group means. To avoid the weighting of the values of the younger subjects in the means, the primary measured data were transformed into standard deviations (SD) from the age-related "expected values" (figure I). The group means of table I were calculated from these SD results. Thus the smaller the group mean, the nearer the results were to the "expected values" (and the "better" the results). Correspondingly, the greater the group mean, the longer the average distance from the "expected values" (and the "worse" the results).
Because of the skewedness of the distribution, and because of the use of the reference values, logarithmic transformations were performed for the results   Since the questionnaire about work experiences was only examined at the end of the follow-up, these results were not included to the statistical calculations, but were kept merely as background information relating to occupational stress.

Results
The results of the ANS function tests are shown as group means in table I. The greater the group mean, the longer the general deviation of the measured values from the "expected values" and the "worse" the results. Consequently, the smaller the group means, the "better" the results.
Generally, the test results of the referents deviated further (-) from the "expected values" in the two fol-lOW-Up measurements. The change of the mean of the deep breathing difference in the hospital reference group from 0.52 to 0.75 between measurements I and III was statistically significant (P = 0.024), as was that of the tachycardia ratio of the bank referents from 0.57 to 0.86 (P = 0.027).
In the intervention groups the measured values changed generally towards (+) the "expected values" during the intervention. In the hospital staff group, the change in the group means of the quiet breathing test and the Valsalva ratio towards the "expected values" between measurements I and II was statistically significant, with P-values of 0,{)42 and 0.025, respectively, as was the change of the Valsalva ratio of the bank group (P = 0.022). The changes in the group means for the Valsalva and tachycardia ratios towards the "expected values" in 322 the hospital intervention group between measurements I and III were also significant (P =0.012 and P = 0.040, respectively), as was that of the Valsalva ratio in the bank intervention group (P = 0.007).
Comparison of the development of the group means (the reference groups verses the intervention groups) according to the MANOVA of repeated measures, or the comparison of the group means at measurement III by the t-test, is shown as P-values in table I. The different development of the means of the two hospital groups from measurements I to III is clear. For the bank subjects it was more modest. The results of measurement II represented levels between measurements I and III.
The importance of motivational support was evident. Nearly all of the members of the hospital intervention group participated in the joint meetings regularly during their workdays. They enjoyed their allowable extra breathing spaces in the middle of strenuous days. Their diary notes about their exercises were also regular and appeared to be reliable. According to them, the total training compliance of the hospital subjects was good throughout the sixmonth follow-up. Evidently, the very positive attitude of the cleaning department also influenced their motivation. For instance, all the time needed for participation in the project, including both the measurements and the continuous relaxation meetings of the intervention group, was included in the worktime of the subjects.
The bank subjects were also eager to begin the project and to learn the relaxation method. However, they were not given the regular training support. Thus, already after a couple of weeks, some of them found that they were no longer able to maintain regular training. The y felt that the "duty" to rela x was in itself stressful. In addition, only some of them were able to keep regular diaries.
According to the report s, so me of the bank subjects continued rather regular relaxation sessions for the first half of the intervention period , but not for the final half. They felt that the worsening economic situation in the banking business demanded their attenti on; they had "more important things" to do, for instance, acquiring more qualification s required for competition for the fewer vacancies. Only a few members of the bank inter vent ion group practiced relaxation training regularly throu ghout the six months. However, even their measured ANS results were modest, as were the result s of the few hospital subjects who preferred to train alone at home without the support meetings. Apparentl y it is not easy to maintain any extra daily behav ioral routine without regular support, and it also seem s that without regular guidance, people do not realiz e how deep relaxation can and should be performed. Deep relaxation is really something other than silent sitting in a chair.
Menstrual cycle and menopause had no significant influence on the test results of the cardiac ANS function.
Work factors were regarded in essentially the same way in both the referenc e and intervention groups. Most of the subj ects, both in the bank and in the hospital, were very worried about their future employment , and only a few were positive in this respect. In the hospital group the perceived strain varied according to the department, the direct supervisor, the attitudes of the other hospital staff, and human relations. Most of the bank subjec ts did not like their work as much as before, and they regarded overtime as particularly stressful. As expected, the extent of the perceived distress varied acco rding to the work post and its status.

Discussion
Both the bank employees and the hospital cleaners experienced psychophysiological distress in their occupations during this time of rationalization in the workplace. Mental stress seemed to influence the health of the subj ects when mea sured as heart rate variab ility in the cardiac ANS functio n tests. Workrelated strain increased during the follow-up (in the referents). However, the easily learned IS-min relaxation method employed in this study, when repeated regularly, was sufficiently effec tive to preserve the health of the subjec ts by normalizing cardiac autonomic functions. Continu ous motivational support seemed to be needed to optimize the effects of the training. Regular guided relaxation meetings and an encourag ing attitude on the part of management greatly strengthened the training motivation of the subjects.
Scand J Work Environ Health 1993, vol 19, no 5 The cardiac ANS function tests used in this study are widely accepted as reliable and reproducible measuring method s ( 18,24) . A combinat ion of tests provides comprehensive and reliable information about the function of the cardiac ANS (18,25). The effect of relaxation in this study was the most evident when tested by the Valsalva maneuver (Valsalva and tachycardia ratios) and deep breathing difference. The reliability of these tests is good in the sense that correct performance during recording s is easy to control. The quiet breathing test and the ortho static test were not as sensitive .
At the beginning of the study the cleaning department of the hospital had recentl y been designated as a self-reliant produ ctivity unit. Preparations were underway to allow for open competitive bidding with large and experienced commerci al cleaning companie s for the contract to clean the hospital in the future. Cutbacks in costs with resulting layoffs had begun, which alarmed the workers. Becau se the hospital subjects felt they needed any possibl e support, they welcomed the project warmly.
The study with the hospital subjects started in the autumn and finished in the spring. The measured values of the referent s deviated further from the "expected values" the longer the time from the previous summer and vacations. This finding was regarded as evidence of inc reased strain and tiredne ss, cumulated during the working winter, and also possibly influenced by the initiated saving actions and insecurity.
The detailed work-related stress factors of the hospital elea ners varied from department to department, socia l relationship s (26,27) and low self-esteem (3) being evidently the most influential factors. The wellor ill-being of the worker s seemed to be dep endent on this type of mental factor. Thu s impro ving the psychological aspects of the work environment might be a possible means with which to prom ote the psychoph ysiological fitness of workers, and their producti vity as well (27).
The intervention with the bank subjects began in the beginnin g of March and finished in the autumn. Initially a nationwide bank strike had ju st ended and the sense of solidarity among the empl oyees was great. But those feelings changed quickly to alarm during the spring (and the study). The following summer was the first durin g which temporary workers were not used to replace workers during vacatio ns. At the same time cutbacks and layoffs started in the bank. Many threatening rumors were circulating and no one knew what lay ahead in the autumn. Th is situation was reflected also in the test results. Despite the fact that summer and vacations had not been over very long at the end of the follo w-up in the autumn, most of the gro up means of the bank referents were at their worst in measurement III.
In all, it has been suggested that the financial savings demanded by current economic conditio ns are being carried out in worklife at the expense of hu-man health (28). Occupational distress which hampers health is a loan from human resources and must be paid back later. The most alarming aspect is that it seems to drag on and on .
The regular daily practice of relaxation normalized the cardiac ANS function of the subjects of the intervention groups in this study, especially in the hospital group. These results are in accordance with those of many other studies (11-15) and suggest that relaxation might be an effective means of lessening psychosomatic strain. Nevertheless, the "efficacy" of relaxation seems to demand determination and self-discipline (with positive attitudes) from the trainer, as well as practical support from management. Coherently organized short re laxation breaks in workplaces, such as the IS-min joint meetings for guided relaxation practice in this study, might mean an inexpensive, bu t proved effective means of enhancing productivity (28). No costly equipment or much time is needed (since there are coffee and smoking breaks anyway), and relaxation provides a return for the time invested in the form of better psychophysiological well-being among the employees (28 , 29) . The subjects of the intervention groups reported that , during the project, their psychosomatic sympto ms diminished and their morale improved (28 ).
Learning the skill of deep relaxation seems to require a long training period. The duration of the intervention period of the pre sent study was rather short. The results of the intervention group were better after six months of regular rela xation than after three months of training. Perhaps still better results would have been possible after a longer period of training, at least with those subjects who seemed to learn relaxation rather slowly. In all , it seems that relaxation should be repeated deeply and regularly over a rather long period to optimize its psychosomatic benefits. However, after the skill has been learned, it can be utilized anytime and anywhere within a few seconds.
Although it seems obvious that stress management programs effectively support the psychosomatic health of workers, there still remains one noteworthy issue. If we assume that occupational strain and related health problems orig inate (mainly or partly) from work factors (9,30,31) , the real improvements should be directed also towards these factors (30,31), at least as a long-term goal. Even if relaxation and the related methods may be effecti ve, their role should be seen primarily as a supportive health care measure. But even before any reconstruction of work as recommended by Kara sek & Theorell (3 1), relaxation and other stre ss management methods can offer, as such, rea sonable possibilities to support both the health of the workers and their productivity (28) . The follow-up period of our study was short, the groups were rather small, and the representativity of the sample was difficult to evaluate. The se aspects restr ict the possibilities to draw general conclusions.
However, we regard the results as promising. We believe that (mental) occupational stress deteriorates the cardiac ANS function (and health), and regular deep relaxation can normalize it. An encouraging attitude on behalf of management and practical support with training seem to be needed to optimize the relaxation benefits at work.