Effects of online group exercises for older adults on physical, psychological and social wellbeing: a randomized pilot trial

Background Intervention programs to promote physical activity in older adults, either in group or home settings, have shown equivalent health outcomes but different results when considering adherence. Group-based interventions seem to achieve higher participation in the long-term. However, there are many factors that can make of group exercises a challenging setting for older adults. A major one, due to the heterogeneity of this particular population, is the difference in the level of skills. In this paper we report on the physical, psychological and social wellbeing outcomes of a technology-based intervention that enable online group exercises in older adults with different levels of skills. Methods A total of 37 older adults between 65 and 87 years old followed a personalized exercise program based on the OTAGO program for fall prevention, for a period of eight weeks. Participants could join online group exercises using a tablet-based application. Participants were assigned either to the Control group, representing the traditional individual home-based training program, or the Social group, representing the online group exercising. Pre- and post- measurements were taken to analyze the physical, psychological and social wellbeing outcomes. Results After the eight-weeks training program there were improvements in both the Social and Control groups in terms of physical outcomes, given the high level of adherence of both groups. Considering the baseline measures, however, the results suggest that while in the Control group fitter individuals tended to adhere more to the training, this was not the case for the Social group, where the initial level had no effect on adherence. For psychological outcomes there were improvements on both groups, regardless of the application used. There was no significant difference between groups in social wellbeing outcomes, both groups seeing a decrease in loneliness despite the presence of social features in the Social group. However, online social interactions have shown to be correlated to the decrease in loneliness in the Social group. Conclusion The results indicate that technology-supported online group-exercising which conceals individual differences in physical skills is effective in motivating and enabling individuals who are less fit to train as much as fitter individuals. This not only indicates the feasibility of training together despite differences in physical skills but also suggests that online exercise might reduce the effect of skills on adherence in a social context. However, results from this pilot are limited to a small sample size and therefore are not conclusive. Longer term interventions with more participants are instead recommended to assess impacts on wellbeing and behavior change.


104
 to investigate if and how online group-exercising and baseline measures of physical, 105 social and psychological well-being influence the adherence of older adults to the training 106 program. 107  to assess the effectiveness of an OTAGO-based exercise program delivered via an online 108 group-exercising -effectiveness measured as the improvements in the physical functions 109 expected by the exercise program. 110 111 Additional objectives were to assess the effect of a social (rather than individual) virtual gym 112 that enables to train in group on psychological and social wellbeing outcomes. 117 Campbell & Robertson, 2003]. Although this specific program was designed for home-based 118 training, a meta-analysis including other exercise programs for fall prevention found that 119 combining group-based and home-based exercises is a strategy used in several effective trials, 120 thus recommending both options to be available [Sherrington et al., 2011]. 121 122 Technology-based interventions have also demonstrated to be effective in increasing and 123 maintaining physical activity (refer to [Müller, 2014;Aalbers et al., 2011] for systematic 124 reviews). Among the technological components that have been explored we can mention: online 125 newsletters [Hageman et al. 2005], personalised emails [Ferney et al., 2009], web-based videos 126 [Irvine et al. 2013;Benavent-Caballer et al., 2015], tablet applications [Silveira et al., 2013a] and 127 video game consoles [Jorgensen et al. 2012]. However, most of the existing intervention studies 128 have focused on individual training, or provided a social context that was limited to forums or 129 chats [Aalbers et al., 2011]. Even exergames, a type of technology that have traditionally 130 provided more immersive experiences (e.g., via MS Kinect and Nintendo Wii), have not been 131 explored in an online group setting. A systematic review on exergames besides reporting on lack 132 of conclusive results on the improvements in physical functioning, reported only one 133 interventions with exercise performed in pairs, but that required physical presence [Molina et al., 134 2014]. 135 136 In summary, there are no interventions exploring online group-exercising in home settings, 137 where training programs were tailored to individuals. In this paper, we report on the feasibility 138 and outcomes of a technology-based intervention in such settings. It complements our previous 139 reports on the same intervention: 140 -Baez et al., [2016], on the application design and human factors. We describe the design 141 rationale and its evolution, and report on the technology acceptance, usage and usability, 142 and the nature of online interactions. 143 -Far et al., [2015], on whether the application motivated participants to adhere to the 144 training program, and whether the virtual gym design motivated older adults to train 145 together in a group. 146 From our previous work we understood that i) the virtual gym was highly usable and accepted, 147 ii) it motivated individuals to follow the training program, iii) and to join the training sessions in 148 groups as opposed to alone. In this paper, however, we focus in understanding how baseline 149 measures of physical, social and psychological well-being affect the adherence to the training 150 program, and how this effect is modified by training in an online social setting. We also report 151 on the effectiveness of the training program in terms of physical outcomes, and additionally on 152 social and psychological wellbeing.
153 Materials & Methods 154 The methods followed in this intervention study has been described in detail in . 155 In this paper, we elaborate in more detail the aspects that are the focus of this paper.  164 The design of the application is based on the metaphor of a virtual gym, mimicking the spaces 165 and services found in a real gym. The main features of the service are: 166 • Reception. The entry point to all the services of the gym. 167 • Locker room. As in a real gym, a space where trainees usually meet each other and 168 get ready for the training classes. In the locker room, users can see each other, interact 169 by means of predefined messages (e.g. "Hi, let's go to the classroom"), and invite 170 members who are not online to join.
• Classroom. The environment where users have access to the training videos. In this 172 space users are not only able to see the Coach, but also each other as avatars, giving 173 the feeling of training together. 174 • Progress report. It displays the progress of the trainee in the training program by 175 means of a growing garden metaphor. 176 • Training schedule. It displays the training schedule for the week, displaying 177 participation of users in each session, and reminding them of the upcoming sessions. 178 • Messaging. Messaging features allow users to exchange public and private messages. 179 Trainees use this feature to communicate with other trainees and the Coach. 180 • Positive and negative reinforcement. After the completion of a training session, 181 positive or negative reinforcement messages are presented to users depending on the 182 number of exercises completed. 183 184 In Gymcentral, personalisation is related to the level of difficulty of the exercise, and it works 215 hours each) focusing on operating the tablet, the use of the main applications and the Gymcentral 216 app. The training took place after the pre-measurements. 217 218 As part of the study kit, participants received a 10.1 inch Sony Xperia tablet with the assigned 219 version of the application installed, the user guide including the names and telephone numbers of 220 the support team, instructions about the use of the tablet and the assigned application, one pair of 221 ankle weights to perform the exercises and a folder to allow the vertical positioning of the tablet. The study took place in Trento, Italy, over a period of 10 weeks from October to December 235 2014. The duration and size of this study is similar to previous technology-supported 236 interventions that have seen significant results in adherence and improvements physical 237 measures (e.g., [Silveira et al., 2013b]). The first week was devoted to technical deployment and 238 application testing, followed by 8 weeks of training and 1 week of post-training measurements. 239 The training program was supervised by a training coach, who could intervene to advance 240 trainees in the exercise program, and to provide technical support upon request. The same level 241 of technical support was made available to both groups. 242 243 The study received ethical approval from the CREATE-NET Ethics Committee on ICT Research 244 Involving Human Beings (Application N. 2014-001). As the study follows a framework for the 245 design of complex interventions in healthcare , at this stage it is 246 considered as a pilot study.
247 Participants 248 We considered eligible for the study participants aged 65 or older, independent-living, self-249 sufficient and with a non-frail, transitionally frail or a mild frailty level. The latter was measured 250 using the Groningen Frailty Indicator [GFI; Steverink et al., 2001], considering eligible those 251 participants scoring lower than 7 in the scale from zero (not frail) to fifteen (very frail). 252 Participants wearing pacemakers were not considered eligible, since the study required the use of 253 a mobility sensor, as well as participants not able to undergo the exercise program according to 254 their family doctor. 255 256 We recruited participants through members of local associations (Ada and Auser) that promote 257 initiatives for elderly persons in Trento, Italy. We sent invitations to the 70 persons that visited 258 the associations more recently. Out of these, 10 persons declined the invitation, 6 were excluded 259 because they lived in an area without 3G or LTE coverage, and 14 did not meet the frailty 260 criteria. In the end, a total of 40 participants between 65 and 87 years old were recruited for the 261 study (29 females and 11 males, mean age = 71, s.d. = 5.7). All participants obtained a formal 262 written approval by their family doctor to allow them to participate in the study. Both doctors 263 and participants received a written outline and explanation of the study and signed the consent 264 before participating. 265 266 267 269 From the initial group of participants, 4 older adults withdrew at different times during the 270 course of the study due to unpredictable health or family problems. One participant was 271 substituted because the withdrawal occurred before the beginning of the study, while the others 272 could not be replaced since they withdrew during the course of the study. For this reason, the 273 results are based on the data from 37 participants (27 females and 9 males, mean age = 71.2, s.d. 274 = 5.8, between 65 and 87 years old). 275 Participants were not told to which group they were assigned or that a different version of the 276 application was being tested.
277 In relation to the technology, less than 20% of the participants had ever used a tablet before, and 278 less than 10% used it regularly. Thus, and as mentioned in the previous subsection, all 279 participants were provided introductory courses on how to use the tablet. 280 281 In Table 1 we summarise the initial measures for both groups. A t-test for independent groups 282 shows no statistical difference between control and experimental groups in terms of the initial 283 measures, except for the ones related to the physical assessment. This issue is addressed later in 284 the analysis. 285  [Gardner et al., 2001], and was adapted with by a professional personal trainer 292 in order to fit the original program into 10 levels of increasing difficulty (Table S1). 293 294 The OTAGO Exercise Program is used worldwide and is one of the most tested fall prevention 295 programs (with four randomised controlled trials and one controlled multi-center trial [Stevens, 296 J. A., 2010]). The program includes muscle strengthening and balance-retraining exercises of 297 increasing levels in terms of duration and repetitions. The duration of the exercise sessions 298 ranged from 30 to 40 minutes, with longer sessions in the higher levels. 299 300 Participants from both the social and control group were assigned an initial level by the Coach 301 based on the pre-test analysis. As a minimum requirement, participants were then asked to 302 participate to two exercise sessions per week. Participants of both groups were able to progress 303 in the exercise program every week, via an automatic level-up suggestion by the system. If  439 As stimulating aspects, two main themes emerged in the social group: training with others and 440 the possibility of following the training program. Other participants expressed the possibility of 441 messaging or challenging themselves. In the control group the possibility of training from home 442 was also a main theme along with the personal satisfaction of doing the exercises. Interestingly, 443 in this group one person reported that "The experience was interesting, though it was a pity that 444 the Coach was in the video and not actually present". 445 As negative aspects, the dominant issue was the intermittent interruptions in the Internet service 446 that occurred at some point during the study, and which affected both groups equally. 447 448 Usefulness of features 449 Participants of the social group continued to use all features of the application throughout the 450 study, although with different perceived usefulness. As reported in Baez et al. (2016), the 451 features that are instrumental to the training were naturally experienced by most of the trainees, 452 and this includes exercising in the classroom, checking out the schedule, and more importantly, 453 training with the company of others. Together these features were highly valued. Persuasion 454 features were also among the most experienced and valued. This includes, following the progress 455 and visualizing their own progress in the garden and, still very positive but to a lesser extent, 456 inviting others to join a training session 457 Interestingly, social interaction features received mixed results. While group chat was used 458 widely, personal messaging was heavily used to interact with the Coach but less with other 459 participants.
460 The results above show not only that participants of the social group did feel as if they were 461 training with others but also that they considered it as one of the most stimulating aspects. This 462 supports the results reported in , where it was observed that participants in an 463 online group setting resulted in a significantly higher number of joint sessions (training together 464 as opposed to training alone) compared to the control group.
465 Online social interactions 466 We briefly summarise the usage of the online social interaction features that were available in 467 the social group: private messages and bulletin board (public messages). The usage was analysed 468 by looking at the database of messages exchanged among participants.
469 Private messages were preferred over the public messages, accounting for 75% of all the 470 messages exchanged (544 messages). The most active user was the Coach, who had to contact 471 the trainees to check on their progress on a weekly basis, followed by the Technician. We 472 illustrate the interactions between all the participants in Fig. S2  , provided interesting insights on the effect of 499 social features on participants' adherence. In this work, we extend on the previous analysis to 500 report on the effect of participants' physical abilities and attitudes towards exercising on their 501 adherence to the training. 502 503 Attrition. The intervention resulted in a 7.5% attrition rate (corresponding to 3 participants), 504 measured in terms of the proportion of participants lost at the end of the study. Reasons behind 505 the withdrawal of these participants were related to unexpected health and family problems or, in 506 one case, because of Internet connection issues which could not be solved. 507 508 Persistence. The overall persistence rate in the two groups was of 76% (SD=22.6%), when 509 considering the total number of sessions available ( Figure 3A). In the social group the 510 persistence rate was 85%, while in the control group it was 64%. A between-subjects analysis of 511 variance was performed to compare the persistence of both groups while controlling for the 512 initial scores in the measures of gait speed and leg muscle strength (physical measures), and the 513 enjoyment of physical activity. The independent variables were grouped in three equally 514 distributed intervals (Low, Medium, High). The analysis showed a significant interaction 515 between group and the initial measures of leg muscle strength (F(2, 23)= 5.966, p=.008, partial 516 eta squared =.342), but no significant interaction with gait speed (F(1, 23)=3.42, p=.08, partial 517 eta squared = .13) nor enjoyment of physical activity (F(2,23)=1.93, p=.17, partial eta squared= 518 .144). In Figure 3B we show the relevant interaction plot. 519 The same analysis, exploring the effect of the initial loneliness score as independent variable 520 (with and without controlling for the other variables) reveal no significant interaction with group. 521 522 523 525 The interaction between group and leg muscle strength shows a higher persistence in participants 526 of the control group that scored higher in the initial leg muscle strength test. For the social group 527 however, there is no significant difference in the persistence of participants based on their initial 528 leg muscle strength score. This suggests that, when training individually, fitter participants tend 529 to adhere more to the training, but when social elements are in place these differences in fitness 530 no longer determine adherence. 531 532 There was also a significant main effect for group (F(1,23) 539 The lower variability of persistence in the social group can be explained by the social features 540 (normative influence, social facilitation, social learning) that might have motivated users to 541 comply with the community norm. In the control group, the higher variability suggests a stronger 542 effect of individual differences due to the lack of social awareness. 543 544 Completeness. The overall completeness rate in the two groups was 90.32% (SD=17.4%), 545 suggesting that participants tended to complete the working sessions once they started. The 546 completeness rate in the social group (M=91.75%, SD=12.46%) was slightly higher compared to 547 that of the control group (M=88.63%, SD=22.24%), although not statistically significant. As in 548 the previous measure, a between-subjects analysis of variance was performed to compare the 549 completeness of both groups while controlling for the initial scores in the measures of gait speed 550 and leg muscle strength, enjoyment of physical activity and loneliness. No interaction was found 551 between group and the initial scores, but a main effect for initial leg muscle strength (F(2,23)= 552 5.075, p=.015, partial eta squared= .306). We attribute this effect to the duration of the video 553 exercises that were assigned to fitter individuals (e.g., required more repetitions), which were 554 longer for higher levels of intensity. 555 556 The higher level of completeness and lower variability in the social group can be explained by 557 the presence of self-monitoring tools (e.g., positive and negative reinforcement) and social 558 facilitation (exercising with others), which were lacking in the training sessions of the control 559 group participants.