Pedisic et al.

BMC Public Health 2014, 14:1249

http://www.biomedcentral.com/1471-2458/14/1249

RESEARCH ARTICLE Open Access

Workplace Sitting Breaks Questionnaire (SITBRQ):

an assessment of concurrent validity and test-retest
reliability
Zeljko Pedisic1,2*, Jason A Bennie2,3, Anna F Timperio4, David A Crawford4, David W Dunstan4,5,6,7,8,9,
Adrian E Bauman2 and Jo Salmon4

Abstract
Background: Breaks in prolonged sitting may have beneficial cardiometabolic and musculoskeletal health
outcomes. Desk-based work settings are an important environment to promote and support breaks in sitting time.
However, few studies have reported the psychometric properties of self-report measures to assess the frequency
and duration of breaks from sitting. This study examined the concurrent validity and test-retest reliability of the
Workplace Sitting Breaks Questionnaire (SITBRQ) designed to assess frequency and duration of breaks in sitting
within desk-based work settings.
Methods: To assess the concurrent validity, a sample of 147 desk-based employees completed the SITBRQ and
wore an Actigraph GT1M accelerometer for seven consecutive days. To establish test-retest reliability, SITBRQ was
administered on two separate occasions 7–14 days apart to a separate sample of 96 desk-based employees.
Results: A low relative agreement with accelerometry (Spearman’s r = 0.24 [95% CI 0.07 - 0.40]) was determined for
self-reported frequency, but not for the duration of sitting breaks (Spearman’s r = 0.05 [95% CI −0.12 - 0.22]).
Adequate reliability was determined for both self-reported frequency (Spearman’s r = 0.71 [95% CI 0.59 - 0.79],
Cohen’s kappa = 0.74 [95% CI 0.64 - 0.84]) and duration of sitting breaks (Spearman’s r = 0.59 [95% CI 0.45 - 0.71],
Cohen’s kappa = 0.61 [95% CI 0.38 - 0.85]).
Conclusion: SITBRQ may be used for assessment of the frequency of sitting breaks within desk-based work settings
with validity and reliability similar to other self-reports in the field of sedentary behaviour research. However, until
adequately improved and re-evaluated, it should not be used to collect data about the duration of breaks in sitting
time.
Keywords: Breaks in sitting time, Physical activity, Sedentary behaviour, Desk-based employees, Light-intensity
physical activity, Reliability, Validity, Accelerometer, Self-report

Background of a number of chronic diseases [2]. Recently, sedentary

Physical inactivity is one of the leading global public behaviour (too much sitting as opposed to too little physical
health issues in developed countries [1]. There is well- activity) has emerged as a candidate independent risk factor
established epidemiological evidence to suggest that a [3]. Several reviews have shown that high volumes of time
minimum of 150 minutes of moderate or 75 minutes of spent sitting or engaged in sedentary behaviour have been
vigorous-intensity physical activity a week, or an equiva- associated with an increased all-cause mortality and chronic
lent combination of both, significantly reduces the risk disease (e.g., diabetes, cardiovascular disease) and obesity
risk [4-6]. In some studies, associations between sedentary
behaviour and all-cause mortality and chronic disease risk
* Correspondence: zeljko.pedisic@kif.hr
1
Faculty of Kinesiology, University of Zagreb, Zagreb, Croatia
occurred irrespective of whether an individual meets the
2
Prevention Research Collaboration, Sydney School of Public Health, The public health physical activity recommendations [7-10].
University of Sydney, Sydney, Australia
Full list of author information is available at the end of the article

© 2014 Pedisic et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain
Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,
unless otherwise stated.
Pedisic et al. BMC Public Health 2014, 14:1249 Page 2 of 9
http://www.biomedcentral.com/1471-2458/14/1249

Researchers have recently investigated the potential for is important to develop valid and reliable assessments of
health benefits associated with light-intensity physical this behaviour. Although self-report measures have been
activity [11,12]. This type of activity is defined as being the most commonly used method in large-scale epidemio-
between 1.8-2.9 metabolic equivalent units of rest logical studies on physical activity and sedentary behavior
(METs) and is typically non-structured and incidental [32,33], almost all studies assessing breaks in sedentary
in nature [13]. Examples of light-intensity physical behavior have used objective measures, such as acceler-
activity include common habitual free-living activities ometers [19,24,34,35], multi-sensor devices [36], inclinom-
such as routine occupational (e.g. standing, retail serving eters [29] and sitting pads [25]. While this may in part be
and food preparation) or domestic tasks (e.g. ironing, due to reduced cost and increased availability of objective
washing up, gardening) [13]. Studies have shown that time devices in the last decade, few self-report instruments are
spent in light-intensity physical activity is highly inversely available and psychometric data are lacking.
correlated (r = −0.95) with time spent in sedentary behav- Several previous studies have developed self-report
iour [14]. Hence, if an individual has low levels of light- instruments that assess occupational physical activity
intensity physical activity, it is likely that he or she is [37-41]. However, these surveys do not ask specifically
highly engaged in sedentary behaviours. about short breaks in sitting time. For example, the
While the evidence base is still developing, data from Occupational Physical Activity Questionnaire (OPAQ)
experimental studies [15,16] and several cross-sectional assesses time spent walking at work and doing heavy labor
observational studies [12,17-19] suggest that higher levels at work [37]. The Occupational Sitting and Physical Activity
of light-intensity physical activity are associated with a Questionnaire (OSPAQ) assesses the proportion of occupa-
reduced risk of cardiometabolic disease biomarkers and tional time spent sitting, standing, walking, and in heavy
being overweight/obese. Importantly, some of these find- labour [38]. The International Physical Activity Question-
ings were present even after adjustment for time spent naire (IPAQ) asks about the weekly frequency and usual
in moderate-to-vigorous physical activity and sedentary time/day spent in vigorous-intensity physical activity,
behaviour. moderate-intensity physical activity and walking at work
These insights have prompted some health agencies [39]. ‘Breaks in occupational sedentary time’ are considered
and professional societies to provide formal documents as any interruption in sitting time during work (e.g., going
outlining some behavioural modification strategies to for a bathroom break, walking to a remote printer, standing
reduce time spent in sedentary behaviour among adults up for a short stretch, standing while talking to a college),
and children, and suggest strategies generally based whilst ‘occupational physical activity’ is usually defined
around increasing volumes of light-intensity physical in terms of longer periods spent in walking, moderate-
activity [20-22]. One of the main strategies is to period- intensity or vigorous-intensity physical activity at work
ically take short breaks in sitting time within the occu- (often in ≥10 minute bouts). Hence, occupational physical
pational and home settings [20-22]. It is suggested, for activity questionnaires may not necessarily capture all in-
example, that workers periodically take a short break in terruptions in sitting time. Moreover, the available occupa-
sitting at least once every 30 minutes, walk to their tional physical activity questionnaires usually capture total
co-workers instead of telephoning or emailing, stand time spent in occupational activities and not the daily
up during meetings/presentations and install height frequency of such activities. It is therefore important to
adjustable sit-stand workstations [22]. Studies have in- develop and validate questionnaires that specifically assess
dicated that taking short breaks may assist in reducing breaks in occupational sedentary time.
detrimental biomechanical consequences of prolonged Three previous studies have assessed measurement
sitting, such as posture-related musculoskeletal disorders properties of self-reported frequency of breaks in occupa-
[23] as well as the risk of cardiometabolic disorders and tional sitting time [42-44]. Clark et al. [42] evaluated an
obesity [15,19,24,25]. interview-administered item (“How many breaks from
In addition to spending approximately half their waking sitting [such as standing up or stretching or taking a short
hours in this setting [26,27], employees in computer- walk] during one hour of sitting would you typically take
centric work environments may be sitting for up to 80% at work?,” with the following response scale; 0, 1, 2, 3, 4,
of their working time [28,29]. It has also been shown that and ≥5) and observed its weak correlation (r = 0.26) with
office workers take significantly fewer breaks in sedentary accelerometer-derived breaks per sedentary hour among
time during working hours compared to non-occupational office workers. Reliability of the item was not assessed.
time [28,30]. Desk-based and computer-centric work envi- Furthermore, Lynch et al. [43] developed a past year
ronments may, therefore, be the key behavioural settings measure of domain-specific sedentary behaviour (SIT-Q),
to support and promote breaks in sitting time [31]. which includes an item about breaks in occupational
With the emerging public health interest around the sitting time (“How often did you ‘break up’ the time you
positive health consequences of breaks in sitting time, it spent sitting in job # 1?,” with the following response scale:
Pedisic et al. BMC Public Health 2014, 14:1249 Page 3 of 9
http://www.biomedcentral.com/1471-2458/14/1249

(a) less than hourly; (b) hourly; (c) half hourly; (d) every was sent to employees to promote the study. This was
10 minutes; (d) every 5 minutes; (f) I did not sit for more followed up by trained research staff visiting each work-
than 30 minutes in a day). The item has shown moderate place to administer study materials to those employees
test-retest reliability (weighted kappa = 0.49), but has not wishing to participate. The participants completed a survey
been tested for validity [43]. Furthermore, Wijndaele et al. and were asked to then wear an accelerometer for the next
[44] changed the reference period of SIT-Q from the past seven consecutive days.
year to the last seven days (SIT-Q-7d), and found poor
test-retest reliability (ICC = 0.26 and 0.12 for the Dutch Reliability sample
and English version respectively) and poor validity tested The present study used a sub-set of participants from a
against activPAL3 monitors (Spearman’s r = 0.06; Dutch larger population survey conducted in metropolitan
version. This shows that more research is needed to Melbourne, Australia [45]. A random sample of 316
further test and improve measurement properties of workplaces from Melbourne, Australia were approached
self-reported frequency of breaks in occupational sitting and 55 organizations agreed to participate (response
time. rate = 17.4%). Contacts within each recruited workplace
Further, the total duration of breaks in occupational were asked to distribute materials to staff whose typical
sitting time is conceptually equal to the total time spent working tasks involved being seated at a desk or worksta-
in light- (including standing quietly), moderate- and tion (such as office administration, data entry and any
vigorous-intensity physical activity. We assumed that other desk-bound occupations). A total of 1467 surveys
asking participants about the duration of breaks in sit- were distributed with 722 returned (response rate = 49.2%).
ting time, instead of asking directly about their physical Participants who completed the questionnaire were asked
activity, might better capture sporadic and short bouts whether they would agree to participate in the test-retest
of light-intensity physical activity. It seems reasonable, reliability study by completing an abbreviated version of
however, that longer breaks in sitting time (e.g. 20 minutes the survey that included SITBRQ, on a second occasion
or more) would be more easily recalled as occupational 7–14 days after completion of the initial survey. Similar
activity, rather than as interruptions to sitting. We, there- time frame was used when assessing test-retest reliability
fore, assumed that measuring the duration of sedentary of self-reported measures of sedentary behaviour in most
breaks via self-reports may only be feasible among partici- previous studies [32]. The between-subject variability in
pants within desk-based work settings that typically do the timeframe of the second survey was inevitable, due to
not engage in longer bouts of physical activity. No previ- the study design. Questionnaires were sent directly to all
ous studies have assessed the reliability and validity of self- participants who agreed to take part in the second survey
reported duration of breaks in sitting time. (n = 96). Completed surveys were returned to the re-
Therefore, the aim of this study was to develop and searchers via a reply-paid envelope.
evaluate concurrent validity and test-retest reliability of
the Workplace Sitting Breaks Questionnaire (SITBRQ; Measures
Additional file 1), which includes items on frequency The Workplace Sitting Breaks Questionnaire (SITBRQ)
and duration of breaks in occupational sitting time. The SITBRQ was developed to assess the frequency and
duration of breaks in sitting time within the context of
Methods desk-based work settings. When constructing self-report
Concurrent validity and reliability of the SITBRQ were instruments to measure complex behaviours, it is recog-
tested in two separate samples (Validity Sample; Reliability nised that the use of relevant cues and examples of the
Sample) between February and April 2009. Written in- behaviours of interest is essential for the design of effect-
formed consent was obtained from the organisations and ive assessment tools [46]. As breaks in sitting time is a
employees involved. The study protocols were approved by relatively new concept, the idea was explored in one-on-
the Deakin University Ethics Committee (EC 207–2009). one interviews with a convenience sample of 33 employees
who typically sit for working tasks, in order to determine
Participants and procedures the most effective terminology. Most described the term
Validity sample ‘short physical activity breaks’ as the most succinct way to
To test validity, a convenience sample of 143 employees describe breaks in sitting time during work hours (unpub-
who worked a minimum of four days a week was re- lished data). The term ‘short physical activity breaks’ was,
cruited. This was a sub-sample of a larger study described therefore, used and defined as ‘any interruption in sitting
elsewhere [28]. In brief, a large organisation located in time during work hours’. To further aid understanding,
metropolitan Melbourne, Australia participated in the we provided a ‘preamble’ to the question on reporting
study. The organisation had various offices and retail out- frequency and duration of short physical activity breaks at
lets in metropolitan Melbourne, and an internal staff email work, which gave examples of this behaviour.
Pedisic et al. BMC Public Health 2014, 14:1249 Page 4 of 9
http://www.biomedcentral.com/1471-2458/14/1249

Participants were asked to report how many breaks in The same approach to determine breaks in sitting time
sitting time they would take during a work hour on a was used in previous studies [24].
typical work day (categorical response options ranged
from 0 to ≥6) and the total time per day typically spent
in short physical activity breaks at work (six categorical Socio-demographic profile
response options were provided: <5, 5–9, 10–19, 20–29, Participants reported their age, sex, level of education, em-
30–59, and ≥60 minutes). The SITBRQ item about the ployment status and main occupation. These were col-
frequency of breaks in sitting time was developed based lapsed as follows: age as 18–29, 30–39, 40–49, 50–59,
on the single-item questionnaire by Clark et al. [42]. The and ≥ 60 years; employment status as full time, and part-
wording of the SITBRQ question is almost identical to time; educational attainment as <12 years, ≥12 years, trade
its source. The corresponding response scale includes 7 or technical, and university or tertiary qualification; and
instead of 6 options that were originally included in occupational domain as managers/administrators, profes-
Clark et al. [42], because we assumed that an additional sionals, associate professional, tradespersons, advanced
category may reduce the number of participants potentially clerical, intermediate clerical, and other (coded according
affected by the measurement ceiling effect. The wording of to the Australian Standard Classification of Occupation
the SITBRQ item about the duration of breaks in sitting (ASCO) coding system) [50].
time was not based on any questionnaires, as no such
items were previously published.
Data analysis
Accelerometry Concurrent validity was assessed by Spearman’s rank
A uniaxial accelerometer (Actigraph GT1M, Pensacola, correlations. There are no specifically intended acceler-
FL, USA) was used to establish the concurrent validity of ometer cut-points for assessment of breaks in sitting
SITBRQ. The Actigraph accelerometers have shown time (i.e., changes from sitting to standing posture) and
acceptable validity against doubly-labelled water [47] and the number of minutes classified as sedentary seems to
have often been used as a concurrent measure to validate be inversely related to the level of sedentary behaviour
physical activity and sedentary behaviours questionnaires threshold (i.e., higher sedentary behaviour thresholds are
[48]. Participants were asked to wear the accelerometer on associated with lower number minutes classified as sed-
their right hip during all waking hours for 7 days (5 work entary) [51]. Accelerometer-based measures can, there-
and 2 non-work days) and to record accelerometer on/off fore, only be considered as relative estimates of breaks
times, as well as work start and finish times in an event in sitting time. Taking this into account, only the rela-
diary [28]. In the present study we only used accelerometer tive agreement between SITBRQ and accelerometer-
data collected during work hours. Only participants who based measures (as expressed by Spearman’s rank
wore accelerometers for five or more hours of their work- correlations) was tested, whilst the absolute agreement
ing time on at least three days were included in the analysis was not hypothesised. Furthermore, Spearman’s rank
(n = 135). Accelerometer data were summarized using SAS correlations were calculated to evaluate the relative
9.1 (SAS Institute, Cary, NC, USA). Previously commonly agreement between test and retest. Quadratic weighted
used cut points were used to classify the accelerometer Cohen’s kappa coefficients of agreement, percentage of
data, with sedentary time defined as <100 counts/minute responses correctly classified, percentage of responses
(cpm) [32], low-intensity as ≥100-1951 cpm and moderate in same or adjacent category, and percentage of highly
to vigorous-intensity physical activity as ≥1952 cpm [49]. misclassified responses were calculated to evaluate the
As shown in Thorp, et al. [28], the employees’ working absolute agreement between test and retest. The meas-
hours were mostly spent sedentary on 77% or 6.6 hours/ ure of relative agreement, therefore, represents how
day). Most of the remaining time was comprised of light- proportionate were participants’ responses on two ad-
intensity physical activity (on 20% or 1.7 hours/day), with ministrations of the questionnaire. The measures of
minimal moderate-to-vigorous physical activity (on 2% or absolute agreement illustrate the proportion of identical
0.2 hours/day) recorded. Light-intensity physical activity and non-identical responses on the test and retest. To
and moderate-to-vigorous physical activity were com- allow for generalization, 95% confidence intervals were
bined to classify time spent in all physical activities as provided for all statistics. The validity and reliability
being ≥100 cpm. All the minutes of accelerometry were assessed by Spearman’s rank correlations, because
with ≥100 cpm were summed to form the total dur- SITBRQ data is ordinal-scaled. The choice of the method
ation of breaks in sedentary time. Every change be- is in accordance with previous studies [32,42,44]. The data
tween <100 cpm and ≥100 cpm was counted as a break analyses were conducted using STATISTICA, version 10
in sitting time. No minimum duration for break or a mini- (StatSoft, Inc., Tulsa, OK, USA) and IBM SPSS Statistics
mum duration of sedentary epochs before a break was set. 21 (SPSS Inc. an IBM Company, Chicago, IL, USA).
Pedisic et al. BMC Public Health 2014, 14:1249 Page 5 of 9
http://www.biomedcentral.com/1471-2458/14/1249

Results Table 1 Sample characteristics, and frequency and duration

Table 1 shows socio-demographic characteristics and of sitting breaks
self-reported frequency and duration of sitting breaks Reliability Validity
from the respective validity and reliability samples. Both sample (%) sample (%)
samples comprised greater proportions of women (63%) (n = 96) (n = 143)
than men (37%). Across both samples, most participants Sex
were aged less than 60 years (96% in both samples), were Men 37.5 37.4
in full-time employment (95% in the validity and 86% in Women 62.5 62.6
the reliability sample) and had a university or tertiary
Age (years)
qualification (62% in the validity and 69% in the reliabil-
18-29 26.0 33.5
ity sample). Approximately one in two participants
reported being managers or professionals (45% in both 30-39 27.4 33.8
samples). Furthermore, over 80% of participants in both 40-49 25.6 19.7
samples reported having between one and three sitting 50-59 16.9 9.2
breaks per working hour. Few participants reported a 60 and over 4.2 2.1
total duration of sitting breaks of ≥60 minutes (3.1% in
Education
the validity and 5.2% in the reliability sample). The aver-
Some high school 8.3 7.1
age number of sedentary breaks per hour and the total
duration of breaks (in minutes), as assessed by accelerom- Year 12 or equivalent 16.7 25.4
eters in the validity sample, was 5.67 ± 1.58 and 124 ± 42 Trade or technical 6.3 9.3
(mean ± standard deviation), respectively. University/tertiary qualification 68.8 61.8
Spearman’s rank correlation showed low relative agree- Occupation
ment between the SITBRQ and accelerometry in estimating
Managers/administrators 23.7 25.2
the frequency of breaks (Spearman’s r = 0.24 [95% CI 0.07 -
Professionals 20.9 19.7
0.40]) (Table 2). No agreement was found between the
SITBRQ and accelerometer-based estimates of the total dur- Associate professional 15.3 16.2
ation of breaks (Spearman’s r = 0.05 [95% CI −0.12 - 0.22]). Trades persons 3.5 0.0
The Spearman’s rank correlation coefficients showed Advanced clerical 17.5 15.9
somewhat higher test-retest reliability of SITBRQ in esti- Intermediate clerical 19.1 23.0
mating frequency of breaks than in estimating the total
Employment Status
duration of breaks (Spearman’s r; 0.71 [95% CI 0.59 - 0.79]
Full-time 86.1 94.6
vs. 0.59 [95% CI 0.45 - 0.71]) (Table 3). According to Landis
and Koch [52] the agreement between categorical responses Part-time 13.9 4.8
in test and retest was substantial (Cohen’s kappa = 0.74 Frequency of sitting breaks*
[95% CI 0.64 - 0.84]) and moderate (Cohen’s kappa = 0.61 0 0.0 1.5
[95% CI 0.38 - 0.85]) for the frequency of breaks and the 1 35.4 26.5
total duration of breaks, respectively. According to the
2 25.0 31.8
reported frequency of breaks, the majority of participants
3 19.8 25.8
(85% [95% CI 78% - 93%]) were classified in the same or
adjacent category between survey administrations. For the 4 6.3 6.8
total duration of breaks, 79% (95% CI 71% - 87%) partici- 5 7.3 7.6
pants selected the same or adjacent response category in ≥6 6.3 0.0
the test and retest. Duration of sitting breaks (minutes)*
<5 14.6 20.9
Discussion
5-9 13.5 15.5
The SITBRQ was designed as a brief self-report instrument
to assess the frequency and duration of breaks in sitting 10 - 19 37.5 21.7
among employees who commonly sit for working tasks. 20 - 29 12.5 15.5
The current study demonstrated low concurrent validity of 30 - 59 16.7 23.3
the SITBRQ for the assessment of frequency of breaks in ≥ 60 5.2 3.1
sitting time and no correlation with accelerometer-based *
Assessed by the Workplace Sitting Breaks Questionnaire (SITBRQ).
estimates of duration of the breaks. The questionnaire
showed good reliability for the assessment of frequency and
duration of breaks in sitting time.
Pedisic et al. BMC Public Health 2014, 14:1249 Page 6 of 9
http://www.biomedcentral.com/1471-2458/14/1249

Table 2 Concurrent validity of the Workplace Sitting Breaks sporadic, which makes them difficult to recall [46]. The
Questionnaire (SITBRQ) against accelerometer-based measure requires participants to sum the time spent in all
measures* breaks during a usual working day. From a cognitive
Questionnaire item Spearman’s rho (95% CI)† perspective, this might be a more demanding task than
Frequency of breaks (breaks/hour) 0.24 (0.07 - 0.40) recalling the usual frequency of breaks per hour [46]. A
Total duration of breaks (minutes/day) 0.05 (−0.12 - 0.22) subsequent analysis revealed that almost all misclassified
*Accelerometer-based measures (total time in ≥100 cpm) were categorized to participants (i.e., those who were not classified in same
reproduce response scales of questionnaire items. categories based on accelerometry and SITBRQ) had higher
†
Spearman’s rank correlation between SITBRQ and accelerometer-based
measures and its 95% confidence interval.
duration of sedentary breaks assessed by accelerometer
when compared to the self-reported estimate (data not
shown). The highest category on the response scale
The correlation between self-reported and accelerometer- (60 minutes or more) was selected by very few partici-
based frequency of breaks in sitting time determined in our pants, while the responses were evenly distributed
sample (Spearman’s r = 0.24) was consistent with that across all other categories. Therefore, it seems unlikely
reported by Clark et al. [42] (Spearman’s r = 0.26) and higher that the relative agreement between SITBRQ and
than the one found by Wijndaele et al. [44] (Spearman’s accelerometer-based estimates was compromised by the
r = 0.06). A review article by Helmerhorst et al. [53] restricted range of the scale. Nevertheless, it might be
showed that the Spearman’s correlation coefficient be- that asking an open-ended question or providing a
tween questionnaire-based and accelerometer-derived response scale with a greater number of categories would
time in sedentary behaviour is usually around 0.23. Hence, increase the relative agreement with accelerometer-based
the relative agreement between the self-reported and estimates. This type of measure could be tested in future
accelerometer-based frequency of breaks in sitting time validation studies. Until adequately improved and re-
determined in our study is similar to other self-reports in evaluated, this item should not be used to collect data
the field of sedentary behaviour research. The low concur- about the duration of breaks in sitting time.
rent validity does not necessarily imply that this item is Participants’ responses showed good relative and abso-
not valid. The concurrent validity was tested by assessing lute agreement across two survey occasions as expressed
a bidirectional correlation between two concurrent instru- by Spearman’s correlations and Cohen’s kappa coefficients,
ments (the SITBRQ and accelerometers). The magnitude respectively. This demonstrates the ability of SITBRQ to
of the correlation, therefore, depended on the validity of reliably rank and classify participants according to their
both instruments, and may have been low because accel- self-reported frequency and duration of breaks in sitting
erometer cut-points are not specifically developed to cap- time. Reliability of both SITBRQ items was similar as for
ture short interruptions in sitting time. It seems that the most other physical activity and sedentary behaviour ques-
SITBRQ may have potential to rank individuals based on tionnaires [53] and can therefore be considered satisfac-
their frequency of breaks in sitting time. However, we tory. The SITBRQ has shown somewhat higher reliability
acknowledge that this assumption needs to be further in assessing frequency of breaks in sitting time than
tested by assessing its criterion validity against a ‘gold SIT-Q [44]. This difference may, however, be explained by
standard’ measure of frequency of breaks in sitting time the shorter interval between two administrations of the
(e.g. using inclinometers worn on the thigh). questionnaire used in our study when compared to Lynch
Our results showed that the SITBRQ does not provide et al. [44] (7–14 days vs. one month). It is possible that
valid estimates of the total duration of breaks in sitting the second administration of the questionnaire in our
time. Light-intensity physical activities that are most often study was under a greater influence by carryover effects
performed in sitting breaks are typically unstructured and due to memory. Furthermore, our participants assessed

Table 3 Test-retest reliability* of the Workplace Sitting Breaks Questionnaire (SITBRQ)

Questionnaire item Spearman’s rho Cohen’s kappa % correctly classified % in same or adjacent % highly misclassified
(95% CI)† (95% CI)‡ (95% CI)§ category (95% CI)|| (95% CI)¶
Frequency of breaks 0.71 (0.59 - 0.79) 0.74 (0.64 - 0.84) 51.0 (41.0 - 61.0) 85.4 (78.4 - 92.5) 14.6 (7.5 - 21.6)
(breaks/hour)
Total duration of breaks 0.59 (0.45 - 0.71) 0.61 (0.38 - 0.85) 46.9 (36.9 - 56.9) 79.2 (71.0 - 87.3) 20.8 (12.7 - 29)
(minutes/day)
*
Test and retest surveys were conducted a maximum of 14 days apart.
†
Spearman’s rank correlation between test and retest and its 95% confidence interval.
‡
Quadratic weighted Cohen’s kappa coefficient of agreement between test and retest and its 95% confidence interval.
§
Percent of participants classified in same categories in test and retest its 95% confidence interval.
||
Percent of participants classified in same or adjacent categories in test and retest its 95% confidence interval.
¶
Percent of participants classified in distant categories in test and retest (two or more categories apart) its 95% confidence interval.
Pedisic et al. BMC Public Health 2014, 14:1249 Page 7 of 9
http://www.biomedcentral.com/1471-2458/14/1249

the frequency of breaks somewhat more reliably than the validity. Furthermore, despite satisfactory reliability, the
duration of breaks. This is in accordance with the study of item related to self-reported duration of breaks in sit-
McCormack et al. [54], who found higher reliability for ting time did not demonstrate acceptable concurrent
the self-reported frequency than for the self-reported validity. Therefore, this item in its current form should
duration of incidental physical activities. not be used to collect data about the duration of breaks
A key limitation of this study was the lack of a true in sitting time. Additional research is needed to further
‘gold standard’ to objectively measure frequency and refine and validate self-report measures of breaks in
breaks in sitting time, namely, accelerometers have sev- sitting time among employees within desk-based work
eral limitations as sedentary behaviour measures. The 9pt?>settings. Future studies might benefit from establishing
accelerometer cut-point used here to assess breaks in the validity of SITBRQ against measures that are more suit-
sitting time (≥100 cpm) is the commonly used threshold able for assessing breaks in sitting time than accelerometers.
for differentiation between sedentary behaviour and phys-
ical activity [32]. However, this cut-point is not specific- Additional file
ally intended for detection of changes from sitting to
standing posture (i.e., breaks in sitting time). Researchers Additional file 1: Workplace Sitting Breaks Questionnaire (SITBRQ).
suggest that behaviours such as fidgeting legs whilst seated
may result in accelerometer readings above 100 cpm
Competing interests
[17,24]. Therefore, the objective and self-report instru- The authors declare that they have no competing interests.
ments used in the present study may have not captured
the same behaviours, with accelerometry potentially over- Authors’ contributions
estimating frequency and duration of breaks by detecting JAB, AFT, DAC, DWD and JS conceptualised the study and its design, and
movements undertaken while seated. To avoid this prob- participated in the data collection; ZP, JAB and AB planned and performed
the statistical analysis; All authors were involved in the interpretation of the
lem, future validation studies should therefore consider findings, contributed to writing of the manuscript, and have read and
using other objective instruments that more specifically approved the final manuscript.
measure the behaviour of interest, such as inclinometers
that are able to detect sit-to-stand transitions and non- Acknowledgments
AT is supported by a Future Leader Fellowship from the National Heart
sitting/lying time. Furthermore, recording accelerometer Foundation of Australia (Award ID 100046). DD is supported by an Australian
data in 1-minute epochs increased the likelihood not to Research Council Future Research Fellowship (FT100100918). JS is supported
capture short sit-stand transitions, which made the discrep- by a NHMRC Principal Research Fellowship (APP1026216). We acknowledge
the support of the Physical Activity Laboratory, Baker IDI Heart and Diabetes
ancies between SITBRQ and accelerometer-based estimates Institute for assistance in the data collection for the validity component of
even more probable. In addition, SITBRQ had been admin- this study.
istered before the accelerometer measures were taken.
Author details
Although SITBRQ asks about breaks in sitting time on a 1
Faculty of Kinesiology, University of Zagreb, Zagreb, Croatia. 2Prevention
typical day, it may be that this has caused a mismatch Research Collaboration, Sydney School of Public Health, The University of
between the timeframe of self-reported and accelerometer- Sydney, Sydney, Australia. 3Institute of Sport, Exercise and Active Living
(ISEAL), Victoria University, Melbourne, Australia. 4Centre for Physical Activity
based estimates, and further lowered their agreement. and Nutrition Research, School of Exercise and Nutrition Sciences, Deakin
Furthermore, the concurrent validity was assessed in a University, Melbourne, Australia. 5The Physical Activity Laboratory, Baker IDI
convenience sample, which may have reduced the Heart and Diabetes Institute, Melbourne, Australia. 6School of Population
Health, The University of Queensland, Brisbane, Australia. 7Department of
generalizability of our findings. Medicine, Monash University, Melbourne, Australia. 8Department of
Strengths of the study include a relatively large sample Epidemiology and Preventative Medicine, Monash University, Melbourne,
size, and a critical evaluation of a specifically designed Australia. 9School of Sport Science, Exercise and Health, The University of
Western Australia, Perth, Australia.
instrument that captures both frequency and duration of
breaks in sitting time within the context of work set- Received: 15 October 2014 Accepted: 24 November 2014
tings. This study was the first to evaluate the concurrent Published: 5 December 2014
validity and reliability of the self-reported duration of
breaks in sitting time. References
1. Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT: Effect of
physical inactivity on major non-communicable diseases worldwide: an
Conclusion analysis of burden of disease and life expectancy. Lancet 2012, 380:219–229.
The SITBRQ item for the assessment of frequency of 2. World Health Organization: Global Recommendations on Physical Activity for
Health. Geneva, CH: World Health Organization; 2010.
breaks in sitting time using showed concurrent validity 3. Owen N, Sparling PB, Healy GN, Dunstan DW, Matthews CE: Sedentary
similar to other self-reports in the field of sedentary be- behavior: emerging evidence for a new health risk. Mayo Clin Proc 2010,
haviour research and satisfactory reliability, indicating 85:1138–1141.
4. Proper KI, Singh AS, Van Mechelen W, Chinapaw MJ: Sedentary behaviors
its potential for utilization in future studies. Neverthe- and health outcomes among adults: a systematic review of prospective
less, additional efforts may be needed to improve its studies. Am J Prev Med 2011, 40:174–182.
Pedisic et al. BMC Public Health 2014, 14:1249 Page 8 of 9
http://www.biomedcentral.com/1471-2458/14/1249

5. Thorp AA, Owen N, Neuhaus M, Dunstan DW: Sedentary behaviors and 27. Pronk N: The problem with too much sitting a workplace conundrum.
subsequent health outcomes in adults: a systematic review of Acsms Health Fitness J 2011, 15:41–43.
longitudinal studies, 1996–2011. Am J Prev Med 2011, 41:207–215. 28. Thorp AA, Healy GN, Winkler E, Clark BK, Gardiner PA, Owen N, Dunstan DW:
6. Wilmot EG, Edwardson CL, Achana FA, Davies MJ, Gorely T, Gray LJ, Khunti Prolonged sedentary time and physical activity in workplace and
K, Yates T, Biddle SJH: Sedentary time in adults and the association with non-work contexts: a cross-sectional study of office, customer service
diabetes, cardiovascular disease and death: systematic review and and call centre employees. Int J Behav Nutr Phys Act 2012, 9:128.
meta-analysis. Diabetologia 2012, 55:2895–2905. 29. Toomingas A, Forsman M, Mathiassen SE, Heiden M, Nilsson T: Variation
7. Katzmarzyk PT, Church TS, Craig CL, Bouchard C: Sitting time and mortality between seated and standing/walking postures among male and female
from all causes, cardiovascular disease, and cancer. Med Sci Sports Exerc call centre operators. BMC Public Health 2012, 12:154.
2009, 41:998–1005. 30. Parry S, Straker L: The contribution of office work to sedentary behaviour
8. Matthews CE, George SM, Moore SC, Bowles HR, Blair A, Park Y, Troiano RP, associated risk. BMC Public Health 2013, 13:296.
Hollenbeck A, Schatzkin A: Amount of time spent in sedentary behaviors 31. Healy GN, Eakin EG, Lamontagne AD, Owen N, Winkler EA, Wiesner G,
and cause-specific mortality in US adults. Am J Clin Nutr 2012, 95:437–445. Gunning L, Neuhaus M, Lawler S, Fjeldsoe BS, Dunstan DW: Reducing
9. Stamatakis E, Chau JY, Pedisic Z, Bauman A, Macniven R, Coombs N, Hamer sitting time in office workers: short-term efficacy of a multicomponent
M: Are sitting occupations associated with increased all-cause, cancer, intervention. Prev Med 2013, 57:43–48.
and cardiovascular disease mortality risk? A pooled analysis of seven 32. Healy GN, Clark BK, Winkler EA, Gardiner PA, Brown WJ, Matthews CE:
british population cohorts. PLoS One 2013, 8:0073753. Measurement of adults’ sedentary time in population-based studies.
10. Van Der Ploeg HP, Chey T, Korda RJ, Banks E, Bauman A: Sitting time and Am J Prev Med 2011, 41:216–227.
all-cause mortality risk in 222 497 Australian adults. Arch Intern Med 2012, 33. Shephard RJ, Aoyagi Y: Measurement of human energy expenditure, with
172:494–500. particular reference to field studies: an historical perspective.
11. Buman MP, Hekler EB, Haskell WL, Pruitt L, Conway TL, Cain KL, Sallis JF, Eur J Appl Physiol 2012, 112:2785–2815.
Saelens BE, Frank LD, King AC: Objective light-intensity physical activity 34. Bankoski A, Harris TB, McClain JJ, Brychta RJ, Caserotti P, Chen KY, Berrigan D,
associations with rated health in older adults. Am J Epidemiol 2010, Troiano RP, Koster A: Sedentary activity associated with metabolic syndrome
172:1155–1165. independent of physical activity. Diabetes Care 2011, 34:497–503.
12. Carson V, Ridgers ND, Howard BJ, Winkler EAH, Healy GN, Owen N, Dunstan 35. Henson J, Yates T, Biddle SJH, Edwardson CL, Khunti K, Wilmot EG, Gray LJ,
DW, Salmon J: Light-intensity physical activity and cardiometabolic Gorely T, Nimmo MA, Davies MJ: Associations of objectively measured
biomarkers in US Adolescents. PLoS One 2013, 8:e71417. sedentary behaviour and physical activity with markers of
13. Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, O’Brien cardiometabolic health. Diabetologia 2013, 56:1012–1020.
WL, Bassett DR Jr, Schmitz KH, Emplaincourt PO, Jacobs DR, Leon AS: 36. Scheers T, Philippaerts R, Lefevre J: SenseWear-determined physical
Compendium of physical activities: an update of activity codes and MET activity and sedentary behavior and metabolic syndrome. Med Sci Sports
intensities. Med Sci Sports Exerc 2000, 32:S498–S504. Exerc 2013, 45:481–489.
14. Owen N, Bauman A, Brown W: Too much sitting: a novel and important 37. Reis JP, Dubose KD, Ainsworth BE, Macera CA, Yore MM: Reliability and
predictor of chronic disease risk? Br J Sports Med 2009, 43:81–83. validity of the occupational physical activity questionnaire. Med Sci Sports
15. Dunstan DW, Kingwell BA, Larsen R, Healy GN, Cerin E, Hamilton MT, Shaw Exerc 2005, 37:2075–2083.
JE, Bertovic DA, Zimmet PZ, Salmon J, Owen N: Breaking up prolonged 38. Chau JY, Van Der Ploeg HP, Dunn S, Kurko J, Bauman AE: Validity of the
sitting reduces postprandial glucose and insulin responses. Diabetes Care occupational sitting and physical activity questionnaire. Med Sci Sports
2012, 35:976–983. Exerc 2012, 44:118–125.
16. Stephens BR, Granados K, Zderic TW, Hamilton MT, Braun B: Effects of 1 day 39. Craig CL, Marshall AL, Sjöström M, Bauman AE, Booth ML, Ainsworth BE,
of inactivity on insulin action in healthy men and women: interaction Pratt M, Ekelund U, Yngve A, Sallis JF, Oja P: International physical activity
with energy intake. Metabolism 2011, 60:941–949. questionnaire: 12-Country reliability and validity. Med Sci Sports Exerc
17. Healy GN, Dunstan DW, Salmon J, Cerin E, Shaw JE, Zimmet PZ, Owen N: 2003, 35:1381–1395.
Objectively measured light-intensity physical activity is independently 40. Wareham NJ, Jakes RW, Rennie KL, Mitchell J, Hennings S, Day NE: Validity
associated with 2-h plasma glucose. Diabetes Care 2007, 30:1384–1389. and repeatability of the EPIC-Norfolk physical activity questionnaire.
18. Pescatello LS, Murphy D, Costanzo D: Low-intensity physical activity Int J Epidemiol 2002, 31:168–174.
benefits blood lipids and lipoproteins in older adults living at home. 41. Baecke JAH, Burema J, Frijters JER: A short questionnaire for the
Age Ageing 2000, 29:433–439. measurement of habitual physical activity in epidemiological studies.
19. Healy GN, Matthews CE, Dunstan DW, Winkler EA, Owen N: Sedentary time Am J Clin Nutr 1982, 36:936–942.
and cardio-metabolic biomarkers in US adults: NHANES 2003–06. 42. Clark BK, Thorp AA, A h Winkler E, Gardiner PA, Healy GN, Owen N, Dunstan
Eur Heart J 2011, 32:590–597. DW: Validity of self-reported measures of workplace sitting time and
20. Brown W, Bauman A, Bull F, Burton N: Development of Evidence-based breaks in sitting time. Med Sci Sports Exerc 2011, 43:1907–1912.
Physical Activity Recommendations for Adults (18–64 years). 43. Lynch BM, Friedenreich CM, Khandwala F, Liu A, Nicholas J, Csizmadi I:
In Australian Government Department of Health. 2012. Development and testing of a past year measure of sedentary behavior:
21. Garber CE, Blissmer B, Deschenes MR, Franklin BA, Lamonte MJ, Lee IM, the SIT-Q. BMC Public Health 2014, 14:899.
Nieman DC, Swain DP: American College of Sports Medicine position 44. Wijndaele K, De Bourdeaudhuij I, Godino JG, Lynch BM, Griffin SJ, Westgate
stand. quantity and quality of exercise for developing and maintaining K, Brage S: Reliability and validity of a domain-specific last 7-d sedentary
cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently time questionnaire. Med Sci Sports Exerc 2014, 46:1248–1260.
healthy adults: guidance for prescribing exercise. Med Sci Sports Exerc 45. Bennie JA, Timperio AF, Crawford DA, Dunstan DW, Salmon JL: Associations
2011, 43:1334–1359. between social ecological factors and self-reported short physical
22. Dunstan DW, Howard B, Healy GN, Owen N: Too much sitting - a health activity breaks during work hours among desk-based employees.
hazard. Diabetes Res Clin Pract 2012, 97:368–376. Prev Med 2011, 53:44–47.
23. Todd AI, Bennett AI, Chrisitie CJ: Physical implications of prolonged sitting 46. Baranowski T: Validity and reliability of self-report measures of physical
in a confined posture - a literature review. J Ergon Soc S Afr 2007, 19:7–21. activity: an information-processing perspective. Res Q Exerc Sport 1988,
24. Healy GN, Dunstan DW, Salmon J, Cerin E, Shaw JE, Zimmet PZ, Owen N: 59:314–327.
Breaks in sedentary time - beneficial associations with metabolic risk. 47. Plasqui G, Bonomi AG, Westerterp KR: Daily physical activity assessment
Diabetes Care 2008, 31:661–666. with accelerometers: new insights and validation studies. Obes Rev 2013,
25. Ryde GC, Brown HE, Peeters GM, Gilson ND, Brown WJ: Desk-based 14:451–462.
occupational sitting patterns: weight-related health outcomes. 48. Troiano RP: Large-scale applications of accelerometers: new frontiers and
Am J Prev Med 2013, 45:448–452. new questions. Med Sci Sports Exerc 2007, 39:1501.
26. Chau JY, van der Ploeg HP, Dunn S, Kurko J, Bauman AE: A tool for 49. Freedson PS, Melanson E, Sirard J: Calibration of the computer science
measuring workers’ sitting time by domain: the Workforce Sitting and applications, Inc. accelerometer. Med Sci Sports Exerc 1998,
Questionnaire. Br J Sports Med 2011, 45:1216–1222. 30:777–781.
Pedisic et al. BMC Public Health 2014, 14:1249 Page 9 of 9
http://www.biomedcentral.com/1471-2458/14/1249

50. ABS: 1220.0 - Australian Standard Classification of Occupations (ASCO) Second

Edition. Canberra: Statistics ABo ed; 1997.
51. Oliver M, Schofield GM, Badland HM, Shepherd J: Utility of accelerometer
thresholds for classifying sitting in office workers. Prev Med 2010, 51:357–360.
52. Landis JR, Koch GG: The measurement of observer agreement for
categorical data. Biometrics 1977, 33:159–174.
53. Helmerhorst HJ, Brage S, Warren J, Besson H, Ekelund U: A systematic
review of reliability and objective criterion-related validity of physical
activity questionnaires. Int J Behav Nutr Phys Act 2012, 9:103.
54. McCormack G, Giles-Corti B, Milligan R: The test-retest reliability of habitual
incidental physical activity. Aust N Z J Public Health 2003, 27:428–433.

doi:10.1186/1471-2458-14-1249
Cite this article as: Pedisic et al.: Workplace Sitting Breaks Questionnaire
(SITBRQ): an assessment of concurrent validity and test-retest reliability. BMC
Public Health 2014 14:1249.

Submit your next manuscript to BioMed Central

and take full advantage of:

• Convenient online submission

• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution

Submit your manuscript at

www.biomedcentral.com/submit