This feasibility study is reported according to the CONSORT guidelines for pilot and feasibility trials (17). We conducted a complex one-year mixed-method, prospective single-arm feasibility study in Tromsø, Norway from September 2017. The study consisted of a six-month intervention programme (October 2017-March 2018) comprising instructor-led high-intensive exercise and dietary- and habit change counselling, followed by a six-month follow-up with daily activity monitoring by a Polar activity tracker (AT) (March 2018 to September 2018). Each participant was interviewed three times through the intervention and follow-up period.
Evaluation criteria used for deciding whether to proceed to a full scale RCT were thematically organized as: 1) recruitment and enrolment and 2) adherence and side-effects. Based on previous experience we expected 25 % positive response to postal invitations, and that few responders would be excluded during telephone and baseline screening. We outlined several success criteria for these aspects (Table 1).
We expected 75 % overall group mean attendance to the exercise group sessions, and higher mean group attendance to the three dietary- and habit change counselling group sessions, respectively. We further considered a drop-out rate up to 20% to be acceptable and expected that less than 10% would drop out due to exercise-induced injury (Table 2).
We monitored screening and study logistics, study team cooperation and data collection, to improve the setup of a future RCT. Finally, we interviewed participants at three time points to obtain their feedback on study participation.
Telephone-screened participants received an invitation letter including study information and several comprehensive questionnaires. A written informed consent to participate in the study was obtained when they met for further screening sequentially at three sites: The Clinical Trial Unit (CTU) at the University Hospital of Northern Norway (UNN), the UiT exercise research laboratory at Alfheim stadium, and at Stamina health clinic downtown Tromsø. Table 4 outlines the eligibility criteria at baseline screening.
* 10-years risk for acute heart attack or stroke, including cardiovascular death (medium to high risk): > 8 % risk for persons at age <65 years, >12 % risk for persons at age 65-74 years.
The study included four visits: Baseline screening (BS), 3 months after start of intervention (V1), 6 months after start of intervention (V2) and 6 months after end of intervention (V3). Several structured questionnaires, clinical and instrument-based examinations, blood tests, urine samples and semi-structured interviews were scheduled at different visits as shown in Table 5.
Table 5 Study activity and data collection schedule
Activity/year
|
Invitations and phone interviews
|
Baseline screening and consent
|
Start exercise
|
2017
|
2018
|
Week in study
|
1-3
|
3-4
|
5
|
6
|
7
|
8
|
9
|
10
|
11
|
12
|
13
|
14
|
15
|
16-17
|
18-19
|
20-27
|
28-29
|
36, 43, 50
|
56-57
|
Measurement point
|
|
BS
|
|
|
|
|
|
|
|
|
|
|
|
|
V1
|
|
V2
|
|
V3
|
Questionnaires
|
|
x
|
|
|
|
|
|
|
|
|
|
|
|
|
x
|
|
x
|
|
|
Food frequency questionnaire (FFQ)
|
|
x
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
x
|
|
|
Diet (3-day registration)
|
|
x
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Measurements, clinical examinations, blood tests, physical function tests
|
|
x
|
|
|
|
|
|
|
|
|
|
|
|
|
x
|
|
x
|
|
x
|
Exercise
|
|
|
x
|
x
|
x
|
x
|
x
|
x
|
x
|
x
|
x
|
x
|
x
|
x
|
x
|
x
|
|
|
|
Individual nutrition counselling
|
|
|
x
|
x
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Nutrition group-session
|
|
|
|
|
|
x
|
|
|
|
x
|
|
|
x
|
|
|
|
|
|
|
Habit change group-session
|
|
|
|
|
|
|
x
|
|
x
|
|
|
x
|
|
|
|
|
|
|
|
Activity monitor manual download
|
|
|
|
x
|
|
|
x
|
|
|
x
|
|
|
x
|
|
x
|
|
x
|
x
|
x
|
Semi-structured interviews
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
x
|
|
|
|
x
|
Focus-group interviews
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
x
|
|
|
Questionnaire data
Information on diet was recorded at BS and V2 by a validated food frequency questionnaire (FFQ) described elsewhere (20). Data were imported into the food and nutrient composition database and calculation system KBS (KBS, version 7.3, database AE14, University of Oslo, Oslo, Norway) where energy, nutrients and food groups were calculated.
We used a questionnaire to obtain information on education, social network, chronic diseases, symptoms and ailments, diseases in the family, use of medication, smoking status, and alcohol consumption. The questionnaire also included full versions or specific items from validated instruments to measure PA by the “Saltin–Grimby Physical Activity Level Scale” (18, 21), the self-reported physical activity questionnaire in the Nord-Trøndelag Health Study (HUNT 1) (22) and the International Physical Activity Questionnaire (IPAQ) (23), sleep by the Bergen Insomnia Scale (24), satisfaction with life by the Satisfaction With Life Scale (25), self-esteem by the Rosenberg Self Esteem Scale (26, 27), self-efficacy by the General Self-efficacy Scale (GSE) (28, 29), mental health by the SCL-10 (30, 31), and Self-perceived health by EQ-5D-5L (32).
Physical capacity testing
All measurements were performed using standard methods by trained test personnel.
Cardiovascular and lung function
Blood pressure, resting heart rate, resting electrocardiogram (ECG), and lung capacity were measured at time points SI, V1, and V2. Blood pressure and resting heart rate (HR) were measured on the right arm of all participants by a Dinamap ProCare 300 monitor (GE Healthcare, Norway). The mean of the last two blood pressure measurements and the last HR measurement were used as measured values. Measurements were recorded three times at one-minute intervals, after two minutes’ seated rest. Spirometry was performed using Diagnostica Spirare spirometry model SPS 330 (Diagnostica AS, Norway).
Cardiorespiratory fitness
Maximal oxygen uptake (V̇O2max) was measured at time point SI, V1, and V2 during walking/running on a treadmill. An incremental test to exhaustion was performed to assess V̇O2max on a motorized treadmill (Woodway GmbH, Weil an Rhein, Germany). To safeguard participants, ECG recordings were conducted at CTU prior to the test visit. Further, the test personnel asked participants two questions before start of the test: 1) "Has your doctor ever told you that you have heart disease that impedes intense physical activity?" and 2) "Have you lately been breathless or have you experienced chest pain when walking uphill?”
During the test the participants wore a Polar RS400 heart rate (HR) monitor (Polar Oy, Finland) connected to Polar H10 HR sensor chest belt (Polar Oy, Finland), and a face mask connected to Cosmed K5, a portable ergo spirometry system (Cosmed SRL, Rome, Italy) set in mixing chamber mode, which was placed on the participants´ back. The test personnel encouraged regularly the participants to continue to exhaustion.
The starting speed and incline (0-4%) for the incremental test to exhaustion were based on a subjective assessment by the test personnel, by evaluating the respiratory exchange ratio (RER) during the warmup. Thereafter the speed was set to either 4 or 5 km·h-1, and the inclination of the treadmill increased with 2% every second minute until exhaustion. Thirty seconds before next increase in incline, the participants answered if they could manage an increase of incline with 2% with a thumb up or down indicating yes or no, respectively. When refusing an increased incline, the participants were instructed to keep walking until they felt exhausted and signal that they wanted to stop. Ventilatory parameters, and pulmonary gas exchange were measured every 10 seconds, and a RER ³ 1.05 was, in combination with a plateau of the V̇O2 –work rate curve, used to determine if the participants reached V̇O2max (33). V̇O2max was calculated as a mean of the three highest consecutive 10 seconds measurements.
Strength testing
We tested participant’s maximum dynamic muscle strength in three exercises at S1 and V2, according to the protocol for testing one repetition maximum (1 RM) (34), which was defined as the heaviest weight the participant could handle: 1) Squat (IT7006 45° Leg Press / Hack Squat), 2) seated pulldown (Technogym selection pro LAT Vertical Traction), and 3) Seated incline chest press (Technogym Pure Strength - Incline chest press).
We tested the participant’s rate of force development (RFD) at BS and V2 by mounting a pressure sensor plate (ALU4 2003, HUR Labs Oy, Kokkola, Finland) on the footplate of the squat testing equipment. The sensor plate was connected to a portable laptop through a USB cable and monitored with the manufacturer´s software (Force platform software suite, HUR Labs Oy, Kokkola, Finland).
Flexibility and balance
Maximal passive range of motion in neck, shoulder, hip, knee, and ankle were measured at time point SI and V2. The range of motion was tested by goniometry. Testing protocol was developed according to Norkin & White (35). Balance was planned to be measured at BS and V2 by “one-foot standing”-test as maximal time standing on one foot with closed eyes (36) as a part of the flexibility test session.
Blood parameters
Non-fasting venous blood samples were collected with standard methods, and consecutively analysed at the Department of Laboratory Medicine, UNN. Haemoglobin, glucose, glycated haemoglobin (HbA1c), total cholesterol, high-density and low-density lipoprotein (HDL-and LDL)-cholesterol, triglycerides, creatinine, creatine kinase (CK), alanine aminotransferase (ALAT), aspartate aminotransferase (ASAT), gamma glutamyl transferase (GGT) and vitamins D2 , D3 and 25-hydroxy vitamin D were analysed at BS, V1, and V2. Thyroxin (FT4) and Thyroid stimulating hormone (TSH) were analysed at BS.
Adiposity
Adiposity was defined by anthropometric measures; BMI (body weight in kilograms divided by height in meters squared (kg/m2)), waist circumference (measured at the umbilical level in centimetres by a measurement tape), and body composition (body fat and lean mass) measured by dual-energy X-ray absorptiometry (DEXA) (Lunar GE Prodigy Advance, GE Medical Systems).
Activity and heart rate measurements
Activity tracker
One week prior to intervention start, all participants received a Polar M430 AT (Polar, Kempele, Finland) to collect PA- and HR data. A Polar user account was created for each participant and included information from baseline screening on BMI and maximum HR. Privacy was maintained by not disclosing the login credentials to participants, only storing non-identifiable information on the account, and by deactivating the AT GPS (global positioning system). Participants were instructed to wear the AT all day and all night for six months, and only take it off during re-charging on Sundays. Participants wore a heart rate sensor chest belt (Polar H10) during the exercise sessions to obtain accurate HR measurements. HR zones followed Polar standards; Very Light (51-60% of HRmax), Light (61-70% of HRmax), Moderate (71-80% HRmax), Hard (81-90 % of HRmax), and Very Hard (91-100% of HRmax). The chest HR- sensors used at exercise sessions were stored at the gym and handed out to the participants before each session. At the end of intervention, all participants were invited to wear the watch and continue to share data for six additional months.
In addition, participants wore the AT without heart rate sensor in daily free living. Each participant contributed daily information on step count, energy expenditure, wear-time and number of minutes in sleep, sedentary behaviour, and light-, medium-, and vigorous PA, respectively. Only data from “valid days”, defined as at least 10 hours of wear time according to Troiano et al. (37), were included in analysis.
Accelerometer
A hip-worn ActiGraph wGT3X-BT (ActiGraph, LLC, Pensacola, United States) accelerometer was used to measure PA and sedentary behaviour for 24 h for eight consecutive days. The participants were instructed to perform their usual daily activities and only remove the accelerometer when performing water activities. ActiLife software (ActiGraph, LLC, Pensacola, United States) version 6.13.3 was used for device initialization, set-up (100 Hz sampling rate), and data download (10 s epoch, 3-axes). Non-wear time was identified using Hecht et al. (38) wear-time algorithm and excluded from analysis. We calculated intensity variables using triaxial activity count (i.e. vector magnitude (VM)) cut-off suggested by Sasaki et al. (39) Kozey-Keadle et al. (40) and Peterson et al. (41). Each minute of activity was classified as sedentary behaviour (≤ 149 VM), light PA (150-2689 VM), moderate PA (2690-6166 VM), or vigorous PA (≥ 6167 VM).
Interviews
The main focus of the study was to investigate how participants experienced the various intervention measures and to reveal potential improvement when planning an
RCT. Individual semi-structured interviews were conducted by a trained researcher at V1 and V3. By using a life-world perspective as described by Kvale and Brinkmann (42), the participants’ experiences and points of view were illuminated (43, 44).
We developed an interview guide with fixed predefined questions to ensure that key topics/questions were covered in the interviews. The questions were not necessarily given in the same order in all interviews, and it was possible to record and/or follow up on other relevant topics that arose during the interviews. Main topics during the interviews were attitudes and subjective norms toward health and activity before study start, motivation and barriers to exercise or to be more active, motivation and barriers to have a more healthy diet, study participation experiences, life style changing factors, effect of intervention on further life style change, and experiences with the intervention.
At V2, the participants gathered for two focus group interviews. The main topics of these focus groups were similar as in the interviews at V1. All interviews were audio recorded.
Data storage
The CTU at UNN used REDCap electronic data capture tools for study data collection and management. REDCap (Research Electronic Data Capture) is a secure, web-based application designed to support data capture for research studies, providing 1) an intuitive interface for validated data entry; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for importing data from external sources (45).
Secured data were transferred to the Department of Community Medicine, UiT while questionnaire data were manually entered into the same secured database.
The intervention
The intervention comprised altogether 22 weeks starting from week 5 in study. During the first 10 intervention weeks, nutritional education sessions based on the Nordic Nutrient Recommendations (46), and psychologist-led group counselling sessions based on “Implementation Intentions”- theory (47), were carried out in addition to group-based exercise.
The participants themselves chose the appropriate time of attendance at the various group sessions. During the last twelve weeks participants engaged in exercise sessions only.
More detailed description of the intervention program:
Exercise focused on endurance, strength, balance, and flexibility, and included two instructor-led sessions per week at the private enterprise “Stamina Trening”, Tromsø. See Table 6 for exercise intervention plan details.
Table 6. Exercise intervention plan
Exercise type (weekday)
|
Week no. (stage)
|
Total session duration (minutes)
|
Exercise intensity (% of max HR/% of 1RM*) + interval length
|
Focus
|
Bike spinning (Thursdays)
|
1-3 (1)
|
<35
|
<80%
|
Introduction to exercise theory, familiarize with exercise program
|
4-5 (2)
|
<42
|
Slightly increased
|
Preparing for high intensity
|
6-11 (3)
|
<50
|
85%/<3 minutes
|
Increase stamina
|
12-18 (4)
|
<55
|
85-92%/<4 minutes (4x4 intervals)
|
Increase stamina
|
Brisk walking uphill (Thursdays)
|
19-22
|
<60
|
85-92%/<4 minutes (4x4 intervals)
|
Adaption to outdoor activity, increase stamina
|
Aerobic hall sessions (Tuesdays)
|
1-3 (1)
|
Aerobic: 20-25 Strength: 20-25 Total: 40-50
|
<80%
|
Introduction to exercise theory, familiarize with exercise program
|
|
4-5 (2)
|
Aerobic: 20-25 Strength: 20-25 Total: 40-50
|
Slightly increased
|
Rhythmic aerobic exercises on step box 45-50s work/15-10s break; 3-4 exercises on each leg x3-4 times. Exercise beat 120-130 bpm**
|
|
6-11 (3)
|
Aerobic: 20-25 Strength: 25-30 Total: 50
|
85%/<3 m Pulse step Borg scale 14-15
|
Rhythmic aerobic exercises on step box 45-50s work / 15-10s break; 3-4 exercises on each leg x 3-4 times. Exercise beat 120-130 bpm. When increased stamina, heighten step/increase weights
|
|
12-22 (4)
|
<50
|
85- 92%/<4 minutes
|
Short intervals ad modum Tabata (20s intensity/10s pause) 4x4 + strength on the lower body
|
Maximum resistance training exercise
|
1-5 (1)
|
10
|
80% of 1 RM
|
Instructor-led hack-squat from extended knee to 90 degrees knee arch. 3x5 repetitions, start at 85% of 1RM. When able to perform > 5reps; increase by 2.5-5 kg. Approx. 3-minute rest. Instructed to perform concentric phase fast, eccentric phase slow
|
6-22 (2)
|
10
|
80% of 1 RM
|
Peer-supervised, episodic instructor supervised
|
Daily activity
|
1-22
|
|
|
Encouraging increase at every instructor-led session
|
*One repetition maximum. ** Beats per minute
Each Tuesday participants did aerobic hall sessions indoors with various bodyweight or light dumbbells exercises, emphasizing endurance, balance, flexibility, and muscular strength. Each Thursday the participants attended indoor bicycle spinning sessions mainly to increase endurance. The indoor spinning- sessions protocol involved 19 supervised and progressively challenging sessions based on a high intensity interval training (HIIT) model, shown superior to moderate intensity exercise for increasing aerobic capacity in healthy older adults (48) and in patients with coronary artery disease (49).
During the last two weeks, spinning was replaced by outdoor exercise, preparing the participants for brisk walking based on the 4x4 interval model (50). Finally, during exercise group sessions, all participants carried out 10 minutes of resistance exercise to improve maximum strength in the leg and hip extensor muscles in the hack-squat apparatus also used for testing. Half of the participants carried out the exercise before the group training started and half of the group after the workouts.
During exercise group sessions the instructors would regularly encourage the participants to increase and to be involved in various daily activities. Each participant’s attendance to different sessions were recorded, as the effects of endurance HIIT may partially depend on how well participants follow the intended exercise protocol. We evaluated protocol adherence during multiple cycling-based interval exercise sessions where participants received “live” instructions using a Technogym projector system. Data on protocol adherence were extracted from 7 sessions evenly distributed throughout the study period.
Dietary intervention was based on Norwegian dietary recommendations published by the Norwegian Directorate of Health and based on the Nordic Nutrition Recommendations 2012 (46). The main objective of the nutrition intervention was to enable participants to achieve long-term reduced energy intake and healthier eating habits. All participants completed a three-day food diary before one individual counselling session with a nutritionist during the first two weeks of the intervention. After the individual session, the nutritionist met with two groups of eight participants each during three sessions weeks 8, 12 and 15 in study. The first focused on general food knowledge, the second on theories for practical cooking, and the third on advice for food purchasing.
Habit change intervention included three group sessions of eight participants each led by a psychologist, in weeks 9, 11 and 14 in study. The psychologist introduced the participants to mental exercises according to “Implementation-intention strategies” (51) to increase their awareness of automated actions and to create new situational actions in a more desirable way linked to their lifestyle goals use of to change behaviour, so-called “action plans”. When properly formulated and memorized, “action plans” are shown to affect participants behaviour by creating new links between specific situations and actions. Participants were encouraged to develop, test, and discuss one individual action plan related to PA, and one related to diet change by linking a critical situation to an action that they wanted to perform in this situation (e.g. "Whenever my grandchildren visit, I will suggest a walk to the close-by playground") (52). The purpose was to prepare participants to behave in line with their new plans when they face these situations in everyday life.
User involvement and steering group
Patient and public involvement (PPI) in health research is important to enhance the quality and relevance of research (53). The project steering group included two user representatives. The representative from The National Association for Heart and Lung Disease (LHL) contributed with patient experience as well as practical advice on study planning and conduct. Second, a representative of the municipality of Tromsø also contributed experiences from a complex lifestyle intervention aimed to reduce sick-leave from work.
Statistical analysis
Feasibility including recruitment, participants’ adherence and side-effects was investigated by comparison with the pre-specified feasibility criteria (Tables 1 and 2). Feasibility criteria for various primary outcomes are reported.
Recruitment aspects and exercise- induced injury drop-out rate were calculated by Wilson score interval to estimate 95% confidence intervals (CI) for proportions (54).
Feasibility criteria for various secondary outcomes are also reported. AT wear-time is reported as percentage of participants with high use of AT. Attendance to various intervention activities and interviews is reported as mean session attendance with 95% CI. For nominal data raw count (number, %) are reported. Deviations from the cycling-based exercise protocol using HR monitors were analysed with coefficients of variance (CVs), where a higher percentage indicates a larger deviation from the intended HR-based exercise plan. Additionally, paired t-tests were used to evaluate changes in protocol adherence between the first and last session. A p-value of <0.05 was considered statistically significant.
A researcher trained in qualitative research methodology performed verbatim transcriptions of the interviews except for the interviews at V3. A professional enterprise, DigForsk AS Kirkenes, performed the verbatim transcription from these interviews. All transcriptions were performed within weeks from the interviews. We used the computer software QSR NVivo 12 Plus (QSR International, Pty Ltd) as a tool for structuring data in the analysis process.