Physically active lifestyles are important for healthy aging, enhancing physical mobility and independence, and reducing risk for many chronic diseases [1,2]. Despite physical activity (PA) guidelines, 94% of Canadians older than 60 years are sedentary for more than 8 hours per day [3], and more than a third of Canadians aged 65 years or older report a mobility disability [4]. Mobility disability is characterized by frequent transitions between states of mobility independence and mobility limitation (disability) [5]. This can include a decline in the frequency of performing certain activities or a modification in the way one performs certain activities, and it is often indicative of poor overall health status [6]. Although declines in indicators of mobility, such as slowing of walking speed (gait speed), is seen with normal aging, such changes predict both survival [7,8] and independence [9].

Increasingly, many people turn to the internet and social media as a source of health information [10-14]. Unfortunately, much of the Web-based health information available is not based on scientific evidence and, therefore, is unlikely to produce the intended health benefits [15,16]. Members of the public may not have the knowledge, skills, or time to sift through and identify credible messages [17-19] and, thus, may be acting on recommendations, which are unlikely to improve their health. Evidence from recent systematic reviews suggests that websites and social media have the potential to improve health behaviors, self-efficacy, and health outcomes in older adults [20], and social media interventions may positively impact health outcomes [21]. However, it is not known if access to high-quality information about maintaining and improving physical mobility results in lifestyle behavior change in older adults.

The McMaster Optimal Aging Portal (the Portal) was launched in English in 2014, and in French in 2017, as a knowledge translation (KT) tool to increase public access to trustworthy health information [22-26]. KT has been defined as “a dynamic and iterative process that includes synthesis, dissemination, exchange and ethically sound application of knowledge to improve the health of Canadians, provide more effective health services and products and strengthen the health care system” [27]. The Portal helps readers to access synthesized evidence-based resources, identify trustworthy messages, and understand scientific findings. Topics related to mobility are of interest to users: the categories “arthritis and joint conditions” and “exercise” are consistently in the monthly top 10 most-accessed lists. On the basis of the monitoring of website and email subscription analytics, users are engaging with the Portal; now we want to know if easy-to-understand, evidence-based messages change what people know and do to stay healthy and mobile.

The purpose of this study was to understand if and how the KT strategies used to disseminate information relevant to increasing PA and maintaining and improving mobility via the Portal impacts knowledge, behavioral intentions, and health among middle-aged and older Canadian adults.

Participant flow through the study is displayed in Figure 1. Of the 523 individuals who responded to our call for participants, 510 provided informed consent and completed baseline questionnaires and were randomized to the intervention group (n=256) or control group (n=254). Participant characteristics are displayed in Table 1. The mean age of the participants was 64.7 years, with the majority female (430/510, 84.3%), well-educated (474/510, 92.9% had completed postsecondary education), and living in urban settings (422/510, 82.7%). There were no baseline differences between groups, with the exception of the proportion of participants who reported a fall in the last 6 months (41/256, 16.0% vs 62/254, 24.4% in the intervention vs control group; P=.02). There were no differences in the number of falls or the proportion of participants who visited a health care provider because of a fall.

There was no difference between the intervention and control groups in the number of participants lost to follow-up. Participants who did not complete the end-of-study (17.6%) or follow-up (31.6%) questionnaires were more likely to have never used the Portal, be employed full time, and live in rural locations than those who completed the study. There were no other differences in participant characteristics or baseline values for study outcomes between those who did and did not complete questionnaires at all 3 time points (data not shown). No adverse events were reported by participants during the study period.

Changes in PA are listed in Table 2. There were no significant between-group differences at the end of the study (P=.09) or follow-up (P=.07). Both groups were more likely to be categorized in a higher PA level using the RAPA at the end of the study or baseline (intervention: odds ratio [OR] 3.35, 95% CI 2.04-5.49; control: OR 1.86, 95% CI 1.14-3.03), with improvements sustained at follow-up compared to baseline (OR 3.27, 95% CI 1.96-5.47; control: OR 1.67, 95% CI 1.01-2.77). There were no between- or within-group differences in the proportion of participants classified as meeting Canada’s PA guidelines at either time point. The proportion of participants that reported they self-monitored PA was higher at the end of the study compared to baseline (intervention: OR 3.56, 95% CI 2.06-6.18; control: OR 3.05, 95% CI 1.76-5.27) and follow-up compared to baseline (intervention: OR 3.33, 95% CI 1.89-5.87; control: OR 2.04, 95% CI 1.17-3.55), but there were no differences observed between the intervention and control groups. A similar pattern was observed for the level of mobility disability using the Manty Preclinical Mobility Disability Scale (Table 2).

There was a significant between-group difference in participants’ attitudes toward mobility-related health behaviors at the end of the study (P=.02) but not at follow-up. Participant’s intentions to participate in mobility-related health behaviors declined slightly among participants in both groups, with a significantly greater decline in the control group (P=.04). There were no significant differences in intentions at follow-up. There were no significant between- or within-group differences for self-rated health or total knowledge score (data not shown).

As part of our planned subgroup analyses, a significant between-group difference was found at both the end of the study (P=.04) and follow-up (P=.02) for level of PA in participants with low self-rated health at baseline. No intervention effect was observed in participants with moderate-high self-rated health. There were no significant differences when the study sample was stratified by previous Portal use (data not shown).

At the end of the intervention period, participants in the intervention group were more likely to report that the Portal influenced their PA behaviors, and that Portal information influenced their decisions more often (3.42 vs 2.73 out of 7; Table 3). There was no difference between groups in the impact of the Portal on monitoring mobility or the proportion of participants who sought information about maintaining or improving mobility from a health care provider or other sources. The majority of participants in both groups reported receiving weekly email alerts from the Portal, with no difference between groups. Approximately one-third of the participants visited the Portal browse page, and 19.5% and 6.1% of participants reported using Facebook or Twitter to access Portal-related materials, respectively. No adverse or unintended events were reported by participants during or after the study period.

In the 3 months following the intervention period, half of the participants in both groups reported using the Portal to look for mobility-related information, with no differences observed between groups. Participants in the intervention group were more likely to report that the Portal had influenced a decision about PA in the last 3 months (3.89 vs 3.43 out of 7; P=.04), whereas the control group was more likely to use the Portal to seek out information on other topics (59.9% vs 49.2%; P<.05). There were no differences between groups in the percentage of participants who continued to receive email alerts or access the Portal through Twitter, Facebook, or the browse page following completion of the study (Table 3).

This study is the first to evaluate the impact of dissemination of evidence-based information about mobility and PA through the Portal on PA and mobility outcomes. Participants in both the targeted KT intervention and self-serve control group reported increased PA after the 12-week intervention, with benefits maintained at 3-month follow-up; however, no significant between-group differences were observed. The lack of difference between groups is not surprising given the high degree of engagement with Portal materials reported by both groups; 89.4% of control group participants reported signing up for the Portal’s general weekly email alerts. Although engagement was lower for social media and Portal browsing, there were no significant differences between the targeted intervention group and control group. Although our KT intervention did focus specifically on topics related to PA and mobility, these topics are among the most common on the Portal itself, and it is likely that the control group was exposed to similar information during the study and poststudy period. Due to the nature of the Portal as an already existing Web-based resource, we were unable to include a true control group in our study. Thus, contamination across the control group may contribute to the lack of significant differences between study groups.

In planned subgroup analyses, we found a significant effect of the intervention in individuals who had low self-rated health at baseline. There are several potential explanations for this finding. It is possible that those with lower self-rated health benefited more from the targeted aspects of the KT intervention and specific content chosen. This suggests that certain subgroups may benefit from different or more tailored KT strategies (eg, medium of message delivery, including behavioral feedback), potentially in line with the barriers to PA that they face. This should be explored in future studies. Given that our study sample was relatively healthy and active at baseline, the small amount of change seen over time may be the result of a ceiling effect; perhaps those with low self-rated health had the greatest potential for change.

A number of behavior change theories suggest that provision of information alone is inadequate to result in long-term behavior change of a sufficient magnitude to affect long-term health outcomes [37]. On the basis of the Theory of Planned Behavior [33], attitudes toward PA and intentions to engage in activity are predictors of PA behavior. In this study, participants’ attitudes toward activity and intentions to engage in PA were significantly different between groups at the end of the study, suggesting that the targeted KT intervention had a stronger effect on these constructs. Portal materials are designed to have actionable messages within content and are specifically targeted at middle-aged and older adults. We hypothesized that this targeting would act on normative and control beliefs of participants, but further tailoring of messaging (eg, dissemination of content specific to participant characteristics or baseline knowledge or preferences) may be necessary to elicit greater behavior change. In a recent study, inner-city minority participants with type 2 diabetes were randomly assigned to an intervention delivered through a Web-based portal, which included self-management modules, health education, and social networking. Importantly, this intervention also included interaction with a telehealth nurse. At the end of the study, participants in the intervention group showed greater knowledge of diabetes and diabetes management, greater self-rated physical and mental health, greater weight loss, and improved diabetes control, although results should be interpreted with caution because of the large loss to follow-up observed in both groups [38]. These findings do, however, support our hypothesis that further tailoring and interaction with participants may increase the effectiveness of our intervention.

Although we did observe a significant within-group difference in PA throughout the study period, the absolute magnitude of the change may be considered small or moderate: an additional 12.6% of intervention group and 6.2% of control group participants were classified in the highest PA at the end of the study compared with baseline. These findings are consistent with a recent Cochrane review of computer-based weight loss or weight maintenance interventions, which found that Web-based interventions were superior to minimal intervention or control; however, they were not as effective as in-person interventions [39]. However, given the relative low cost, ease of delivery using existing Portal materials, and scalability of an intervention such as this, we believe that the small absolute change observed in this study has the potential to contribute to a meaningful difference at a population level.

An important limitation to our study is the reliance on self-report data for PA and mobility disability. Although we used a previously developed and validated tool, it is known that individuals tend to self-report higher levels of PA [40]. Due to the lack of blinding of study participants, it is possible that the intervention group had a higher degree of self-report bias; however, given the high engagement with the Portal materials in both groups, particularly around PA and mobility-related content, we believe that any overestimation of PA was similar between groups. We chose to use a self-report tool from a feasibility standpoint to be able to include a broad sample of participants across Canada. Future work could consider low-cost methods such as smartphone tracking to gather some objectively measured data.

Our study sample was relatively homogenous, consisting of relatively healthy (59.4% of participants rated their health as “Excellent” or “Very Good” at baseline), well-educated, urban-dwelling adults. Demographics of our study sample are similar to those of general Portal users previously reported by our study team [22], although our study sample was approximately 5 years younger and had a higher proportion of females. This is not surprising as approximately one-third of the study participants reported being regular Portal users at baseline, with another third reporting using the Portal occasionally. This is consistent with findings from a recent systematic review, which found that individuals with lower education as well as racial and ethnic minorities are typically less likely to use health portals [41]. More work is needed to understand how to engage these underserved groups, who may have potentially more to gain from a KT intervention such as this.

Although the Portal has been successful in engaging citizens and health care professionals, its use has not yet been evaluated with respect to changes in knowledge or behaviors. An understanding of how participants engage with both the Portal and the KT strategies is essential for ensuring the content and delivery of information through the Portal, and other health information websites will be most effective at encouraging behavior change and ultimately improving health. As highlighted by Grimshaw et al [42], the current evidence-base to guide the choice of effective KT strategies aimed at consumers to improve health outcomes is still incomplete [42]. These study findings have relevance for both individuals who use Web-based heath information resources and organizations that develop and provide it. On the basis of our findings, the KT strategies used in this study may result in improved intentions and health behaviors in particular subgroups and thus have the potential to impact a number of health outcomes, including mobility and functional independence over a longer follow-up period. More work is needed to understand which groups may benefit most from a low-cost, easily scalable intervention such as this.