Synchronous Group-Based Tele-Exercise versus Community-Based Interventions: Effects on Physical Functioning and Adherence in Older Adults at Risk of Falls - A Randomized Controlled Trial

Study design

This was a 3-month single-blind, randomized controlled trial (RCT) comparing synchronous group-based tele-exercise training (TE) with face-to-face community-based exercise training (CB) for community-dwelling older people at risk of falls. More detailed information, including details of the outcome measures and intrinsic differences between the two groups, can be obtained from our study protocol.¹⁹ Reporting follows the CONSORT 2010 statement.²⁰

Research ethics

Approval for the study was granted by the Ethics Committee of the College of Professional and Continuing Education (RC/ETH/H/0029). The study was registered in the Chinese Clinical Trial Registry (ChiCTR2200063370) on 5 September 2022. In accordance with the WMA declaration of Helsinki, informed consent was obtained from all participants.

Sample size calculation

Sample size calculations were conducted using G*Power software, assuming an effect size of 0.367 based on a previous study by Benavent-Caballer, et al²¹ that focused on the Time Up and Go Test. With a confidence level of 95%, a power of 80%, and an estimated attrition rate of 20% (including mortality) ²², a minimum sample size of 92 participants was determined to be necessary.

Participants and eligibility criteria

The participants were recruited from 10 elderly community centres located in various locations across Hong Kong. The inclusion and exclusion criteria are listed in table 1.

Equity, diversity, and inclusion statement

Our study encompassed all eligible cases based on the specified inclusion and exclusion criteria, without intentional exclusion of individuals from marginalized communities. We ensured the inclusion of participants across diverse genders, socioeconomic backgrounds, and technological proficiency levels. The author team comprised two women and seven men from various disciplines, including exercise physiology, health economics, medicine, exercise sociology, occupational therapy, and telemedicine. It also included two junior scholars. Data collection methods were consistent across all cases, irrespective of educational, or socioeconomic variations within the communities. Gender-specific findings were discussed in the subsequent analysis.

Randomization and blinding

Ten community centres will be coded and then randomly assigned to either the tele-exercise group (TE) or community-based group (CB) with a 1:1 allocation in a concealed, blinded fashion to minimize cross-over contamination between the groups. Randomization procedure and allocation sequence was supervised by an independent assistant. The project assistant enrolled participants in the community centres. Trained outcome assessors, who were blinded to group assignments, performed the assessments. Participants were aware of their unique training modes but had no contact with participants from other groups due to different locations. Instructors and researchers were not blinded and were instructed not to discuss group assignments with the participants and outcome assessors.

Intervention programmes

Both interventions have been described in detail in the published protocol.¹⁹ All participants engaged in exercise training sessions three times a week for a duration of three months, either through tele-exercise (TE) or face-to-face at the community centres (CB). The participants in both groups were divided into classes consisting of 8 to 11 individuals. The exercise programme followed a structured format, which included a warm-up, Otago exercise programme (OEP),²³ "workouts on-demand," and a cool-down.

Outcome measures

Baseline assessments and post-intervention assessments at the three-month were conducted by trained and blinded assessors at the community centres. Participants received HK$100 gift vouchers at 3 months for completing the training and assessments. In addition to the assessments, demographic information and background variables such as sex, age, and history of falls were collected. Cognitive function at baseline was evaluated using the Mini-Mental State Examination (MMSE).²⁴ The Chinese version of the Physical Activity Scale for the Elderly (PASE-C) was utilized to assess the physical activity level of older adults.²⁵

Data analysis

Baseline characteristics including age, adherence rate, and dropout rate, were analyzed using an independent t-test. An intention-to-treat analysis, following the guidelines for randomized controlled trials,²⁹ was employed, including all originally randomized participants in the study. For the outcome measures, a linear mixed models with time and group as the fixed effect and within person correlation modelled as a random effect. The differences between categorical variables such as education level and % participants with possible sarcopenia were analyzed by Chi-square test. A significance level of 0.05 was set for all tests. The statistical analysis was carried out using SPSS, version 26.0 (IBM Co., Armonk, NY).

Sensitivity analyses were conducted to validate the statistical findings. These included utilizing multiple imputation to address missing data, which confirmed consistent results with the original analysis.

Deviations from the registered trial protocol

Instead of the initially planned 92 participants, we recruited 91 participants, factoring in an anticipated attrition rate of 20% to account for potential dropouts. Despite this deviation, the trial remains adequately powered. As the follow-up period has yet to be completed, we have only reported the results for baseline and the 3-month timepoint. In terms of statistical analysis, we employed a linear mixed model instead of a two-way repeated measures ANOVA. This decision was made because a linear mixed model is better suited to handle unbalanced data, enables capturing individual differences, reduces within-subject variability.

Flow of participants

During the screening period from October 2021 to June 2022, a total of 212 older adults were identified for preliminary screening. Out of these, 120 individuals were excluded based on the inclusion and exclusion criteria of the study. Eventually, 91 participants successfully met the inclusion criteria. The TE group comprised 45 participants, while the CB group consisted of 46 participants (Figure 1).

• Download figure
• Open in new tab

Figure 1

Flow of Participants.

Participant characteristics

The baseline characteristics of the participants did not show significant differences between the CB and TE groups, as summarized in Table 2.

Technological accessibility and technological support for the TE participants

In terms of smartphone ownership, the TE group exhibited a higher percentage at 84.4%, while the CB group reported 71.7% (p<0.05). Similarly, in terms of home Wi-Fi ownership, the TE group also had a higher percentage at 71.1%, compared to the CB group at 58.7%, but this difference did not reach a significant level. Additionally, only 17.8% of TE participants had previous experience with tele-exercise (Table 2).

Baseline home visits were conducted for all participants before the intervention, with an average of 3.7 visits per participant. Additional visits addressed technological issues and provided support, totaling 260 visits (including baseline visits) for the 45 participants in the TE group, averaging 5.8 visits per participant (range: 0-12 visits).

Primary outcomes- falls incidence, fall risks and sarcopenia status

There was no significant difference between the groups in the falls incidence. Prior to the intervention, the average number of falls per participant was 1.57 ± 0.83 over the past 12 months in the CB and 1.78 ± 0.77 in the TE. The majority of falls were classified as minor incidents (CB: 97.2% vs TE: 97.8%).³⁰ During the intervention, the CB group reported no falls, whereas one participant in the TE group reported experiencing a fall outside of the lessons, which resulted in a minor injury (Table 3). The FES-I scores improved significantly in both groups, indicating a reduction in fear of falling. (p<0.001) (Table 4).

There were no significant differences between the two groups or between the two time points in the proportion of participants with possible sarcopenia, sarcopenia, and severe sarcopenia. However, there was a trend towards a decrease in participants with possible sarcopenia and severe sarcopenia in the CB group over time (Table 3).

Primary outcomes- physical function and activity level

Although no significant difference was observed between the CB and TE interventions, most outcomes showed improvement after the 3-month intervention (Table 4). The 6-meter walk test, the TUG scores and the BBS scores improved significantly in both groups over the 3-month period (p<0.05).

SARC-CalF scores and the FRT scores demonstrated significant improvement in the CB group compared with its baseline (p<0.05). Handgrip strength, measured in all participants, did not show significant changes between or within groups. However, within the female subgroup, CB participants showed a significant increase in handgrip strength compared with its baseline (p<0.001). Activity levels, assessed by the Physical Activity Scale for the Elderly (PASE-C), significantly improved in both groups (p<0.001).

Primary outcomes- body composition changes among female participants

Due to the different standards for body composition between genders and the significant proportion of female participants in both groups, the body composition changes among female participants (CB n=41; TE n=39) were analyzed and presented. The results showed minimal differences between the CB and the TE (Table 5). Calf circumference, BMI and appendicular skeletal muscle mass index (ASMI) remained relatively stable in both groups over the 3-month period. However, body weight exhibited a significant decrease in the TE group (p<0.05).

Secondary outcomes- adherence rate, dropout rate and PACES

Adherence rates were high in both groups, with the CB group averaging 90.0% adherence and the TE group averaging 85.0%. The dropout rate was 0% in the CB group, while the TE group had a dropout rate of 4.4%. The scores on the PACES were 108.6 in the CB group and 108.2 in the TE group, out of a total of 126 (Table 2).

Adverse events reporting

One participant in the TE group reported general tiredness and pain during screening phase, but her doctor found no issues after thorough examinations and tests. Living alone and rarely engaging in social activities, she insisted on participating in our programme to maintain her physical functioning.

During the tele-exercise class, she reported tiredness and knee and lower back pain at the beginning. The instructor advised her to rest when needed and practice at an easier level. With the frequent reminders to lower the intensity, her problems subsided. She completed the programme with an adherence rate of 72.2%.

Key findings

To our knowledge, this study is the first rigorously designed RCT with sufficient statistical power to investigate the effectiveness of the synchronous group-based tele-exercise compared with the community-based group in older adults at risk of falls. In contrast to previous trials^{14 15}, our study includes a comprehensive range of outcomes such as fall efficacy, fall incidence, sarcopenia status, physical function, activity level, body composition, exercise adherence, dropout, enjoyment, adverse events, technological accessibility and their education backgrounds. The study findings indicate that both the CB and TE interventions were effective in decreasing fear of falling, and enhancing physical function including mobility, balance and gait speed, among the participants. These results regarding the effectiveness of the modified OEP are consistent with the findings of systematic review conducted by Martins, et al³¹ and Wu, et al³² The positive outcomes can be attributed to the fact that our intervention adhered to the recommended effective programme requirements identified by Wu, et al³², with an average duration of 45 minutes, three times per week for 12 weeks.

The study employed synchronous tele-exercise, utilizing video-linked appointments supervised by healthcare professionals, and emphasized three main strategies: self-efficacy, self-confidence, and empowerment, to enhance the exercise experience of the participants.^{19 33 34} Immediate feedback after each class facilitated timely personalized guidance, addressing participants’ concerns. All classes were led by the same knowledgeable instructor who had over 10 years of experience in teaching older adults and people with chronic diseases. Our training programme effectively mitigated the challenges associated with remote exercise sessions, as evidenced by the lack of significant differences between the two groups in physical function.

The study demonstrated high adherence and enjoyment among participants in both the CB and TE interventions, with low dropout rates and high PACES scores. CB achieved 90% adherence and TE reached 85%, surpassing the approximately 70% reported in a systematic review of community-based exercise programmes.³⁵ As recommended, an 80% adherence rate is considered feasible.³⁶ Consistent with previous studies, community-dwelling older adults expressed high satisfaction with online exercise training.^37–39 The success of the tele-exercise programme can be attributed to factors such as professional support, individualized technology orientation sessions, and standby technical assistance.^{39 40} Our programme also provided prompt technical assistance, averaging 5.8 home visits per participant throughout the 36 lessons, accounting for approximately 16.1% of the sessions. These measures ensured prompt problem-solving, exercise safety, and the absence of serious adverse event, indicating a high margin of exercise safety with the synchronous online delivery of modified OEP.

Upper body strength improvements

While the OEP primarily focuses on strength and balance exercises for the lower extremities, it has been shown to prevent a decline in handgrip strength when compared to no intervention control⁴¹ and conventional care.⁴² However, recent meta-analysis reviews did not find clear benefits of the OEP on upper body strength in older adults.³² These analyses did not analyze the changes based on gender, as there are gender differences in the norm of handgrip strength. Nevertheless, our study observed significant improvements in upper body strength exclusively among female participants within the CB group. The primary purpose of engaging the upper limb muscles was to provide support and maintain balance during exercises. While this aspect may appear minor, it holds significant importance for older adults who are at risk of falls. Previous research has demonstrated that even small volumes of resistance exercise, performed briefly and infrequently, can lead to substantial increases in muscular strength.⁴³ The improvement in handgrip strength in the CB may explain the decreasing trend of possible sarcopenia and severe sarcopenia among the CB participants over time.

Impact on body weight and fat

Overall, the impact of the interventions on body composition parameters was modest, with slight variations between CB and TE groups. However, subgroup analysis revealed that female participants in the TE group experienced a significant decrease in body weight after three months, compared to baseline. The significant decrease in body weight in our study may be due to increased physical activity levels, as indicated by higher PASE-C scores. Regular participation in Otago-based exercise can serve as an effective energy expenditure method for older adults at risk of falls, particularly for those who did not have established exercise habits and maintained very low activity levels. Unlike Duckham, et al⁴⁴ ,who demonstrated a reduction in fat mass with an Otago-based exercise program, and Hong, et al⁴⁵, who found improvements in lower limb muscle mass through resistance exercise delivered via video conferencing, our results showed a divergence. It’s worth noting that Duckham, et al⁴⁴ ’s program had twice the duration of ours, lasting for 24 weeks. Additionally, Hong, et al⁴⁵ incorporated nutrition education alongside exercise, which is crucial for preventing and managing sarcopenia. ⁴⁶ This disparity in approach could explain the significant reduction in fat mass and increase in muscle mass observed in their study but not in ours.

Bridging technological and education gaps in tele-exercise

Most tele-exercise studies primarily recruited individuals with existing internet access at home, favoring those who were technologically competent.^{11–13 47} A strength of our study was that technological competence and device ownership were not inclusion criteria. We provided participants with the necessary devices, welcoming individuals with various technological backgrounds. Our sample consisted of less educated individuals, with over 50% completed primary education or below, and only one TE participant completing university. Our findings indicated that older adults, regardless of education and technological proficiency, could enjoy tele-exercise when accompanied by well-designed programmes, technical support, and timely feedback. Furthermore, the utilization of commonly available telecommunications software in the trial, instead of custom-built systems¹³, enhanced the generalizability of the findings.⁴⁸

Technological exposure and potential selection bias

Our findings revealed a significantly higher rate of smartphone ownership in the TE group (84.4%) compared to the CB (71.7%) (p<0.05), indicating greater technological exposure among TE participants. Additionally, the TE group had a higher proportion of participants with Wi-Fi at home (72.1% vs. CB: 58.7%, N.S.). The TE group also exhibited slightly higher levels of education, suggesting potential selection bias towards technologically experienced and higher-educated individuals. It is important to note that three participants declined participation before the intervention phase due to family objections and concerns about unsafe home environments. This highlights the presence of skepticism towards the use of tele-exercise among older adults, indicating that some individuals and their families may have reservations about engaging in exercise programmes remotely.

Limitations

One noteworthy concern with tele-exercise is the potential impact on the proper execution of exercises, especially for older adults with specific health conditions or mobility issues. Due to intrinsic differences between the delivery modes of CB and TE, the instructor could only provide verbal corrections for participants’ movements, without the ability to physically assist them.¹⁹ It is worth noting that conducting tele-exercise sessions in the limited space of most Hong Kong residences presents challenges. Census and Statistics Department⁴⁹ indicate that the median floor areas for domestic households and individuals in Hong Kong are 40 and 16 m², respectively. This spatial constraint could impact the ability to capture a full view of participants while they were standing during tele-exercise sessions. To overcome this issue, we made efforts during home visits to find the optimal placement and angle for tablets with built-in cameras to maximize visibility during the exercise sessions. To ensure understanding of the exercise programme and promote exercise safety, all participants were required to attend the first class at the community centre. This allowed for in-person instructions and reminders before transitioning to tele-exercise sessions.¹⁹ It is important to note that the success of this program may not be transferable to other exercises that involve more dynamic movements such as standing, sitting, and lying down, within a single lesson. From a technological standpoint, it would be beneficial for the technology company to explore the possibility of incorporating tablets with wide-angle front cameras or establishing a seamless connection between a tablet and an external camera without any delays. These enhancements would help improve visibility during tele-exercise sessions.