This study was a 6-week pre-post pilot trial. The study protocol was approved by the University of California, San Francisco (UCSF), Institutional Review Board prior to participant recruitment and enrollment, and all participants signed an informed consent. Participants were recruited from community talks, university research databases of older adults who had previously agreed to be contacted for future studies, and dental and physical therapy clinics at the UCSF Medical Center. Eligibility criteria were age ≥65 years, kyphometer-derived kyphosis measurement ≥40° (±5°), ability to walk ≥1 block independently without an assistive device, and access to a mobile phone. Participants were excluded for cognitive impairment (inability to draw a normal clock or recall any words on the Mini-Cog [24]). We recruited a small sample to determine feasibility and acceptability and provide preliminary estimates of the effects and SD in clinical measures of kyphosis, HRQoL, and physical function in older community-dwelling adults with hyperkyphosis.

Interested participants were initially screened by telephone, and those who met the preliminary inclusion criteria were scheduled for a face-to-face screening/baseline visit where baseline kyphosis was measured. Participants meeting all eligibility criteria were enrolled in the study and later attended a 30-minute face-to-face session (intervention) with the study coordinator who provided an overview of the 6-week program and a study manual that coincided with the study video clips including pictures and instructions for exercise and good posture during activities of daily living. This session included two components: instructions and practice in the technical aspect of logging in to the university and library websites to view the video clips and instructions and practice in responding to the text messaging prompts. Each participant was given a unique password and log-in instructions to the university and library websites.

Baseline demographic characteristics of the enrolled participants were summarized using mean, SD, and range for continuous measures. Tabulations and percentages were used to summarize categorical variables. We used descriptive statistics of mean with SD and median to characterize process measures required to demonstrate adherence to the intervention. Kyphosis, physical function, and HRQoL scores at baseline and 6 weeks postintervention were summarized with means and SDs. Paired t tests were used to assess the effects of the intervention on changes in the scores from baseline to the end of the 6-week intervention period. The results were also reported with confidence intervals and P values. All analyses were conducted using SAS software, version 9.4 (SAS Institute Inc, Cary, NC).

A total of 64 potential participants were recruited between January and May 2018, of which 29 met the preliminary eligibility criteria by telephone and were scheduled for an in-person clinic/baseline visit. Twelve potential participants did not meet the kyphosis eligibility criteria (kyphosis ≤40° ± 5°), and 17 (58.6%) met all eligibility criteria and were enrolled in the study (Figure 4). The 12 participants who completed the intervention and follow-up visit had a mean age of 71.6 (SD 4.9) years, and 92% had college, professional, or graduate degrees (Table 2). As per the baseline self-efficacy scales, participants were 97% confident they would practice good posture at least three times a day, 97% were confident that they would watch a daily video, and 98% were confident that they would reply to a daily text. Five participants did not complete the study. One dropped out for medical reasons, two dropped out during the first 2 weeks due to frustration with the two-step log-in process, one lost interest after 2 weeks, and one completed all aspects of the intervention but did not return for the 6-week postintervention testing. The mean age of the 5 participants (two male, three female) who did not complete the 6-week follow-up visit was 71.5 years, and 80% had college, professional, or graduate degrees (one unknown). The one participant who completed the intervention but did not return for the 6-week follow-up visit reported 50% confidence on the self-efficacy scales at baseline.

We explored the feasibility of subject recruitment, retention, and acceptability of an exercise and posture training program sent as video clip links and text messaging prompts via a mobile phone to older adults with hyperkyphosis. Only 8 (12.5%) participants who were screened for the study did not qualify due to no access to a mobile phone, highlighting prior reports that technology use is rapidly increasing in older populations [17]. Of those who did not complete the 6-week postintervention visit, two dropped out during the first 2 weeks because of frustration with the two-step log-in procedures; we learned that a two-step process is too cumbersome, even for those who completed the study. Overall, participant acceptance and satisfaction with the intervention was positive. Adherence to the intervention was high among those who completed the 6-week intervention. Based on participant feedback, we learned that daily text messaging reminders were too frequent and participant preference for frequency of text messaging reminders is an important consideration for future studies. Furthermore, the mobile phone screen was too small for easy viewing, and future modifications should include access to the video for viewing on a larger computer screen and an accessible log-in without the two-step process. To investigate the effects of a technology-based intervention on kyphosis and kyphosis progression in the future, we plan to conduct a longer intervention with regular interactive video webinars to keep participants engaged, and provide online demonstrations and feedback to participants. These webinars could be accessed from individual personal computers that provide a larger screen and avoid the cumbersome two-step log-in process that our participants found difficult.

Kyphosis decreased by 8° (95% CI –12 to –5), OTW distance decreased by 1.9 cm (95% CI –3.3 to –0.7), and PASE score increased by 29 points (95% CI 3 to 54), which indicates improvement in clinical measures of kyphosis and physical activity, although we did not have a large enough sample to assess the potential efficacy of this program on progression of kyphosis, OTW distance, and physical activity in older adults with hyperkyphosis. Results of this study suggest that delivering an exercise and posture self-management program via technology is promising and deserves further investigation.

Our results are consistent with studies showing that self-management programs that focus on day-to-day management of chronic diseases significantly improve heath behaviors and health status [42,43]. At baseline, our participants reported a high likelihood of adherence to the proposed self-management intervention, an indication of high self-efficacy that has been associated with a significantly higher likelihood of good outcomes among older adults with knee osteoarthritis [44,45]. High self-efficacy has also been shown to predict successful integration of other healthy behaviors and enhance sustainability of high levels of self-care [46,47].

Results of our study are also consistent with those of a previous systematic review that reported that text messaging reminders enhanced participants’ abilities to self-manage their chronic condition (asthma, diabetes, or hypertension) [48]. Text messaging reminders have demonstrated efficacy in improving adherence to a variety of health behaviors across multiple domains [49-51]. Moreover, participants who perceived improvements in self-management preferred reminders via mobile phone messaging over email reminders, highlighting the acceptance and usability of mobile phone technology in older populations [48]. Another systematic review of smart technology interventions aimed at facilitating, supporting, and sustaining self-management through behavioral change in people with chronic obstructive pulmonary disease, concluded that the use of technology improved HRQoL and physical activity compared to face-to-face or digital/written support [52] and continued use of a smart technology intervention improves sustainability of behavior change over time.

The improvement in kyphosis in our study exceeds that reported in previous randomized controlled trials that tested the efficacy of in-person exercise and posture training interventions in older adults with hyperkyphosis over longer periods of time [14,15]. We used trained testers at the University’s Clinical and Translational Science Institute’s physical performance laboratory, provided additional training for our outcome measurements, and ensured high within- and between-tester reliability prior to the study. Although there were large SDs in the kyphosis measurements and a very small sample size, the change in kyphosis exceeded the standard error of the measurement. It is possible that our technology-based educational intervention improved motor control and provided participants self-management tools and greater autonomy for improving posture as compared to prior in-person interventions.

There are no prior studies that specifically target OTW distance; however, the 1.9-cm change in OTW distance observed in our study did not exceed the smallest detectable difference of 3.2 cm according to Bland-Altman criteria previously reported among adults with ankylosing spondylitis [34]. In contrast, the PASE scores of physical activity improved by 29 (SD 40) points, which is a robust 27% improvement from the baseline mean score of 107 (SD 61) points to 138 (SD 55) points at the 6-week visit. The mean score for older adults aged ≥65 years was 102.9 (SD 64.1) points, suggesting that educating participants about their posture may also increase their participation in physical activities compared to their age-matched peers [38]. Participants reported greater frequency and duration of walking and performing more yard work and home repair after the intervention. It is possible, although unlikely, that this additional physical activity contributed to a change in kyphosis, but we cannot discount the possibility that the skills learned from the intervention are responsible for both the increase in physical activity and improved kyphosis. 

This pilot study may be the first of its kind to investigate the feasibility of delivering a technology-based kyphosis-specific exercise and posture-training program by mobile phone over 6 weeks in older community-dwelling adults with hyperkyphosis. These programs have previously required in-person training for 3-6 months. However, there are several limitations. First, this was a pilot study without a control group, and the results may be larger than a between-group comparison in a controlled trial. Second, we excluded participants who were not English speakers and did not have access to a mobile phone, tablet, or computer. In addition, most of our participants were white and had college degrees or graduate degrees, which limits our ability to generalize the preliminary results to the overall older adult population with hyperkyphosis. Third, our sample size was small with large SDs in the effects; however, these data will be helpful in designing a future randomized controlled trial in a large sample. Fourth, failure to complete all aspects of the study was high, although one participant dropped out for medical reasons unrelated to the study, three dropped out within the first 2 weeks, and one completed all aspects of the study except the 6-week testing. This drop-out rate is within the range of 6% to 34% reported in exercise interventions in older persons. In addition, the early drop-out rate observed in our study is consistent with that in previous studies reporting the highest number of dropouts in the first 3 months of longer-term studies [53]. Fifth, our sample was highly educated and implementing this study in a lower-educated population may be challenging. However, a systematic review that described characteristics of community-dwelling older adults that influence acceptance of technology listed the familiarity of older adults with modern technology, the fit between the housing type and technology, and the compatibility of the technology with adults’ cultural background, rather than educational level, as limiting factors [18]. Future randomized studies to determine the efficacy of this technology among larger and more diverse populations are needed.

Technology-based exercise and posture training using video clip viewing and text messaging reminders is feasible and acceptable in a small cohort of older adults with hyperkyphosis. Adherence to video viewing is excellent and adherence to practicing exercise at least three times a day warrants further study to optimize adherence and the optimal dose. Future trials should assess the benefits of more individualized feedback with posture training videos and customize daily text reminder prompts according to participant preference. Technology-based exercise and posture training in older adults with hyperkyphosis warrants further study as a potential self-management program for age-related hyperkyphosis that may be more easily disseminated than in-person training.