Abstract
Introduction
As a consequence of the ageing population and increasing prevalence of obesity, the number of individuals with stroke has shown a substantial growth in the last years (Mitchell et al., 2015). Stroke is a leading cause of long term disability in adults and has great impact on the quality of life (Jaracz et al., 2014). Individuals with stroke often face upper extremity dysfunction, with only 5% recovering their complete upper extremity function (Kwakkel et al., 2003). This decrease in function can be caused by spasticity, loss of dexterity, sensory loss, and loss of muscle strength (Tsu et al., 2014). The latter appears to make the highest contribution to limitations in activities in daily life (Ada et al., 2006; Canning et al., 2004; Bohannon et al., 1991). Ada et al. (2006) stated that strength training in individuals with stroke has a beneficial effect on both muscle strength and activity, and recommended to implement strength training in the stroke rehabilitation.
Earlier, strength training in individuals with stroke was believed to increase muscle spasticity (Bobath et al., 1990). However, several studies have shown that strength training can be given to individuals with stroke without increasing spasticity, or even suggested the contrary (Ada et al., 2006; Flasbjer et al., 2012; Bourbonnais et al., 2002; Abdollahi et al., 2015). At present, it remains unclear which strengthening intervention is most feasible and effective. Researchers pointed out that eccentric strength is more preserved than concentric strength following a stroke, and that eccentric strength training is more effective in improving leg strength than concentric strength training (Engardt et al., 1995; Eng et al., 2009; Clark et al., 2013). In addition, eccentric exercises appear to be well tolerated in hemiplegic patients since they require a smaller degree of energy expenditure (Hammami et al., 2012). These findings suggest that eccentric training could be the most suitable strength training for individuals with stroke.
However, very expensive devices such as isokinetic dynamometers are often used in eccentric strength interventions (Engardt et al., 1995; Clark et al., 2013; Lee et al., 2013). The dynamometer can only be used by one patient at a time. Furthermore, not many physiotherapy centers have such a device at their disposal, making it impossible for individuals with stroke to train in their own environment. Less expensive options for eccentric strength training are rubber exercise bands or weights that can be used at home. These options could make eccentric strength training more feasible and applicable, but their effectiveness in individuals with stroke still has to be examined.
Besides eccentric strength training, studies on task-oriented training also showed promising results. For example, Yang et al. (2006) concluded that strength improves due to the fact that a task-oriented strength training could carry over into improvement in function. Similar results were found in another study, where it was concluded that a two-week long task-oriented training program is effective towards improving both performance during daily activities and upper limb strength (Park et al., 2015). Dynamic, task specific force production is critical to functional motor performance (Clark et al., 2006). Therefore, task-oriented training should be part of rehabilitation of upper limb function in individuals with stroke.
There is evidence suggesting that a combination of strength training and task-oriented training is more effective than task-oriented training alone for improving upper limb function and muscle strength. A six week intervention was performed on individuals with stroke with mild impairment to recover function of the upper limb (Da Silva et al., 2015). It was found that, although both groups improved, the group who received both strength training and task-oriented training showed larger improvements than the group who performed task-oriented training alone. It may be possible that an minimal amount of strength is needed for task-oriented strength training to have beneficial effects. Therefore it would be interesting to test an intervention in which both strength training and task-oriented training are performed.
Strength training and task-oriented training often have a repetitive nature. Training interventions in which movements have to be performed repeatedly may become unattractive and demotivating over time, which could decrease the patient’s engagement and interest towards the intervention. Computer assisted training where computer games are used to perform task-oriented movements may make training more appealing and improve the patients therapy adherence. In a recent study, a training with a motion-based computer game controller was performed by individuals with chronic stroke (King et al., 2012). The authors reported that, although not supervised by a therapist, all participants showed engagement with the intervention. This suggests that the use of computer assisted training could keep the patient motivated and dedicated to their training.
The purpose of this study is to contribute to the preliminary evidence for the effectiveness and feasibility of an eight week intervention combining eccentric and task-oriented strength training. Therefore the following research question is addressed: to what extent does a combination of eccentric and task-oriented strength training improve upper limb function and strength in individuals with chronic stroke? It is hypothesized that both function and strength will increase after a combination of eccentric and task-oriented strength training. The secondary goal is to find whether the order of the eccentric and task-oriented strength training results in a different effect of the training. It is hypothesized that eccentric strength training followed by task-oriented strength training has a larger effect on upper limb function and muscle strength than vice versa. Thirdly, the feasibility of both training programs is examined, to determine whether it can be implemented in the rehabilitation of individuals with stroke on a largescale.
Methods
The study was approved by the Medical Ethical Committee of the University Medical Center Groningen (METC2013/435), each subject signed a written informed consent before the start of the measurements, which were conducted according to the declaration of Helsinki.
Participants
Eleven individuals with stroke were recruited on a voluntary basis, through their attending physician, from the Rehabilitation Center Beatrixoord in Haren, The Netherlands. The following inclusion criteria were used: understanding of the Dutch language; a clinical definite diagnosis of stroke; at least six months post stroke; a Mini-Mental State Examination (MMSE) score of at least 21 (cut-off score of no cognitive impairment) (Lopez et al., 2005); reduced arm function reported by the physician; the ability to visit the rehabilitation center. Potential participants were excluded from the study if they presented shoulder problems such as frozen shoulder or treatment specifically focused on the upper limb at the time of recruitment. The characteristics of the participants are shown in Table 1.
Participant characteristics
Participant characteristics
Abbreviations: P = participant; MMSE = Mini-Mental State Examination; EST-TOST = Eccentric strength training – Task-oriented strength training group; TOST-EST = Task-oriented strength training group - Eccentric strength training. aAdministered at admission of the rehabilitation center. bNot administered due to phatic problems.
All participants received four weeks of eccentric strength training (EST) and four weeks of task-oriented strength training (TOST) in a cross-over design. Participants were allocated in order of the received informed consent to either an intervention group receiving eccentric strength training first followed by task-oriented strength training (EST-TOST group) or to an intervention group receiving TOST first, followed by EST (TOST-EST group). The time line of the study is shown in Table 2. In the first (T0), fifth (T1) and tenth (T2) week the measurements were performed. The outcome measures were the Action Research Arm Test (ARAT), strength measured with the hand-held dynamometer (HHD) (IRL, New-Zealand) and the Intrinsic Motivation Inventory (IMI) (Lyle et al., 1981; Janssen et al., 2009; selfdeterminationtheory.org, 2014). The IMI was administered at T1 and T2, as it is a questionnaire about intrinsic motivation of the participant during the intervention. The interventions were performed in week one until four and week six until nine. During each phase of the intervention, the training was delivered three times weekly. Both the EST and TOST intervention were performed once a week in the rehabilitation center under supervision of the investigator and two times a week at home without supervision. In the first week the investigators visited the participants at home to explain the training and provide assistance in technical set-up and with the exercises. If the participant needed more assistance at home the investigator would supervise the training until the participant was able to train independently. The participant kept a diary about the date and time, the intensity, frequency and duration of the intervention.
Timetable of the study
Timetable of the study
Abbreviations: EST- TOST group: Eccentric strength training – Task-oriented strength training group; TOST-EST group: Task-oriented strength training - Eccentric strength training group; HHD: hand-held dynamometer; IMI: Intrinsic Motivation Inventory.
The eccentric strength training consisted of exercises with lightweight dumbbells and rubber bands (CandoProducts, Fabrication Enterprises Inc., USA) to train the muscles of the shoulder, elbow and wrist. In Table 3 a description of each exercise is shown. The unaffected arm was used to assist the affected arm in concentric direction, and then the affected arm slowly moved back to the initial position resulting in an eccentric movement. The intensity (the weight of the dumbbell and the resistance of the rubber band) and duration (amount of repetitions) depended on the physical fitness of the participant and was determined by the investigator. In the first training session the investigator decided which dumbbell and rubber band was appropriate based on feedback of the participant. The amount of repetitions was decided in a similar way. At home, the participant trained with the same intensity and duration. At the beginning of each training session in the rehabilitation center it was decided if the intensity and/or duration could beincreased.
Exercises of the eccentric strength training for the muscles of the shoulder, elbow
and wrist. The arrows point in the direction of the unilateral eccentric movement
Exercises of the eccentric strength training for the muscles of the shoulder, elbow and wrist. The arrows point in the direction of the unilateral eccentric movement
The task-oriented strength training consisted of bilateral upper-limb training using a movement-based game controller (Able X) (Im-Able Ltd; New-Zealand). The Able X was designed to be used by individuals with stroke or other neurological impairments, so the games are easy to interact with and fast reaction speed is not required. The Able X is a game controller incorporated into a handlebar and enables the participant to make bilateral exercises by allowing the unaffected arm to support and assist the affected arm (Hijmans et al., 2011). A weight was attached to the handle bar to create a task-oriented strength training instead of just a regular task-oriented training. The weight was applied to the side of the handlebar that was held by the affected arm. If the participant was able to use the Able X trigger button with their affected hand, the controller was placed in that hand, otherwise the controller was placed in the unaffected hand. The participant played games on a PC using the movement-based game controller for 30 minutes. Participants were allowed to pause a game when necessary. The participant sat on a chair and held the handlebar in different horizontal positions in front of the body to train all the joints of the upper-limb. The participant trained the wrist, elbow and shoulder extension and flexion during the intervention. The participants were instructed to use the ‘Rehab Routine’ function of the Able X. This is a computer customized routine, where the participant starts with a short game to measure their movement capability. The program then automatically leads the participant through an exercise routine, which is based on the measurement and their scores in each subsequent game. In the first session, the investigator determined the amount of weight applied to the handle bar based on feedback from the participant. The weight remained the same during the training sessions at home. At the beginning of each training session in the rehabilitation center there was decided if the weight could be increased.
Assessment
The ARAT was administered to measure upper limb function. The ARAT is a 19-item measure divided into four sub-tests: grasp, grip, pinch, and gross arm movement, and has a maximum score of 57 points. The ARAT scored an excellent test-retest and inter/intrarater reliability in individuals with chronic stroke (Platz et al., 2005; Van der Lee er al., 2001). Also, the ARAT is substantially more responsive to improvements in upper limb function in individuals with chronic stroke than, for example, the Fugl-Meyer assessment scale (Van der Lee et al., 2001).
To evaluate if strength was increased during the eccentric and task-oriented strength training, strength was measured with a HHD. The HHD is proven to be a reliable instrument to measure strength in individuals with brain damage (Bohannon et al., 1986; Jordan et al., 2014; Sampson et al., 2012; Suzuki et al., 2011; Riddle et al., 1989). Strength was measured, in Newton (N), of the flexors, extensors, adductors, abductors, external rotators and internal rotators of the shoulder, the flexors and extensors of the elbow and the palmar flexors and dorsiflexors of the wrist. During the measurements participants performed an isometric contraction to measure the maximal strength. This was measured twice for each movement and the highest score was used in the analysis. The highest score was chosen instead of the mean of the two measurements, because the maximal strength was chosen as an outcome measure(Bohannon et al., 1986; Riddle et al., 1989). The average amount of strength of the shoulder, elbow and wrist was calculated by adding up the measured strength of the joint movements and dividing it by the amount of measurements. For example, for the elbow the strength of flexion and extension were added and the value was divided by two.
Feasibility was determined with the IMI, the adherence and the dropout rates. The IMI, shown in the attachment, consists of five sub-scales; interest/enjoyment, perceived competence, effort/importance, perceived choice and value/usefulness. A higher score indicates a higher intrinsic motivation. The scores range from one (complete disagreement with the statement) until seven (complete agreement with the statement). The IMI is a validated questionnaire designed to assess participants’ subjective experience related to a target activity in laboratory experiments(Choi et al., 2011; McAuley et al., 1989). Also, the IMI has been used previously in studies with stroke individuals (Sampson et al., 2012; Choi et al., 2011; Colombo et al., 2007). The total score of the IMI was used to determine the motivation of the participants. The subscales effort/importance and value/usefulness were used to determine the importance and value of the training program experienced by the participants.
Statistics
The data contained continuous variables and were presented quantitatively. The following statistics, performed using SPSS 20.0, were used to get a better understanding of the data. Nonparametric tests were performed because of the small amount of participants in this study. A Bonferroni correction was used to counteract the problem of multiple comparisons. Statistics were used for the total score of the ARAT and the total strength of the shoulder, elbow and wrist. Therefore, the alpha used in this study was 0.05/4 = 0.0125.
To examine the effect of the interventions on the ARAT and strength over the eight weeks of training the Friedman’s test was used. To calculate the effect sizes, first the Wilcoxon signed-rank test was used to determine the z-score between T0 and T2. The used equation to calculate effect size was: rT0 - T2 = Z/. An effect size of r = 0.10 is a small effect, r = 0.30 is a medium effect, r = 0.50 is a large effect and r = 0.70 is a very large effect (Cohen et al., 1988).
To test for baseline differences the Wilcoxon rank-sum test was used. The Wilcoxon signed rank test was used to test for differences within the EST or TOST intervention or within the EST-TOST and TOST-EST group. To test for differences between the EST and TOST intervention and between the TOST-EST and EST-TOST group the Wilcoxon rank-sum test was used with the weighted mean differences.
Feasibility of the intervention was evaluated by the scores of the IMI, the adherence and dropout rates. The internal consistency of the IMI was calculated using Cronbach’s (1951) alpha coefficient (Cronbach et al., 1951). An internal consistency higher than 0.7 is reported to be acceptable (George et al., 2003).
Results
In total eleven participants were included in the study of which five participants were allocated to the EST-TOST group, and six to the TOST-EST group. In the last group, one participant dropped out in the second week. All other participants successfully completed the study.
During the eccentric strength training the intensity varied among the participants between no dumbbell and two kilo dumbbells, and extra-extra light and light rubber bands. The duration ranged from 2 × 5 repetitions to 3 × 15 repetitions per joint movement, which took approximately between 30 and 60 minutes. The intensity and/or duration of the intervention increased for every participant during the four weeks. During the task-oriented strength training the weight applied to the handlebar varied between no weights at all and 500 grams. The weight applied to the handlebar increased for every participant during the four weeks of intervention. Every participant could complete the 30 minutes of training. No participant mentioned shoulder problems, physical injuries or muscle pain in their affected arm after eccentric and task-oriented strength training.
Upper-limb function and strength
First, no significant differences at baseline were found with the Wilcoxon rank-sum test for upper limb function (p = 0.251) and for shoulder (p = 0.175), elbow (p = 0.175) and wrist (p = 0.530) strength between the EST-TOST and TOST-EST group. Table 4 shows the results of the Friedman’s ANOVA for upper limb function measured with the ARAT and strength, measured with the HHD, of the complete intervention (both groups together). Significant increases were found for upper limb function (p = 0.010) and for shoulder (p = 0.000) and elbow (p = 0.003) strength while no difference was found in wrist strength (p = 0.368). The effect of upper limb function was very large with r = –0.75. Also, the effects for shoulder and elbow strength were very large with r = –0.89 and r = –0.85, respectively. Wrist strength had a large effect with r = –0.51.
Changes in upper limb function (ARAT) and strength (HHD) calculated with the
Friedman’s ANOVA (N = 10). The means and standard deviations of T0, T1 and T2 are
included, even as the calculated effect size
Changes in upper limb function (ARAT) and strength (HHD) calculated with the Friedman’s ANOVA (N = 10). The means and standard deviations of T0, T1 and T2 are included, even as the calculated effect size
Abbreviations: SD = Standard deviation; p = p-value; ES = Effect size.
Table 5 shows the differences in upper limb function and strength over time within both the EST and TOST intervention and the difference in change between these interventions. The mean differences and standard deviations are shown for both groups. A trend in positive direction in upper limb function was found for the EST and TOST intervention with respectively p = 0.035 and p = 0.028. The effects were large and very large for respectively the EST (r = –0.67) and TOST (r = –0.70) intervention. A significant increase in strength was found in the TOST intervention for shoulder (p = 0.007). Furthermore, a trend in positive direction was found for shoulder (p = 0.028) in the EST intervention and for elbow (p = 0.017) in the TOST intervention. In the EST intervention, the effects were very large (r = –0.89) for shoulder, large for elbow (r = –0.60) and small for wrist (r = –0.27) strength. In the TOST intervention, the effect sizes were very large for both shoulder (r = –0.85) and elbow (r = –0.76) strength. Wrist strength had a small effect with r = –0.21.
Differences in upper limb function (ARAT) and strength (HHD) within and between the EST and TOST group calculated with respectively the Wilcoxon signed rank and rank-sum test (N = 10). The means and standard deviations of the EST and TOST group are included, as well as the calculated effect sizes
Abbreviations: SD = Standard deviation; p = p-value; ES = Effect size; diff = difference.
No significant differences were found between the EST and TOST group in ARAT and strength.
Table 6 shows the differences in upper limb function (ARAT) and strength (HHD) within and between the EST-TOST and TOST-EST group. Also, the calculated effect sizes are shown.
No significant increases are found in upper limb function and strength in the EST-TOST and TOST-EST group. For upper limb function a trend is found in the EST-TOST group with p = 0.043. This is a very large effect with r = –0.90. For strength, trends are found for shoulder (p = 0.043) in the EST-TOST group and for shoulder (p = 0.042) and elbow (p = 0.043) in the TOST-EST group. The effect sizes in the EST-TOST group for shoulder, elbow and wrist are respectively very large (r = –0.90), large (r = –0.78) and medium (r = –0.30). The effect sizes in the TOST-EST group for shoulder, elbow and wrist are all very large with respectively r = –0.91, r = –0.90 and r = –0.82.
Differences in upper limb function (ARAT) and strength (HHD) within and between
EST-TOST (N = 5) and TOST-EST (N = 5) group calculated with respectively the Wilcoxon
signed rank and rank-sum test. The means and standard deviations of the EST-TOST and
TOST-EST group are included, even as the calculated effect sizes
Differences in upper limb function (ARAT) and strength (HHD) within and between EST-TOST (N = 5) and TOST-EST (N = 5) group calculated with respectively the Wilcoxon signed rank and rank-sum test. The means and standard deviations of the EST-TOST and TOST-EST group are included, even as the calculated effect sizes
Abbreviations: SD = Standard deviation; p = p-value; ES = Effect size; diff = difference.
No significant differences are found between the EST-TOST and TOST-EST group, although there is a trend for upper limb function (p = 0.015). The means of the EST-TOST and TOST-EST group show that the significant differences are in favor of the EST-TOST group of the total score of the shoulder, elbow and wrist are all very large with respectively r = –0.91, r = –0.90 and r = –0.82.
The feasibility of the EST and TOST was determined with the IMI, the adherence and the dropout rates. Table 7 shows the scores of the IMI for the EST and TOST intervention and the calculated Cronbach’s (1951) alpha. The internal consistency is questionable for the sub scale Perceived Choice of the EST intervention (α= 0.64) and Perceived Competence of the TOST intervention (α= 0.69). The other subscales are between acceptable for Perceived Competence in the EST intervention and excellent for Value/Usefulness in the EST intervention. The total score of the IMI is in both the EST and TOST intervention good with α= 0.89 (George et al., 2003).
Scores of the IMI questionnaire administered after the EST and TOST training. The
scores range from 1 (complete disagreement) till 7 (complete agreement). The
calculated Cronbach’s alpha is also reported
Scores of the IMI questionnaire administered after the EST and TOST training. The scores range from 1 (complete disagreement) till 7 (complete agreement). The calculated Cronbach’s alpha is also reported
Abbreviations: Perc. = Percentage; α= Cronbach’s α.
Scores of the subscales range in the EST group between 5.08 for Perceived Competence and 6.51 for Value/Usefulness. In the TOST group the scores range between 5.27 for Perceived Competence and 6.16 for Value/Usefulness. The total score of the IMI is for both interventions 5.69.
In this study, one participant withdrew after the second week of the TOST intervention because he believed the training had no value for him. All other participants successfully completed the total training program. One participant forgot one training session in the second training week of the TOST intervention at home. All the other participants performed the total amount of 24 training sessions. On a few occasions one participant was unable to visit the rehabilitation center. When this situation occurred the participant trained at home three times a week instead of once a week in the rehabilitation center and twice at home.
Upper limb function and strength
The findings of this study support the hypothesis that a combination of eccentric and task-oriented strength training will increase function and strength in individuals with chronic stroke. These results are promising, pointing towards the implementation of such a training in the rehabilitation of individuals with stroke. Previous studies often reported an increase in strength but failed to find an effect in function (Eng et al., 2004; Morris et al., 2004). In the current study large effects in both strength and function are shown, which indicates that a combination of eccentric and task-oriented strength training is an effective training program to increase both strength and function. For example, compared to the strength training provided to individuals with chronic stroke in the study of Sampson et al. (2012), strength increased much more in the current study. The training program in that study consisted of four sessions a week for six weeks and increased shoulder (without adduction) and elbow strength with respectively 11 N and 8 N (Sampson et al., 2012). In the current study the eight week program, provided three times a week, resulted in an increased shoulder (without adduction) and elbow strength by respectively 23 N and 27 N. Even four weeks of eccentric strength training resulted in an increased elbow strength by 12 N and four weeks of task-oriented strength training resulted in increased shoulder and elbow strengths by 18 N and 16 N respectively. Furthermore, the results of the current study are in line with the findings of another study where the same movement based game controller was used (Hijmans et al., 2011).
Order of the strength training
In this study, a trend was found for strength in both the EST-TOST and TOST-EST group and for upper limb function in the EST-TOST group. In the EST-TOST group the ARAT score increased on average by 12.8 points, whereas the increase of the ARAT score in the TOST-EST group was 1.8 points. The Minimally Clinically Important Difference (MCID) of the ARAT is 5.7 (Van der Lee et al., 2001), indicating that the increase in the EST-TOST group is a clinically important difference and in the TOST-EST group it is not. This is in line with the expectation that task-oriented strength training is more beneficial when preceded by a regular strength training. However, no difference in the increase in strength and upper limb function was found between the EST and TOST group. This indicates that both training programs were equally effective in increasing strength and upper limb function. Possibly, the difference in effect between the EST-TOST and TOST-EST group can be explained by the differences at baseline. Although no significant difference at baseline was found, it appears that the participants of the EST-TOST group had a higher score on the ARAT and larger strength measured with the HHD at baseline than the TOST-EST group despite the random assignment to both groups. These differences at baseline indicate that the intervention provided in this study could be more effective in individuals with stroke who start with a higher strength and function of the upper limb. This is in line with the stating of Rimmer et al. (2001) that the effect of the training can be different in people with more severe stroke, e.g. less strength and function (Rimmer et al., 2001). However this specific aspect of the training should be addressed in future research.
Feasibility
The feasibility of this study was determined using the IMI, the adherence and the dropout rates. The average total score of the IMI for both the EST and TOST was 81%. Previous studies in individuals with stroke reported total scores on the IMI of 86% (Sampson et al., 2012) and 87.4% (Jordan et al., 2014) as highly motivated. The reported IMI scores of these studies are almost similar to the total IMI score in the current study, indicating that the participants were very motivated to engage in the eccentric and task-oriented strength training program. The computer assisted training was not perceived as more motivating than the regular strength training, as was suggested by King et al. (2012). The high scores on the subscales effort/importance and value/usefulness for both EST and TOST indicate that the participants experienced the training as valuable and important. The participants put a lot of effort into the intervention and were willing to do the training again because they thought it was useful to increase strength and mobility of the upper limb. In addition, the participants mentioned that the combination of two types of strength training kept the training enjoyable due to the variety of exercises.
The adherence rate in this study was high and there was only one dropout, which is in line with a study of Maclean et al. (2002), stating that highly motivated patients are more likely to participate and continue a rehabilitation program. The reason one participant did not continue the study was because he believed the training program had no value for him.
The findings of the IMI, the high adherence rate and the fact that there was only one dropout suggest that the provided strength training program is feasible to implement in a rehabilitation program.
Future directions and limitations
The protocol in this pilot study would be sufficient to use in the rehabilitation of individuals with stroke. However, a limitation of the current protocol is the inability to check whether the training was executed as prescribed because the participants trained twice a week at home without supervision. Full supervision would be preferred, although using a diary is a much less labor intensive and suitable alternative. Another limitation of the current protocol is the ceiling effect of the ARAT; a few participants reached the maximum score in the ARAT after intervention (two in the EST-TOST and one in the TOST-EST group). This could have resulted in an underestimation of the results.
Conclusion
A combination of eccentric and task-oriented strength training is an effective and feasible training method to increase upper limb function and strength in individuals with chronic stroke. The training was low cost and kept the participants motivated. The current results give an indication that TOST should be preceded by EST. No conclusion can be drawn on whether the order of the training results in a different effect in strength and/or function.
Conflict of interest
The authors declare that there is no conflict of interest.
Funding
This study was supported by the foundation ‘Beatrixoord Noord-Nederland’, contract number: 210.150.
Footnotes
Acknowledgments
The authors thank the foundation ‘Beatrixoord Noord-Nederland’ for funding this project and the participants who willingly gave their time during the intervention. Also, the authors would like to thank Dr. Henk Meulenbelt for all his help with the participant recruitment and Renate Hensen who supported with the execution of the intervention and measurements.