Feasibility, Acceptability and Effects of a Home-Based Exercise Program Using a Gerontechnology on Physical Capacities After a Minor Injury in Community-Living Older Adults: A Pilot Study

Abstract

Context

Several studies have demonstrated that physical activity can help limit decline in functional capacities of older adults. Nevertheless, many adults aged 65 and over are inactive.

Objective

To explore the feasibility, the acceptability and the effects of a home-based exercise program (HEP) using a motion capture gerontechnology in independent community-living older adults at risk of function decline.

Design

Interventionnal clinical trial.

Participants

Sixteen previously independent individuals aged 65 and older recruited at the Emergency Department after being treated for a minor injury and discharged home were assigned to a home-based exercise program group (HEP=8) or to a control group (CONTR=8). Twelve participants completed the study, 6 in each group

Setting

Canadian Community-dwelling in Montreal area.

Intervention

The HEP group engaged in a twelve-week physical activity intervention using a gerontechnology while the CONTR group continued with discharge plan from ED.

Measurements

Participants were evaluated for functional status using validated questionnaires and objective physical measures at baseline, three and six months later. Feasibility and acceptability of the HEP was assessed using data reports from the gerontechnology and from self-reported assessments.

Results

There was no differences between groups at baseline except for the fallrelated self-efficacy: HEP=8.33/28±1.51 vs CONTR=7/28±0 p=0.022. The HEP was found to be feasible and acceptable (adherence rate at 86% and average quality of movements at 87.5%). Significant improvement in walking speed on 4m was observed three months after baseline for HEP vs CONTR group (+0.25 vs +0.05 m/sec, p=0.025). Effects remained at follow-up. Only CONTR group resulted in a significant increase in SF-36 global score.

Conclusion

This twelve-week HEP intervention using the Jintronix® gerontechnology is feasible, acceptable and safe for community-living older adults who sustained a minor injury. This intervention could increase walking speed, the most important predictor of adverse events in the elderly population, and that the improvement could be maintained over time.

Introduction

In Canada, older adults (aged 65 and over) represent 16% of the population (1) and this proportion will continue to increase in the next decades. The same trend is observed worldwide. Among older Canadian community-dwelling older adults, 24% are considered frail while 32% are considered pre-frail (2). Frailty is an evolving state of vulnerability leading to adverse events, such as functional losses, disability and eventually death. The direct consequences of these events are an increase load on healthcare systems, especially on emergency departments (ED). In fact, more than 20% of visits at EDs are made by seniors from which 17% are injury-related (3). While most of these patients (79%) are returned home after medical consultation (3), it has recently been reported that, following a minor injury, older adults functionality may decline by 15 to 17% in activities of daily living (ADL) and instrumental activities of daily living (IADL) (4). Moreover, poorer scores on mobility assessments and falls efficacy scales in those injured older adults are reported at baseline and up to 6 months post-injury (4). Considering that a large number of communitydwelling older adults admitted to ED with minor injury were found to be frail or pre-frail at the time of consultation (5), and that minor injuries may accelerate the path to disability, interventions to prevent a functional decline following such adverse events have to be implemented.

Physical activity interventions have proven to be effective in improving or maintaining functional capacities of frail and nonfrail older adults (6), and in reducing the risk of major mobility disabilities in a frail population (7). Nevertheless, more than 50% of older adults are inactive in their leisure time, spending most of their time awake in sedentary activities, such as sitting or watching television (8). Many barriers to exercise are stated by seniors who are not physically active, the most significant being a poor health, the fear of falling or getting injured and the lack of motivation or enjoyment (9). In order to overcome these barriers, exercise programs supervised by health professionals, either in laboratory, clinical or community settings have been implemented. These programs have been shown to be efficient, however they tend to be resource consuming (10). Moreover, traditional approaches may not necessarily be suitable for frail individuals who have comorbidities, who commonly experience feelings of exhaustion, or who are not able to transport themselves outside their home (11). This population could benefit from other type of exercise programs.

Adapted home-based exercise programs (HEP) are foreseen as a potential solution to address some of these issues and allow older adults, including frail people, to benefit from increased physical activities. HEP present many advantages such high adherence rates (12), increased feelings of empowerment over lifestyle changes, positive results in preventing falls and an efficiency rate similar to highly supervised programs in improving ADL functional capacities and mobility (13). However, HEP also present some limitations. For instance, information on adherence and compliance to programs mostly relies on self-reports (14), thus making it difficult for the kinesiologists to individually adapt and optimize programs. Adherence and compliance may also be more difficult to achieve in people with lower self-motivation or self-efficacy feelings that are associated with increased health problems (15) and require close follow-up and feedbacks which cannot be extensively provided in traditional HEP. A recent study indicated that HEP tailored interventions, with individually adapted goals, provide more benefits than a HEP general activity goal (16). However, personalized interventions remained difficult to provide without constant monitoring and regular supervision of participants.

New emergent technological systems, especially gerontechnologies (i.e. those adapted to elderly people) can potentially overcome the limits of traditional HEP by offering distance supervision and follow-up. Fisrt of all, motion-sensor portable devices (accelerometers and gyroscopes) are useful tools to monitor and increase physical activity level in older adults. However, O'Brien, Troutman-Jordan et al. (17) found that these devices do not contribute to building self-efficacy feelings because of lack of feedbacks on motion quality. Furthermore, exercises through video games, called exergames, also provide guidance and monitoring and show good potential as an adapted HEP tool for older adults (18). According to Skjæret, Nawaz et al. (18) exergame interventions result in comparable adherence rates and improvement in functional capacities as traditional interventions. Gerontechnologies have been shown to be effective in improving mental health and cognitive functioning (19), fitness and balance (20) in older adults. Furthermore, because exergames are considered entertaining and are perceived as games rather than exercises, they can enhance motivation and reduce the negative perceptions associated with exercise programs such as potential tiredness, pain or boredom. Moreover, the potential features of exergames, like the adaptability to individuals' limitations (range of motion, speed, intensity, etc.), the challenge and progression of games as well as the provision of automated reminders and feedbacks can contribute to the feeling of control and autonomy (21) and therefore increase self-efficacy feelings.

However, not all exergaming systems are equally adapted to the elderly. In their review, Van Diest, Lamoth (22) reported that there are currently three types of devices on the market; those based on inertial sensors often integrated in wireless hand-held controllers, those based on pressure sensors requiring a board or a mat and those based on camera systems using motion capture. The latter present a clear advantage over those based on inertial and pressure sensors since camera systems do not require equipment that might interfere with older adults' ability to perform exercises (22), a factor that can facilitate compliance and adherence. Nevertheless, the vast majority of studies have been conducted using the Nintendo Wii handheld controller and balance board, i.e. systems based on inertial and pressure sensors (18). The paucity of studies using a gerontechnology based on a motion capture camera system for physical activity programs with older adults does not allow for any conclusion on the effectiveness of such interventions. Furthermore, the feasibility of implementing a gerontechnology such as a HEP as well as its acceptability in communitydwelling older adults at risk of functional decline remain unclear, as most studies were conducted in laboratory/clinical settings (18).

The main purpose of this study was to explore the feasibility and the acceptability of a HEP using a motion capture gerontechnology in community-living older adults following a minor injury. We also aimed at verifying the effects of an intervention using such gerontechnology. We selected the Jintronix® technology, which is a rehabilitation software using the Microsoft Kinect® motion capture system and offering a variety of exercises and games adaptable in speed, duration, precision, range of motion and number of repetitions. Not only does this system provide live feedback to participants, it also reports through its automatic data reporting system on quantity and quality of movements to the clinician. Based on previous studies using exergames (23, 24, 25), our primary hypothesis was that a twelve-week HEP intervention using this gerontechnology was feasible, acceptable and safe for community-living older adults. Our secondary hypothesis was that this physical activity intervention could reduce the decline in physical functions following a minor injury.

Methods

Study Population and Procedures

Participants were recruited at the Emergency Department (ED) of Hôpital du Sacré-Coeur de Montréal after being evaluated for a minor injury and discharged back home within 48-hours (4). To be included in this study, participants had to (1) be 65 years old and over, (2) had suffered a minor injury that brought them to the ED, (3) be independent in all 7 Activity of Daily living (ADL) previous to injury (4) live in the community and (5) be released from the ED. Participants were not included if they could not give their informed consent, could not communicate in French, and clinically appeared to have cognitive disorder. Occasional use of a walking aid was accepted. Potential participants were authorized by the Emergency physician to undertake a physical activity program. They gave consent to be referred to a kinesiologist and were later contacted for final consent and complete assessment within seven days. Participants were randomly divided in two groups; the intervention group (HEP) or the control group (CONTR). A randomised list was created by the software SPSS version 24 to determine to which group each participant would be assigned depending on their recruited order. Only the research nurses had access to this list, which was kept safely at the hospital. The HEP group engaged in a 12-week physical activity program using a gerontechnology. The CONTR group received no intervention and no special monitoring during this study, but may have been recommended with a traditional treatment plan by the medical staff at ED.

After the first screening at ED, participants were visited at home by a kinesiologist. After obtaining the participant consent, sociodemographic information was collected (ethnicity, marital status, occupation and level of education achieved). Participants were then evaluated for baseline values within 7 days of their ED visit [baseline (T0)], three months later [three-month evaluation (T1)] and six months from baseline [follow-up evaluation (T2)], using validated questionnaires in an interview setting and physical tests and measures. Both follow-up were conducted at the participant's home. All procedures were approved by three research ethic boards: Centre de Recherche du Centre hospitalier Universitaire de Québec, Université du Québec à Montréal and Hôpital du Sacré-Coeur de Montréal.

Intervention

The HEP group was asked to complete a twelve-week intervention program, at home, using the gerontechnology provided. Participants were required to complete two sessions per week. They were free to decide the day and the time to perform the exercise program, but had to take a minimum of one day off between sessions. The exercise program consisted of a warm-up of 5 minutes, nine aerobic exercises lasting 20 minutes, nine resistance & balance exercises lasting 20 to 25 minutes and a cool down of 10 minutes. Each exercise session lasted approximately 50 to 55 minutes depending on the speed and number of repetitions of various activities in the resistance & balance section.

Within one week of the baseline evaluation, the gerontechnology was installed at the participant's house of the HEP group. The technology used is a portable system working with the Microsoft Kinect® motion capture system and a software named Jintronix®, which is available online. Jintronix® allows the kinesiologist to create an adapted physical activity program based on available exercises and exergames. The physical activity program was developed in accordance with general guidelines for the prescription of exercises to older adults (26, 27). Through a TV or a computer screen, users are guided with continuous visual and audio cues while performing exercises. Some are built as traditional exercise movements, like walking on the spot, repeated sitting and standing, high knees, butt kicks and lateral launches while others are developed as exergames, such as skiing downhill (combination of squats and lateral weight transfers), kicking a soccer ball in a net (leg extension), climbing (one-leg stand), step on the moles (sideway, backward and forward steps), space race (shoulder abduction/adduction) or going through a labyrinth with a ball (lateral and forward trunk movements). The Jintronix® system provides live feedbacks allowing the user to correct movements, improve performance and progress to higher levels. It also provides feedbacks to the clinician, in this case the kinesiologist, through a report on quantity and quality of movements performed by users.

There were two visits to install the technology and explain the program. Participants were progressively brought to independently use the technology. A kinesiologist visited the participant to supervise 6 out of the 24 sessions (session 1, 2, 4, 6, 12 and 18) and was available for support for the remaining sessions. In addition, the kinesiologist ensured a follow-up in each participant's Jintronix® account and constantly adjusted the level of difficulty of exercises to ensure proper individual adaptations and progress.

Feasibility and compliance measures

Feasibility was assessed using the automatic data reporting system from the Jintronix® available for each session as well as each resistance and balance exercise. Adherence rate was calculated by the proportion of sessions completed out of the total prescribed sessions, i.e. 24 (28). The information on adherence was readily available as each session was recorded in the participant's Jintronix® account.

Two aspects of compliance were assessed: (1) quantity of movements (number of executed repetitions) and (2) quality of movements (performance according to the level of difficulty predetermined). This information was obtained by extracting data directly recorded by the system. Furthermore, the level of difficulty of each exercise was adjusted by the kinesiologist throughout the twelve-week intervention according to the participant's progress. In order to determine the feasibility of the intervention, the minimal overall quality of movements was set at 75% while the minimal quantity of movements was set at 80% of expected repetitions in completed sessions.

Acceptability was evaluated using a four level analog scale for perceived level of difficulty (PLD) and personal level of enjoyment (PLE). Participants were required to complete a logbook at the end of each session, attributing a PLD and a PLE score for each exercise of the aerobic and resistance & balance portions (18 exercises for a total of 36 entries). For PLD, a fourlevel color scale was used. This method was chosen because it eases the assessment of perceived exertion in older adults [29]. Participants were asked to record their PLD for each exercise of the aerobic and the resistance & balance portions by circling one of four colored circles representing their perceived level of difficulty of each exercise: easy (green), moderate (yellow), difficult (orange) or very difficult (red). Compared to the Borg 6-to-20 scale for ratings of perceived exertion, these ratings would correspond approximately to very light/fairly light (easy), somewhat hard (moderate), hard (difficult) and very hard (very difficult), i.e. a score going from 9 to 17 (30).

A similar type of scale was used to record their PLE. Participants were asked to record their PLE for each exercise of the aerobic and the resistance & balance portions by circling one of four pictogram faces representing their enjoyment (how they liked each exercise): not at all, a little, well or a lot. The use of this assessment tool to measure acceptability was inspired by the physical activity enjoyment scale (31, 32). However, it was reduced to only one item and simplified to four possible answers to ensure compliancy of participants in rating the exercises given the total number of entries to log at the end of every session.

Our goal was to have the majority of participants considering the exercises as being easy or moderate for the PLD and rating them as being enjoyed a lot or well enjoyed for the PLE. Meldrum, Glennon et al. (33) used a range of 70% to 90% as an acceptable to a high level of usability to assess an exergame program. Therefore, we established that the Jintronix® program would be acceptable if a proportion of 80% of all exercises were rated by participants within our targets for both the PLD and the PLE.

Measures

Functional performance in basic daily activities was assessed using the Katz Index of independence in ADL [34] and an adapted version (7/8 questions) of the Lawton and Brody Instrumental ADL scale (35). Frailty was assessed using the Study Osteoporotic Fractures (SOF) Index, a 3-point scale based on three components (unintentional weight loss, selfreported reduced energy level and reduced lower limbs strength measured by the 5 times sit-to-stand test). Participants with no problematic items are considered robusts, those with one or two problematic items are considered prefail and frail, respectively (36).

Fear of falling was assessed using the Short Fall Efficacy Scale-International (Short FES-I) [37]. The level of concern is measured by a 4 level scale (1=not at all concerned to 4= very concerned), a higher score meaning a greater fear of falling.

Physical activity level was determined using the Rapid Assessment of Physical Activity (RAPA). The RAPA is a 9-item questionnaire that estimates the level and intensity of physical activity. According to the score, the person is classified from sedentary to active (38).

Finally, health-related quality of life was evaluated using the Short Form (36) Health Survey (SF-36), which assesses selfperceived physical and mental health status (39, 40). The SF-36 is reliable and valid in the older adults population when selfadministered and in an interview-based setting (41).

Anthropometric Measurement and Body Composition Assessment

Anthropometric measures were taken according to standardized techniques (42, 43). Height was measured using a rigid tape against the wall. Waist, hip and calf circumferences as well as leg length were measured with a flexible measuring tape.

Body weight, % fat mass and % muscular mass were measured using a bioelectrical impedance (BIA) full body sensor body composition monitor and scale (Omron Model HBF-510W).

Body mass index [BMI = BW/height (m²)], waist-to-hip ratio and other health and fitness indicators were then calculated.

Physical Function Assessment

Maximum voluntary handgrip strength was measured using a hand dynamometer (Lafayette Hand dynamometer Model 78010). This upper body strength assessment test is reliable, inexpensive and has been validated (44, 45). Participants performed three trials with each hand, alternating left and right. The maximal score was used to calculate dynapenia index (Handgrip strength/Body weight). Relative handgrip strength is a strong strength index and a good indicator of functional impairments (46).

Functional mobility was assessed using with the 3 meter Timed-Up-and-Go (TUG) test and the Short Physical Performance Battery (SPPB). The TUG is valid and reliable (47) and is a good predictor of falls in community-dwelling older adults [48]. The SPPB is comprised of three tests: gait speed, repeat chair stands and standing balance. Each test scores from 0 (inability to complete the test) to 4 points (highest performance level) according to previously established quartiles of performance (49). The SPPB score is associated with subsequent disability and mortality in the elderly population (49, 50). A score of 10 or less has been suggested as a cut-off point to identify people with significantly higher odds of having mobility disabilities (51).

Health status and cognition

A health status questionnaire was used to record comorbidities (hypertension, diabetes, Parkinson's disease, cancer, stroke, or others diseases) and medication intake. Cognitive status was assessed using the validated Montreal Cognitive Assessment (MoCA) (52).

Statistical analysis

Data are presented as mean (x), ± standard deviation (SD). Normality was evaluated using a Kurtosis test. Because of our small sample size, non-parametric tests were used. Nonparametric (Man-Whitney) tests were conducted to compare HEP and CONTR groups at baseline (see Table 1). Then, non-parametric paired t-tests (Wilcoxon) were conducted to compare variables within a same group before and after intervention. Finally, non-parametric (Man-Whitney) tests were conducted to compare the delta change (%) between groups, following the intervention. A p value ≤ 0.05 was considered statistically significant. Furthermore, a p ≤ 0.1 was considered a tendency to statistical significance. All analysis were performed using SPSS 22.0.

Table 1.

Baseline Participants’ Characteristics

Variable	HEP (n=6)N(%) or mean ± SD	CON (n=6)N(%) or mean ± SD	p value
Age, mean (SD)	73.17 ± 2.93	76 ± 6.51	0.469
Gender, (female; (%))	83.3	100
Education level, (<University; (%))	66.7	83.3
Body composition
BMI (kg/m2)	30.42 ± 6.50	27.98 ± 6.22	0.337
% Fat mass	42.08 ± 6.09	36.20 ± 7.86	0.221
% Muscular mass	24.34 ± 2.78	22.98 ± 5.62	0.624
MoCA-Cognitive function (/30)	25.20 ± 2.49	26.20 ± 1.92	0.750
SOF index (x/3)	3 ± 0	3 ± 0	1.00
FES (x/28)	8.33 ± 1.51	7 ± 0	0.022*
IADL (x/7)	6.67 ± 0.52	6 ± 1.53	0.702
Number Co-Morbidities	4 ± 3.22	4.33 ± 1.97	0.569
RAPA-Physical Activity Level (/10)	3 ± 2.53	3.5 ± 1.64	0.807
Walking speed (m/s)	0.86 ± 0.20	0.80 ± 0.11	0.872
Functional capacities
SPPB-Total (/12)	8.83 ± 1.17	7.67 ± 1.37	0.138
Handgrip/body weight	0.267 ± 0.049	0.234 ± 0.106	0.522
3 meter Timed-Up-and-Go (sec)	13.79 ± 3.51	14.52 ± 5.03	1.00

^*=p < 0.05; BMI : Body Mass Index; SOF : Study of Osteoporotic Fractures; FES: Fall Efficacy Scale; IADL; Instrumental Activity of Daily Living; RAPA: Rapid Assessment of Physical Activity

Results

Participants

A total of 16 volunteers, previously independent, individuals (2 men and 14 women) aged 65 years and over (75 ± 5.45 years; ranging from 66 to 83) recruited at ED were included and assessed at baseline, 8 in the intervention group (HEP) and 8 in the control group (CONTR). Four participants withdrew (age: 77 ± 6.58; ranging from 67 to 84) due to lack of motivation (2) or interest in pursuing the study (2). Altogether, 12 participants completed the study (1 men and 11 women), 6 per group. Ten participants consulted the ED after a fall and two for other reasons. Diagnosis were as follows: 5 sprains, 3 contusions, 1 wound, 1 fracture, 1 head trauma without brain injury and 1 dorso lumbar pain. Participants' characteristics are described in Table 1.

No difference were observed between HEP and CONTR groups at baseline except for fall-related self-efficacy measured by the Short FES-I. The HEP group showed greater fear of falls than the CONTR group. Results are shown in Table 1.

Feasibility

The HEP group completed 123.25 out of the 144 sessions planned, resulting in an adherence rate of 86%. However, one participant was able to perform only 19% of the program due to severe osteoarthritis. Nevertheless, this person insisted on using the technology and completing the program as much as possible. If this individual was excluded, the adherence rate would be 99% for the remaining five participants. HEP participants executed 27,154 out of 28,149 repetitions in the resistance & balance portion of the program of the completed sessions, resulting in a compliance rate of 96%. The individual lowest compliance rate was 85% while the highest was 100% of expected number of repetitions.

The average quality of movements in the resistance & balance portion of the program of the completed sessions, for each participant, ranged from 80% to 91% with a mean of 87.5%. The lowest score of 80% was obtained by the participant who has severe osteoarthritis and major difficulty in performing the exercises. Moreover, as shown in Figure 1, a group quality level above 80% was reached throughout the sessions despite a constant increase in the level of difficulty set by the kinesiologist.

Figure 1.

Study design and project overview

Acceptability

According to the participants' PLD scores, the program was considered easy for 76% of all exercises, moderate for 14% of them, leaving only 10% of all exercises being perceived as difficult (5%) or very difficult (5%), as shown in Figure 2A. Only one participant scored well above this 10% average, indicating that 39% of all exercises were difficult (13%) or very difficult (25%). It must be noted that this participant reported the second highest number of comorbidities (total of 7 health problems).

Figure 2.

Average quality level reached by the HEP group by session and average level of difficulty set by the kinesiologist

The PLD scores also revealed that the exercises of the aerobic portion of the program were considered the easiest with 81% of them being perceived as easy, 11% as moderate, 3% as difficult and 5% as very difficult (data not shown). In the resistance & balance portion, participants rated the exercises as follows: 71% were easy, 17% were moderate, 6% were difficult and 6% were very difficult (data not shown).

Finally, despite a constant increase in the set level of difficulty, a constant decrease in the proportion of exercises perceived as very difficult was observed, going from a proportion of 10% of all exercises during week 1 to 3, to 5% during week 4 to 6 and to 3% during week 7 to 12. The proportion of exercises that were perceived as easy or moderate reached over 80% at every session throughout the 12 weeks (data not shown).

According to the participants' PLE scores, participants enjoyed the exercises a lot in a proportion of 68% and well in a proportion of 24%, leaving only 8% of all exercises being enjoyed a little (5%) or not at all (3%), as shown in Figure 2B.

The PLE scores also revealed that the exercises of the resistance & balance portion of the program were considered the most enjoyable, with exercises as being enjoyed a lot in a proportion of 71%, well enjoyed in a proportion of 21%, enjoyed a little in a proportion of 6% and not enjoyed at all in a proportion of 2%. Scores were only slightly different in the aerobic portion. Participants rated their level of enjoyment for the aerobic exercises as follows: 66% were enjoyed a lot, 25% were well enjoyed, 5% were enjoyed a little and 3% were very difficult (data not shown).

Finally, the PLE was relatively constant throughout the 24 sessions. The proportion of exercises rated as being enjoyed a lot or were well enjoyed reached 90% all the time except for the first sessions of week 4 and 7 where 89% and 88% of all exercises were rated as such (data not shown).

Effects

Within-group differences: Evaluation results at T0 and T1 are shown in Table 2. From T0 to T1, both HEP and CONTR groups decreased their bodily pain, according to the SF-36 sub-score. This decrease was significant for the HEP group and tended to be significant for the CONTR group. Only the CONTR group resulted in a significant increase in the SF-36 global score. The HEP group significantly improved its one-leg stand and its walking speed on 4 m, as shown in Figure 3. From T0 to T2, both HEP and CONTR groups improved their walking speed on 4 m and their TUG. The significant increase in the SF-36 global score observed in the CONTR group remained at T2 (data not shown).

Table 2.

Within-group changes from T0 to T1 and T0 to T2

Variables			HEP					CON
	T0	T1	p	T2	p	T0	T1	p	T2	p
Total Fat Mass (%)	42.08 ± 6.09	44.17 ± 9.25	0.686	43.45 ± 6.81	0.686	36.20 ± 7.86	38.50 ± 8.86	0.068†.¶	42.38 ± 6.13	0.068†.#
RAPA- Physical Activity Level	3.00 ± 2.53	5.67 ± 0.82	0.068 †.¶	4.83 ± 1.60	0.102	3.5±1.64	2.0±2.28	0.167	3.83 ± 1.17	0.595
SF-36 Quality of Life Total score	60.61 ± 9.14	71.59 ± 14.87	0.173	70.38 ± 17.34	0.345	66.51 ± 12.70	85.03 ± 4.98	0.043*††.	87.61 ± 6.32	0.028*‡
SF-36 Bodily pain component	19.18 ± 11.80	60.00 ± 35.67	0.046*††.	47.08 ± 31.80	0.078†.#	39.58 ± 19.07	72.00 ± 26.06	0.08†.¶	84.58 ± 18.06	0.027*‡
One-leg stand (s)	6.00 ± 2.95	17.90 ± 20.75	0.028*††.	8.51 ± 6.63	0.463	5.38 ± 2.47	6.66 ± 8.55	0.917	8.94 ± 14.05	0.893
Walking speed (m/s)	0.86 ± 0.20	1.12 ± 0.24	0.028*††.	1.13 ± 0.12	0.028*‡	0.80 ± 0.11	0.85 ± 0.10	0.225	0.93 ± 0.10	0.028*‡
Handgrip/BW	0.267 ± 0.05	0.278 ± 0.095	0.753	0.286 ± 0.064	0.345	0.234 ± 0.11	0.207 ± 0.096	0.116	0.224 ± 0.091	0.600
Time up and go (sec)	13.79 ± 3.51	10.28 ± 2,72	0.116	9.65 ± 1.76	0.046*‡	14.52 ± 5.03	12.94 ± 2.71	0.249	10.13 ± 1.75	0.028*‡

^*=p < 0.05;

^†.= p < 0.1;

^{††.}= result is significantly different within the same group at T1 vs. T0;

^¶= result tends to be different within the same group at T1 vs. T0;

^‡= result is significantly different within the same group at T2 vs. T0;

^#= result tends to be different within the same group at T2 vs. T0

Figure 3.

Perceived Level of Difficulty (PLD) (Fig 3A) and Personal level of enjoyment (PLE) (Fig 3B) for all exercises (aerobic and resistance & balance combined) according to the proportion of ratings given by HEP participants

Between-group differences: Changes between results at T0 and T1 are shown in Table 2. From T0 to T1, HEP group increased its PA level compared to CONTR group, as expected. However, the HEP group tended to become more frail, reducing its frailty index (SOF) compared to CONTR group. Finally, comparing the HEP group to the CONTR group, a tendency to improvement in the one-leg stand duration and a significant improvement in the walking speed on 4 m were observed in favour of the HEP group. From T0 to T2, despite an improvement observed in the walking speed on 4 m within both groups, this improvement was significantly more important in the HEP group, as shown in Figure 4.

Table 3.

Between-group differences (Δ) from T0 to T1 and T0 to T2

Variables	HEP		CON		p Δ
	Δ T0-T1	Δ T0-T2	Δ T0-T1	Δ T0-T2	p Δ T0-T1	p Δ T0-T2
Total Fat Mass (%)	3.06 ± 8.04(%)	4.25 ± 16.41 (%)	4.63 ± 1.63(%)	17.04 ± 9.79 (%)	0.221	0.142
RAPA- Physical Activity Level	29.63 ± 31.95(%)	20.37 ± 23.74 (%)	-16.67± 24.09 (%)	3.70 ± 21.85 (%)	0.035*§	0.829
SF-36 Quality of Life Total score	12.50 ± 20.92(%)	14.51 ± 33.72 (%)	16.67 (34.16) (%)	34.60 ± 18.40 (%)	0.584	0.109
SF-36 Bodily pain component	40.83 ± 37.41(%)	27.92 ± 34.07 (%)	16.50 ± 20.66(%)	174.40 ± 166.91(%)	0.582	0.855
One-leg stand (s)	11.90 ± 18.50	2.51 ± 5.30	2.16 ±8.41	4.54 ± 15.47	0.078†.||	0.584
Walking speed (m/s)	0.25 ± 0.12	0.27 ± 0.10	0.05 ± 0.09	0.13 ± 0.05	0.025 *§	0.025*,**
Handgrip/BW	2.51 ± 21.88 (%)	7.15 ± 14.25 (%)	-10.65 ± 14.57(%)	-1.08 ± 15.52 (%)	0.337	0.262
Time up and go (sec)	-3.5 ± 4.13	-4.14 ± 3.71	-1.57 ± 3.04	-4.39 ± 15.47	0.262	0.873

^*=p < 0.05;

^†.= p < 0.1;

^§= difference from T0 to T1 is significantly different between HEP and CON groups;

^||= difference from T0 to T1 tends to be different between HEP and CON groups;

^**= difference from T0 to T2 is significantly different between HEP and CON groups; §§ = difference from T0 to T2 tends to be different between HEP and CON groups

Figure 4A.

Within and between groups differences in walking speed on 4 meters Figure 4A: Within-group difference in walking speed from baseline (T0) to three-month evaluation (T1)

Discussion

This study aimed at exploring the feasibility, the acceptability and the effects of a HEP using the Jintronix® rehabilitation system in a group of previously independent community-living older adults who sustained a minor injury. In accordance with our hypothesis, this twelve-week HEP intervention using the Jintronix® rehabilitation system was found to be feasible, acceptable and safe. Feasibility was measured by the adherence and the compliance to the prescribed program. Firstly, the adherence to the program was high, with 5 out of 6 participants completing 99% of their exercise sessions. Only one participant had a very low adherence rate of 19%. Lower adherence to exercise programs being associated with poor health status and low physical ability (28), this person's profile could explain this low rate. Even though this person's physical ability did not change much over the twelve-week intervention period, with a SPPB score going from 8/12 to 9/12, his health status seemed to have languished, with the number of reported health problems going from one to four from baseline to the three-month evaluation.

Secondly, participants executed 96% of expected repetitions with a mean quality level of 87.5% (calculated on sessions actually realized), even though direct supervision was offered only 25% of the time (6 out 24 sessions). These results are in line with Schutzer and Graves (9)'s review indicating that older adults' compliance to exercise programs can be achieved through low cost and simple prompt methods such as telephonesupervised follow-ups and motivational interventions. Moreover, feedbacks provided by the Jintronix® system as well as the constant adaptation of individual difficulty level, that progressed from an average of 4/10 to 6.2/10 throughout the 24 sessions, probably contributed to the enhancement of participants' self-efficacy feelings, and consequently to their motivation to comply with the prescribed program. Indeed, self-efficacy is known as strong predictor of compliance to exercise in the elderly population (9, 53) and gradual activity progression is recognized as a contributing factor to motivation (53). One of the most interesting features of using a technology for exercising is the possibility to collect objective information, available through the computer system, on adherence and compliance. However, this has not been thoroughly examined in past studies, even less in home settings (18), making it difficult to position this trial. Nevertheless, these results present a great potential for adequate monitoring and supervision of participants to physical activity interventions.

Acceptability was measured by the perceived level of difficulty (PLD) and the personal level of enjoyment (PLE) of participants. In fact, equally important to objective measures of feasibility are the participants' feelings and perceptions over the program and the technology. Allender, Cowburn (54) reported that older adults might restrain from practicing physical activity because they are unsure of the adequacy and relevancy of exercise programs with their needs, especially their age, thus making acceptability an important factor of success. The ratings obtained in this study reveal that 90% of all exercises were considered easy or moderate which is well above our 80% target. Even though these results can hardly be compared to other studies since they are usually not reported in this manner, they suggest that the program was adapted to the participants' needs. Indeed, according to the American College of Sports Medicine and the American Heart Association, aiming at a vigorous level of intensity of exercise might not be suitable for all older adults, but the simple fact of reducing the time spent in sedentary activity may provide health benefits (55). Furthermore, the ratings obtained for PLE indicate that 92% of all exercises were enjoyed a lot or well enjoyed, which is also well above our 80% target. These results are in line with Meldrum, Glennon et al. (33) who reached an acceptability level based on enjoyment of 82% in a study using the Nintendo Wii® Fit Plus. Enjoyment being considered one of the main reasons for older adults to participate in physical activities (54), the high PLE observed in this study was most likely a contributing factor to high adherence and compliance (9). Finally, acceptable level of difficulty and high level of enjoyment are crucial for eventual large scale implementation of physical activity programs.

Figure 4B.

Between-group difference in walking speed from baseline (T0) to three-month evaluation (T1) and from Baseline (T0) to follow up (T2)

As for the effects this HEP, they are considered positive and promising. Indeed, results showed that this HEP was effective in improving walking speed. This parameter is one of the most important predictor of adverse events, namely mobility loss and mortality, in the elderly. On the one hand, the HEP group went from a walking speed of 0.86 m/s, which is closed to the 0.8 m/s cutpoint associated with higher risk of adverse events (56), to a 1.12 m/s at T1 and 1.13 m/s at T2, which are results above the 1.0 m/s cutpoint associated with lower risk of health events and better survival rate (56). On the other hand, the observed increases in walking speed of 0.25 m/s from T0 to T1 and 0.27 m/s from T0 to T2 are above the minimal clinically important difference of 0.1 - 0.17 m/s (57). Surprisingly, we observed a significant increase in the CONTR group walking speed between T0 and T2. We should remember that all participants were recruited after a minor injury in baseline evaluation was conducted within 7 days of their ED visit. Their walking speed could have been affected by their injury. Then, the significant increase in their walking speed could be explained by the fact that six months later, they have recover from their injury and feel better. However, contrary to our hypothesis, the intervention did not prevent decline in all physical functions, as seen in the results for the frailty index. Surprisingly, only the HEP group tended to reduce its score, i.e. becoming more frail, a change solely explained by a decline in the energy level component; the two other components (weight loss and ability to raise from a chair) remained the same from baseline to the three-month evaluation (results not shown). Nevertheless, other results such as the one-leg stand, an important predictor of falls in the elderly population (58), significantly increased within the HEP group from T0 to T1 and tended to be significantly different from the CONTR group at the three-month evaluation. Baseline evaluation positioned both groups closed to the 5 second cutpoint associated with higher risk of falls, but only the HEP group well exceeded this cutpoint three months later.

Finally and most importantly, no adverse events were reported throughout the trial. Overall, our results are consistent with those of Gerling, Livingston (21) who reported that exergames are suitable for the elderly, as long as they are adapted to their individuals needs, and can be used as an enjoyable way of staying active.

This study is not without limitations. The small sample size may have interfered with the results of both groups. Moreover, the Jintronix® rehabiliation system being under development at the time of the trial, only the resistance & balance portion of the program could be objectively assessed in quantity and quality. Finally, the design of the study prevented us from accurately assessing the effects of the HEP on quality of life as perceived physical and mental health status was greatly affected by the very recent injury sustained by the participant, particularly the pain component which was, by design, expected to improve from baseline to the three-month evaluation, even without an intervention.

Limitations

The major limitation of this study is the sample size. The small number of participants in each group limits the results and the statistical power. Also, the fact that the data primarily apply to women (11 women, 1 man) is a limitation of this study. Moreover, no follow-up data has been collected for the participants who withdrew which could be considered a limitation. The results regarding acceptability may be biased by the fact that only data of those who remained in the study were analyzed. Finally, the Hawthorne effect is also one limitation of this study. Effectively, the fact that the participants knew they were evaluated during this study could have biased their results since they might have more desire perform an evaluation setting.

Conclusion

Overall, the results of using this gerontechnology for community-dwelling older adults are promising. We demonstrated that this twelve-week HEP intervention using the Jintronix® gerontechnology is feasible, acceptable and safe for community-living older adults who sustained a minor injury. We also demonstrated that this intervention could increase walking speed, the most important predictor of adverse events in the elderly population, and that the improvement could be maintained over time. Results suggest that other functional outcome measures such as one-leg stand, TUG and SPPB could also improve with a larger number of participants. Thus, this specific HEP should be validated in a larger population and in other conditions in order to better assess the potential effects of such intervention in the prevention of inactivity and, consequently, the loss of autonomy in elderly population.

Conflict of Interest

The authors have no conflicts of interest in association with this study.

Ethical standard

All the interventions on human subjects published in this article are in accordance with international ethical standards supported by this journal.