In 2019, nearly one-quarter of US adults reported using wearable health and activity trackers (wearables) [1-3], which is approximately twice the reported use in 2015 [4]. Wearables are mobile health (mHealth) technologies worn on the body that can detect, record, and report information about behaviors (eg, step count and dietary intake) and health indicators (eg, heart rate and calories burned) [5]. Wearables may improve health and behavioral outcomes, such as physical activity participation [1,6] and weight status [7], by prompting users to set health goals, providing automated personalized feedback about health and activity data, motivating healthy habit formation, and encouraging social data sharing and competition [6-9].

Connected devices, such as mHealth apps on smartphones or computer tablets, enable users to share health information from wearables with others, such as health care providers and family or friends. Sharing wearable data may improve the health and behavioral outcomes of users by generating social support and improving their ability to manage their health [9-12], increasing patient–provider engagement, and facilitating individualized counseling and clinical decision making [12-15]. Studies also suggest that wearable data sharing among behavioral intervention participants may increase intervention effectiveness [16].

When used as a health communication tool, the potential of wearable technologies to improve health may not be fully realized without an understanding of the willingness to share wearable data, particularly with providers. Reasons for not sharing wearable data with providers include lack of awareness of the social sharing features of wearables, uncertainty about the relevance or usefulness of the data to providers, low expectation of supportive feedback, and concerns about privacy or control over data shared from the device [12-14]. Past studies of patient data sharing from various mHealth technologies also suggest that willingness to share wearable data may vary by individual characteristics, such as sociodemographics (eg, sex, age, race, or ethnicity), health-related factors (eg, weight status or having a chronic condition), and trust [17-19]. However, the correlates of willingness to share data from wearable health and activity trackers, specifically, need further examination, as the characteristics of mHealth users vary by type of device used for health and behavioral tracking [20].

Following evidence that the use of and engagement with digital health can influence health and communication behavior [21-23], digital health behaviors may be additional factors associated with the willingness to share wearable data. For example, individuals who use mHealth technologies such as smartphones or tablets to help communicate with providers may have greater technology self-efficacy [21], which may influence their willingness to share health information from wearable devices. Frequency of wearable use, one aspect of engagement with mHealth technologies for health and behavioral tracking [22], could also be a factor associated with willingness to share data from these devices. Moreover, other digital health behaviors, such as sharing health information on social networking sites (SNSs; eg, Facebook) or within web-based health communities (eg, online cancer support groups), may be associated with willingness to share wearable data if social sharing of health information across digital media is broadly perceived as useful or beneficial [9,11,23]. However, to the best of our knowledge, such relationships between digital health behaviors and willingness to share data from wearables have not been examined in a nationally representative sample of US adults.

A better understanding of the factors that influence willingness to share data from wearables could have implications for the use of patient-generated data in clinical practice [24], particularly given the growing number of interventions that use wearable devices to track health and activity [25] and recent calls for integration and use of these data in interventions [26,27]. Identifying the sociodemographic, health, and behavioral correlates of willingness to share data from wearables could provide insights on the population subgroups that are most or least likely to engage with, and benefit from, wearable interventions or multicomponent behavioral interventions that involve wearable use. In addition, exploring the correlates of willingness to share wearable data with providers, as well as with family or friends, may identify different drivers of willingness to share data with different types of recipients.

This study has 2 primary aims to address gaps in the literature. The first aim is to describe the sociodemographic and health-related correlates of the reported willingness of wearable users to share data with health care providers and with family or friends. The second aim is to investigate the relationship between different digital health behaviors (ie, use of smartphones or tablets to help communicate with providers, frequency of wearable use, sharing health information on SNSs, and participation in a web-based health community) and the willingness of users to share wearable data with health care providers and with family or friends.

Data from the 2019 Health Information National Trends Survey (HINTS) of the National Cancer Institute were analyzed. HINTS is a nationally representative, probability-based cross-sectional survey. Self-administered questionnaires were completed by adult, civilian, noninstitutionalized individuals (N=5438) between January and April 2019 (Multimedia Appendix 1). Respondents completed mailed paper questionnaires (paper-only group) or completed the questionnaire on the web as part of a push-to-web pilot study. Individuals participating in the web pilot were randomly assigned to a web-option group (choice of responding by paper or web) or a web-bonus group (choice of responding by paper or web, with a US $10 bonus incentive for responding via web). There were no significant differences in response rates for the paper-only group (30.2%), the web-option group (29.6%), and the web-bonus group (31.5%). Additional information about HINTS data, resources, and methodology has been described elsewhere [28], and information specific to HINTS 5 Cycle 3 (2019) can be found in publicly available methods reports [29].

To be included in the analytic sample, respondents had to report the use of a wearable device to track their health or activity. Thus, respondents were included if they selected “yes” (vs “no”) in response to the item “In the past 12 months, have you used an electronic wearable device to monitor or track your health or activity? For example, a Fitbit, Apple Watch, or Garmin Vivofit.” In addition, respondents must have reported data for the outcome variables of interest, as described in the following sections.

Frequencies, weighted percentages, and chi-square statistics were calculated to describe the distribution of US adults who reported using a wearable device to track health or activity. Binomial logistic regression analysis was conducted to examine correlations between individual characteristics (sociodemographic, health-related, and digital health behavior variables) and willingness to share wearable data with health care providers, and correlations between individual characteristics and willingness to share wearable data with family or friends. In total, 2 regression models were constructed.

For the model predicting willingness to share wearable data with providers, all sociodemographic and health-related variables (excluding trust in health information from family or friends) were entered first to address the first aim of this study. Because digital health behaviors were factors of particular interest, the second aim was addressed by adding digital health behavior variables stepwise in the following order: frequency of wearable use, use of mHealth technologies to help communicate with providers, use of SNSs to share health information, and participation in a web-based health community. The order in which variables were added to the model was based on the extent of supporting literature [21-23] that suggests a potential association between the respective digital health behavior and willingness to share wearable data, such that variables with a greater evidence base were added to the model first. Pseudo R² was examined after adding each digital health behavior variable to the model to determine how much variability could be explained by each of these key predictor variables. For the model predicting willingness to share wearable data with family or friends, covariates were entered using a similar stepwise approach; however, variables pertaining to interactions with providers were excluded (ie, use of mHealth technologies to help communicate with providers, having a regular health care provider, and trust in health information from a physician), whereas trust in health information from family or friends was added.

Statistical analysis was conducted using SAS (version 9.4; SAS Institute). Complete-case analysis with listwise deletion was used for the regression models. Group differences by survey modality (paper-only, web-option, and web-bonus) for the outcome variables of interest were assessed using the jackknife replication variance estimation method, applying a final sample weight and replicate weights created using the Rizzo method [33]. Because group differences were not significant in our analysis of outcomes by modality, the full-sample weight was applied to calculate population estimates for the combined sample without controlling for group differences by survey modality. Replicate weights were also used to compute SEs of estimates using the jackknife replication method for the combined sample without controlling for group differences by survey modality.

The analytic sample comprised 1300 wearable users. Analysis of weighted data showed that women (55.03%) and men (44.97%) each constituted approximately half of the sample. The majority were under 50 years of age (64.1%), urban residents (88.07%), non-Hispanic White (64.69%), and reported having an education beyond high school (85%). Individuals with annual household income under US $75,000 comprised 44.02% of the sample, with the remaining 55.98% having an annual household income of US $75,000 or more (Table 1). The characteristics of the HINTS analytic sample can be referenced alongside the characteristics of the analytic sample of wearable users in Multimedia Appendix 2. Similar to other health-based surveys, HINTS respondents tend to be female, older, non-Hispanic White, urban-dwelling, and more educated, and have a higher annual household income than the general population [28].

Most individuals reported having a good health status (89.65%), higher health self-efficacy (77.15%), and a regular health care provider (64.73%). More individuals reported higher (vs lower) trust in health information from a physician (75.55%) and lower (vs higher) trust in health information from family or friends (90.89%). A majority of individuals had a BMI≥24.9 (69.44%), and just over half reported having one or more chronic conditions (53.71%). Approximately half of the individuals reported a higher (48.54%) versus lower (51.46%) level of physical activity.

Most individuals included in the analytic sample reported using their wearables “every day” or “almost every day” (72%). They were relatively evenly divided on the use of other mHealth technologies (eg, smartphones and tablets) to help communicate with providers (47.93% “yes” vs 52.07% “no”). A minority of individuals reported sharing health information on SNSs (19.54%) or participating in a web-based health community (11.95%).

A small number of wearable users had missing data regarding willingness to share wearable data with providers (n=18) and willingness to share wearable data with family or friends (n=22); therefore, the analytic sample comprised 1282 respondents for willingness to share wearable data with providers and 1278 respondents for willingness to share wearable data with family or friends. A majority of individuals reported that they would be willing to share wearable data with health care providers (81.86%) and with family or friends (69.1%). In the bivariate analyses, willingness to share wearable data with providers was significantly associated with trust in health information from a physician and level of physical activity. Willingness to share wearable data with family or friends was significantly associated with perceived health status, health self-efficacy, and level of physical activity (Table 1).

Willingness to share wearable data with providers was also significantly associated with each of the 4 measured digital health behaviors: frequency of wearable use, use of mHealth technologies to help communicate with providers, use of SNSs to share health information, and participation in a web-based health community. Only the use of SNSs to share health information was significantly correlated with reported willingness to share wearable data with family or friends (Table 2).

Regression analysis showed that individuals who reported higher (vs lower) health self-efficacy (odds ratio [OR] 1.97, 95% CI 1.11-3.51), higher (vs lower) trust in health information from a physician (OR 1.98, 95% CI 1.12-3.49), and higher (vs lower) levels of physical activity (OR 2.00, 95% CI 1.21-3.31) had significantly greater odds of reported willingness to share wearable data with providers. Among the digital health behaviors, higher (vs lower) frequency of wearable use (OR 2.15, 95% CI 1.35-3.43) and use of mHealth technologies to help communicate with providers (OR 1.99, 95% CI 1.09-3.63) were significantly associated with willingness to share wearable data with providers (Table 3). On the basis of pseudo R² values, the model fit improved with the addition of each digital health behavior variable.

In the regression analysis, only individuals who reported higher (vs lower) levels of physical activity had higher odds of reported willingness to share wearable data with family or friends (OR 1.70, 95% CI 1.02-2.84; P=.04). Of the 3 digital health behaviors included in the model, none were significantly associated with willingness to share wearable data with family or friends (data not shown). As in the first model, the model fit improved with the addition of each digital health behavior variable.

The purpose of this study was to describe the willingness to share health information collected on wearable health and activity trackers with health care providers and family or friends in a nationally representative sample of US adult wearable users. The findings of this study suggest that most individuals who used wearables were willing to share data generated from these devices with providers (approximately 80%), as well as with family or friends (approximately 70%); however, willingness to share this information varied with behavior-related factors. Health self-efficacy, trust in providers as an information source, frequency of wearable use, use of other mHealth technologies to help communicate with providers, and being physically active appeared to be key factors that influenced willingness to share wearable data with providers. Being physically active also appeared to play an important role in willingness to share data from wearables with family or friends, whereas other factors such as sociodemographics, health-related characteristics, and digital health behaviors played a less prominent role.

These findings contribute to the literature by identifying individual characteristics associated with willingness to share data from wearable health and activity trackers in the adult population and distinguishing the correlates of willingness to share on the basis of the recipient of the data. Interestingly, our study revealed no differential willingness to share according to the sociodemographic characteristics of wearable users. Although the HINTS response rate was relatively low for all survey modalities (approximately 30%), differences by survey modality group (paper, web, and web-bonus) were not significant for the response rate and for the outcome variables of interest.

The findings of this study suggest that willingness to exchange health- and activity-related information with providers via mHealth technologies may be increasing. For example, in 2013, approximately 50% of US adults who used smartphones, tablets, or other mobile devices reported that they would be “somewhat” or “very” willing to use these technologies to exchange health information about lifestyle behaviors with a provider [18]. Similar to the results of this study, Hyde et al [2] found that approximately 76% of adults reported willingness to share data from wearable health and activity monitors or fitness trackers with providers.

In contrast to prior studies examining willingness or sharing of data from mHealth technologies, factors such as sex, age, weight status [17], race and ethnicity [17,34], income [18,19], and education [18] were not significantly associated with willingness to share wearable data with providers. In addition, having a regular health care provider and having a chronic condition, such as diabetes or hypertension, were not key predictors of willingness to share wearable data. Past studies of US adults have found these to be significant correlates of reported data sharing from electronic medical devices (eg, glucometers and blood pressure monitors) [19]. Therefore, along with prior research [20], the findings of this study demonstrate the importance of examining behavioral predictors and outcomes of patient-generated data sharing for different types of technology.

In this study, level of physical activity and health self-efficacy were significant health-related correlates of willingness to share wearable data with providers. Previous research has shown that wearable users tend to be more physically active than the general population [1]; however, the association between physical activity and willingness of wearable users to share data with providers has been unclear to date [2]. Therefore, our study results contribute to the literature that, among wearable users, those with higher versus lower levels of physical activity may be more willing to share their data with providers. Moreover, although research suggests that using a wearable device may increase health self-efficacy [21,35], our findings suggest that users with relatively high versus low health self-efficacy may be more willing to share their wearable data. Because wearable users with lower levels of physical activity or lower health self-efficacy may benefit the most from sharing wearable data with receptive providers (eg, individualized counseling), future mHealth intervention studies could include these factors in intervention design and explore how to overcome these potential barriers to data sharing. An increasingly acknowledged digital divide that has arisen from disparate health or behavioral outcomes among technology users [36] makes such research particularly important.

The results of this study also showed that trust in health information from providers is a strong predictor of willingness to share wearable data with them. Previous studies of US adults also found an association between trust in providers and willingness to exchange lifestyle behavior information via mHealth technologies, such as smartphones or tablets [18]. To increase willingness to share wearable data among those with lower levels of trust in providers as information sources, future mHealth intervention studies could explore ways to build trust in health information exchange within the patient–provider–technology relationship.

The second aim of our study was to investigate the relationship between digital health behaviors and willingness of users to share wearable data. The study findings showed that those who reported using their wearables every day or almost every day were more likely to report willingness to share data with providers than those who used them less often. Those who reported using (vs not using) smartphones or tablets to help communicate with providers were also more likely to report willingness to share. Consistent with prior research [21,22,25], these results suggest that greater use and technology self-efficacy, specifically in the context of health care and the patient–provider relationship, may increase the intention to share wearable health information with providers. These may be important targets for future intervention research focused on increasing health information exchange with providers via wearables.

By contrast, using SNSs to share health information and participating in a web-based health community were not significantly correlated with willingness to share health information from wearables with providers. These findings suggest that health information–sharing behaviors may vary based on the context (eg, health care setting or online support group), the audience or recipient (eg, health care providers or peers), and the technology through which the information is shared. Because SNSs and web-based health communities may be helpful to individuals through the visibility, availability, control, and reach they offer [37,38], future mHealth intervention studies that aim to improve wearable data sharing with providers could consider how to incorporate these factors into the intervention design.

One of the limitations of this study is the reliance on self-reported and cross-sectional data. In addition, our study was limited by the inability to distinguish between various types of wearable health and activity trackers, which can vary considerably in their functionality. Due to limitations of the data set, we also could not assess other factors that potentially affect willingness to share wearable data with others, such as technology self-efficacy or concerns about privacy or data security.

This study contributes to understanding the willingness of US adults to share data from wearable health and activity trackers with health care providers and family or friends. Several behavior-related factors were independently associated with willingness to share wearable data with providers, including level of physical activity, health self-efficacy, information-related trust in providers, frequency of wearable use, and use of mHealth technologies to help communicate with providers. Only level of physical activity was significantly associated with willingness to share wearable data with family or friends, controlling for other factors. Future behavioral surveillance research could assess attitudes associated with willingness to share wearable data, as well as factors that may influence these attitudes (eg, concerns about privacy), given the strong relationship between attitudes and behavioral intention [39]. In addition, given that attitudes about mHealth technologies and use of patient-generated data from wearables involve both patients and providers, researchers could use participatory action approaches that include these stakeholders in intervention design and implementation. When used as a communication tool, the potential of wearables to improve population health may not be fully realized without attention to these individual and relational factors.