Accepted Article
Physical activity and personal factors associated with nurse
1
resilience in intensive care units
Fiona Yu1, Alana Cavadino2, Lisa Mackay3, Kim Ward4, Anna King5, Melody Smith6
Ph.D., Candidate, MN (HONS), School of Nursing, Faculty of Medical and Health Science, University of Auckland, RN in Intensive
Care Unit, Waikato Hospital, Waikato, New Zealand. E-mail: syu037@aucklanduni.ac.nz
2
Ph. D., Biostatistician, School of Population Health, Faculty of Medical and Health Science, University of Auckland, Auckland, New
Zealand. E-mail: a.cavadino@auckland.ac.nz
3
Ph.D., Lecturer, School of Sport and Recreation, Auckland University of Technology, Auckland, New Zealand.
E-mail: lisa.mackay@aut.ac.nz
4
Ph.D., RN, Lecturer, School of Nursing, Faculty of Medical and Health Science, University of Auckland, Auckland, New Zealand. E-
mail: k.ward@auckland.ac.nz
5
Ph.D., BNurs (HONS), RN, Lecturer, School of Nursing, Faculty of Medical and Health Science, University of Auckland, Auckland,
New Zealand. E-mail: a.king@auckland.ac.nz
6
Ph.D., Associate Professor, Co-Associate Head (Research), School of Nursing, Faculty of Medical and Health Science, University of
Auckland, Auckland, New Zealand. E-mail: melody.smith@auckland.ac.nz
Correspondence should be addressed to Fiona Yu, 85 Park Road, Grafton, Auckland, New Zealand (E-mail:
syu037@aucklanduni.ac.nz).
Key words: Axivity AX3 accelerometer, dynamic standing, ICU, job demands and resources model,
leisure-time physical activity, moderate to vigorous physical activity, MVPA, occupational physical activity,
recovery, resilience
A short running title: Physical activity and nurse resilience
Conflict of Interest Statement: The authors declare no conflict of interest.
Funding: Melody Smith was supported by the University of Auckland and Sir Charles Hercus Research
Fellowship (grant number 17/013). Fiona Yu was supported by the University of Auckland Doctoral
Scholarship.
Authors contribution statement:
Fiona Yu: Methodology, Formal analysis, Investigation, Resources, Data curation, Writing-Original draft
preparation, Reviewing and Editing, Visualisation
Alana Cavadino: Reviewing and Editing
Lisa Mackay: Resources, Reviewing and Editing
Kim Ward: Reviewing and Editing, Supervision
Anna King: Reviewing and Editing, Supervision
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/JOCN.15338
This article is protected by copyright. All rights reserved
Accepted Article
Melody Smith: Conceptualisation, Methodology, Resources, Reviewing and Editing, Visualisation,
Supervision, Project administration
This article is protected by copyright. All rights reserved
Accepted Article
1
2
MISS FIONA YU (Orcid ID : 0000-0001-8747-3545)
3
4
5
Article type
: Original Article
6
7
8
Physical activity and personal factors associated with nurse
9
resilience in intensive care units
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Key words: Axivity AX3 accelerometer, dynamic standing, ICU, job demands and resources model,
28
leisure-time physical activity, moderate to vigorous physical activity, MVPA, occupational physical
29
activity, recovery, resilience
30
31
Abstract
32
Aim and objectives: The study aimed to assess intensive care nurses’ resilience, and identify
33
associated personal factors and physical activity behaviours using a job demands-recovery
34
framework.
35
Background: Currently, there is inconsistent evidence as to whether nurse resilience is associated
36
with personal factors, or with physical activity at work or during leisure-time.
37
Design: A cross-sectional study was conducted with nurses from four intensive care units in Auckland,
38
New Zealand.
This article is protected by copyright. All rights reserved
Methods: An online survey was conducted to collect nurses’ personal information and assess their
40
resilience levels using the Connor-Davidson Resilience Scale 25. Participants were nurses working at
41
least 32 hours fortnightly and providing direct patient care. Physical activity was objectively measured
42
using a pair of accelerometers worn on the back and thigh over four consecutive days (two workdays
43
followed by two non-workdays). Bivariable and multivariable regression were used to identify personal
44
factors and physical activity behaviours associated with resilience. (Followed the STROBE checklist)
45
Results: A total of 93 nurses were included in the study. The participants’ average resilience level was
46
low. Resilience was positively associated with the objectively measured physical job demand factors:
47
occupational physical activity, moderate to vigorous physical activity at work, and dynamic standing at
48
work. Resilience was negatively associated with one objectively measured recovery factor: sleep
49
during leisure-time. In multivariable modelling, being married and moderate to vigorous physical
50
activity at work were positively associated with resilience, while not having religious beliefs and sleep
51
during leisure-time were negatively associated with resilience.
52
Conclusions: Resilient nurses have a greater tolerance to high physical activity at work and lower
53
sleep duration during leisure-time. Strategies are needed to improve intensive care nurses’ resilience
54
levels.
55
Relevance to clinical practice: Results may help managers gain a better understanding of the ICU
56
nurses’ characteristics associated with resilience, leading them to develop strategies for improving
57
ICU nurse resilience.
Accepted Article
39
58
59
What does this paper contribute to the wider global clinical
60
community?
61
Impact statement:
62

This study is the first to objectively measure nurses’ physical activity utilising a job demands-
63
recovery framework and to identify the associations between resilience and nurses’ physical
64
activity at work/leisure-time. Dynamic standing, as a novel concept and a new way of
65
measuring physical activity at work, is introduced in this study.
66

67
68
69

Married or religious nurses were found to have higher levels of resilience, compared to single
or non-religious nurses.
Resilient nurses had higher levels of physical activity at work and lower sleep duration during
leisure-time.
70
71
Introduction
72
Health promotion is of great importance for organisations as well as for nurses to support physical
73
and psychological wellbeing, ultimately improving work performance and staff retention. Prolonged
74
nursing shortages result in heavy workloads that lead to increased work stress and turnover, and this
75
profoundly affects nurses physically and mentally (Chan, Tam, Lung, Wong, & Chau, 2013; da Costa
76
& Pinto, 2017). Evidence shows that nurses from intensive care units (ICU) experience a high risk of
77
developing psychological problems, and resilience helps attenuate the effects of adverse outcomes.
This article is protected by copyright. All rights reserved
Mealer et al. (2012) collated data from 744 American ICU nurses studied in 2010 and found that 80%
79
had one or more symptoms of burnout syndrome (emotional exhaustion, depersonalisation, or lack of
80
personal accomplishment) (Mealer et al., 2012). The study also reported that highly resilient nurses
81
(22% of the 744 participants) had a lower profile of burnout syndrome, posttraumatic stress disorder,
82
anxiety or depression (Mealer et al., 2012). Similarly, another study by Rushton, Batcheller,
83
Schroeder, and Donohue (2015) surveyed 114 nurses from six ICUs at four American hospitals and
84
identified that highly resilient nurses exhibited lower stress levels, and that resilience helped protect
85
them from burnout. Increased personal resilience may help nurses improve adaptation to stressful
86
work environments (Hart, Brannan, & De Chesnay, 2014). It may also reduce nurse vulnerability
87
(Jackson, Firtko, & Edenborough, 2007) and mitigate the effects of emotional dissonance (Delgado,
88
Upton, Ranse, Furness, & Foster, 2017). Therefore, understanding resilience is crucial for intensive
89
care nurses to improve their ability and capacity to deal with stressful and challenging situations.
Accepted Article
78
90
91
Scholars have identified that resilience encompasses various characteristics that have been explored
92
in different ways. Resilience is described as an individual’s ability to recuperate and capacity for
93
proactive protection (Youssef & Luthans, 2007). Resilience is also defined as a personality profile of
94
self-control and conscientiousness (Fisk & Dionisi, 2010), and a dynamic process or an innate life
95
force of motivation (Grafton, Gillespie, & Henderson, 2010). A recent study has identified five
96
resilience
97
multidimensional concept, mental immunity, and recovery ability (Aburn, Gott, & Hoare, 2015). Within
98
the context of intensive care nursing speciality, resilience has been explained as nurses’ cognitive
99
flexibility, coping ability and adaptability (Mealer et al., 2012).
characteristics
in
general:
overcoming
adversity
process,
adaptation
capacity,
100
101
Background
102
Resilience and job demands-recovery
103
Using the Job Demands and Resources Model (Bakker & Demerouti, 2007), researchers identified
104
that resilience mitigated the effects of job demands, thus helping nurses recover quickly from work
105
adversity (Yu, Raphael, Mackay, Smith, & King, 2019). Increased job demands can negatively affect
106
nurses physically and psychologically, while sufficient recovery can enhance their wellbeing and work
107
performance (Bakker & Demerouti, 2007). Numerous studies identified the associations between
108
resilience and the negative psychological impact of job demands, such as stress, burnout, fatigue,
109
anxiety/depression, posttraumatic stress disorder and workplace bullying (Yu et al., 2019). There is
110
very little in the literature to connect how physical activity fits within the Job Demands-Resources
111
Model and nurses’ resilience. In particular, a systematic review has identified that only a few studies
112
have objectively measured physical activity in relation to nurses’ job demands (Chappel, Verswijveren,
113
Aisbett, Considine, & Ridgers, 2017).
114
115
Resilience, activity, and recovery
116
Occupational / leisure-time physical activity and recovery
This article is protected by copyright. All rights reserved
Regular exercise can help enhance levels of psychological resilience, thus improving well-being
118
(Ozkara, Kalkavan, Alemdag, & Alemdag, 2016; Silverman & Deuster, 2014). Indeed, studies with
119
nurse populations have shown that the promotion of leisure-time physical activity can attenuate the
120
negative outcomes of job demands, such as back pain, premature death, stress, and burnout
121
(McCarthy, Wills, & Crowley, 2018; Reed et al., 2018; Schluter, Turner, Huntington, Bain, & McClure,
122
2011). Further, research with undergraduate student populations has highlighted the link between
123
resilience and leisure-time physical activity (Hegberg & Tone, 2014; Lines et al., 2018).
Accepted Article
117
124
125
However, it is unclear whether resilience is associated with nurses’ physical activity at work and/or
126
leisure-time. Nurses’ occupational physical activity primarily involves standing and slow walking
127
behaviours, and they engage in a substantially higher amount of light-intensity activity along with
128
moderate-intensity nursing tasks (Chappel et al., 2017). Furthermore, high job demands, long work
129
hours, and frequently rotating shift work patterns are the main barriers preventing nurses from actively
130
exercising during leisure-time (Caruso, 2014; Reed et al., 2018; Schluter et al., 2011).
131
132
It is also unclear whether sleep, as an important recovery factor, is related to nurse resilience. Sleep
133
quality, as an essential health indicator, has a significant impact on nurses’ work performance and
134
personal life (Pérez-Fuentes, Molero Jurado, Simón Márquez, & Gázquez Linares, 2019). Sleep
135
disturbance is related to increased age, family dependents, unhealthy eating, lack of physical activity,
136
and low emotional intelligence (Pérez-Fuentes et al., 2019). Resilience is a protective resource and
137
can help mitigate the negative effect of perceived stress on sleep disturbance, thus improving sleep
138
quality (Liu et al., 2016).
139
140
Physical activity measurement and concepts
141
One issue that has hindered progress in this area is how physical activity has been measured and
142
conceptualised. For example, there has been a general lack of objective measurement of physical
143
activity in relation to nurse resilience. Objective measures, such as using accelerometry, are preferred
144
over subjective measures (e.g. surveys), as they eliminate issues with self-report bias, recall or social
145
desirability, which may overestimate physical activity (Chappel et al., 2017). Accelerometry is a
146
technique for quantifying human movement patterns using accelerometer-based systems.
147
Accelerometers can be worn on the wrist, ankle, hip or on the lower back. The application of multiple
148
accelerometer units allows for accurate determination of physical activity behaviour (such as sitting,
149
standing, moving, or lying) and for estimating the intensity of movement (such as moderate to
150
vigorous physical activities). The time-stamped feature of accelerometry also enables the extraction of
151
specific time-periods for data analysis.
152
153
In the context of this research, comprehending work-related physical activity “demands” and leisure-
154
time “recovery” is important in order to understand the different pathways to resilience. Moderate to
155
vigorous intensity physical activity (such as brisk walking or jogging) is of particular interest. This is
156
because studies have identified the beneficial outcomes of moderate to vigorous physical activity on
This article is protected by copyright. All rights reserved
health, such as improving metabolic rate and decreasing mortality risk (Ekelund et al., 2019; Saint-
158
Maurice, Troiano, Berrigan, Kraus, & Matthews, 2018). Evidence has also highlighted the link
159
between resilience and moderate to vigorous physical activity in studies with different populations
160
(Hegberg & Tone, 2014; Lines et al., 2018; Wermelinger Ávila, Corrêa, Lucchetti, & Lucchetti, 2018).
161
However, all these studies linking moderate to vigorous physical activity to health outcomes utilised
162
subjective measurement.
Accepted Article
157
163
164
The use of accelerometry to measure free living activity behaviours is progressing rapidly. One
165
advancement of interest is the use of multiple accelerometers to detect postures, using algorithms to
166
predict specific activity behaviours. This is opposed to generic measures of physical activity intensity.
167
For example, one recent development of relevance to nursing physical activity behaviours is the
168
conceptualisation and objective measurement of “dynamic standing”. It is defined as “standing with
169
slight movement” which could include normal daily activities such as cooking, washing, or vacuuming
170
(Narayanan, Stewart, & Mackay, 2019). This type of movement aligns well with ICU nurse physical
171
activities as described above. These could include providing hygiene care (oral care and bed baths),
172
turning patients to relieve pressure areas, removing chest drains or pacing wires, preparing
173
medication for infusion pumps, dialysing patients and so on. To date, the concept of dynamic standing
174
has not been explored in workplace health research on nurses.
175
176
There is a clear knowledge gap in understanding relationships between resilience and nurses’
177
physical activity at work (job demands) or during leisure-time (recovery). Accordingly, a research
178
question has been raised for this study: Is resilience related to nurses’ physical activity at work (job
179
demands) and/or during leisure-time (recovery)? Hence, this study aimed to examine ICU nurses’
180
resilience levels in relation to personal factors and physical activity behaviours using a job demands-
181
recovery framework.
182
183
Methods
184
This study, titled “Physical activity and personal factors associated with nurse resilience in intensive
185
care units” was reported following the STROBE checklist (Supplementary File 1).
186
187
Design
188
Setting and ethical approval
189
A study-specific framework of job demands-recovery was developed for contextualising the study aim
190
(Figure 1). The framework categorises nurses’ occupational physical activity, moderate to vigorous
191
physical activity, and dynamic standing as “job demands” during workdays. Conversely, it classifies
192
leisure-time physical activity, moderate to vigorous physical activity, sedentary time, and sleep as
193
“recovery” during non-workdays. Additionally, it considers how resilience is related to the physical job
194
demands in the health impairment process as well as recovery in the motivational process.
195
[Please insert Figure 1 about here]
This article is protected by copyright. All rights reserved
A cross-sectional multi-centre study was designed and conducted in three tertiary teaching hospitals
197
within the Auckland region. Auckland is New Zealand’s largest city with 1.6 million residents, which is
198
approximately one-third of the nation’s population. The three hospitals are serviced by three different
199
District Health Boards that are responsible for public health services in the greater Auckland area.
200
One of these hospitals has two ICUs, while the others have one each. These three hospitals have a
201
collective workforce of approximately 500 ICU nurses and nurse managers (with varying workloads
202
and schedules).
Accepted Article
196
203
204
Four Intensive Care Units from these three hospitals were invited to participate in the study via email
205
between May and June 2019, and the data were then collected from July to October 2019. Ethics
206
approval for the study was granted by the Auckland Health Research Ethics Committee (Ref. 000070).
207
Approval was also obtained from each District Health Board and research site. Participation was
208
voluntary and informed, with each participant required to sign a consent form before answering an
209
online survey and wearing a pair of accelerometers.
210
211
Inclusion/exclusion criteria
212
The study targeted ICU nurses who worked at least 32 hours fortnightly and were involved in direct
213
patient care. It excluded anyone allergic to the Elastoplast medical dressings that were used to attach
214
the accelerometers. It also excluded nurse coordinators, as their workloads differ from ICU nurses
215
who provide direct patient care. The study further required the participants to wear a pair of
216
accelerometers continuously for four consecutive days (two workdays followed by two non-workdays).
217
218
Recruitment and study size
219
Several strategies were utilised for recruitment. Nominated local investigators (independent of the
220
study) assisted the researcher in advertising the research. A $200 prize draw was offered at the
221
completion of data collection at each unit, and each participant received a report of their four-day
222
physical activity. Approximately 500 ICU nurses and nurse managers work in the three selected public
223
hospitals. A targeted sample size of 100 participants was initially identified, based on a previous study
224
which recruited 103 participants from 576 office workers (participation rate 17.8%) in Perth, Western
225
Australia (Tobin, Leavy, & Jancey, 2016). In this earlier study, a total of 103 participants provided
226
sufficient power to identify significant differences in respondents’ physical activity behaviours using an
227
activPAL activity monitor (Tobin, Leavy, & Jancey, 2016). The targeted sample size of 100
228
participants for the current study was further checked using G*Power. At a significance level of 5%, a
229
sample of 100 participants provided 90% power to detect an effect size as small as f² = 0.107 in linear
230
regression analysis, or a “medium” effect size.” (Cohen, 1988; Faul, Erdfelder, Lang, & Buchner,
231
2007).
232
233
Data collection
234
Protocol and measures
235
Survey
This article is protected by copyright. All rights reserved
An online survey was used to measure ICU nurse resilience (25 questions) and personal factors
237
(sociodemographic information, work factors, and subjectively assessed health behaviours; 20
238
questions). Sociodemographic information included age, sex, marital status, religious beliefs, ethnicity,
239
family dependents, highest qualification attained. It also collected information regarding shift patterns,
240
fortnightly work hours, frequency of working night shifts, years of nursing experience, and years of
241
ICU nursing experience. The health behaviours encompassed frequency of physical activity per week,
242
cigarettes smoked per day, cups of coffee consumed per day, frequency of alcohol consumption,
243
usual sleep duration per 24-hour period, sleep medication use, sleep quality over the last 30 days,
244
and general health status. These questions were mainly adapted from the New Zealand Health
245
Survey (Ministry of Health, 2018).
Accepted Article
236
246
247
Resilience was measured using the Connor-Davidson Resilience Scale 25 (Connor & Davidson,
248
2003). This scale assesses resilience from five aspects: personal competence and persistence,
249
negative outcome tolerance, adaptation to change, self-control, and spiritual influences. The answers
250
employ a five-point Likert Scale with 0 representing “not true at all”, 1 “rarely true”, 2 “sometimes true”,
251
3 “often true”, and 4 “true nearly all the time” (Connor & Davidson, 2003). The total scores range from
252
0 to 100, with higher scores indicating greater resilience (Connor & Davidson, 2003). Accordingly, the
253
resilience levels are categorised as “lowest (0-73)”, “low-medium (74-82)”, “high-medium (83-90)” and
254
“high (91-100)” (Connor & Davidson, 2003). The Cronbach’s alpha was 0.93 in the study of Connor
255
and Davidson (2003), indicating that the reliability of this tool makes it pertinent to the current
256
research.
257
258
Objective assessment of physical activity
259
Each participant was issued with a pair of Axivity AX3, triaxial accelerometers, incorporating a
260
temperature sensor and a real-time clock, to wear continuously for four consecutive days (two
261
workdays followed by two non-workdays). Accelerometers were distributed to participants with a
262
unique numerical code used to match the recorded data. One unit was affixed to each participant’s
263
lower back (offset from the spine) and the other onto the anterior aspect of their thigh, using
264
hypoallergenic medical dressings. Figure 2 shows the image of an Axivity AX3 accelerometer, and the
265
locations where a pair of accelerometers were affixed to a participant. OMGUI software (Version
266
1.0.0.30, Open Movement, Newcastle University, UK) was used to configure the accelerometers
267
before wearing, and to download the physical activity data after removal.
268
[Please insert Figure 2 about here]
269
Machine-learning has been utilised to develop algorithms using the data collected from the Axivity
270
AX3 accelerometers. This process enables detection and classification of postures (such as sitting,
271
standing, lying, or moving) and movement intensity (such as sedentary, light, moderate or vigorous
272
levels of physical activity) (Stewart et al., 2018). Movement intensity can also be estimated as being
273
sedentary, or being of a light, moderate or vigorous nature. The protocol for accelerometer wear
274
(dual-placement) and data processing has been validated by previous studies (Duncan et al., 2018;
275
Schneller et al., 2017; Stewart et al., 2018).
This article is protected by copyright. All rights reserved
Accepted Article
276
277
Variables
278
Physical activity behavioural variables were classified following the job demands-recovery framework
279
and are described below:
280
Job demands variables:
281

282
283

284
285
286
287



290
291

292
293
behaviours over two 12-hour shifts.
Moderate to vigorous physical activity comprises moderate intensity and vigorous intensities
of movement over two 12-hour shifts.
Dynamic standing refers to standing with slight movements over two 12-hour shifts.
Recovery variables
288
289
Occupational physical activity consists of standing, dynamic standing, walking, and running

Leisure-time physical activity encompasses dynamic standing, walking and running
behaviours over two non-workdays.
Moderate to vigorous physical activity includes moderate intensity and vigorous intensities of
movement over two non-workdays.
Sedentary time comprises the sum of sitting and lying (excluding sleeping) behaviours over
two non-workdays.
Sleep refers to sleeping over two non-workdays.
294
295
Statistical analyses
296
MATLAB (release 2017b, The MathWorks, Inc., MA, USA) was used to convert the Axivity AX3
297
accelerometer raw data into daily 6-part movement behaviours (lying, sitting, standing, dynamic
298
standing, walking and running) over 24-hours. SPSS version 25 was utilised to analyse the
299
accelerometer data and the online survey information. Descriptive statistics were used to identify
300
participants’ demographic characteristics (frequency and percentage), resilience levels (mean and
301
standard deviation), and physical activity (mean and standard deviation). Shapiro-Wilk test,
302
independent samples t-test, and Chi-square were also undertaken to determine participants’ inclusion
303
and exclusion from analyses. Any participant who failed to wear their accelerometers for time-periods
304
exceeding 104.4 minutes in total were excluded from the study, as missing data can affect the
305
accuracy of overall physical activity measurement.
306
307
Bivariable analysis (linear regression) was performed to identify potential associations between
308
resilience and each independent variable. Variables associated with resilience at p < 0.1 were then
309
included in a multivariable linear regression analysis. Stepwise multivariable regression was
310
employed to remove variables with the largest non-significant p-value from the group step by step in
311
order to achieve the best fit model in predicting resilience. Models were repeated with additional
312
adjustment for ICU unit, to account for potential differences in average levels of resilience between
313
the different ICU settings. Variance inflation factor, tolerance, and coefficient correlations were
314
assessed for collinearity diagnostics. A p-value < 0.05 was set as the level of significance for this
315
study.
This article is protected by copyright. All rights reserved
Accepted Article
316
317
Results
318
Participation rate and compliance rate
319
A total of 374 ICU nurses met the inclusion criteria and were invited to participate in the study. Of
320
these, 132 (35.3%) agreed to participate and signed the consent forms. Of these 132 nurses, 107
321
completed the online survey and wore the accelerometers for four consecutive days. The remaining
322
25 participants were excluded due to unavailable shift schedules (15), injuries (3), or personal
323
reasons (7). Of those 107 participants, one person’s data was lost due to a faulty accelerometer,
324
while 13 failed to continuously wear their accelerometers over the required four consecutive days.
325
The time periods that their accelerometers were unattached varied between 188 minutes and 1,368
326
minutes. The average time-period the accelerometers were not worn was 104.4 minutes (Figure S1 in
327
supporting information). Therefore, the 13 participants were excluded from the study, as the total time
328
their accelerometers were not worn for exceeded 104.4 minutes.
329
330
As a result, 93 of the 106 participants wore the accelerometers continuously over the four consecutive
331
days, giving a compliance rate of 87.7%. A flow diagram (Figure 3) explains the process of the
332
participants’ inclusion and exclusion. There were no statistically significant differences between
333
participants who were included and excluded in terms of resilience, age, ethnicity, self-reported
334
physical activity, and objectively measured moderate to vigorous physical activity at workdays/leisure-
335
time (Table S1 in supporting information). Using the Shapiro-Wilk test, the normality of the distribution
336
of the 93 participants’ resilience scores distribution was confirmed (p = 0.608, Figure S2 in supporting
337
information).
338
[Please insert Figure 3 about here]
339
Sample characteristics
340
Resilience
341
A total of 93 participants were recruited from four intensive care units (ICU1, ICU2, ICU3, ICU4). Of
342
the 93 respondents, 31 (33.3%) and 27 (29.0%) were recruited from ICU1 and ICU2 respectively,
343
while 23 (24.7%) were from ICU3, and the remaining 12 (13.0%) were from ICU4. Table 1
344
summarises the participants’ resilience scores and levels, socio-demographic information, work
345
factors, self-reported health behaviours, and objectively assessed physical activity. The mean
346
resilience score for the total sample was 73.0 ±9.6, indicating the average resilience level was in the
347
lowest category (resilience score 0-73) (Connor & Davidson, 2003). Of the four units, ICU2 had the
348
highest average resilience score at 75.7 ±9.2, while ICU4 had the lowest score of 70.0 ±6.9. Of the 93
349
participants, 55.9% had the lowest levels (0-73), and 29.0% demonstrated low-medium levels (74-82),
350
while only 5.4% exhibited high resilience levels (91-100), and 9.7% scored at high-medium levels (83-
351
90) (Connor & Davidson, 2003). The mean resilience level in each subgroup for each unit and the
352
total sample is shown in Table S2 (in supporting information): Demographic and work-related
353
characteristics by resilience.
354
[Please insert Table 1 about here]
This article is protected by copyright. All rights reserved
Participants’ mean age was 33.9 ±9.6 years old. ICU1 had the youngest group (30.7 ±6.4), while
356
ICU2 contained the oldest (38.1 ±11.6). Of the 93 nurses, 72.0% were between 20 and 34 years old,
357
three quarters (73.1%) were female, 59.1% were unmarried, and 52.7% were religious. Overall, 46.2%
358
of the 93 participants identified as being of European ethnicity, 37.6% had one or more family
359
dependents, and 62.4% attained a postgraduate qualification. It is noted that ICU1 had the highest
360
percentage of non-Europeans and non-family dependents (71.0% of 31). A majority (86.0%) worked
361
fulltime, 8.6% worked night shifts permanently, 73.1% did night shifts every two weeks, and 18.3%
362
had monthly rotated night shifts or only worked days. Two-thirds (66.7%) had worked more than five
363
years in nursing, while 39.8% had experience in the ICU specialty longer than five years. ICU2
364
employed more experienced nurses than other units, as 51.9% of 27 ICU2 participants had 11 years
365
or more of nursing experience, and 40.7% had worked in ICU over 11 years or more.
Accepted Article
355
366
367
Health behaviours
368
Eight-five percent of nurses reported exercising one to five times per week, while only 6.5% never
369
exercised, and 8.5% did not know or preferred not to answer (Table 1). Almost all (98.9%) participants
370
were non-smokers, 22.6% did not drink coffee, and 26.9% did not consume alcohol. Almost two-thirds
371
(72.0%) of respondents had seven or more hours sleep, 86.0% did not use any sleep medication,
372
76.3% had “very good” or “fairly good” sleep quality over the last 30 days, and 63.4% stated that their
373
general health status was “good or fair”. Of the four units, ICU1 had the highest percentage of
374
reporting “fairly good” sleep quality over the last 30 days (71.0% of 31 participants) and “good or fair”
375
general health status (74.2% of 31 participants).
376
377
Physical activity
378
Table 1 shows the participants’ mean physical activity levels per 12-hour shift and per non-workday
379
for each unit and the total sample. The mean level of moderate to vigorous physical activity for the 93
380
participants was 0.8 ±0.6 hours at work (per 12-hour shift) and 0.7 ±0.4 hours in leisure-time (per non-
381
workday). Nurses from ICU1 had the highest mean level of moderate to vigorous physical activity at
382
work (1.0 ±0.7 hours), while ICU3 showed the highest mean level of moderate to vigorous physical
383
activity during leisure time (0.8 ±0.5 hours). The mean level of dynamic standing for the total sample
384
was 2.8 ±0.8 hours per 12-hour shift.
385
386
Table 1 also highlights that the average occupational physical activity level (i.e. non-sedentary time)
387
was 8.9 ±1.2 hours for the total sample per 12-hour shift, and the mean value was similar at each unit.
388
The mean level of leisure-time physical activity (per non-workday) was 3.2 ±1.3 hours from the total
389
sample. There was no vigorous-intensity physical activity measured at any of the units during work
390
hours and during leisure-time. Mean levels for sedentary time and sleep during leisure-time (per non-
391
workday) were 9.9 ±2.2 hours and 8.6 ±1.8 hours respectively, while 3.2 ±1.1 hours was the mean
392
value for sedentary time and 0.3 ±0.5 hours for sleep per 12-hour shift.
393
394
Main results
This article is protected by copyright. All rights reserved
Bivariable analysis
396
Table 2 summaries the results of linear regressions between resilience and the independent variables.
397
The table shows that resilience was significantly associated with sex (male) (ß = -0.3, 95% confidence
398
interval -9.8 to -1.1, p = 0.015), marital status (being married) (ß = 0.3, 95% confidence interval 2.6 to
399
10.2, p = 0.001), religious beliefs (non-religiousness) (ß = -0.4, 95% confidence interval -11.7 to -4.5,
400
p < 0.001), and ethnicity (European) (ß = 0.2, 95% confidence interval 0.6 to 8.4, p = 0.024). It also
401
shows that resilience had an association with the job demand behavioural variables as follows:
402
moderate to vigorous physical activity (ß = 0.2, 95% confidence interval 0.5% to 6.4%, p = 0.021),
403
dynamic standing (ß = 0.3, 95% confidence interval 0.7% to 4.8%, p = 0.010), and occupational
404
physical activity (ß = 0.2, 95% confidence interval 0 to 2.7%, p = 0.047). In addition, the table
405
illustrates that sleep (ß = -0.2, 95% confidence interval -1.9% to -0.1%, p = 0.026), as only one of the
406
recovery variables, was related to resilience.
Accepted Article
395
407
408
Table 2 shows the change in the outcome (resilience score) was associated with a one unit or
409
category change in each predictor. Using “marital status” as an example with “single” as the reference
410
group, those classified as “being married” had, on average, resilience scores that were 6.4 points
411
higher than those who were “single”. Similarly, for religiosity, compared with those classified in the
412
“non-religious group”, those classified as having religious beliefs had, on average, resilience scores
413
that were 8.1 points higher. Their correlation coefficients with resilience for these variables were 0.3
414
(being married) and -0.4 (non-religiousness), indicating that these two predictors have a medium or
415
large significant effect on predicting resilience (±0.1, ±0.3, ±0.5 represent a small, medium, or large
416
effect, respectively (Field, 2014)). Similarly, moderate to vigorous physical activity (ß = 0.2),
417
occupational physical activity (ß= 0.2), dynamic standing (ß = 0.3) and sleep (ß = -0.2) all had a
418
medium significant effect on predicting resilience.”
419
[Please insert Table 2 about here]
420
Multivariable analysis
421
Four assumptions were made before performing the multivariable regression. Firstly, resilience had a
422
linear relationship with each independent variable. Secondly, it was assumed that residuals of the
423
regression were normally distributed. Thirdly, the independent variables were not highly correlated.
424
Lastly, the data should follow a homoscedastic pattern, or the variance is equal.
425
426
Multivariable regression was performed between resilience and the selected variables (if p < 0.1)
427
according to the linear regression results shown in Table 2. These multivariable results were
428
observed for the individual variables when all others were held at their respective means. These
429
selected variables were sex, marital status, religious beliefs, ethnicity, family dependents, frequency
430
of physical activity per week, job demand variables (moderate to vigorous physical activity, dynamic
431
standing, and occupational physical activity), and a recovery variable (sleep). Multicollinearity
432
between these variables was assessed before performing the analysis. Variance inflation factors (VIF)
433
ranged from 1.1 to 2.5, tolerance (T) ranged from 0.4 to 0.9, and correlation coefficients (r) ranged
This article is protected by copyright. All rights reserved
from 0.0 to 0.7. The results indicated that there was no multicollinearity between these variables (VIF
435
< 10, T > 0.2, and r < 0.8).
Accepted Article
434
436
437
A seven-step multivariable linear regression (with and without ICU as a fixed factor) was carried out
438
by removing the variable (with the largest non-significant p value) from the group at each step. The
439
final model is summarised in Table 3. Table 3 shows that the significant predictors for resilience were
440
marital status (being married) (ß = 0.2, 95% confidence interval 1.1 to 8.0, p = 0.011), religious beliefs
441
(non-religiousness) (ß = -0.3, 95% confidence interval -10.0 to -3.1, p < 0.001), moderate to vigorous
442
physical activity over two 12-hour shifts (job demands factor) (ß = 0.2, 95% confidence interval 0.0 to
443
0.1, p = 0.021), and sleep during two non-workdays (recovery factor) (ß = -0.223, 95% confidence
444
interval 0.0 to -0.0, p = 0.013).
445
[Please insert Table 3 about here]
446
The variance inflation factor (VIF), collinearity tolerance (T), and coefficient correlations (R) ranged
447
from 1.0 to 1.1 (VIF <10), 0.9 to 1.0 (T>0.2), and 0.0 to 0.4 (r <0.8) respectively. The range of the
448
absolute standardised ß values was from 0.2 to 0.3, suggesting that religious belief (ß = -0.3) was the
449
most important predictor in the model. The value of R was 32.5%, indicating these four significant
450
predictors accounted for 32.5% of the variation in resilience. There was no association between ICU
451
site and resilience, and adding ICU as a fixed factor to the model did not improve the model fit or have
452
an impact on the other four predictors. Additionally, independent samples t-tests were performed to
453
compare the differences in the mean resilience levels between ICUs, and Table S3 (in supporting
454
information) shows the results. No significant differences in resilience were observed between ICUs;
455
thus no further analysis by units was undertaken.
2
456
457
Table S2 (in supporting information) shows the mean resilience levels for the subgroups of the
458
identified significant demographic factors: marital status and religious beliefs. The table highlights that
459
mean resilience level for single nurses was 70.4 ±8.3, lower than those of the married (76.8 ±10.1).
460
Mean resilience score for nurses who reported having religious beliefs was 76.8 ±9.4, compared with
461
a mean score of 68.7 ±7.9 for nurses who reported not having religious beliefs.
462
463
Discussion
464
This study measured ICU nurses’ resilience levels and the associated personal and physical activity
465
behavioural factors across four main ICU sites in Auckland, in New Zealand. This study is the first to
466
objectively measure nurses’ physical activity utilising a job demands-recovery framework and to
467
introduce a new concept of “dynamic standing” as a nursing activity behaviour for analysis. Data
468
describe a low resilience level in this study cohort. Marital status (being married vs. being single) and
469
religious beliefs (religiousness or non-religiousness) were found to be related to resilience. Within the
470
job demands-recovery framework, resilient nurses had higher levels of physical workloads and lower
471
duration of sleep in leisure time. Marital status, religious beliefs, moderate to vigorous physical activity
472
at work, and sleep during leisure-time were associated with resilience in the final multivariable model.
473
Sex and ethnicity were no longer associated with resilience after adjustment in this final model.
This article is protected by copyright. All rights reserved
Additionally, a test for multiplicity was not performed due to the exploratory nature of this study
475
(StatsImprove, 2019).
Accepted Article
474
476
477
Resilience and personal factors
478
The study found that nurses, who were married or religious, exhibited higher levels of resilience,
479
respectively compared to the single or non-religious nurses. This may be because married or religious
480
nurses have more social connectedness, as resilience is positively related to social support (Yu et al.,
481
2019). The finding is in support of a previous study which found that the married nurses reported
482
higher levels of life satisfaction and less emotional turmoil (Ang et al., 2018). In contrast, another
483
study concluded that religious beliefs did not play a key role in improving nurse resilience (Hsieh,
484
Hung, Wang, Ma, & Chang, 2016).
485
486
Our data support marital status (being married) as a protective factor for resilience. A settled
487
partnership may enable a couple to support each other to cope with stress from work and daily life,
488
thus improving resilience (Melvin, Gross, Hayat, Jennings, & Campbell, 2012). Improved resilience in
489
the context of high-quality marital relationships are reflected in studies of highly resilient US soldiers
490
(Melvin et al., 2012) and of Chinese HIV patients (Huang, Zhang, & Yu, 2018). Indeed, a high-quality
491
marital relationship was an essential factor in tolerating stress and increasing personal resilience
492
resources (Huang et al., 2018).
493
494
Our findings align with previous research where religiosity was positively associated with resilience
495
(Choi & Hastings, 2019; Fradelos et al., 2018). Religiousness is often regarded as a source of
496
strength for buffering the effects of adversity (Pargament & Cummings, 2010). Nurses believing in a
497
God are convinced that their God can provide them with shelter or a miracle to deal with stressful
498
work environments (Perera, Pandey, & Srivastava, 2018). This religious coping strategy may be
499
accompanied by greater self-efficacy (Pargament & Cummings, 2010), thus resulting in a higher level
500
of resilience (Yu et al., 2019). The religious coping strategy is of particular importance to ICU nurses,
501
as the highly stressful ICU work environment may lead them to have irrational beliefs, such as low
502
frustration tolerance, or condemnation, that negatively contribute to resilience (Pargament &
503
Cummings, 2010).
504
505
Resilience and job demands-recovery
506
A nursing workload refers to the number of tasks that a nurse performs (Alghamdi, 2016). Accordingly,
507
in this study, occupational physical activity, moderate to vigorous physical activity and dynamic
508
standing during workdays can all be regarded as a nurse’s physical workload (or physical demands).
509
Workload has been shown to directly influence the job outcome (Bogaert, Clarke, Willems, &
510
Mondelaers, 2012). Therefore, understanding the relationships between resilience and physical
511
demands can help improve a nurse’s ability to achieve better work outcomes.
512
This article is protected by copyright. All rights reserved
The study found that increased resilience was associated with higher levels of occupational physical
514
activity and with moderate to vigorous physical activity at work. The findings align with previous
515
research where there was a link between resilience and participants’ self-reported physical activity
516
(Hegberg & Tone, 2014; Lines et al., 2018; Wermelinger Ávila et al., 2018). The findings indicate that
517
nurses who are able to accommodate a high physical workload may be resilient and corroborates
518
other studies (Lanz & Bruk-Lee, 2017; Mealer et al., 2012). In previous research, resilience appeared
519
to have a mediating role on the relationship between stress and job outcomes, and highly resilient
520
nurses had a higher ability to control themselves when faced with high workloads (Lanz & Bruk-Lee,
521
2017). Similarly, negative outcomes from the stressful ICU environment had less impact on highly
522
resilient nurses (Mealer et al., 2012). Nevertheless, future research should be conducted to further
523
identify the associations between resilience and physical work activity, and the results would give
524
more credence to the current study findings.
Accepted Article
513
525
526
Additionally, the positive relationship between resilience and dynamic standing indicates that resilient
527
nurses spend a greater proportion of their working time in dynamic standing. As dynamic standing is
528
representative of typical nursing tasks, this may be indicative of greater productivity at work. However,
529
further research is necessary to explore dynamic standing in relation to resilience under ICU clinical
530
settings.
531
532
Unexpectedly, the study identified that sleep was the only recovery factor negatively associated with
533
resilience. The finding suggests that the nurses who sleep longer may have lower resilience levels, or
534
resilient nurses may have a lower sleep duration. The finding is in line with the conclusion from a
535
previous longitudinal study that surveyed 903 Chinese students from 16 universities and colleges in
536
Hong Kong and Macau (Wong et al., 2013). The study found that sleep duration and quality had a
537
direct impact on the satisfaction of physical wellbeing and indirect (via mood) effect of psychological
538
health (Wong et al., 2013). However, this finding is not supported by the current study which found
539
that sleep quality was not related to nurses’ resilience according to the online survey results. There
540
may be more underlying factors related to this current finding that sleep during leisure-time was
541
associated with resilience. Therefore, further research is required to understand the relationship
542
between resilience and sleep recovery for ICU nurses.
543
544
Strengths and limitations
545
This study has a number of strengths, which include the objective measurement of physical activity,
546
use of job demands-resources theory to drive analyses, comparatively high participation rate, and use
547
of established survey measures. It also takes a comprehensive approach to understanding resilience
548
in a unique workforce. Moreover, a high compliance rate (87.7%) of accelerometer wear highly
549
strengthens this study.
550
551
However, this study is not without its limitations; its cross-sectional nature means causality cannot be
552
determined. Larger multicentre studies are warranted to corroborate findings presented here.
This article is protected by copyright. All rights reserved
Additionally, the resilience scale used is from a self-reported questionnaire, so the results may have
554
been affected by the participants’ interests and time-period available in which to answer the questions.
555
However, objectively measured physical activity data minimises bias in reporting physical job
556
demands.
Accepted Article
553
557
558
The participants’ recollection, personal interests, and social desirability were the primary potential
559
sources of bias in this study. Some participants may have overreported their health behaviours when
560
answering the online survey. For example, they may have overestimated or underestimated their
561
frequency of physical activity per week, sleep hours, sleep quality during the past month, or health
562
status. Some participants may not have answered the questions carefully during the resilience
563
assessment. Additionally, some participants may have filed inaccurate reports while wearing
564
accelerometers. To reduce the potential sources of bias, the researcher arranged a generous time-
565
period for participants to answer the online survey and reiterated that the success of the study
566
depended on the accuracy of the information from them.
567
568
The study offered a $200 prize draw to each unit as well as a four-day activity graph to each
569
participant, and these may have stimulated the ICU nurses’ interest in participation. It is possible
570
some reactivity may have occurred in response to the research; however, participants were
571
encouraged to undertake their usual physical activities. Nonetheless, changes to usual behaviours
572
may have occurred during the testing period.
573
574
The results may also have been affected by the number of days the participants worked or had off
575
prior to the day when their accelerometers were fitted. Most nurses from the research sites are shift
576
workers with randomly assigned rosters. Some, working full-time, may have three or four 12-hour
577
shifts within a week, but may only be required to do two 12-hour shifts the following week. The study
578
was thus designed to require the participants to wear a pair of accelerometer over four consecutive
579
days (two workdays followed two non-workdays) to maximise recording of their work and leisure-time
580
physical activity. During a testing week, some participants had three or four 12-hour consecutive shifts,
581
while others only worked two continuous 12-hour shifts. Therefore, those who worked one or two 12-
582
hour shift(s) already before their accelerometers were fitted were not as refreshed as someone who
583
had their accelerometers fitted on their first day back at work after a few days off.
584
585
Furthermore, the results may have been affected by the participants’ workload, as some of them
586
might not have had an ICU patient to care for while wearing the accelerometers. Extending the time
587
that the accelerometers are worn to seven days instead of four may have provided more reliable
588
assessments of job demands. However, the protocol for two workdays followed by two non-workdays
589
was chosen to minimise participant variability of job demands and recovery periods.
590
591
Some participants reported itchiness on the skin areas where the accelerometers were fitted, while
592
some found it difficult to sleep while wearing these devices. Therefore, they removed the
This article is protected by copyright. All rights reserved
accelerometers and reaffixed them when they felt comfortable. However, some of these participants
594
were excluded when their non-wear times exceeded 104.4 minutes, and this may have also affected
595
the participation rate in this study (although noting no significant differences in key variables between
596
those included and excluded were found). Finally, the generalisability of this study is limited due to the
597
ICU-based participants, as the ICU clinical setting is different from other nursing roles. Further
598
research with nurses working in non-ICU areas, or with the ICU nurses from the same research sites
599
at a later date, may improve the external validity of the study.
Accepted Article
593
600
601
Conclusions
602
The low level of nurse resilience in this study indicates that it is necessary to develop resilience
603
training programs. Married or religious nurses were found to have higher levels of resilience,
604
compared to singles or non-religious nurses. Increased resilience was found to be related to higher
605
levels of nurses’ occupational physical activity, moderate to vigorous physical activity, and dynamic
606
standing for job demands, but it was associated with less sleep duration in leisure-time for recovery.
607
Marital status (being married), religiosity (having religious beliefs), and moderate to vigorous physical
608
activity at work were protective factors in predicting resilience, while sleep on non-workdays was
609
negatively associated with resilience. The findings also suggest that highly resilient nurses have a
610
greater ability to accommodate a substantial amount of physical activity at work and require less sleep
611
duration during leisure-time for recovery. Future research with a larger sample size is needed to
612
further analyse the associations between nurse resilience and physical activity at work and during
613
leisure-time.
614
615
Relevant to clinical practice
616
The findings may help managers gain a better understanding of the ICU nurses’ characteristics
617
associated with resilience, leading them to develop strategies to improve ICU nurse resilience. For
618
example, setting up a network, such as a social club or website, for ICU nurses may increase their
619
social connectedness preventing them developing mental health problems like depression (Santini,
620
Koyanagi, Tyrovolas, Mason, & Haro, 2015). Mindfulness and self-compassion training accompanied
621
with positive religious coping strategies, such as spiritual connectedness, meditation or praying may
622
also help increase nurse resilience (Perera et al., 2018).
623
624
The findings imply that enhancing resilience is vital for ICU nurses to tolerate their physical job
625
demands. Therefore, it is imperative to develop resilience training programs and workplace support
626
for ICU nurses. Previous studies have shown that mindfulness-based stress reduction, event-
627
triggered counselling sessions, and exercise are effective for improving resilience (Mealer et al.,
628
2014). Mindfulness-based stress reduction techniques such as yoga or meditation, can be taught to
629
maintain or improve resilience levels. Combining experienced nurses with new staff members in
630
event-triggered counselling sessions can be utilised to promote resilience. Exercise programs,
631
tailored to meet personal interests, can be delivered by mobile phone, email or website. For example,
632
a self-care poster campaign along with personalised physical activity feedback could be used to
This article is protected by copyright. All rights reserved
encourage nurses to interact with colleagues, set personal goals, and self-monitor progress (Raney &
634
Zanten, 2019). Additionally, several sessions of progressive muscle relaxation combined with music
635
could be offered to ICU nurses to help reduce their stress and fatigue levels, thus improving resilience
636
(Ozgundondu & Metin, 2019). Overall, all these strategies may help ICU nurses increase their
637
resilience levels, thus accommodating high physical workload and enhancing wellbeing. Developed
638
strategies may assist nurses to maintain a healthy profile and improve patient quality of care.
Accepted Article
633
639
640
Funding
641
MS was supported by the University of Auckland and Sir Charles Hercus Research Fellowship (grant
642
number 17/013). FY was supported by the University of Auckland Doctoral Scholarship.
643
This article is protected by copyright. All rights reserved
References
645
Aburn, G., Gott, M., & Hoare, K. (2015). What is resilience? An integrative review of the empirical
Accepted Article
644
646
647
648
649
literature. Journal of Advanced Nursing, 72(5), 980-1000. doi:10.1111/jan.12888
Alghamdi, M. G. (2016). Nursing workload: A concept analysis. Journal of Nursing Management, 24,
449-457. doi:10.1111/jonm.12354
Ang, S. Y., Uthaman, T., Ayre, T. C., Mordiffi, S. Z., Ang, E., & Lopez, V. (2018). Association between
650
demographics and resilience: A cross-sectional study among nurses in Singapore. International
651
Nursing Review, 65(3), 459-466. doi:10.1111/inr.12441
652
653
Bakker, A. B., & Demerouti, E. (2007). The job demands-resources model: State of the art. Journal of
Managerial Psychology, 22(3), 309-328. doi:10.1108/02683940710733115
654
Bogaert, P. V., Clarke, S., Willems, R., & Mondelaers, M. (2012). Nurse practice environment,
655
workload, burnout, job outcomes, and quality of care in psychiatric hospitals: A structural
656
equation model approach. Journal of Advanced Nursing, 69(7), 1515–1524.
657
doi:10.1111/jan.12010
658
659
660
Caruso, C. C. (2014). Negative impacts of shiftwork and long work hours. Rehabilitation Nursing,
39(1), 16-25. doi:10.1002/rnj.107
Chan, Z. C. Y., Tam, W. S., Lung, M. K. Y., Wong, W. Y., & Chau, C. W. (2013). A systematic
661
literature review of nurse shortage and the intention to leave. Journal of Nursing Management,
662
21, 605-613. doi:10.1111/j.1365-2834.2012.01437.x
663
Chappel, S. E., Verswijveren, S. J. J. M., Aisbett, B., Considine, J., & Ridgers, N. D. (2017). Nurses’
664
occupational physical activity levels: A systematic review. International Journal of Nursing
665
Studies, 73, 52-62. doi:10.1016/j.ijnurstu.2017.05.006
666
Choi, S. A., & Hastings, J. F. (2019). Religion, spirituality, coping, and resilience among African
667
Americans with diabetes. Journal of Religion & Spirituality in Social Work: Social Thought, 38(1)
668
doi:10.1080/15426432.2018.1524735
669
670
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ:
Lawrence Earlbaum Associates.
671
Connor, K. M., & Davidson, J. R. T. (2003). Development of a new resilience scale: The Connor-
672
Davidson resilience scale (CD-RISC). Depression and Anxiety 18:76–82 (2003), 18, 76-82.
673
doi:10.1002/da.10113
This article is protected by copyright. All rights reserved
da Costa, B., & Pinto, I. (2017). Stress, burnout and coping in health professionals: A literature review.
Accepted Article
674
675
676
Journal of Psychology and Brain Studies, 1(1:4), 1-8.
Delgado, C., Upton, D., Ranse, K., Furness, T., & Foster, K. (2017). Nurses' resilience and the
677
emotional labour of nursing work: An integrative review of empirical literature. International
678
Journal of Nursing Studies, 70, 71-88. doi:10.1016/j.ijnurstu.2017.02.008
679
Duncan, S., Stewart, T., Mackay, L., Neville, J., Narayanan, A., Walker, C., . . . Morton, S. (2018).
680
Wear-time compliance with a dual-accelerometer system for capturing 24-h behavioural profiles
681
in children and adults. International Journal of Environmental Research and Public Health, 15(7),
682
1296. doi:10.3390/ijerph15071296
683
Ekelund, U., Tarp, J., Steene-Johannessen, J., Hansen, B. H., Jeeris, B., Fagerland, M. W., . . . Lee, I.
684
(2019). Dose-response associations between accelerometry measured physical activity and
685
sedentary time and all cause mortality: Systematic review and harmonised meta-analysis. Bmj,
686
366(I4570) doi:10.1136/bmj.l4570
687
Faul, F., Erdfelder, E., & Lang, A., & Buchner, A. (2007). G*Power 3: A flexible statistical power
688
analysis program for the social, behavioral, and biomedical sciences. Behavior Research
689
Methods, 39, 175–191. doi:10.3758/BF03193146
690
691
Field, A. (2014). Discovering statistics using IBM SPSS statistics (4th ed.). London: SAGE.
692
693
Fisk, G. M., & Dionisi, A. M. (2010). Chapter 7: Building and sustaining resilience in organizational
694
settings: The critical role of emotion regulation. Research on Emotion in Organizations, 6, 167-
695
188. doi:10.1108/S1746-9791(2010)0000006011
696
Fradelos, E. C., Latsou, D., Mitsi, D., Tsaras, K., Lekka, D., Lavdaniti, M., . . . Papathanasiou, I. V.
697
(2018). Assessment of the relation between religiosity, mental health, and psychological
698
resilience in breast cancer patients. Contemporary Oncology, 22(3), 172–177.
699
doi:10.5114/wo.2018.78947
700
701
702
703
704
Grafton, E., Gillespie, B., & Henderson, S. (2010). Resilience: The power within. Oncology Nursing
Forum, 37(6), 698-705. doi:10.1188/10.ONF.698-705
Hart, P. L., Brannan, J. D., & De Chesnay, M. (2014). Resilience in nurses: An integrative review.
Journal of Nursing Management, 22(6), 720-734. doi:10.1111/j.1365-2834.2012.01485.x
Hegberg, N. J., & Tone, E. B. (2014). Physical activity and stress resilience: Considering those at-risk
705
for developing mental health problems. Mental Health and Physical Activity,
706
doi:10.1016/j.mhpa.2014.10.001
This article is protected by copyright. All rights reserved
Hsieh, H. F., Hung, Y. T., Wang, H. H., Ma, S. C., & Chang, S. C. (2016). Factors of resilience in
Accepted Article
707
708
emergency department nurses who have experienced workplace violence in Taiwan. Journal of
709
Nursing Scholarship, 48(1), 23-30. doi:10.1111/jnu.12177
710
Huang, J., Zhang, J., & Yu, N. X. (2018). Close relationships, individual resilience resources, and
711
well-being among people living with HIV/AIDS in rural China. AIDs Care, 30(55), S49–S57.
712
doi:10.1080/09540121.2018.1496222
713
Jackson, D., Firtko, A., & Edenborough, M. (2007). Personal resilience as a strategy for surviving and
714
thriving in the face of workplace adversity: A literature review. Journal of Advanced Nursing,
715
60(1), 1-9. doi:10.1111/j.1365-2648.2007.04412.x
716
Lanz, J. J., & Bruk-Lee, V. (2017). Resilience as a moderator of the indirect effects of conflict and
717
workload on job outcomes among nurses. Journal of Advanced Nursing, 73(12), 2973-2986.
718
doi:10.1111/jan.13383
719
Lines, R. L. J., Ducker, K. J., Ntoumanis, N., Thøgersen-Ntoumani, C., Fletcher, D., McGarry, S., &
720
Gucciardi, D. F. (2018). Stress, physical activity, and resilience resources: Tests of direct and
721
moderation effects in young adults. Sport, Exercise, and Performance Psychology,
722
doi:10.1037/spy0000152
723
Liu, X., Liu, C., Tian, X., Zou, G., Li, G., Kong, L., & Li, P. (2016). Associations of perceived stress,
724
resilience and social support with sleep disturbance among community-dwelling adults. Stress
725
and Health, 32(5), 578-586. doi:10.1002/smi.2664
726
McCarthy, V. J. C., Wills, T., & Crowley, S. (2018). Nurses, age, job demands and physical activity at
727
work and at leisure: A cross-sectional study. Applied Nursing Research, 40, 116-121.
728
doi:10.1016/j.apnr.2018.01.010
729
Mealer, M., Conrad, D., Evans, J., Jooste, K., Solyntjes, J., Rothbaum, B., & Moss, M. (2014).
730
Feasibility and acceptability of a resilience training program for intensive care unit nurses.
731
American Journal of Critical Care, 23(6), 97. doi:10.4037/ajcc2014747
732
Mealer, M., Jones, J., Newman, J., McFann, K. K., Rothbaum, B., & Moss, M. (2012). The presence
733
of resilience is associated with a healthier psychological profile in intensive care unit (ICU)
734
nurses: Results of a national survey. International Journal of Nursing Studies, 49(3), 292-299.
735
doi:10.1016/j.ijnurstu.2011.09.015
736
Melvin, K. C., Gross, D., Hayat, M. J., Jennings, B. M., & Campbell, J. C. (2012). Couple functioning
737
and post-traumatic stress symptoms in US army couples: The role of resilience. Research in
738
Nursing & Health, 35, 164-177. doi:10.1002/nur.21459
This article is protected by copyright. All rights reserved
Ministry of Health. (2018). New Zealand health survey. Retrieved from https://www.health.govt.nz/nz-
Accepted Article
739
740
741
health-statistics/national-collections-and-surveys/surveys/new-zealand-health-survey
Narayanan, A., Stewart, T., & Mackay, L. (2019). A dual-accelerometer system for detecting human
742
movement in a free-living environment. Medicine and Science in Sports and Exercise,
743
doi:10.1249/MSS.0000000000002107
744
Ozgundondu, B., & Metin, Z. G. (2019). Effects of progressive muscle relaxation combined with music
745
on stress, fatigue, and coping styles among intensive care nurses. Intensive & Critical Care
746
Nursing, 54, 54-63. doi:10.1016/j.iccn.2019.07.007
747
748
749
Ozkara, A. B., Kalkavan, A., Alemdag, S., & Alemdag, C. (2016). The role of physical activity in
psychological resilience. Baltic Journal of Sport & Health Sciences, 3(102), 24-29.
Pargament, K. I., & Cummings, J. (2010). Anchored by faith: Religion as a resilience factor. In J. W.
750
Reich, A. J. Zautra & J. S. Hall (Eds.), Handbook of adult resilience (pp. 193-210). New York:
751
The Guilford Press.
752
Perera, C. K., Pandey, R., & Srivastava, A. K. (2018). Role of religion and spirituality in stress
753
management among nurses. Psychological Studies, 63(2), 187-199. doi:10.1007/s12646-018-
754
0454-x
755
Pérez-Fuentes, M. C., Molero Jurado, M. M., Simón Márquez, M. M., & Gázquez Linares, J. J. (2019).
756
Analysis of sociodemographic and psychological variables involved in sleep quality in nurses.
757
International Journal of Environmental Research and Public Health, 16(20), 3846.
758
doi:10.3390/ijerph16203846
759
Raney, M., & Zanten, E. V. (2019). Self-care posters serve as a low-cost option for physical activity
760
promotion of hospital nurses. Health Promotion Practice, 20(3), 354-362.
761
doi:10.1177/1524839918763585
762
Reed, J. L., Prince, S. A., Pipe, A. L., Attallah, S., Adamo, K. B., Tulloch, H. E., . . . Reid, R. D. (2018).
763
Influence of the workplace on physical activity and cardiometabolic health: Results of the multi-
764
centre cross-sectional Champlain nurses’ study. International Journal of Nursing Studies, 81, 49-
765
60. doi:10.1016/j.ijnurstu.2018.02.001
766
Rushton, C. H., Batcheller, J., Schroeder, K., & Donohue, P. (2015). Burnout and resilience among
767
nurses practicing in high-intensity settings. American Journal of Critical Care, 24(5), 412-420.
768
doi:10.4037/ajcc2015291
This article is protected by copyright. All rights reserved
Saint-Maurice, P. F., Troiano, R. P., Berrigan, D., Kraus, W. E., & Matthews, C. E. (2018). Volume of
770
light versus moderate-to-vigorous physical activity: Similar benefits for all-cause mortality?
771
Journal of the American Heart Association, 7(7), e008815. doi:10.1161/JAHA.118.008815
772
Santini, Z. I., Koyanagi, A., Tyrovolas, S., Mason, C., & Haro, J. M. (2015). The association between
Accepted Article
769
773
social relationships and depression: A systematic review. Journal of Affective Disorders, 175, 53-
774
65. doi:10.1016/j.jad.2014.12.049
775
Schluter, P. J., Turner, C., Huntington, A. D., Bain, C. J., & McClure, R. J. (2011). Work/life balance
776
and health: The nurses and midwives e-cohort study. International Nursing Review, 58(1), 28-36.
777
doi:10.1111/j.1466-7657.2010.00849.x
778
Schneller, M. B., Bentsen, P., Nielsen, G., Brond, J. C., Ried-Larsen, M., Mygind, E., & Schipperijn, J.
779
(2017). Measuring children’s physical activity: Compliance using skin-taped accelerometers.
780
Medicine & Science in Sports & Exercise, 1261-1269. doi:10.1249/MSS.0000000000001222
781
782
Silverman, M. N., & Deuster, P. A. (2014). Biological mechanisms underlying the role of physical
fitness in health and resilience. Interface Focus, 4(5) doi:10.1098/rsfs.2014.0040
783
Statslmprove. (2019). Multiple testing: When should we adjust for multiplicity? Retrieved from
784
https://www.statsimprove.com/en/multiple-testing-when-should-we-adjust-for-multiplicity/
785
Stewart, T., Narayanan, A., Hedayatrad, L., Neville, J., Mackay, L. M., & Duncan, S. (2018). A dual-
786
accelerometer system for classifying physical activity in children and adults. Medicine & Science
787
in Sports & Exercise, doi:10.1249/MSS.0000000000001717
788
Tobin, R., Leavy, J., & Jancey, J. (2016). Uprising: An examination of sit-stand workstations, mental
789
health and work ability in sedentary office workers, in Western Australia.55, 359-371.
790
doi:10.3233/WOR-162410
791
Wermelinger Ávila, M. P., Corrêa, J. C., Lucchetti, A. L. G., & Lucchetti, G. (2018). The role of
792
physical activity in the association between resilience and mental health in older adults. Journal
793
of Aging and Physical Activity, 26, 248-253. doi:10.1123/japa.2016-0332
794
Wong, M. L., Lau, E. Y. Y., Wan, J. H. Y., Cheung, S. F., Hui, C. H., & Mok, D. S. Y. (2013). The
795
interplay between sleep and mood in predicting academic functioning, physical health and
796
psychological health: A longitudinal study. Journal of Psychosomatic Research, 74, 271-277.
797
doi:10.1016/j.jpsychores.2012.08.014
798
Youssef, C. M., & Luthans, F. (2007). Positive organizational behavior in the workplace:
799
The impact of hope, optimism, and resilience. Journal of Management, 33(5), 774-800.
800
doi:10.1177/0149206307305562
This article is protected by copyright. All rights reserved
Yu, F., Raphael, D., Mackay, L., Smith, M., & King, A. (2019). Personal and work-related factors
Accepted Article
801
802
associated with nurse resilience: A systematic review. International Journal of Nursing Studies,
803
93, 129-140. doi:10.1016/j.ijnurstu.2019.02.014
804
805
806
807
808
809
810
811
812
813
814
815
816
This article is protected by copyright. All rights reserved
Accepted Article
Tables and figures
Figure 1: The job demands-recovery framework (adapted from Bakker & Demerouti, 2007)
Thigh
Dimensions: 23mm (width) x 32.5mm (length) x
7.6mm (height); Weight: 11g. Further details can be
found at https://axivity.com
Figure 2: Accelerometer (Axivity AX3) image and locations when affixed to participants
This article is protected by copyright. All rights reserved
Lower Back
Accepted Article
Figure 3: A flow diagram for participant recruitment, participation, eligibility, and inclusion
This article is protected by copyright. All rights reserved
Accepted Article
Table 1: Resilience, personal factors, and physical activity for ICU1, ICU2, ICU3, ICU4 and the total sample
ICU
ICU1
ICU2
ICU3
ICU4
Total
sample
Variables
1
Resilience score (Mean (SD))
71.7 (9.5)
75.7 (9.2)
73.1 (11.0)
70.0 (6.9)
73.0 (9.6)
Average resilience level 2
Lowest
Low-
Lowest
Lowest
Lowest
medium
Mean age (SD)
30.7 (6.4)
38.1 (11.6)
33.2 (9.0)
34.2 (10.8)
33.9 (9.6)
Number of participants N (%)
31 (33.3)
27 (29.0)
23 (24.7)
12 (13.0)
93 (100)
Age group
20-34
27 (87.1)
14 (51.9)
17 (73.9)
9 (75.0)
67 (72.0)
35+
4 (12.9)
13 (48.1)
6 (26.1)
3 (25.0)
26 (28.0)
Male
10 (32.3)
7 (25.9)
6 (26.1)
2 (16.7)
25 (26.9)
Female
21 (67.7)
20 (74.1)
17 (73.9)
10 (83.3)
68 (73.1)
Single
22 (71.0)
14 (51.9)
12 (52.2)
7 (58.3)
55 (59.1)
Married
9 (29.0)
13 (48.1)
11 (47.8)
5 (41.7)
38 (40.9)
Yes
19 (61.3)
15 (55.6)
9 (39.1)
6 (50.0)
49 (52.7)
No
12 (38.7)
12 (44.4)
14 (60.9)
6 (50.0)
44 (47.3)
European
9 (29.0)
15 (55.6)
12 (52.2)
7 (58.3)
43 (46.2)
Other
22 (71.0)
12 (44.4)
11 (47.8)
5 (41.7)
50 (53.8)
Family
0
22 (71.0)
16 (59.3)
13 (56.5)
7 (58.3)
58 (62.4)
dependents
1+
9 (29.0)
11 (40.7)
10 (43.5)
5 (41.7)
35 (37.6)
Highest
Undergraduate
16 (51.6)
7 (25.9)
7 (30.4)
5 (41.7)
35 (37.6)
qualification
Postgraduate
15 (48.4)
20 (74.1)
16 (69.6)
7 (58.3)
58 (62.4)
Work part-time or
Part-time
1 (3.2)
8 (29.6)
3 (13.0)
1 (8.3)
13 (14.0)
fulltime
Fulltime
30 (96.8)
19 (70.4)
20 (87.0)
11 (91.7)
80 (86.0)
Work night shifts
Permanently
3 (9.7)
4 (14.8)
1 (4.3)
0
8 (8.6)
Every two weeks
25 (80.6)
17 (63.0)
14 (60.9)
12 (100)
68 (73.1)
Sex
Marital status
Religious beliefs
Ethnicity
attained
Every month/ Don’t work night shifts
3 (9.7)
6 (22.2)
8 (34.8)
0
17 (18.3)
Years of nursing
Under 2
2 (6.5)
2 (7.4)
2 (8.7)
2 (16.7)
8 (8.6)
experience
2 to 5
10 (32.3)
5 (18.5)
8 (34.8)
0
23 (24.7)
6 to 10
14 (45.2)
6 (22.2)
6 (26.1)
6 (50.0)
32 (34.4)
11+
5 (16.1)
14 (51.9)
7 (30.4)
4 (33.3)
30 (32.3)
Years of ICU
Under 2
6 (19.4)
5 (18.5)
6 (26.1)
4 (33.3)
21 (22.6)
nursing
2 to 5
16 (51.6)
7 (25.9)
11 (47.8)
1 (8.3)
35 (37.6)
experience
6 to 10
8 (25.8)
4 (14.8)
3 (13.0)
4 (33.3)
19 (20.4)
11+
1 (3.2)
11 (40.7)
3 (13.0)
3 (25.0)
18 (19.4)
Frequency of
1-2
9 (29.0)
11 (40.7)
5 (21.7)
0
25 (26.9)
physical activities
3-4
12 (38.7)
8 (29.6)
11 (47.8)
5 (41.7)
36 (38.7)
per week
5+
4 (12.9)
5 (18.5)
3 (13.0)
6 (50.0)
18 (19.4)
Never
3 (9.7)
1 (3.7)
2 (8.7)
0
6 (6.5)
Don’t know/I prefer to not answer
3 (9.7)
2 (7.4)
2 (8.7)
1 (8.3)
8 (8.5)
Cigarettes
None
30 (96.8)
27 (100.0)
23 (100)
12 (100)
92 (98.9)
smoked per day
Less than 5
1 (3.2)
0
0
0
1 (1.1)
Cups of coffee
None
8 (25.8)
5 (18.5)
3 (13.0)
5 (41.7)
21 (22.6)
consumed per
1-2
17 (54.8)
13 (48.1)
17 (73.9)
6 (50.0)
53 (57.0)
day
3+
6 (19.4)
9 (33.3)
3 (13.0)
1 (8.3)
19 (20.4)
Frequency of
None
11 (35.5)
4 (14.8)
5 (21.7)
5 (41.7)
25 (26.9)
alcohol
Monthly or less
13 (41.9)
11 (40.7)
11 (47.8)
4 (33.3)
39 (41.9)
consumption
Up to 4 times a month
5 (16.1)
7 (25.9)
6 (26.1)
2 (16.7)
20 (21.5)
4 or more times a week
2 (6.5)
5 (18.5)
1 (4.3)
1 (8.3)
9 (9.7)
This article is protected by copyright. All rights reserved
5-6
12 (38.7)
5 (18.5)
4 (17.4)
1 (8.3)
22 (23.7)
duration per 24-
7+
18 (58.1)
21 (77.8)
17 (73.9)
11 (91.7)
67 (72.0)
hour period
Don’t know
1 (3.2)
1 (3.7)
2 (8.7)
0
4 (4.3)
Sleep medication
No
29 (93.5)
25 (92.6)
17 (73.9)
9 (75.0)
80 (86.0)
use
Yes
2 (6.5)
2 (7.4)
6 (26.1)
3 (25.0)
13 (14.0)
Sleep quality
Very good
2 (6.5)
3 (11.1)
3 (13.0)
3 (25.0)
11 (11.8)
over the last 30
Fairly good
22 (71.0)
18 (66.7)
13 (56.5)
7 (58.3)
60 (64.5)
days
Fairly bad/Very bad
7 (22.6)
6 (22.2)
7 (30.4)
2 (16.7)
22 (23.7)
General health
Excellent / Very good
8 (25.8)
14 (51.9)
8 (34.8)
4 (33.3)
34 (36.6)
status
Good / Fair
23 (74.2)
13 (48.1)
15 (65.2)
8 (66.7)
59 (63.4)
Job demands7
Sedentary
7.2 (0.9)
7.3 (1.0)
7.7 (0.9)
7.8 (0.6)
7.4 (0.9)
Physical work
Light
4.1 (0.6)
4.3 (0.9)
3.9 (0.5)
3.4 (0.5)
4.1 (0.7)
activity, intensity,
Moderate
1.0 (0.7)
0.8 (0.4)
0.7 (0.6)
0.7 (0.2)
0.8 (0.5)
sedentary time,
Vigorous
0
0
0
0
0
and sleep
MVPA3
1.0 (0.7)
0.8 (0.4)
0.7 (0.6)
0.7 (0.2)
0.8 (0.6)
duration per 12-
Sitting
2.7 (0.8)
3.3 (1.5)
2.8 (0.8)
3.6 (0.8)
3.0 (1.1)
hour shift
Standing
4.5 (0.9)
4.4 (1.0)
4.8 (0.8)
4.4 (0.7)
4.5 (0.9)
Dynamic standing
3.4 (0.6)
3.6 (0.9)
3.2 (0.8)
2.9 (4.4)
2.8 (0.8)
Lying (excluding sleeping)
0.2 (0.4)
0.1 (0.1)
0.2 (0.4)
0.2 (0.2)
0.1 (0.3)
Walking
1.1 (0.4)
1.1 (0.3)
1.0 (0.3)
1.1 (0.4)
1.1 (0.4)
Running
0
0
0
0
0
Accepted Article
Usual sleep
Mean (SD)
Unit: hours
Occupational PA
4
9.0 (1.1)
8.3 (1.6)
9.0 (1.1)
8.4 (0.8)
8.9 (1.2)
Sedentary time5
2.9 (1.0)
3.3 (1.4)
3.0 (0.9)
3.9 (0.8)
3.2 (1.1)
Sleep
0.5 (0.6)
0.1 (0.2)
0.4 (0.7)
0.2 (0.2)
0.3 (0.5)
Recovery
Sedentary
19.6 (2.6)
19.3 (2.8)
19.6 (1.8)
19.2 (2.0)
19.5 (2.4)
Leisure- time
Light
3.0 (1.1)
3.1 (1.2)
3.4 (1.3)
3.4 (0.9)
3.2 (1.2)
physical activity,
Moderate
0.7 (0.5)
0.7 (0.5)
0.8 (0.5)
0.7 (0.3)
0.7 (0.4)
intensity,
Vigorous
0
0
0
0
0
sedentary time,
MVPA3
0.7 (0.5)
0.7 (0.5)
0.8 (0.5)
0.7 (0.3)
0.7 (0.4)
and sleep
Sitting
7.3 (2.2)
8.6 (2.1)
8.0 (2.2)
6.9 (1.4)
7.8 (2.2)
duration per non-
Standing
1.8 (0.8)
2.0 (0.8)
2.3 (1.1)
2.4 (1.0)
2.0 (0.9)
Dynamic standing
2.2 (1.1)
2.2 (1.1)
2.6 (1.3)
2.6 (0.9)
2.3 (1.1)
Lying (excluding sleeping)
0
0
0
0
0
Walking
0.7 (0.4)
0.8 (0.5)
0.8 (0.4)
0.9 (0.5)
0.8 (0.4)
Running
0
0.0 (0.1)
0.0 (0.1)
0.1 (0.2)
0.0 (0.1)
Leisure time PA5
2.9 (1.3)
3.0 (1.3)
3.4 (1.5)
3.5 (1.1)
3.2 (1.3)
Sedentary time6
9.9 (2.2)
10.0 (2.1)
10.0 (2.6)
9.8 (2.1)
9.9 (2.2)
Sleep
8.9 (2.1)
8.5 (1.8)
8.4 (1.6)
8.3 (1.8)
8.6 (1.8)
7
workday
Mean (SD)
Unit: hours
1
Resilience score: Ranging from 0 to 100, with higher scores reflecting greater resilience.
2
Resilience level: Lowest (0-73), low-medium (74-82), high-medium (83-90), and high (91-100) (Connor & Davidson, 2003). Of the 93
participants, 55.9% had the lowest levels (0-73), and 29.0% demonstrated low-medium levels (74-82), while only 5.4% exhibited high
resilience levels (91-100), and 9.7% scored at high-medium levels (83-90).
3
MVPA stands for moderate to vigorous physical activity, consisting of moderate-intensity and vigorous-intensity physical activity.
4
Occupational PA stands for occupational physical activity, consisting of standing, dynamic standing, walking, and running
behaviours over two 12-hour shifts.
5
Leisure time PA stands for leisure time physical activity, consisting of dynamic standing, walking and running behaviours over two non-
workdays.
6
Sedentary time comprises the sum of sitting and lying (excluding sleeping) behaviours over two 12-hour shifts or two non-workdays.
7
The unit for each variable in the job demands and recovery is “hours”.
This article is protected by copyright. All rights reserved
Accepted Article
Table 2: Associations between resilience and personal /physical activity behavioural factors
Factors
Unstandardised
Coefficients
Standardised
Coefficients B
Std. Error
Coefficient ß
t
Age group
20-34
1.5
35+
Reference
Male
Reference
Female
-5.4
2.2
0.1
0.7
2.2
-0.3
-2.5
Married
6.4
Single
Reference
1.9
0.3
3.4
No
-8.1
1.8
-0.4
-4.5
-2.9 to 5.9
0.015
-9.8 to -1.1
Yes
Reference
European
Reference
Other
4.5
1.9
0.2
2.3
0
3.4
2.0
0.2
1.7
1+
Reference
Undergraduate
Reference
Postgraduate
-1.3
0.001
2.6 to 10.2
0.000
Ethnicity
0.000
-11.7 to -4.5
0.024
Family dependents
0.024
0.6 to 8.4
0.097
Highest qualification
0.097
-0.6 to 7.4
0.544
2.1
-0.1
-0.6
Work part-time or
0.544
-5.3 to 2.8
0.623
Part-time
Reference
Fulltime
1.4
2.9
0.1
0.5
Work night shifts
0.623
-4.2 to 7.1
0.388
Permanently
Reference
Every two weeks
3.1
3.6
0.1
0.9
0.393
-4.0 to 10.2
Every month / Don’t
-4.4%
4.1
-0.2%
-1.1%
0.991
-8.2 to 8.1
work night shifts
Years of nursing
0.166
Less than 2
Reference
2 to 5
-2.0
3.9
-0.1
-0.5
0.612
-9.7 to 5.7
6 to 10
2.2
3.7
0.1
0.6
0.555
-5.2 to 9.7
11+
3.8
3.8
0.2
1.0
0.320
-3.7 to 11.3
Years of ICU
nursing experience
0.498
0.001
Religious beliefs
experience
Interval
0.015
Marital status
fulltime
95% Confidence
p value
0.498
Sex
attained
Significant
0.457
Under 2
Reference
2 to 5
-3.2
2.7
-0.2
-1.2
0.229
-8.5 to 2.1
6 to 10
-0.9
3.0
-3.9%
-0.3
0.762
-6.9 to 5.1
11+
0.8
3.1
3.2%
0.2
0.806
-5.4 to 6.7
physical activities
1-2
12.3
4.2
0.6
2.9
0.005
3.9 to 20.7
per week
3-4
8.1
4.1
0.4
2.0
0.052
-0.1 to 16.3
5+
8.1
4.4
0.3
1.8
0.070
-0.7 to 16.8
Never
Reference
Don’t know/I prefer
7.5
5.0
0.2
1.5
0.138
-2.5 to 17.5
Frequency of
0.062
to not answer
Cups of coffee
consumed per day
0.526
None
Reference
1-2
-2.8
2.5
-0.1
-1.1
0.264
-7.7 to 2.1
3+
-1.6
3.0
-0.1
-0.5
0.603
-7.6 to 4.5
Frequency of
alcohol consumption
0.754
None
Reference
Monthly or less
0.9
2.5
4.5%
0.3
0.728
-4.1 to 5.8
Up to 4 times a
-1.8
2.9
-0.1
-0.6
0.535
-7.6 to 3.9
month
This article is protected by copyright. All rights reserved
Accepted Article
4 or more times a
-1.5
3.8
-4.6%
-0.4
Usual sleep duration
per 24-hour period
0.612
Reference
7+
-2.1
2.4
-0.1
-0.9
0.371
-6.8 to 2.6
Don’t know
-3.7
5.2
-0.1
-0.7
0.484
-14.1 to 6.3
0.784
No
-0.8
Yes
Reference
2.9
-2.9%
-0.3
Very good
Reference
Fairly good
4.1
3.1
0.2
1.3
0.193
-2.1 to 10.4
Fairly bad/Very bad
2.5
3.5
0.1
0.7
0.474
-4.5 to 9.6
Sleep quality over
the last 30 days
0.784
-6.5 to 4.9
0.395
General health
status
-9.0 to 6.0
5-6
Sleep medication
use
0.692
week
0.280
Excellent /Very good
Reference
Good / Fair
-2.2
2.1
-0.1
-1.1
0.280
-6.3 to 1.8
1
Job demands
MVPA
3.5%
1.5%
0.2
2.3
0.021
0.5% to 6.4%
during two 12-h
Dynamic standing
2.8%
1.1%
0.3
2.6
0.010
0.7% to 4.8%
shifts
Occupational PA2
1.4%
0.7%
0.2
2.0
0.047
0.0 to 2.7%
Recovery during two
MVPA
2.9%
1.9%
0.2
1.6
0.117
-0.8% to 6.6%
non-workdays
Leisure time PA3
0.8%
0.6%
0.1
1.3
0.193
-0.4% to 0.9%
(leisure-time)
Sedentary time4
0.1%
0.4%
3.8%
0.4
0.717
-0.6% to 0.9%
Sleep
-1.0%
0.4%
-0.2
-2.3
0.026
-1.9% to -0.1%
Note: The low values from this table (which become zero if rounded to one digit) have been presented as percentages.
1
MVPA stands for moderate to vigorous physical activity. PA stands for physical activity.
2
Occupational PA consists of standing, dynamic standing, walking, and running behaviours during two 12-h shifts.
3
Leisure-time PA encompasses dynamic standing, walking, and running behaviours during two non-workdays.
4
Sedentary time includes sitting and lying (excluding sleeping)behaviours over two non-workdays.
This article is protected by copyright. All rights reserved
Accepted Article
Table 3: Summary of the final model in resilience for multivariable analysis
B
Variables
2
Standard
Standard-
Error
ised Beta
t
p
95%
Collinearity
Variance
Confidence
Tolerance
Inflation
Interval
Factor
R
Marital status
4.6
1.8
0.2
2.6
0.011
1.1 to 8.04
0.9
1.1
=32.5%
Religious
-6.5
1.7
-0.3
-3.8
0.000
-10.0 to -3.1
0.9
1.1
0.0
0.0
0.2
2.4
0.021
0.0 to 0.1
0.9
1.0
-0.0
0.0
-0.2
-2.5
0.013
-0.0 to -0.0
0.9
1.0
beliefs
MVPA* over
two 12-hour
shifts
Sleep during
two nonworkdays
(leisure-time)
*MVPA stands for moderate to vigorous physical activity.
This article is protected by copyright. All rights reserved