Use parameters refer to the duration of the intervention, the duration of each session, the number of sessions, and their frequency. In this review, the duration of the intervention varied from 5 weeks to 18 months. Only 18% (2/11) of the interventions reported the duration of each session [37,39,40]. It would have been helpful to know the duration of each session in the other studies to understand how much time older adults need to spend to complete the sessions. The number of sessions varied from 5 (1/11, 9%) [41] to 16 (1/11, 9%) [39,40]. Although the total number of sessions was not specified, in some interventions, the authors proposed an intensity of intervention such as 5-minute daily sessions (1/11, 9%) [37] or free access during the entire study period (2/11, 18%) [36,51,52]. Otherwise, in 45% (5/11) of the interventions, the number of sessions was not reported [12,19,20,38,42,44-50], which does not allow us to understand how many web-based intervention sessions older adults require. The session frequency was daily [37], twice per week [39,40], weekly [41], and once every 2 weeks [43]. In 45% (5/11) of the interventions, the frequency of sessions was not reported [12,19,20,38,42,44-50]. These were the same studies that did not report the number of sessions.

BCTs are the strategies used in interventions to promote behavior change, such as feedback, action planning, or instruction [21]. On the basis of the taxonomy by Michie et al [21] comprising 93 BCTs, we identified 15 different BCTs used in the web-based interventions. The number of BCTs used varied from 1 to 9, with an average of 4.5, meaning that all the interventions except 1 (10/11, 91%) [42] combined multiple BCTs. As the combination of BCTs was varied, it is difficult to establish links between the most used or effective combinations and the results obtained and discern the contribution made by each of them. However, we identified a trend in the combination of instruction, self-monitoring, and feedback (7/11, 64%) despite the fact that this combination was mostly coupled with one or more other BCTs such as action planning or goal setting.

All the interventions (11/11, 100%) used instructions on how to perform the behavior. Most of the interventions included feedback on the targeted health behavior (9/11, 82%) [12,19,20,36,37,39-41,43-45,48-50] and self-monitoring of the behavior (8/11, 73%) [12,19,20,36-41,43,48-52]. Other BCTs used included action planning (5/11, 45%) [36,41,43-47], goal setting (4/11, 36%) [12,19,20,37,41,43,48-50], problem solving (4/11, 36%) [12,19,20,37,41,46-50], awareness (1/11, 9%) [46,47], verbal persuasion (1/11, 9%) [39,40], commitment (1/11, 9%), self-regulation (1/11, 9%) [44,45], prompts and cues (1/11, 9%), rewards (1/11, 9%), social comparison (1/11, 9%), and relapse prevention (1/11, 9%) [43]. Although common techniques were identified, they were rarely explained by the authors, which makes their definition and application unclear. In fact, only 18% (2/11) of the interventions provided detailed discussions on how the BCTs were used to generate the desired change [41,43]. For example, Alley et al [43] detailed that action planning was used by asking questions about the participants’ actions in terms of what, when, and where to perform the behavior. Several other BCTs from the taxonomy by Michie et al [21], such as distraction, self-affirmation, or scheduled consequences, were not identified in this review. For a summary of the BCTs used in each web-based intervention, see Multimedia Appendix 4 [12,19,20,36-52].

The delivery modes include every mode in addition to the web component grouped as follows: automated function (eg, video, automated tailored feedback, and automated following messages such as reminders or encouragement), communicative function (eg, chat session, peer-to-peer access, and “ask the expert facility”), and additional modes (eg, email, phone calls, and videoconferencing).

The coding scheme developed by Michie and Prestwich [33] allows for an assessment of the extent to which the interventions are theory-based. In total, 55% (6/11) of the interventions were based on at least one theory. A total of 18% (2/11) of the interventions were based on 1 theory, whereas 36% (4/11) were based on 2 to 5 theories. The theories used were the theory of planned behavior (2/11, 18%), social cognitive theory (2/11, 18%), precaution adoption process (1/11, 9%), integrated model for change (1/11, 9%), self-regulation theory (3/11, 27%), transtheoretical model (2/11, 18%), self-determination theory (2/11, 18%), motivational interviewing (1/11, 9%), I-Change Model (1/11, 9%), and health action process approach (1/11, 9%).

Among the theory-based interventions, most of the studies (5/6, 83%) did not explicitly state how the theory was used to develop the intervention or how it was integrated into the intervention to lead to the desired change. Therefore, it is difficult to understand the contribution made by the theories used and link them to the results obtained. This was despite the fact that the authors of the papers on all the theory-based interventions specified that they wanted to act on constructs of the theory related to the targeted behavior change, such as self-efficacy, motivation, or attitudes, yet they did not provide definitions of the constructs or explanations of how they operationalized them. Definitions and explanations of the constructs would be needed to understand how the intervention attempted to act on them and lead to behavior change [53]. In only 18% (2/11) of the interventions [41,43] did the authors link theory constructs with the BCTs used. For example, Alley et al [43] indicated that the BCTs instruction and feedback were used to change the attitudes of the participants. In other studies (9/11, 82%), no connection was made between the BCTs and theory constructs when this would have helped us understand how the chosen BCTs would lead to the desired change in terms of the constructs targeted. Otherwise, the authors of the papers on 18% (2/11) of the interventions measured a construct of the theory, such as self-efficacy, as an outcome of the study [12,43]. The authors of the papers on 9% (1/11) of the interventions specified that the theory’s constructs were used to tailor the intervention to the participants such that the content of the advice depended on the intrinsic motivation of the participant [44,45]. Table 1 presents a summary of the components of the interventions surveyed.

Of the 6 studies that evaluated the effects of web-based interventions on PA in older adults, 4 (67%) found positive outcomes on PA after the intervention, including increasing weekly minutes of moderate to vigorous PA and greater likelihood of performing self-reported cycling [37,41,45,46]. Other positive effects were also reported on blood pressure, lipid levels, BMI, smoking cessation, self-efficacy [12], and participants’ knowledge and skills to adopt a healthy lifestyle [38]. Among the 35% (7/20) of studies that conducted a qualitative evaluation, 5 themes emerged from the thematic analysis, which are detailed in the following sections: tailoring, motivation, support, barriers, and perceptions.

Several studies (4/7, 57%) addressed the concept of tailored web-based interventions. Indeed, older adults mentioned that they appreciated participating in a web-based intervention tailored to their limitations and preferences [20,42]. Some participants mentioned that the exercises proposed in the web-based intervention were too easy and repetitive, not adapted to their environment [51], or of limited value owing to their medical condition [36]. This suggests the need for web-based interventions tailored to older adults’ preferences, environments, and conditions.

Motivation appeared to be central to behavior change among older adults as most studies (6/7, 86%) addressed it. Participants stated that web-based interventions should help increase their motivation for change [19,51]. Identifying their own motivation is important for older adults to maintain behavior change [36], and the sources of such motivation varied, including personal benefits and health improvement [19,36]. Some participants mentioned that the coach’s positive message could help boost their motivation [20,51] and that being motivated helped them continue using the intervention and vice versa [20]. Other participants argued that behavior change, such as being more physically active, was not a goal in itself but rather that they were motivated by other health benefits such as remaining independent as long as possible [42]. For these reasons, it would appear necessary to explore the individual motivations of each older adult to facilitate change.

The theme of support was reported in several studies (5/7, 71%). Older adults mentioned that they need support to achieve their health goals [19] and that the platform could provide continuous support [20], especially as training alone at home requires discipline [42]. Some participants mentioned that they appreciated having discussions with a coach throughout the web-based intervention [19,40]. Some participants also argued that a first meeting with the coach was necessary to develop a relationship of trust and then facilitate change and that the coach played an important role in stimulating the initial use of the web-based intervention and sustaining it [20]. In other words, participants who feel connected with the coach are more likely to keep using the platform and continue pursuing goals for lifestyle changes [20]. In addition, participants in the study by de Souto Barreto et al [51] would have appreciated having more contact with a member of the research team, suggesting that, as older adults, they would have been favorable to having the support of a coach. Conversely, peer interactions were less valued and used by older adults [40].

Some barriers were consistently identified in the studies (4/7, 57%) on the use of web-based interventions among older adults. First, older adults mentioned barriers to the web-based interventions, such as a lack of computer skills [19,20] or using an old computer [42]. Difficulties encountered in computer use or limited internet skills could discourage older adults from using a web-based intervention [20]. Some participants also mentioned that they sometimes lacked the discipline required to exercise alone at home [36,42]. Without the support of a coach, older adults feared getting hurt [42], which could be a barrier to behavior change.

Behavior change among older adults seemed to be influenced by their perceptions of their age and the benefits of change. Indeed, older adults who do not perceive a need to improve their lifestyles or who do not prioritize it because of their advanced age are less likely to use a web-based intervention and, therefore, engage in lifestyle changes [20]. By contrast, perceiving that behavior change could lead to health benefits positively influences older adults toward using the intervention [19,20]. However, this theme emerged in only 29% (2/7) of the studies, which is why it may have had less impact than the other themes.

This scoping review sought to explore the extent of the available literature on web-based interventions as a way to promote healthy lifestyles among people aged ≥65 years. In total, 11 different interventions discussed in 20 published articles were included in this review. Almost all the articles (19/20, 95%) were published in the last 5 years, which indicates growth in the development and evaluation of this type of intervention among older people. This is consistent with the increased use of the web by older adults in recent years [6-8] and the urgent need to deploy cost-effective strategies to facilitate access to health care [13-15].

As found in other studies [22,54], our results show that the studies included predominantly young older adults, with few that took an interest in the “oldest old” such as persons aged ≥85 years. As the literature is so limited on web-based interventions involving people of more advanced age (eg, ≥85 years) and as the components and effects of web-based interventions may differ for this population, further studies are needed across the aging spectrum. This scoping review found that web-based interventions among older adults are mainly focused on increasing PA. This high prevalence could be explained by the fact that older adults are considered the most sedentary age group [55] and that the benefits of PA are considerable for this population [56]. Given that other lifestyle habits such as diet, stress, and alcohol consumption [2] would also be favorable to the health of older adults, more studies should be conducted to evaluate the effects of web-based interventions on these habits in this population.

The web-based interventions included in this review had various components. The interventions were diverse in terms of their use parameters (ie, duration, number of sessions, completion time, and frequency). In almost all the interventions (10/11, 91%), at least one detail regarding the use parameters was omitted, making it difficult to understand the intensity of the interventions offered to older adults. As noted in the studies examining preferences toward web-based interventions, older adults prefer a few 30-minute sessions [57] or shorter 10-minute sessions on a regular basis every 2 or 3 days [57,58]. Among adults, other studies have shown that web-based interventions that are more intensive [24], that allow for longer durations, such as 60-minute sessions or more, and that propose a total number of sessions of >3 [59] are more effective at producing behavior change [24,59]. In this review, because of a lack of detailed information on use parameters, it was difficult to identify any trend in use parameters that were more relevant to supporting change among older adults. Further research is needed to better investigate the optimal use parameters (ie, duration, number of sessions, completion time, and frequency) of web-based interventions for older adults.

In this review, we found that the most used BCTs were instruction, feedback, and self-monitoring, which is similar to the findings of other studies that explored the use of BCTs in web-based interventions [22,60]. However, BCTs were used in combination without explaining why these choices were made and how they were operationalized. For this reason, it was difficult to discern the contribution of each BCT to the results obtained [61] and how they could have led to change [53]. BCTs are the active ingredients in an intervention to effect behavior change. Therefore, it is crucial for authors to be explicit about their choice of BCT combinations to understand how interventions produce their effects [62]. Although some interventions (2/11, 18%) observed that some BCTs such as self-regulation could be effective for adults and not for older adults [63,64], further research is needed to understand which BCTs are more appropriate to support older adults in their adoption of healthy lifestyles. In particular, further studies are needed to explore which combinations of BCTs could optimize the intervention’s impact and how each BCT interacts with the others within an intervention to produce behavior change among older adults [62].

The results of this review show the diverse range of delivery modes used in web-based interventions. Some included an electronic diary (2/11, 18%), quiz (1/11, 9%), or videos (4/11, 36%) as well as supplementary modes such as phone calls (2/11, 18%), face-to-face meetings (2/11, 18%), and email reminders (3/11, 27%). All the interventions (11/11, 100%) provided instructions on how to perform the behavior, which other authors have pointed out as the core of most web-based interventions [65]. Many (5/11, 45%) proposed automated feedback, which would be one of the most effective delivery modes leading to behavior change in adults [26]. Only 9% (1/11) of the interventions included a forum with peers, and it was underused by participants, which is inconsistent with other studies in which adults aged ≥50 years showed high use [66] and appreciation [67]. Few interventions (3/11, 27%) offered a chat with a coach, which is similar to the findings of other studies on web-based interventions among adults [22,59]. In the interventions that did offer a chat with a coach, participants were offered an opportunity to communicate with a coach if needed rather than for constant support. The actual nature, dose, and type of coaching provided by this coach was poorly reported by the studies. However, as pointed out by other authors, the constant support of a coach throughout a web-based intervention could replace the sense of interpersonal connectedness found in in-person interventions [68], which older adults seek [18,69]. A systematic review also found that web-based interventions that include human support are more effective at behavior change among middle-aged and older adults than stand-alone interventions [24]. That being said, although a certain delivery mode could be more appropriate for older adults [27], the literature on this subject is limited. More studies, such as meta-analyses, are needed to identify which delivery modes are more effective at inducing behavior change among older adults. This could guide the design of future interventions. Further studies should also investigate web-based interventions that give older adults the choice to participate in a group forum, as well as different forms of coaching by a professional throughout the web-based interventions. More studies are needed to examine the nature of the role played by the coach throughout a web-based intervention as well as explore the dosing and type of support needed to help older adults adopt healthy lifestyles. For us, it seems clear that this coaching could be provided by a health professional such as a nurse as it is a nurse’s role to support people in health promotion and older adults appreciate developing a trusting relationship with a nurse [69].

In this review, few of the interventions (6/11, 55%) were based on a theory. The most common theory reported was the self-regulation theory. This differs from other reviews, in which one mainly finds the social cognitive theory in behavior change interventions among middle-aged [70] and older adults [71]. Among the theory-based interventions, most of the studies (5/6, 83%) did not report how the theory was used to design an intervention that would lead to the desired change. For this reason, it is difficult to draw conclusions regarding the theories and the results obtained. Theories help explain why and how behavior change occurs and provide guidance on the potential determinants to be targeted by the intervention to induce behavior change [62]. In addition, designing interventions based on theories allows us to link the theoretical determinants of behavior change with intervention components [33], know which BCTs to use [62], and ensure that the intervention will lead to behavior change. Indeed, it is well known that theory-based interventions are more effective than non–theory-based interventions [72], and this has also been demonstrated in a population of older adults [71]. Future web-based interventions to promote healthy lifestyles among older adults should be based on theory, and researchers should clearly state how theories guide the development of their interventions. Further studies are needed to compare interventions based on different theories in terms of the effects identified on the lifestyles of older adults.

As reported in previous studies [65,70], we observed a favorable trend in the use of web-based interventions to increase PA among older adults. This review found that web-based interventions can also have positive effects on blood pressure, lipid levels, BMI, smoking cessation, self-efficacy and knowledge, and the skills needed to adopt a healthy lifestyle. This is consistent with research findings on the benefits of web-based interventions [24,25].

As a result of our analysis, 5 themes emerged that appear to be central to web-based lifestyle change interventions among older adults: tailoring, motivation, support, barriers, and perceptions. As has been pointed out by many authors [29,73-75], the results of this review show that motivation is one of the most important factors influencing the lifestyle habits of older adults. As motivation is an intrinsic factor (ie, each person must identify their own), increasing individual motivation among older adults may facilitate behavior change [76]. In this sense, future web-based interventions among older adults should target this determinant as a way to help them adopt healthy lifestyle habits [26]. This review also found that older adults appear to appreciate interventions that include support from a coach, which also supports their motivation for change and engagement with the intervention. These findings are consistent with other studies in which older adults mentioned that support and the development of a relationship of trust are necessary in behavior change interventions [18,29,77]. This finding may inform the development of future web-based interventions intended to promote healthy lifestyles among older adults by including the support of a coach.

In line with the results of other studies [57,78], this review highlighted the fact that older adults would prefer interventions that are tailored to their preferences and conditions. Indeed, it appears that tailored web-based interventions can make older adults more engaged in behavior change [79] and lead to better recall of information [80]. Previous studies have suggested that tailored web-based interventions are more effective at inducing behavior change than generic interventions in a middle-aged adult population [25,26]. For older adults, designing a tailored web-based intervention appears to be even more important considering the heterogeneity of this population and the various challenges associated with aging, including comorbidities and frailty, which are experienced differently by older adults [27]. Consistent with the findings of other authors [57,79], this review found that older adults may face barriers to using web-based interventions, such as lack of computer skills and difficulties using the technology. For the development of future web-based interventions, it would appear necessary to consider the barriers that older adults face in using technology and find ways to overcome them. Including access to a coach through the web platform for initial and ongoing guidance could help reduce such barriers and, in turn, avoid discouragement among older adults committed to change [79]. In addition, the findings of this review indicate that older adults’ lifestyle habits are influenced by their perceptions of change in old age, as reported in other studies [73,75,79]. It would appear necessary to explore older adults’ perceptions of change and its benefits in future studies to promote change in this population.

In summary, we believe that the results of this review provide a better understanding of the components of web-based interventions that can lead to behavior change among older adults. In the studies identified, we found an overrepresentation of interventions focused on the PA behavior of older adults and conclude that other studies should be conducted to assess the effects on other lifestyle habits. The results of this review lead us to believe that authors should provide a more in-depth description of their interventions’ components, including the use parameters, BCTs, delivery modes, and theories used, to understand what is favorable to the adoption of a healthy lifestyle among other adults, how this is achieved, and how it could have influenced participants in behavior change. In particular, further studies should be carried out to understand how BCTs are used in an intervention, the impact of each of these BCTs, and the influence of the diverse delivery modes used on behavior change among older adults. Future web-based interventions should be based on one or more theories, and authors should indicate how these theories are used in the intervention to induce change. The results of this review suggest that further studies of web-based interventions to promote a healthy lifestyle in older adults should include support from a coach to develop a relationship of trust, seek to increase motivation among older adults, be tailored to older adults’ conditions, help them reduce barriers to using technology, and modify their perceptions of effecting change at their age. We propose that future web-based interventions be coconstructed with older adults to better identify their needs and what they seek, particularly with regard to support from a professional.