Content uploaded by Jarek Krajewski
Author content
All content in this area was uploaded by Jarek Krajewski on Feb 02, 2018
Content may be subject to copyright.
Relaxation-induced cortisol changes within lunch
breaks – an experimental longitudinal worksite
field study
Jarek Krajewski
1
*, Martin Sauerland
2
and Rainer Wieland
3
1
Experimental Work Psychology, University of Wuppertal, Wuppertal, Germany
2
Work and Social Psychology, University of Landau, Landau, Germany
3
Work and Organizational Psychology, University of Wuppertal, Wuppertal,
Germany
The aim of the worksite study presented here is to elucidate the cortisol reducing
impact of different ways of spending lunch breaks. With the help of the so-called silent
room cabin concept it was possible to induce a relaxation opportunity that provides
visual and territorial privacy. In order to evaluate its proposed effects, 14 call centre
agents were distributed to either 20 min progressive muscle relaxation (PMR) or small
talk break group. Participants were analysed in a controlled trial for a period of
6 months (1 day each month with five daily measurements at awakening, awakening
þ30 min, start of lunch break, end of lunch break, and bedtime) using saliva cortisol
measurements as a stress indicator. Results indicated that only the PMR break reduced
awakening, lunchtime, and bedtime cortisol response. Although further intervention
research is required, our results suggest that post-lunch PMR may sustainably reduce
participants’ cortisol states in real worksite settings.
The hypothalamic–pituitary–adrenal (HPA) axis is primarily activated when the body
responds to physical and mental stress (Miller & O’Callaghan, 2002); it is responsible for
the secretion of the stress hormone cortisol. It is hypothesized that prolonged activation
of this axis can suppress certain immune functions, can be detrimental to health and
increase the risk of disease (Rabin, 2005) or of faster disease progression (Sephton,
Sapolsky, Kraemer, & Spiegel, 2000). Furthermore, changes in the cortisol rhythm are
discussed in the aetiology of a number of diseases including heart disease, atopic
neurodermitis, and osteoporosis (Manelli & Giustina, 2000).
Numerous studies have indicated that the HPA axis may be specifically activated
under conditions of social-evaluative threat of the sort that appear daily in
professional worksite settings. Since work-related stress has been linked to a wide
* Correspondence should be addressed to Dr Jarek Krajewski, Experimental Business Psychology, Schumpeter School of
Business and Economics, University of Wuppertal (Germany), Gaussstraße 20, 42097 Wuppertal, Germany
(e-mail: krajewsk@uni-wuppertal.de).
JOOP 1285—4/1/2010—SANKARS—357577
The
British
Psychological
Society
1
Journal of Occupational and Organizational Psychology (2010), 00, 1–14
q2010 The British Psychological Society
www.bpsjournals.co.uk
DOI:10.1348/096317910X485458
spectrum of negative health outcomes and impaired well-being (e.g. Wegge,
Van Dick, Fisher, Wecking, & Moltzen, 2006), the development of effective worksite
stress countermeasures is a central task for applied stress research. A promising
approach for stress related interventions is the effective restoration of spent
resources within recovery sections (e.g. Binnewies, Sonnentag, & Mojza, in press;
Fritz & Sonnentag, 2005; Sonnentag, Binnewies, & Mojza, 2008; Trougakos, Beal,
Green, & Weiss, 2008). However, little attention has been paid to the longest and
thus probably the most influential, of all breaks within the workday: the lunch
break. Thus, optimizing the recovery impact of lunch breaks may be a promising
path for solving problems of high stress and the resulting impact on performance,
health, and quality of life.
Recent research on relaxation techniques has indirectly provided some ideas for
developing recovery intensive lunch break routines. One of the most widely used
techniques – progressive muscle relaxation (PMR) – has been shown to be an effective
countermeasure in reducing stress in experiments conducted in laboratory and clinical
settings (e.g. Schneider et al., 2005). PMR aims at enabling subjects to achieve physical
and mental relaxation using exercises to tense and release 16 different muscle groups
(legs, arms, shoulders, face, chest, etc.). The commonly used subform, abbreviated
progressive relaxation training, is derived from Jacobsen’s original PMR and routinely
used, both clinically and in research (Carlson & Hoyle, 1993).
Laboratory settings have produced many well-documented recovery effects of PMR
on the cardiovascular, neuromuscular, electrodermal, autonomous, and central nervous
systems. Furthermore, PMR shows effects on a wide range of psychosomatic disorders
(e.g. high blood pressure, sleep disturbance, asthma, rheumatic complaints, atopical
neurodermitis – see e.g. Rainforth et al., 2007), as well as on psychological variables
such as increased positive moods and physical well-being (Lohaus, Klein-Heßling,
Vo¨gele, & Kuhn-Hennighausen, 2001). Moreover, it also increases pain thresholds and
decreases inner tension and stress (Emery, France, Harris, Norman, & Vanarsdalen,
2008; Lolak, Connors, Sheridan, & Wise, 2008; Shapiro & Lehrer, 1980). As demonstrated
by Pawlow and Jones (2002, 2005), and Nickel et al. (2005) PMR even reduces the
endocrinological stress marker cortisol.
A theoretical framework explaining such stress-reducing effects of PMR lunch breaks
is provided by the cognitive-behavioural model of relaxation (Smith, 1988; Smith,
Amutio, Anderson, & Aria, 1996). This model suggests that focusing (the ability to
maintain focus on simple stimuli), tension relief (positive sensations associated with
reduced cognitive and somatic arousal), and passive disengagement (the ability to stop
unnecessary goal directed and analytic activity; cf. Sonnentag, Mojza, Binnewies, &
Scholl, 2008) are the basic components of all forms of effective relaxation. Similarly,
Meijman and Mulder (1998) have determined that effective recovery is enhanced by low
task-related physical, mental, and emotional demands and low external stressor
frequency and intensity. Moreover, drawing on literature from Trenberth and Dewe
(2002), break time should guarantee distraction from work-related ruminative thought,
for instance by strongly focusing on involving tasks (Cropley & Purvis, 2003).
Furthermore, physical distance from the workplace and the resulting detachment are
relevant to effective recovery processes in non-working time (Hartig, Johansson, &
Kylin, 2007). Additionally, as demonstrated by Fastenmeier, Gstalter, and Lehnig (2003),
obligatory activities seem to have a reduced recovery value. Nearly, all the theoretical
and empirical claims made for recovery intensive lunch break routines can be taken into
account with PMR lunch breaks.
2Jarek Krajewski et al.
JOOP 1285—4/1/2010—SANKARS—357577
Previous empirical research has independently demonstrated the stress-reducing
short-term effects of PMR. However, few attempts have been made to capture the
stress-reducing effects of PMR for more than a few hours. Furthermore, studies have so
far only shown stress reducing effects of PMR in artificial experimental or clinical
settings. To the best of our knowledge, no study has been conducted to replicate
laboratory findings within the context of real worksites, adding such information as
chronic strain indicators (cortisol awakening responses, CARs), involving employees
(instead of students) as participants, and testing the long-term stability of the proposed
recovery effects over half a year in a real life setting. To enable this kind of research,
sustainable organizational solutions for implementing PMR into daily worksite routines
had to be offered. The major difficulties associated with incorporating PMR were solved
by the infrastructural framework of the silent room.
Infrastructural framework for implementing relaxation techniques into worksite
settings
As shown above, numerous studies have documented the recovery potential of
systematic relaxation techniques in non-worksite research fields. In fact, implementing
these procedures into organizational contexts faces several problems. One serious
problem is related to the general security and privacy needs of deep relaxation, which
requires eye closure and a horizontal lying position. Another problem refers to the
professional setting in which relaxation activities take place. To satisfy these needs a
relaxation setting should ensure visual, auditory, and territorial privacy. In order to fulfil
the described demands in a worksite setting a room-in-room concept, called ‘silent
room’ has been developed. The core features of this conception consist of lockable
cabins and medical daybeds (see Figure 1). The silent room is an intimacy maintaining
and stressor-free place that protects privacy by noise subdued and opaque cabins
coupled with a hygienic dental-medical appearance. Moreover, a flexible acoustic
4
1.81.8
2
3
2
3
41
55
1
2.5
Figure 1. Inner view of the silent room cabin module (length and width are displayed in metres).
1, medical daybed; 2, music system with PMR instruction; 3, eye mask, alarm clock; 4, noise subdued
floor; 5, lockable cabin door.
Relaxation-induced cortisol changes 3
JOOP 1285—4/1/2010—SANKARS—357577
system offering standardized PMR instructions is integrated into the room. By installing
the silent room in a call centre, it was possible for us to integrate PMR into daily worksite
lunch break routines.
The scarcity of experimental evidence in this area of worksite implementation of
PMR breaks highlights the need for more detailed research. Hence, the focus of the
present study is to analyse the immediate (þ10 min), spillover effects (þ10 h; see e.g.
Pawlow & Jones, 2005), and the ‘quasi-chronic’ changes of cortisol over a longer period
(þ42 h; chronic stress marker CAR; Nickel et al., 2005) due to PMR lunch breaks within
daily worksite settings. Thus, it is hypothesized that PMR breaks reduce cortisol states
more efficiently than the usual small talk (ST) breaks.
Method
Participants
All participants (14 call centre agents) took part voluntarily. Due to the study’s focus on
measuring cortisol in typical call centre employees, the following inclusion criteria had
to be met: (a) had worked as call centre agents for more than 6 months, with a regular 5
day and 40 h workweek and a fixed daily work schedule from 8:00 to 17:00; and (b) had
no prior experience in systematic depth relaxation procedures. Of the 14 participants,
seven age-and-gender matched pairs were created (range ^4 years). Each member of
a pair was randomly assigned to either an ST or PMR group. Both groups consisted of
4 male and 3 female participants.
Procedure
After identifying lunch breaks as important recovery occasions within a pre-survey
interested parties were invited to an informative meeting. Here, employees were asked
to participate in an experimental study. All subjects were screened in a short personal
interview in order to assure that they corresponded to our criteria of selection.
Participants were informed about the purpose of the study. After that, participants
received a short summary with important information including a ‘don’t’ list (e.g. not to
drink any alcoholic beverages 6 h prior to sampling time) before taking part in our study.
These instructions were given in order to reduce the influence of possible confounders
related to cortisol measurement. We received written informed consent from all
participants. In return for participation, reports about individual stress profiles, as well
as the overall findings, were promised. Participating employees were not compensated
for their services. In a first appointment, the procedure and the use of the material
(collection devices, compliance monitor, diary, and questionnaires) were explained to
them and practiced intensively.
At the beginning of a 6 months period (from September to April), 14 call centre
agents were randomly allocated to experimental lunch break groups: either (a) 20 min of
PMR, or (b) 20 min ST break. The lunch break was scheduled between 12:00 and 13:00.
The experimental PMR session (12:15–12:45) of the break took place (for the PMR
group) in a noise-subdued, dimly-lighted (10 lux), opaque, lockable cabin, called ‘silent
room’, where participants wore eye masks. PMR instructions were given via wireless
headphones (including calm instrumental background music) while participants lay on
medical daybeds. The ST break was located in the company’s staff room and conducted
with three to four self-chosen colleagues. Instructions were given to follow the usual
4Jarek Krajewski et al.
JOOP 1285—4/1/2010—SANKARS—357577
choice of ST topics. Since cortisol measurement can be influenced by food intake no
lunch was consumed on measurement days during the break. However, small snacks
were allowed when participants were back at their desk (13:05–14:00). On the other
hand, snacks were consumed between 12:00 and 12:30 on non-measurement days
within the 6 months period.
Participants were instructed to use their allocated break every workday over a period
of 6 months. Exceptionally, there were no PMR sessions on the day before the
measurements to exclude direct spill over effects to the following day and determine
‘quasi-chronic’ changes of cortisol (see Figure 2). Cortisol measurements were taken
on 8 fixed days (d0¼20:25 months, d1¼þ0:25 months, d2¼þ1:25 months,
d3¼þ2:25 months, d4¼þ3:25 months, d5¼þ4:25 months, d6¼þ5:25 months,
d7¼þ6:25 months). The pre-measurement d
0
(ST for both groups) served as baseline.
On each measurement day samples of cortisol were (self-) administered at awakening
and awakening þ30 min then 11:55, 13:05, and bedtime. Altogether, this procedure
resulted in a Treatment (between-subject factor: ST, PMRÞ£Time (within-subject factor:
t1¼awakening, t2¼awakening þ30 min, t3¼11 : 55, t4¼13 : 05, t5¼bedtimeÞ£
Measurement Day (within-subject factor: d
0
,d
1
,…,d
7
) design.
Measurements
To collect saliva samples, participants used a device called a ‘salivette’ (Sarstedt;
Nu
¨mbrecht, Germany). Participants woke up as usual (at times ranging from 6:00 to
7:15) roused by their own alarm clocks. Participants were instructed to provide five
samples over the course of a normal workday. The appropriate time for the saliva
collection (especially for the post-lunch time 13:05 measurement) was chosen with
regard to the delayed response of cortisol to acute stress. Previous studies had shown
that salivary cortisol reaches its maximum about 10–20 min after acute stress reaction
(Hammerfald et al., 2006).
LT
LTRWT
Time of day
10:00 12:0008:00 22:0020:0018:0016:0014:00
06:00
RWT
ST
RWT
RWT
t
5
t
4
t
2
t
3
t
1
Process of experiment [months]
42
ST
d
0
d
4
d
3
d
2
d
7
d
6
d
5
60 531
d
1
P
P
Figure 2. Time schedule of the experiment. (RWT, regular work task; P, PMR break; ST, small talk
break; LT, leisure time; t
1
–t
5
, time of cortisol measurement; d
0
–d
7
, measurement day).
Relaxation-induced cortisol changes 5
JOOP 1285—4/1/2010—SANKARS—357577
In order to avoid interference with experimental effects, participants were
instructed to refrain from eating, smoking, strenuous physical exercise, brushing teeth,
and consuming acidic drinks or caffeine for at least 1 h prior to testing, and from
consuming alcohol for at least 6 h prior to testing. None of the samples had to be
excluded due to non-compliant behaviour.
Regular saliva sampling by participants is prone to measurement error due to a lack
of compliance in taking the sample at the prescribed time. To monitor compliance with
the salivary cortisol collection protocol, we removed the sampling swabs from their
original plastic tubes and put them in an electronic drug exposure monitor (Smart Caps,
eDEMTM; Aardex Ltd., Switzerland). This monitor recorded the time at which the box
containing the cotton rolls was opened. The employment of this device strengthened
the compliance of the participants and prevented invalid cortisol profiles in
non-compliant participants (Kudielka, Broderick, & Kirschbaum, 2003). Sufficient
compliance was defined as taking the saliva sample (i.e. opening the cap of the
compliance monitor) within a time frame of 5 min before and after the prescribed time.
Applying this criterion, the compliance rate was 98.6%, which is in accordance with the
data from Kudielka et al. (2003). Salivary free cortisol concentrations were determined
employing a chemiluminescence assay with high sensitivity and inter-assay and
intra-assay variations ,10%(University of Du
¨sseldorf, Germany).
Cortisol parameter
The CAR in saliva is increasingly regarded as a non-invasive and reliable method for
detecting subtle changes in the HPA axis. It allows repeated assessment and has been
shown to have a high intra-individual stability (Hellhammer et al., 2007). Furthermore,
an enhanced CAR has been found in healthy subjects under chronic stress either
manifested as high workload or social stress (Wu
¨st, Federenko, Hellhammer, &
Kirschbaum, 2000). CAR
delta
was determined by calculating the difference between
awakening þ30 min and awakening, and CAR
mean
by the mean of the two awakening
samples. Furthermore, lunch break cortisol (LBC) reductions indicating immediate
effects of PMR and ST were calculated as the difference of 13 :05 211 :55 cortisol
values. Spillover effects of PMR and ST were determined by collecting bedtime cortisol
values. In sum, the participants provided 543 cortisol samples (97%; 560 total samples,
14 participants £8days£5 samples per day). The missing data of a specific sample (e.g.
participant 6, t
5
,d
3
) were replaced by the total mean of the corresponding time of day
and measurement day over all remaining participants (mean of all participants in t
5
,d
3
).
Manipulation check
In addition to the application of Smart Caps, we checked the compliance and quality of
relaxation method realization by means of (a) a checklist of relaxation symptoms
(for PMR breaks), which involved, e.g. questions about the feeling of heaviness in the
16 different muscle groups; (b) informal questioning by a neutral person (was not
associated with the study); and (c) a masked observation sample (by a peer colleague)
on the measurement days. Non-compliant behaviour (non-adherence to the lunch
break-mode) could be extrapolated from the above-mentioned questioning and
observation. Furthermore, reporting less than 50% of the relaxation symptoms served as
an exclusion criterion. None of the participants fell below this criterion. Moreover, the
results of the informal questioning showed that the average percentage of PMR breaks
6Jarek Krajewski et al.
JOOP 1285—4/1/2010—SANKARS—357577
on the measurement days was 100% and the average number of PMR breaks during
6-month measurement period was 3.6 per week (SD ¼0:2). The major reasons for not
conducting a PMR break on non-measurement days were private obligations (banking,
administrative tasks), social obligations (solving within-group conflicts, emotional
support for colleagues), and falling asleep while practising PMR (under 20% of the
whole PMR break trials, but none on measurement days). Due to the very high number
and accurate collection of samples, high compliance with the sampling protocol, a high
level of general compliance could be assumed.
Statistical analysis
Before performing analyses, the raw cortisol values were positively skewed and
log-transformed to approximate normal distributions. However, in order to be
physiologically meaningful, Figure 3 shows absolute cortisol values. One three-way
repeated measure analysis of variance (ANOVA) was carried out to examine the
interaction effects of treatment (PMR, ST), time (t
3
,t
4
), and measurement day (d
0
–d
7
)on
pre- and post-lunchtime cortisol. Moreover, three ANOVAs were computed to determine
the interaction effects of treatment (PMR, ST) and measurement day (d
0
–d
7
) on the
remaining cortisol parameters. To control for possible effects of age, sex, intake of
contraceptive medication, wake-up times, and sleep duration, these variables were
10
5
0
40
30
20
10
0
30
20
10
0
–5
–10
d
0
10
8
6
4
2
0
d
1
d
2
d
3
d
4
Day
LBC [nmol/l]CAR mean [nmol/l]
BED [nmol/l]CAR delta [nmol/l]
Day
d
5
d
6
d
7
d
0
d
1
d
2
d
3
d
4
d
5
d
6
d
7
d
0
d
1
d
2
d
3
d
4
d
5
d
6
d
7
d
0
d
1
d
2
d
3
d
4
d
5
d
6
d
7
Day Day
Figure 3. Cortisol parameter (LBC, BED, CAR
mean
, and CAR
delta
) of the pre- (d
0
) and seven post-
measurement days (d
1
–d
7
) for PMR and ST break. Data are shown as mean ^SD. PMR break, triangle
marker; ST break, circle marker.
Relaxation-induced cortisol changes 7
JOOP 1285—4/1/2010—SANKARS—357577
included as covariates in the ANOVAs. Moreover, unpaired ttests were used to examine
short- (d
1
) and long-term (mean of d
6
and d
7
) differences between ST and PMR groups.
A significance level of p#:05 was used.
Furthermore, we conducted an a priori power analysis for repeated measure ANOVA
to determine the statistical power. Based on an estimated medium effect size, an alpha
level of .05, and a sample size of seven subjects per cell (total sample size:
8£5£2£7¼560), it was computed that the power exceeds the necessary 80% for
the significance tests. The statistical analyses were conducted with SPSS for Windows
release 17 (SPSS, Inc., Chicago, Illinois, USA).
Results
Sample characteristics and preliminary analyses
Table 1 shows the main characteristics of the sample separately for the PMR and ST
groups. The groups were statistically indistinguishable with regard to several cortisol
influencing variables (age, gender, ethnicity, use of contraceptives, awakening time, and
sleep duration), and cortisol sample values of the pre-measurement day. These findings
support the interpretation of treatment group differences as being caused by the
experimental factor treatment.
Cortisol indicators
Effects on post-lunch break cortisol
The LBC effects (cortisol level at 13:05 minus cortisol level at 11:55, LBC) revealed
significant different time courses for the PMR and ST groups as depicted in Figure 3.
The results obtained from a Treatment (between-subject factor: ST, PMRÞ£Time
(within-subject factor: t3¼11 : 55, t4¼13 : 05Þ£Measurement Day (within-subject
factor: d
0
,d
1
,…,d
7
) threee-way interaction effect, Fð7;96Þ¼2:56, p,:05, indicate the
Table 1. Sample characteristics (N¼14)
PMR (N¼7) ST (N¼7) p
Age [years] 34.71 (7.43) 42.00 (9.51) ns
Female [%] 57.1 57.1 ns
BMI [kg/m
2
] 24.26 (1.71) 21.90 (2.10) ns
Caucasian ethnicity 100 100 ns
Contraceptives [%] 42.8 42.8 ns
Awakening time [h] 6.40 (0.24) 6.31 (0.29) ns
Sleep duration [h] 8.17 (0.56) 7.95 (0.34) ns
Cortisol t
1
[nmol/l] 16.53 (5.21) 15.80 (5.46) ns
Cortisol t
2
[nmol/l] 23.83 (7.00) 24.16 (7.28) ns
Cortisol t
3
[nmol/l] 7.14 (3.66) 5.63 (3.70) ns
Cortisol t
4
[nmol/l] 8.29 (4.15) 7.74 (4.19) ns
Cortisol t
5
[nmol/l] 6.01 (2.16) 4.70 (1.53) ns
Q5 CAR
delta
[nmol/l] 7.30 (3.34) 8.361 (5.18) ns
CAR
mean
[nmol/l] 0.18 (5.94) 9.97 (5.89) ns
LBC [nmol/l] 1.14 (1.84) 2.11 (2.47) ns
Note. Cortisol t
1
, awakening; t
2
, awakening þ30 min; t
3
, 11:55; t
4
, 13:05; t
5
, bedtime.
8Jarek Krajewski et al.
JOOP 1285—4/1/2010—SANKARS—357577
divergent pre- and post-lunchtime cortisol values of PMR and ST group. LBC effects can
be observed for both short-term (d
1
) and long-term (mean of d
6
and d
7
) perspective,
t(12) ¼25.92, p,.001, t(12) ¼24.97, p,.001, respectively.
Effects on bedtime cortisol
A two-way ANOVA (2 Treatment £8 Measurement Days) showed systematic differences
in bedtime cortisol levels between the PMR and ST groups, Fð7;96Þ¼1:99, p,:10
(see Figure 3). The bedtime cortisol effects (BED) reached significance within both
short-term and long-term perspective, tð12Þ¼23:38, p,:01, tð12Þ¼23:21, p,:01,
respectively.
Effects on CAR
The CAR
mean
revealed no significant difference in time course for the PMR and ST
groups, as depicted in Figure 3, Fð7;96Þ¼1:15, p.:10. Accordingly, no short-term or
long-term effects were found, tð12Þ¼0:22, ns, tð12Þ¼21:44, ns, respectively.
In contrast to this result, the CAR
delta
values showed – referring to the significance
criterion of the two-way ANOVA (2 Treatment £8 Measurement Day) treatment-by-
measurement day interaction effect – a substantial effect, Fð7;96Þ¼3:73, p,:001
(see Figure 3). In contrast to the pattern observed above, CAR
delta
values reached
significance not from short term but from long-term perspective, tð12Þ¼0:29, ns,
tð12Þ¼24:34, p,:001.
Discussion
The aim of the 6-month experimental worksite study presented here is to elucidate the
cortisol reducing impact of different ways of spending lunch breaks. We expected that
PMR lunch breaks would elicit smaller cortisol responses than ST lunch breaks, and the
results do indeed document the cortisol decreasing effect of PMR lunch breaks.
The main finding apparent in the data is the strong reduction of lunchtime and
awakening cortisol states in response to the PMR break. Cortisol states at bedtime seem
to be less influenced by the chosen type of lunch breaks. The highest cortisol reduction
and largest effects were found for immediate post-lunch break cortisol states.
This corresponds to laboratory-based results concerning immediate cortisol reduction
due to PMR stress reduction (Pawlow & Jones, 2005). A theoretical framework
explaining such stress reducing effects of PMR has already been provided (see Meijman
& Mulder, 1998; Smith, 1988; Trenberth & Dewe, 2002). In contrast to the lunchtime
and bedtime effects, a reduced CAR
delta
can not be observed in the short run (after 0.25
months), only in the long run (after 5–6 months). According to Fries, Dettenborn, and
Kirschbaum (2009), who link CAR to prospective and anticipated demands of an
upcoming day, we assume that chronic stress as measured by CAR
delta
was not reduced
immediately since attitudes towards anticipated workload of the upcoming day are
unlikely to change in short term. This result corresponds with the resisting inertia of
CAR
delta
, which reflect rather long-term psychophysiological processes, and is
associated with chronic stress states (Thorn, Hucklebridge, Evans, & Clow, 2006).
Besides, the above-mentioned potential recovery effects of PMR, the specific
characteristics of the silent room-based worksite implementation selected for this
experiment could be responsible for the results. Based on long-term implementation,
Relaxation-induced cortisol changes 9
JOOP 1285—4/1/2010—SANKARS—357577
the study at the same time provided familiarization with PMR procedures and its silent
room setting for participants. These factors might have enabled them to experience
deep and effective relaxation during the lunch break and thus explain the effect sizes
obtained by the PMR-based breaks. In contrast, the ST break might have shown
increased cortisol levels due to interpersonal and situational characteristics such as
social-evaluative threats, unpredictability, uncontrollability, and the anticipation of
negative consequences. Freely choosing a preferred small talk partner may, on the other
hand, have diminished these effects and led to the observed nearly normal circadian
rhythm as expected from the literature.
In general, our results correspond to the hypothesis made at the beginning. Similar
results concerning the cortisol reducing effects of PMR have been found in artificial
laboratory contexts. However, this is to the best of our knowledge the first report on a
longitudinal implementation of systematic relaxation techniques in a real work setting
with daily lunch break routines.
One limitation of this study refers to the cortisol-based approach that was selected.
It is well-known that cortisol measurement faces several confounder-related threats with
reference to validity. However, cortisol levels throughout the experiment fell within the
expected range for normal, healthy adults showing a normal circadian rhythm, as
expected from the literature (see Westermann, Demir, & Herbst, 2004). Furthermore,
the observed compliance (derived from Smart Caps and self-report measures) confirms
the reliability of the measurements. Nevertheless, caution is warranted in the
interpretation of these data. Hence, future research might attempt to use other
non-obtrusive (electro-physiological, acoustic, or behavioural) stress measures.
In general, methodological difficulties may only disturb slightly the realization of the
experimental PMR break or manipulate its conditions. Participants’ compliance can be
considered high. This suggestion is supported by the fact that ST breaks serve as the
most common and natural form of lunch break. Participants’ compliance in the PMR
break condition was confirmed by random observations and informal questioning at the
end of the experiment. Nevertheless, it may be a matter of debate whether the observed
cortisol reducing effect resulted from placebo effects, characteristics of the ‘silent room’
(e.g. silence, darkness), from short periods of napping during PMR or from pure PMR.
Napping itself has already proved its effectiveness in industrial settings (Takahashi,
Nakata, Haratani, Ogawa, & Arito, 2004). Furthermore, imitation of the PMR break might
have occurred in the leisure time of the ST group. But informal interviews gave no hint
of this, and even if this imitation had occurred, the real difference between bedtime and
awakening cortisol of ST and PMR would have been underestimated. A further
uncertainty refers to the explanation BED and CAR
delta
results, which might be induced
by mediator effects of changed activity pattern (see Geurts & Sonnentag, 2006) or
irritation level at home rather than directly influenced by PMR breaks. Moreover,
Hawthorne effects could be responsible for the results. Even, if the latter might be less
probable due to the long experimental period of 6 months and the fact that cortisol (in
contrast to performance levels) cannot be enhanced by voluntary effort. Nevertheless,
due to habituation effects it remains unclear to what extent we can extrapolate the
6 months effects on to real long-term effects (,6 months). In addition, when extending
to a longer period seasonal effects might become more obvious. Again, this potential
confounder would only change the absolute cortisol level; the relative distance between
ST and PMR should remain stable.
Although in comparison to large-scale cross-sectional correlation designs the sample
size in this study is quite small, it is still within the typical range of experimental
10 Jarek Krajewski et al.
JOOP 1285—4/1/2010—SANKARS—357577
worksite field studies (see Takahashi et al., 2004). Moreover, repeated measurements
ensured the robustness, internal validity, and significance of the results. Furthermore, no
group differences in cortisol influencing variables or cortisol sample values of the
pre-measurement day could be found, which confirms the a priori equivalence and
baseline-specific superiority of PMR experimental groups.
The present study was carried out in a real, but small-sized worksite. In order to
judge external validity properly, it is evident that clarification concerning the ability to
generalize the results is needed. The extent to which we can extrapolate from this
call-centre context to other professional sectors remains unclear. With the limiting
factors described above in mind, our present findings should be viewed as preliminary
ones that warrant more controlled research. Some possible starting points and questions
for future research might be concerned with improving measurement instruments for
intervention studies (What physiological, behavioural, and acoustic instruments can
detect stress non-obtrusively and without interrupting the primary work task?;
Krajewski, Batliner & Golz, 2009) or optimizing the recovery value of worksite lunch
breaks (Which combination of different break usages as, e.g. napping or
pharmaceuticals improves the intensity and sustainability of the recovery process
best?; Wesensten, Killgore, & Balkin, 2005).
In sum, the longitudinal field-experiment results indicate that PMR lunch breaks may
reduce cortisol states significantly. Additionally, the current study extends prior research
by addressing cortisol reduction problems in real work settings. Finally, this study builds
evidence suggesting the acceptance, sustainability, and success of the silent room as an
implementation module enabling PMR within daily lunch break routines.
References
Binnewies, C., Sonnentag, S., & Mojza, E. J. (in press). Recovery during the weekend and
fluctuations in weekly job performance: A week-level study examining intra-individual
relationships. Journal of Occupational and Organizational Psychology.
Q1
Carlson, C. R., & Hoyle, R. H. (1993). Efficacy of abbreviated progressive muscle relaxation
training: A quantitative review of behavioural medicine research. Journal of Consulting and
Clinical Psychology,61, 1059–1067.
Q2
Cropley, M., & Purvis, M. L. J. (2003). Psychological contracting: Processes of contract formation
during interviews between nannies and their ‘employers’. Journal of Occupational and
Organizational Psychology,76, 213–241. doi:10.1348/096317903765913713.
Emery, C. F., France, C. R., Harris, J., Norman, G., & Vanarsdalen, C. (2008). Effects of progressive
muscle relaxation training on nociceptive flexion reflex threshold in healthy young adults:
A randomized trial. Pain,138, 375–379.
Fastenmeier, W., Gstalter, H., & Lehnig, U. (2003). Was empfinden Menschen als Freizeit? –
Emotionale Bedeutung und Definitionen [What is leisure time? Emotional meaning and
definitions]. In Institut fu
¨r Mobilita¨tsforschung (Ed.). Motive und Handlungsansa
¨tze im
Freizeitverkehr. Berlin: Springer Verlag.
Q3
Fries, E., Dettenborn, L., & Kirschbaum, C. (2009). The cortisol awakening response (CAR):
Facts and future directions. International Journal of Psychophysiology,72, 67–73.
Fritz, C., & Sonnentag, S. (2005). Recovery, health, and job performance: Effects of weekend
experiences. Journal of Occupational Health Psychology,10, 187–199.
Geurts, S. A. E., & Sonnentag, S. (2006). Recovery as an explanatory mechanism in the relation
between acute stress reactions and chronic health impairment. Scandinavian Journal of
Work, Environment and Health,32, 482–492.
Relaxation-induced cortisol changes 11
JOOP 1285—4/1/2010—SANKARS—357577
Hammerfald, K., Eberle, C., Grau, M., Kinsperger, A., Zimmermann, A., Ehlert, U., & Gaab, J.
(2006). Persistent effects of cognitive-behavioral stress management on cortisol responses to
acute stress in healthy subjects – a randomized controlled trial. Psychoneuroendocrinology,
31, 333–339.
Hartig, T., Johansson, G., & Kylin, C. (2007). The telework tradeoff: Stress mitigation vs.
constrained restoration. Applied Psychology: An International Review,56, 231–253.
Hellhammer, J., Fries, E., Schweisthal, O. W., Schlotz, W., Stone, A. A., & Hagemann, D. (2007).
Several dailymeasurements are necessary to reliably assess the cortisol rise after awakening:
State and trait components. Psychoneuroendocrinology,32, 80–86.
Krajewski, J., Batliner, A., & Golz, M. (2009). Acoustic sleepiness detection – framework and
validation of a speech adapted pattern recognition approach. Behavior Research Methods,41,
795–804.
Kudielka, B. M., Broderick, J. E., & Kirschbaum, C. (2003). Compliance with saliva sampling
protocols: Electronic monitoring reveals invalid cortisol daytime profiles in noncompliant
subjects. Psychosomatic Medicine,65, 313–319.
Lohaus, A., Klein-Heßling, J., Vo¨gele, C., & Kuhn-Hennighausen, C. (2001). Relaxation in children:
Effects on physiological measures. British Journal of Health Psychology,6, 197–206.
Lolak, S., Connors, G. L., Sheridan, M. J., & Wise, T. N. (2008). Effects of progressive muscle
relaxation training on anxiety and depression in patients enrolled in an outpatient pulmonary
rehabilitation program. Psychotherapy and Psychosomatics,77, 119–121.
Magkos, F., & Kavouras, A. (2004). Caffeine and ephedrine, physiological, metabolic and
performance enhancing effects. Sports Medicine,34, 871–889.
Q4
Manelli, F., & Giustina, A. (2000). Glucocorticoid-induced osteoporosis. Trends Endocrinology
Metabolism,11, 79–85.
Meijman, T. F., & Mulder, G. (1998). Psychological aspects of workload. In P. J. D. Drenth,
H. Thierry, & C. J. des Wolff (Eds.), Handbook of work and organizational psychology
(Vol. 2, pp. 5–33). Hove: Psychology Press.
Miller, D. B., & O’Callaghan, J. P. (2002). Neuroendocrine aspects of the response to stress.
Metabolism,51, 5–10.
Nickel, C., Lahmann, C., Tritt, K., Loew, T. H., Rother, W. K., & Nickel, M. K. (2005). Stressed
aggressive adolescents benefit from progressive muscle relaxation: A random, prospective,
controlled trial. Stress and Health,21, 169–175.
Pawlow, L. A., & Jones, G. E. (2002). The impact of abbreviated progressive muscle relaxation on
salivary cortisol. Biological Psychology,60, 1–16.
Pawlow, L. A., & Jones, G. E. (2005). The impact of abbreviated progressive muscle relaxation of
salivary cortisol and salivary immunoglobulin. Journal of Applied Psychophysiology and
Biofeedback,30, 375–387.
Rabin, B. S. (2005). Stressor-induced alteration of health across the life span: There is more to it
than immunology. Clinical and Applied Immunology,5, 207–224.
Rainforth, M. V., Schneider, R. H., Nidich, S. I., King, C. G., Salerno, J. W., & Anderson, J. W. (2007).
Stress reduction programs in patients with elevated blood pressure: A systematic review and
meta-analysis. Current Hypertension Reports,9, 520–528.
Schneider, R. H., Alexander, C., Staggers, F., Orme-Johnson, D. W., Rainforth, M., Salerno, J., …
Nidich, S. I. (2005). A randomized controlled trial of stress reduction in the treatment of
hypertension in African Americans during one year. American Journal of Hypertension, 18,
88–98.
Sephton, S. E., Sapolsky, R. M., Kraemer, H. C., & Spiegel, D. (2000). Diurnal cortisol rhythm as a
predictor of breast cancer survival. Journal of the National Cancer Institut,92, 994–1000.
Shapiro, S., & Lehrer, P. M. (1980). Psychophysiological effects of autogenic training and
progressive relaxation. Biofeedback and Self-Regulation,5, 249–255.
Smith, J. C. (1988). Steps toward a cognitive-behavioral model of relaxation. Biofeedback and
Self-Regulation,13, 307–329.
12 Jarek Krajewski et al.
JOOP 1285—4/1/2010—SANKARS—357577
Smith, J. C., Amutio, A., Anderson, J. P., & Aria, L. A. (1996). Relaxation: Mapping an uncharted
world. Biofeedback and Self-Regulation,21, 63–90.
Sonnentag, S., Binnewies, C., & Mojza, E. J. (2008). Did you have a nice evening? A day-level study
on recovery experiences, sleep and affect. Journal of Applied Psychology,93, 674–684.
Sonnentag, S., Mojza, E. J., Binnewies, C., & Scholl, A. (2008). Being engaged at work and detached
at home: A week-level study on work engagement, psychological detachment and affect.
Work and Stress,22, 257–276.
Takahashi, M., Nakata, A., Haratani, T., Ogawa, Y., & Arito, H. (2004). Post-lunch nap as a worksite
intervention to promote alertness on the job. Ergonomics,47, 1003–1013.
Thorn, L., Hucklebridge, F., Evans, P., & Clow, A. (2006). Suspected non-adherence and weekend
versus week day differences in the awakening cortisol response. Psychoneuroendocrinology,
31, 1009–1018.
Trougakos, J. P., Beal, D. J., Green, S. G., & Weiss, H. M. (2008). Making the break count: An episodic
examination of recovery activities, emotional experiences, and positive affective displays.
Academy of Management Journal,51, 131–146.
Trenberth, L., & Dewe, P. (2002). The importance of leisure as a means of coping with work
related stress: An exploratory study. Counselling Psychology Quarterly,15, 59–72.
Wegge, J., Van Dick, R., Fisher, G., Wecking, C., & Moltzen, K. (2006). Work motivation,
organisational identification, and well-being in call centre work. Work and Stress,20, 60–83.
Wesensten, N. J., Killgore, W. D. S., & Balkin, T. J. (2005). Performance and alertness effects of
caffeine, dextroamphetamine, and modafinil during sleep deprivation. Journal of Sleep
Research,14, 255–266.
Westermann, J., Demir, A., & Herbst, V. (2004). Determination of cortisol in saliva and serum by
a luminescence-enhanced enzyme immunoassay. Journal of Clinical Laboratory Analysis,50,
11–24.
Wu
¨st, S., Federenko, I., Hellhammer, D. H., & Kirschbaum, C. (2000). Genetic factors, perceived
chronic stress, and the free cortisol response to awakening. Psychoneuroendocrinology,25,
707–720.
Received 13 May 2009; revised version received 2 November 2009
Relaxation-induced cortisol changes 13
JOOP 1285—4/1/2010—SANKARS—357577
Author Queries
JOB NUMBER: 1285
JOURNAL: JOOP
Q1 Please update the publication details in reference Binnewies et al. (in press).
Q2 If possible, please supply DOI numbers for the journal articles in the list, as per
current APA style.
Q3 Please supply the page ranges for the reference Fastenmeier et al. (2003).
Q4 Reference Magkos and Kavouras (2004) is provided in the list but not cited in
the text. Please supply citation details or delete the reference from the
reference list.
Q5 Please check the edit of the value (8.361) in Table 1.
14 Jarek Krajewski et al.
JOOP 1285—4/1/2010—SANKARS—357577
