Physical activity after cardiac arrest; protocol of a sub-study in the Targeted Hypothermia versus Targeted Normothermia after Out-of-Hospital Cardiac Arrest trial (TTM2)

Highlights • The results of this study will provide novel information about physical activity among OHCA-survivors.• Physical activity after a cardiac arrest is currently an important knowledge-gap.• The results of this study have a potential to improve support and rehabilitation for OHCA-survivors.


Introduction
Coronary artery disease (CAD) is the most common cause of out-ofhospital cardiac arrest (OHCA) 1,2 . One of the major risk factors for CAD is physical inactivity 3,4 . Physical activity is a preventive strategy for CAD, and physical exercise training is an important part of cardiac rehabilitation 5,6 . International guidelines recommend moderate intensity physical activity for 30 min, 5 days weekly for primary prevention 4,7 and moderate to vigorous physical activity for at least 20 min, 3À5 times weekly for secondary prevention of CAD 5À12 . Many individuals who had a myocardial infarction (MI) are not physically active to these recommended levels 13 . Previous studies have identified poor functional capacity 14 , cognitive impairment 15 , depression 16 kinesiophobia (fear of movement) 17 and fatigue 18 as factors negatively affecting the level of physical activity and adherence to exercise-based rehabilitation among CAD-patients 19 . OHCA-survivors often experience additional physical, cognitive and emotional problems related to brain injury secondary to the circulatory arrest 20À24 . They may therefore have higher risk of physical inactivity, compared to other populations of CAD-patients, for whom current exercise-based cardiac rehabilitation programs have been designed. In a previous study, OHCA-survivors with cognitive impairment had lower exercise capacity 25 . Another study on self-reported cardiacanxiety among CA-survivors showed that 11À15% avoid physical activities due to concerns about their heart 26 . Other factors related to physical activity are age and gender 24,27 .
The Targeted Hypothermia versus Targeted Normothermia after Out-of-Hospital Cardiac Arrest trial (TTM2-trial, clinicaltrials.gov NCT02908308) is an international multi-center study comparing targeted temperature of 33℃ post-arrest to targeted normothermia with early treatment of fever 28 . This study is a cross-sectional prospective sub-study within the TTM2-trial focusing on physical activity.
The primary aim of this study is to investigate whether OHCAsurvivors have lower levels of self-reported physical activity compared to a non-CA control group who had acute MI. Additional aims are to explore potential predictors of physical inactivity (older age, female gender, problems with general physical function, global cognition, mental processing speed/attention, anxiety symptoms, depression symptoms, kinesiophobia and fatigue), and to investigate the relationship between self-reported and objectively measured physical activity among OHCA-survivors.
Our hypotheses are: 1 OHCA-survivors will have lower levels of self-reported physical activity compared to MI-controls. 2 Older age, female gender, problems with general physical function, global cognition and mental processing speed/attention, anxiety symptoms, depression symptoms, kinesiophobia and fatigue will be predictors of lower levels of self-reported physical activity amongst OHCA-survivors. 3 Self-reported levels of physical activity among OHCA-survivors will show at least moderate agreement with objectively measured levels of physical activity.

Study population
The TTM2 main trial has included 1900 adult OHCA-survivors of a presumed cardiac origin 28 . This sub-study is performed at 8 of the 61 TTM2 sites in Sweden (n = 5), Denmark (n = 1) and the United Kingdom (n = 2). Inclusion and exclusion criteria for the TTM2 trial have been described 28 . Additional exclusion criteria for this sub-study are >80 years of age, a Clinical Frailty Score 29 of 8 or 9 before the CA, a dementia diagnosis, inability to speak the local language well enough to complete the tests without an interpreter, active drug abuse, and wheel-chair bound. The same exclusion criteria apply for the MIcontrols with the addition of having suffered a CA.

Study procedure
All survivors in the main TTM2-trial are invited to a structured face-toface follow-up visit 6 months after their OHCA where information on health-related quality of life, overall functional outcome, return-towork, general physical function, global cognition, mental processing speed/attention and cardiovascular risk collected 30 . As part of the cardiovascular risk assessment, OHCA-survivors answer two selfreported questions about their physical activities (Table 1) based on recommended levels for primary 4,7 and secondary prevention of CAD 5À12 . At the end of the main trial follow-up, participants at selected TTM2-sites are provided written and verbal information about this substudy (clinicatrials.gov NCT03543332) and are invited to participate.
For those who consent, an additional follow-up meeting is scheduled within approximately 4 weeks, i.e. 6À7 months after the CA. Before this meeting, questionnaires regarding anxiety symptoms, depression symptoms, kinesiophobia and fatigue are distributed to the participant by regular mail. At the meeting, participants return the questionnaires and are provided an accelerometer to wear for one week as an objective measure of physical activity. During the same week, participants are asked to register their physical exercise training in a diary. When the entire accelerometer-registration is completed, they answer the two self-reported questions about physical activity once again and return them in a pre-paid envelope together with the accelerometer and the training diary ( Table 2).
MI-controls are recruited from a cohort of participants with confirmed and treated acute ST-elevation or Non-ST-elevation MI who received coronary angiography, but without occurrence of CA. They are matched according to the following criteria in descending order: 1) country 2) date of OHCA/MI 3) gender and 4) age. If a MI-control participant declines the second-best match is approached to obtain a 1:1 ratio between OHCA-survivors and MI-controls. MI-controls are recruited from one site in each country.
The MI-controls are invited to a single face-to-face visit approximately 6À7 months after their MI and receive questionnaires regarding anxiety symptoms, depression symptoms, kinesiophobia and fatigue two weeks before this meeting by regular mail. At the follow-up visit, they return the questionnaires and perform the same tests on general physical function, global cognition and mental processing speed/attention as the OHCA-survivors complete in the main TTM2 trial follow-up (Table 2). Information about cardiovascular risk factors, including the two self-reported questions regarding physical activity are also collected.
The first participant of this study was included in July 2018. So far, 105 OHCA-survivors and 88 MI-controls are included. This study is coordinated with another TTM2 sub-study investigating neuropsychological function (clinicaltrials.gov NCT03543371) 31 .

Outcome measures
An overview of the different outcome measures is provided in Table 2. The main difference between OHCA-survivors and MI-controls is that the MI-controls are only seen once at 7 months after the MI. OHCA- survivors will complete the measures both at the main TTM2 follow-up, at the sub-study meeting, during and after which they wear the accelerometer (see Table 2). MI-controls do not wear the accelerometer or complete the training diary.

Self-report physical activity
The two self-reported questions on physical activity used in the main TTM2 follow-up (Table 1) will allow for categorization of the OHCAsurvivors into three ordered groups of physical activity: physical activity below recommended levels for primary prevention, physical activity to a level recommended for primary prevention, or physical activity to a level recommended for primary prevention and secondary prevention for CAD 4À12 ( Table 3). The self-reported questions were designed for the TTM2-trial and their psychometric properties have not been tested.

General physical function
The Timed Stands Test (TST) is an objective measure of lower extremity strength, used as a rapid measure of general physical function 32,33 . The participant is asked to rise 10 times to straight upright position from sitting, as quickly as possible, while the time is measured to the nearest 10th of a second. A shorter time indicates a better functioning of the lower extremity muscles. Age and gender adjusted norm values are available 32 . The test will be interrupted after 60 s to avoid participants' exhaustion and maximum time assigned to participants unable to complete 10 stands. Participants will be categorized into three groups according to performance: 1) normal level 32 , 2) lower than expected given the age-and gender cut-offs but within 60 s and 3) do not manage within 60 s or at all. TST has been validated for rheumatoid arteritis and other chronic diseases 33 but not for CAD-patients or OHCA-survivors.

Global cognition
The Montreal Cognitive Assessment (MoCA) is used and recommended for cognitive screening after CA 34 . The MoCA contains 11 sub-tasks of several cognitive domains as executive functioning, short-term memory and delayed recall combined into a total maximum score of 30. We use the full version 7.1À7.3 and add 1 point for participants who have 12 years or fewer of formal education according to the manual 35 . A score !26 points is considered to be within the average range.

Mental processing speed/attention
Processing speed/attention will be assessed with the Symbol Digit Modalities Test (SDMT) 36 . It was found to discriminate between OHCA-survivors and MI-controls in a previous study 21 . The total amount of correct answers within 90 s are calculated and transformed into age and education adjusted z-scores. Scores 1.5 SD below the mean indicate impaired function 37 .

Anxiety and depression symptoms
Information on emotional difficulties will be collected using the Hospital Anxiety and Depression Scale (HADS) 38 . HADS is a selfreport questionnaire validated in CAD-patients 39,40 and previously administered to CA-survivors 22 . It consists of 14 questions, 7 related to anxiety symptoms (HADS-A), and 7 related to depression symptoms (HADS-D). Each item is scored from 0 to 3 and a total sum score for each subscale (anxiety symptoms and depression symptoms) is calculated (range 0À21) with a cut-off of >7 indicating clinically significant difficulties 38 .

Kinesiophobia
The Tampa Scale for Kinesiophobia Heart (TSK Heart) will be used to evaluate self-reported kinesiophobia. TSK Heart is a modified version of the original Tampa Scale 41 aimed at detecting kinesiophobia in patients with CAD. The questionnaire includes 17 statements/ questions added into a total score that range from 17 to 68. Values >37 represent a high level of self-reported kinesiophobia 17 . TSK Heart has adequate reliability and validity for patients with CAD 42 .

Fatigue
The Multidimensional Fatigue Inventory (MFI-20) is a 20-item selfreported questionnaire that measures fatigue over the last few days in five dimensions: general fatigue, physical fatigue, mental fatigue, reduced activity and reduced motivation. Sum-scores are calculated for each dimension separately and range from 4 to 20, higher scores indicates a higher level of fatigue 43 . MFI-20 has previously been used in a sample of MI-patients 18 , but not for patients with OHCA.

Physical activity, objective assessment
The OHCA-survivors physical activity will be objectively evaluated by the accelerometer Actigraph wGT3X-BT (Actigraph, Florida, USA), found to be an instrument with good validity to assess physical activity among persons with chronic diseases including CAD 44 . The Actigraph wGT3X-BT measures wear time, steps and intensity of physical activity. This data can be used to calculate the duration and frequency of physical activity in different intensities. The OHCAsurvivors will wear the Actigraph for 7 consecutive days. The analysis will include all participants with a minimum of 4 valid days of assessment 45 . A valid day will be defined as a minimum of 10 h wear Table 3 -Categorization on physical activity through self-reported questions and accelerometer.

Self-reported questions
Accelerometer Actigraph time. The Actigraph is programmed to start at 7.00 AM on the day after the sub-study follow-up meeting and stops exactly 7 days later. The accelerometer will be placed over the right hip with an elastic belt and removed when the participants are sleeping or in contact with water 44

À48
. The participants will receive verbal and written information, including information on how to contact study team if necessary during this week, to minimize the risk of missing data.
To analyze the information collected by the accelerometer, the ActiLife software version 6.16.3 will be used with the following settings: normal filter, 30 Hz sampling frequency and 60 s epoch length 45 . As there is currently no specific accelerometer cut points for OHCA-survivors, the commonly used accelerometer cut points from Freedson Adult 49 will be used to analyze minutes of physical activity in moderate and vigorous intensity. The results will be used to categorize the OHCA-survivors into three ordered groups of physical activity (Table 3).

Training diary of physical exercise training
Participants will be asked to fill in a pre-printed training diary of physical exercise, structured as an activity log, including activities they consider as physical exercise training. This is defined as physical activity that is planned, structured and repetitive and aims to improve or maintain physical fitness 50 . Information about what constitutes physical activity and exercise is given to the participant verbally during the follow-up appointment and explained in writing in the training diary. The aim of the training diary is to add information on activities that cannot be measured by the accelerometer e.g. swimming. Participants will specify the type of activity, duration and rate the level of exertion according to the Borg Rating of Perceived Exertion Scale (RPE). The RPE measures exertion on a scale from 6 to 20, where 6 denotes "no exertion at all" and 20 denotes "maximal exertion" 51 .

Socio-demographical and medical characteristics for the participants
Socio-demographic and medical characteristics for the OHCAsurvivors will be obtained from the TTM2 database. The same information will be obtained from the MI-controls at the follow-up meeting (Table 4).

Ethics
The TTM2-trial (EPN-reference number 2015/228) 28 and this substudy (EPN-reference number 2017/933) was approved by the  Regional Ethics Committee at Lund University, Sweden and corresponding ethics committees in the United Kingdom and Denmark. Written informed consent to participate in the sub-study is obtained before the sub-study follow-up for both OHCA-survivors and MI-controls.

Statistical analysis plan
Data will be presented according to the STROBE-guidelines. Descriptive information and patient characteristics for OHCAsurvivors and MI-controls (Table 4) will be presented, with potential differences between the two groups described. Information on missing data will be provided.

Analyses for hypothesis 1
The OHCA-survivors' self-reported physical activity, obtained at the main TTM2-trial follow-up, will be compared to the self-reported physical activity of the MI-controls, obtained at the sub-study meeting.
The three ordered categories of physical activity will be used in the analyses (Table 3). Numbers and percentages of participants in each category will be presented and potential differences in level of physical activity between OHCA-survivors and MI-controls will be analyzed using the non-parametric Mann Whitney U test.

Analyses for hypothesis 2
Physical activity will for these analyses be dichotomized into two categories; not fulfilling the primary prevention or fulfilling at least primary prevention recommendations. Single and multivariable logistic regressions will be used to identify predictors of not fulfilling the primary prevention recommendations. Pre-specified potential predictors are: older age, female gender, problems with general physical function (TST), global cognition (MoCA), mental processing speed/attention (SDMT z-scores), anxiety symptoms (HADS-A), depression symptoms (HADS-D), kinesiophobia (TSK-heart) and fatigue . Effect sizes will be reported as odds ratios.

Analyses for hypothesis 3
In the OHCA-group, retrospective self-reported levels of physical activity will be compared to objective measurements obtained during the same week as recorded by the accelerometer. Both datasets will be categorized as described in Table 3 to allow cross tabulation and analysis. Percent agreement and Cohen's kappa will be calculated. At least moderate agreement is by Cohen's kappa defined as 0.41À0.60 "moderate", 0.61À0.75 "substantial" and 0.80À1.00 "almost perfect" 52,53 .

Sample size calculation
The Swedish national registry of MI-patients 54 use identical two selfreported questions on physical activity as in this study. In our local registry data, 20% of the MI-patients were below recommended levels for primary prevention of physical activity, 52% were at a physical activity level recommended for primary prevention and 28% were physically active to a level recommended for both primary and secondary prevention of CAD.
In lack of published data on physical activity for OHCA-survivors, we used unpublished data on self-reported physical activity (SF-36v2, Physical Function-subscale) from a previous trial 21 . This led to an assumption that 30% of the OHCA-survivors were below recommended levels for primary prevention and that 12% were physically active to a level recommended for primary and secondary prevention of CAD.
Using these two assumed distributions of ordered data and a Mann-Whitney U test, 105 participants are required in each group for a statistical power >80% to detect differences in physical activity between the OHCA survivors and MI-controls at p = 0.05. Assuming a loss of 5%, we aim to include 110 participants in each group.
Due to the explorative design of this study, p-values <0.05 will be used to indicate potentially important differences.

Discussion
We describe a sub-study to a large clinical trial in which physical activity is assessed by two self-reported questions as a part of the OHCA-survivors' cardiovascular risk. This sub-study was designed to complement the main trial by the addition of an objective measure of physical activity and a control group with similar level of CAD, but without the problems related to hypoxic-ischemic brain injury and/or the post-intensive care stay. To our knowledge, this is the first multicenter study specifically focusing on physical activity after OHCA.
The current recommendations for physical activity after CAD are based on self-reports and consensus decisions 3À12 , but information about how well self-reported questions correspond to the actual performed physical activity among OHCA-survivors is unknown. Modern techniques have enabled quantification of physical activity in an objective way.
Little comparable data exist for accelerometer data for the cardiac population 55 , none for OHCA. Accelerometer cut points for different intensities of physical activity may differ between patient populations, influenced by age and exercise capacity. We have chosen the most used accelerometer cut points available 49 but we recognize that they are based on young healthy adults, and were found to overestimate sedentary-light intensity physical activity and underestimate moderate-vigorous intensity physical activity in older adults 56 . The accelerometer does not measure the intensity of all activities for example resistance training. To add some descriptive information to lessen this limitation participants are encouraged to report all their physical exercise training in the training diary.
Our matching criteria for the MI-controls were developed to enable recruitment of participants with similar degree of CAD and pre-event related risk-factors as our OHCA-survivors since this is the group for whom current cardiac rehabilitation programs have been designed. We do not know if current cardiac rehabilitation programs are appropriate for OHCA-survivors. Admittedly, we lack information on the participants' physical and cognitive functional status before the OHCA/MI. We aim to present descriptive information regarding the two groups as reported in Table 4.

Conclusion
This study will provide important information on the levels of physical activity after OHCA in relation to recommended levels for primary and secondary prevention of CAD, and explore predictors of importance for physical inactivity. In addition, our results will provide information on how to measure physical activity in OHCA-survivors.

Credit author statement
All authors have made substantial contributions to the manuscript. KH, GL, ÅBT, TC and NN designed the sub-study. KH is the principal investigator. KH, TC, GL, ÅT, SU, EBN, NN and HF are the coordinating team. CR, HK and MW are national investigators. LGØ, AMG, TRK, MM, MS and HH are site investigators. KH, EBN, MS, MM and HH conducted the follow-up meetings. SU is the trial statistician. KH, GL, ÅBT and TC drafted the manuscript. All authors revised and gave final approval to the manuscript.