This paper is in the following e-collection/theme issue:

Formative Evaluation of Digital Health Interventions (2220) mHealth for Wellness, Behavior Change and Prevention (2752) Wearable Devices and Sensors (822) Prevention and Health Promotion (899)

Published on in Vol 7 (2023)

Preprints (earlier versions) of this paper are available at https://preprints.jmir.org/preprint/42087, first published .
Predicting Habitual Use of Wearable Health Devices Among Middle-aged Individuals With Metabolic Syndrome Risk Factors in South Korea: Cross-sectional Study

Predicting Habitual Use of Wearable Health Devices Among Middle-aged Individuals With Metabolic Syndrome Risk Factors in South Korea: Cross-sectional Study

Predicting Habitual Use of Wearable Health Devices Among Middle-aged Individuals With Metabolic Syndrome Risk Factors in South Korea: Cross-sectional Study

Authors of this article:

Jaeyoung Ha1 Author Orcid Image ;   Jungmi Park1 Author Orcid Image ;   Sangyi Lee1 Author Orcid Image ;   Jeong Lee2 Author Orcid Image ;   Jin-Young Choi1 Author Orcid Image ;   Junhyoung Kim1 Author Orcid Image ;   Sung-il Cho1, 3 Author Orcid Image ;   Gyeong-Suk Jeon4 Author Orcid Image
Article Authors Cited by (4) Tweetations Metrics

    Original Paper

    1Department of Public Health Science, Graduate School of Public Health, Seoul National University, Seoul, Republic of Korea

    2Department of Nursing, Chodang University, Muan, Republic of Korea

    3Institute of Health and Environment, Seoul National University, Seoul, Republic of Korea

    4Department of Nursing, Mokpo National University, Muan, Republic of Korea

    Corresponding Author:

    Gyeong-Suk Jeon, PhD

    Department of Nursing

    Mokpo National University

    1666 Yeongsan-ro, Dorim-ri

    Cheonggye-myeon

    Muan, 58554

    Republic of Korea

    Phone: 82 61 450 2675

    Fax:82 61 450 2679

    Email: gsj64@naver.com


    Background: Prevention of the risk factors for metabolic syndrome (MetS) in middle-aged individuals is an important public health issue. Technology-mediated interventions, such as wearable health devices, can aid in lifestyle modification, but they require habitual use to sustain healthy behavior. However, the underlying mechanisms and predictors of habitual use of wearable health devices among middle-aged individuals remain unclear.

    Objective: We investigated the predictors of habitual use of wearable health devices among middle-aged individuals with risk factors for MetS.

    Methods: We proposed a combined theoretical model based on the health belief model, the Unified Technology of Acceptance and Use of Technology 2, and perceived risk. We conducted a web-based survey of 300 middle-aged individuals with MetS between September 3 and 7, 2021. We validated the model using structural equation modeling.

    Results: The model explained 86.6% of the variance in the habitual use of wearable health devices. The goodness-of-fit indices revealed that the proposed model has a desirable fit with the data. Performance expectancy was the core variable explaining the habitual use of wearable devices. The direct effect of the performance expectancy on habitual use of wearable devices was greater (β=.537, P<.001) than that of intention to continue use (β=.439, P<.001), and the total effect estimate of the performance expectancy was 0.909 (P<.001), including the indirect effect (β=.372, P=.03) on habitual use of wearable devices via intention to continue use. Furthermore, performance expectancy was influenced by health motivation (β=.497, P<.001), effort expectancy (β=.558, P<.001), and risk perception (β=.137, P=.02). Perceived vulnerability (β=.562, P<.001) and perceived severity (β=.243, P=.008) contributed to health motivation.

    Conclusions: The results suggest the importance of the users’ performance expectations for wearable health devices for the intention of continued use for self-health management and habituation. Based on our results, developers and health care practitioners should find better ways to meet the performance expectations of middle-aged individuals with MetS risk factors. They also should generate device use easier and find a way to encourage users’ health motivation, thereby reducing users’ effort expectancy and resulting in a reasonable performance expectancy of the wearable health device, to induce users’ habitual use behaviors.

    JMIR Form Res 2023;7:e42087

    doi:10.2196/42087

    Keywords

    wearable health devices; HBM; UTAUT2; habitual use; metabolic syndrome; wearables; digital intervention; health promotion; predictors of use; acceptability 


    Metabolic syndrome (MetS) is characterized by the presence of 3 of the following 5 conditions: abdominal obesity, hypertension, hypertriglyceridemia, hyperglycemia, and high serum low-density lipoprotein [1]. The prevalence of MetS is increasing worldwide, leading to a high burden of several diseases, including cardiovascular disease (CVD) [2,3]. In particular, lifestyle changes associated with urbanization and industrialization in the Asia-Pacific region, including South Korea, have led to a remarkable increase in the prevalence of MetS [4-6]. MetS most commonly affects middle-aged individuals. According to the Korean Statistical Information Service, in 2019, overall 57% of individuals with MetS risk factors were aged 40-65 years [7]. Effective management and prevention of MetS in the middle-aged population are essential to reducing the increasing burden of noncommunicable diseases with aging. MetS may lead to adverse health outcomes, such as the development of CVD and even death [8]. However, it is easily modified through behavioral changes in health [9].

    Several studies have evaluated the effectiveness of technology-mediated interventions (eg, wearable health devices, smartphone apps, mobile health, and eHealth) to modify the lifestyles of individuals with MetS [10]. Some studies specifically found the positive effects of wearable activity trackers (WATs) on behavioral change compared with conventional intervention through a randomized controlled trial [11] and with other types of technology (eg, mobile apps) [12]. WATs are particularly effective in improving physical activity by monitoring and recording physical activity [13]. A meta-analysis revealed that WAT-based interventions significantly increase physical activity and improve health indicators (waist circumference, systolic blood pressure, and low-density lipoprotein cholesterol level) in individuals with chronic diseases [13]. WATs offer various personalized features that are convenient for users, such as setting individual goals, tracking progress, monitoring health, and providing feedback, which is regarded as essential to motivating lifestyle change [14]. Although specific technical features, such as usability, aesthetics, reliability and accuracy of measured data, and price affect the preference of wearable types, studies suggest that the overall characteristics of wearable devices that offer personalized use have a greater impact on lifestyle changes than specific features [15,16].

    For effective self-health management using wearable health devices to manage risk factors for MetS prevention, a more in-depth understanding of the psychosocial process of users from adoption to use of the devices continuously and habitually is required. Studies on the acceptance of the use of wearable devices for health promotion have found that acceptance and abandonment of wearable devices are mediated by social, cognitive, and psychological factors [15]. For example, the technology acceptance model (TAM) suggests that technology acceptance and use are mainly based on perceived usefulness (ie, perceived effectiveness and efficiency) and perceived ease of use (ie, the perceived cost of device use) [17]. Venkatesh et al [18] developed the Unified Theory of Acceptance and Use of Technology (UTAUT) based on 8 models, including TAM and the theory of planned behavior, to identify the drivers of acceptance of technology. The extended version of UTAUT (UTAUT2) includes 3 additional constructs (hedonic motivation, price value, and habit) that make it more user-centered [19].

    However, compared to the many theories and models that have developed and advanced on the initial acceptance of wearable health care devices, there is no unified theory or model for the postadoption stage [20]. Some have combined UTAUT2 with other health behavior models to explain the continued use of technology after initial adoption, as it cannot be explained by a single model [21-23]. Wei et al [23] combined the UTAUT2 and the health belief model (HBM) to identify the predictors of adoption and continued use of a diet and fitness mobile app. Many studies of the continued use of mobile platforms have focused on satisfaction, perceived benefit, and trust [20]. Recent studies have evaluated emerging important constructs, such as perceived privacy risk (ie, risk perception) [23], confirmation, and satisfaction [24]. With advancements in the internet and social media, the perceived risk among consumers is an important consideration for fitness mobile apps [25,26]. Some studies have considered differences in individual characteristics, such as self-efficacy [27] or the creative application of the devices to foster environments to make them routine [28]. Furthermore, according to the qualitative studies of habit formation in wearable device usage among older adults [28,29], strong motivation was an important factor for the long-term usage of wearable health care devices for self-health management.

    The middle-aged group with risk factors for MetS is likely to have unique factors in adopting and continuing the use of wearable health care devices that differ from those of the general population. This is due to the conflicting predictors of continued device use, such as the finding that long-term use is less likely for individuals with chronic conditions [29] but is more likely for those with strong motivation and a meaningful initiation [28,29]. Therefore, it is necessary to develop a valid theoretical model suitable for devices for individuals aged >40 years with risk factors for MetS. We set up a hypothetical model incorporating strong health motivation, perceived privacy risk, which has risen as an important factor, and habitual use as the behavioral outcome and tested it with empirical data.


    Model Construction

    We combined UTAUT2 and HBM with the perceived risk regarding personal privacy and habitual use (Figure 1). We adopted certain constructs of HBM to predict individuals’ motivation to engage in healthy behaviors. Perceived vulnerability was defined as the perceived potential for transitioning to CVD if the risk factors for MetS are not controlled. Perceived severity was defined as the socioeconomic burden expected in cases of CVD development. The perceived threat was defined as the sum of perceived vulnerability and perceived severity, and indicated the perceived need to perform healthy behaviors to prevent a severe consequence [30]. We defined health motivation (ie, the desire to follow treatment instructions and belief therein) as the degree of motivation for self-health management. We assumed that high perceived severity and perceived vulnerability would lead to strong health motivation.

    H1: Individuals with high perceived severity have a strong motivation for health management.
    H2: Individuals with high perceived vulnerability have a strong motivation for health management.

    We integrated health motivation into the performance expectancy domain of the UTAUT2 model, which indicates the acceptance of wearable devices for health management. According to HBM, individuals with health motivation perceive greater benefits (higher performance expectancy in UTAUT2) than barriers to health behaviors and self-manage their health, which in our case is indicated by the acceptance of wearable health devices [31]. We assumed that individuals with strong health motivation would have high-performance expectancy, that is, they would expect that using wearable devices to manage health is highly productive.

    H3: Health motivation positively influences performance expectancy.

    Technological advances have led to increased awareness of privacy risks [32] and other risks involved in the adoption and acceptance of wearable devices. Therefore, we added the perception of privacy-related risk to the research model [33]. Risk perception was first proposed by Bauer et al [33] as “a combination of uncertainty and seriousness of outcome involved” [34]. Its effects on performance expectancy are controversial. Some studies have reported that risk perception negatively affects performance expectancy, whereas high personal security during device use is associated with higher performance expectancy [35]. Risk perception may negatively influence performance expectancy, particularly for wearable devices used for health management. Although the perceived risk may be related to several factors associated with the use of information technology [36], we focused on privacy risk, that is, the potential loss of control over personal information such as when the personal information of individuals is used without their knowledge or permission [37] because health management using wearable devices requires the collection of personal information. We assumed that a high-risk perception would result in low-performance expectancy.

    H4: Risk perception negatively influences performance expectancy.

    The main components of UTAUT2 are performance expectancy (ie, an individual’s degree of expectation that they can achieve their goals effectively through the use of technology) and effort expectancy (ie, the degree of expectations regarding the cost or time required to effectively use the technology), which are similar concepts to perceived usefulness and perceived ease of use, respectively, in TAM [17]. However, TAM and UTAUT2 involve conflicting explanations regarding whether effort expectancy directly affects intention to use [38]. According to TAM, familiarity with device use leads to expectations about device use and promotes the continued use of technology (context-agnostic). According to UTAUT2, the intention of continued device use is immediately acquired through familiarity with its use (technology-agnostic) [38]. Based on the latter, we hypothesized that effect expectancy may affect performance expectancy and the intention of continued use. This key concept of the model suggests that the evaluation of the use of wearable health care devices (ie, when performance expectancy exceeds effort expectancy) would lead to a decision on device acceptance [17]. Similar to Son et al [39], we developed items to measure effort expectancy (ie, the degree of perceived ease of use). We hypothesized that effort expectancy would affect performance expectancy and the intention of continued use.

    H5: Perceived effort expectancy positively influences performance expectancy
    H6: Perceived effort expectancy positively influences the intention of continued use

    Previous studies of UTAUT2 have reported high intentions for continued use with high-performance expectancies [23,40,41]. Few individuals reported that wearable health devices did not meet their expectations [24], and 1 study reported that expectations about device performance had a positive relationship with its habitual use [42]. Therefore, we hypothesized that high-performance expectancy may have a positive effect on the intention of continued and habitual use.

    H7: Performance expectancy positively influences intention of continued use.
    H8: Performance expectancy positively influences habitual use.

    In accordance with the evaluation-intention-usage relationship, the intention of continued use positively influences habitual use [43]. We hypothesized that the intention of continued use may lead to habitual use of wearable health devices.

    H9: Intention of continued use positively influences habitual use.
    ‎
    Figure 1. Proposed research model. EE: effort expectancy; HB: habitual use; HM: health motivation; IC: intention of continued use; PE: performance expectancy; PS: perceived severity; PV: perceived vulnerability; RP: risk perception.

    Measurements

    The survey tool consisted of 3 parts. The first part evaluated demographic characteristics, including age, sex, occupation, and risk factors for MetS (ie, hypertension, hyperglycemia, hyperlipidemia, and abdominal obesity). The second part contained questions about 3 characteristics of wearable health devices, including the period of use, device type, and main features of the devices. In the final part, items related to each construct were measured.

    The measurement items were adopted from previous studies that evaluated health information technology using HBM and UTAUT2: 24 items were derived from HBM and UTAUT2 [44], and 2 items related to risk perception were adopted from Moore and Benbasat [45]. To measure habitual use as a successful behavioral change outcome, we adopted 3 items from UTAUT2 [44]. Because after reviewing 2 qualitative studies evaluating habit formation outcomes [46,47], we found they asked the same questions, such as whether participants felt that their use of devices had become a habit and if the behavioral changes were programmed into their routines. To ensure the validity and reliability of measurement items, we adopted translated items from peer-reviewed Korean studies that used identical models and similar subjects. We translated certain measurement items that were not previously translated. The items were validated by 3 researchers who are fluent in Korean and English. The questionnaire was evaluated in a pilot web-based survey of 17 volunteer graduate students in public health programs, and any ambiguous or confusing questions were modified. Items were scored on a 5-point Likert scale (1=strongly disagree, 2=disagree, 3=neutral, 4=agree, 5=strongly agree).

    Data Collection

    A web-based survey used questionnaires consisting of 24 items to evaluate 300 individuals between September 3 and September 7, 2021. We investigated the health motivation and use of wearable devices for health management in middle-aged individuals with risk factors for MetS. For this, we included participants aged ≥40 years with ≥1 risk factor for MetS (hypertension, hyperglycemia, hyperlipidemia, and abdominal obesity) and experience with using wearable devices for health management. The web-based survey was conducted by a web-based panel and field research company. In total, 300 participants (150 females and 150 males) with ≥1 risk factor for MetS were recruited. Table 1 lists the general characteristics of the participants.

    Table 1. Constructs and questions related to the measurement items.
    ConstructQuestionnaire
    Perceived severity [48]

    		PS1aIf I develop a cardio-cerebrovascular disease, there will be a social impact.

    		PS2If I develop a cardio-cerebrovascular disease, there will be an economic impact.

    		PS3If I develop a cardio-cerebrovascular disease, there will be an impact on my daily life.
    Perceived vulnerability [49]

    		PV1bI may develop a cardio-cerebrovascular disease if my health risk factors (hypertension, hyperglycemia, hyperlipidemia, and abdominal obesity) are not managed.

    		PV2My risk of disease may increase if my neighbors have a cardio-cerebrovascular disease.
    Health motivation [50]

    		HM1cI try to manage my health risk factors (hypertension, hyperglycemia, hyperlipidemia, and abdominal obesity) in my daily life.

    		HM2I try to maintain a healthy lifestyle.

    		HM3The health risk factors (hypertension, hyperglycemia, hyperlipidemia, and abdominal obesity) need to be managed to prevent cardio-cerebrovascular diseases (hypertension, diabetes, and hyperlipidemia).

    		HM4I have an interest in maintaining a healthy lifestyle.
    Effort expectancy [39]

    		EE1dIt is or will be easy to use a wearable device (eg, smartwatch and smart glasses)

    		EE2It is not or will not be very difficult to use a wearable device (eg, smartwatch and smart glasses) proficiently.

    		EE3It is or will be easy to understand how wearable devices (eg, smartwatch and smart glasses) work.

    		EE4It is or will be easy to learn the use of wearable devices (eg, smartwatch and smart glasses).
    Performance expectancy [23]

    		PE1eThe use of wearable devices can improve health status monitoring and management efficiency.

    		PE2The use of wearable devices improves my health.

    		PE3The use of wearable devices helps me achieve my health goals earlier.
    Risk perception (Privacy) [51]

    		RP1fIt would be risky to allow wearable devices to record personal information.

    		RP2Recording personal information on the wearable devices would lead to unexpected problems.
    Intention of continued use [23]

    		IC1gI plan to continue using wearable devices to record my health status.

    		IC2I will continue to use wearable devices to record my health status.

    		IC3I intend to continue using wearable devices to monitor and manage my health status.
    Habitual use [44]

    		HB1hIt feels natural for me to use wearable devices.

    		HB2I must regularly use wearable devices to monitor my health status.

    		HB3I have developed a habit of using wearable devices to monitor my health status.

    aPS: perceived severity.

    bPV: perceived vulnerability.

    cHM: health motivation.

    dEE: effort expectancy.

    ePE: performance expectancy.

    fRP: risk perception.

    gIC: intention of continued use.

    hHB: habitual use.

    Ethics Approval

    The study protocol was approved by the Institutional Review Board of Mokpo National University (approval no: MNUIRB-210625-SB-009-01). All participants provided verbal consent for the web-based survey.

    Statistical Analysis

    We used structural equation modeling to identify the predictors of habitual use of a wearable health device using the hypothesis model. We evaluated the measurement model and then tested each path of the research model. The reliability and validity of the measurement model were also determined. Descriptive statistics were analyzed using RStudio (version 4.1.2; RStudio Team). Subsequent analyses were performed using SPSS software (version 26; IBM Corp). The reliability of constructs was evaluated using Cronbach α and composite reliability. Cronbach α coefficients of .7-.9 and .5-.7 indicated high and moderate reliability, respectively [52]. A composite reliability of >0.7 indicated good reliability of the construct variables [53]. We examined the validity of construct and latent variables using convergent and discriminant validity. Convergent validity was confirmed if the average variance extracted (AVE) value was ≥0.5 [54]. Discriminant validity was confirmed if the correlations between the 2 constructs were less than the square root of the AVE value [55,56]. The strength of the correlation between each pair of variables was measured using the Pearson correlation test. The structural model fit was used to test the reliability and validity.

    Mediation analysis was performed to investigate the direct and indirect effects of performance expectancy on habitual use; the indirect effect was mediated through the intention of continued use. The bootstrapping method was used to determine the significance of the mediating effect, with a 2000-time bootstrapping sample and 95% CI.


    Demographics and Characteristics Related to Wearable Health Device Use

    Table 2 lists the demographics and characteristics related to wearable health device use. Half of the respondents were males, and the other half were females. All were aged 40-69 years, and the proportions of individuals aged 40-49, 50-59, and 60-69 years were similar (34.0%, 35.3%, and 30.7%, respectively). Most respondents worked as machine operators and assemblers (45.7%), followed by managers and professionals (16.7%), and others (16.0%). Hypertension, hyperglycemia, hyperlipidemia, and abdominal obesity were present in 28.1%, 27.9%, 23.5%, and 20.6% of respondents, respectively. In total, 36.7% of respondents had used wearable devices for longer than a year. The most frequently used device was the “Galaxy Watch” (46.9%), followed by the “Mi Band” (24.7%). The most frequently used functions of the devices were monitoring heart rate and blood pressure (34.4%), followed by physical activity (32.3%).

    Table 2. Demographics and characteristics related to wearable health device use (N=300).
    CharacteristicsParticipants, n (%)
    Demographic characteristics

    		Gender


    		Males150 (50.0)


    		Females150 (50.0)

    		Age group (years)


    		40-49102 (34.0)


    		50-59106 (35.3)


    		60-6992 (30.7)

    		Occupation


    		Managers and professionals50 (16.7)


    		Machine operators and assemblers137 (45.7)


    		Service workers26 (8.7)


    		Craft and related trades workers8 (2.7)


    		Agriculture, forestry, and fishery workers3 (1.0)


    		Sales workers28 (9.3)


    		Others48 (16.0)

    		Health status(multiple choices may be applicable)


    		Hypertension172 (28.1)


    		Hyperglycemia171 (27.9)


    		Hyperlipidemia144 (23.5)


    		Abdominal obesity126 (20.6)
    Characteristics related to wearable health device use

    		Period of use


    		<1 year190 (63.3)


    		≥1 year110 (36.7)

    		Device type(multiple choices may be applicable)


    		Apple Watch (Apple)42 (10.9)


    		Galaxy Watch (Samsung)180 (46.9)


    		Galaxy Fit (Samsung)41 (10.7)


    		Mi Band (Xiaomi)95 (24.7)


    		Fitbit watch or tracker (Fitbit)20 (5.2)


    		Others6 (1.6)

    		Main functions(multiple choices may be applicable)


    		Physical activity226 (32.3)


    		Sleep pattern131 (18.7)


    		Heart rate and blood pressure241 (34.4)


    		Stress level102 (14.6)

    Measurement Model

    Table 3 presents the reliability of the measurement model. The item loadings were >0.40 (0.56-0.89) for all items, with a composite reliability of >0.70 (0.85-0.96). The Cronbach α of each factor was >.6 (.61-.84), and the overall Cronbach α was .89. Convergent validity was verified based on the AVE value of >0.5 (0.73-0.89). Table 4 lists the discriminant and construct validity of the measurement model. The discriminant validity was tested using the Fornell-Larcker criterion. The square root of the AVE value (shown in italics) was higher than its highest correlation with other constructs. Pearson correlations with latent variables were measured to evaluate the construct validity. The correlation values were lower than the root of AVE written diagonally on the matrix.

    Table 3. Construct and composite reliability.
    Construct and measured variablesEstimateaSEConstruct reliabilityAVEbCRcCronbach α

    		ßdβe




    Perceived severity



    		0.8230.933.803

    		PS11.841.000Reference



    		PS2.846.699.07611.134



    		PS3.842.753.06612.785


    Perceived vulnerability



    		0.7480.853.609

    		PV11.782.000Reference



    		PV2.779.562.1146.824


    Health motivation



    		0.7710.931.819

    		HM11.711.000Reference



    		HM2.912.670.06813.343



    		HM3.951.724.1009.509



    		HM4.894.679.08810.137


    Effort expectancy



    		0.8240.949.844

    		EE11.819.000Reference



    		EE2.917.701.07612.138



    		EE3.914.773.06813.369



    		EE4.961.746.07313.201


    Performance expectancy



    		0.7260.887.756

    		PE11.678.000Reference



    		PE2.990.678.09210.748



    		PE3.789.559.07410.644


    Risk perception



    		0.7810.876.787

    		RP11.894.000Reference



    		RP2.870.726.06114.157


    Intention of continued use



    		0.8880.960.886

    		IC11.865.000Reference



    		IC2.998.829.05518.222



    		IC3.955.821.05218.222


    Habitual use



    		0.7720.910.812

    		HB11.805.000Reference



    		HB2.854.719.06313.510



    		HB3.915.728.06514.054


    aP value was considered significant at <.001.

    bAVE: average variance extracted.

    cCR: composite reliability.

    dß : crude path coefficient β

    eβ : standardized path coefficient β

    Table 4. Discriminant validitya,b.
    VariablePScPVdHMeEEfPEgRPhICiHBj
    PS0.907N/AkN/AN/AN/AN/AN/AN/A
    PV0.5710.865N/AN/AN/AN/AN/AN/A
    HM0.4730.5880.878N/AN/AN/AN/AN/A
    EE0.2650.3260.4230.908N/AN/AN/AN/A
    PE0.3100.5200.5390.6030.852N/AN/AN/A
    RP–0.277–0.125–0.129–0.0530.0100.884N/AN/A
    IC0.3170.4230.5530.4850.755–0.0500.942N/A
    HB0.1950.4510.5820.6070.7980.0680.8770.879

    aSquare root of average variance extracted value of each variable and correlation coefficient matrix.

    bItalicized values represent square root of the average variance extracted value; values below them indicate the correlation coefficients.

    cPS: perceived severity.

    dPV: perceived vulnerability.

    eHM: health motivation.

    fEE: effort expectancy.

    gPE: performance expectancy.

    hRP: risk perception.

    iIC: intention of continued use.

    jHB: habitual use.

    kN/A: not applicable.

    Structural Model

    After examining the reliability and validity of the measurement model, the relationships between each construct were tested. The hypotheses were tested by measuring the coefficients of the paths. Table 5 lists the unadjusted and adjusted coefficients and the significance of the results along with the SE. For H1 and H2, the standardized path coefficients of health motivation with perceived severity (H1: β=.243, P=.008) and perceived vulnerability (H2: β=.562, P<.001) were positive. Health motivation (H3: β=.497, P<.001), risk perception (H4: β=.137, P=.02), and effort expectancy (H5: β=.558, P<.001) had significant positive effects on performance expectancy, while effort expectancy did not have any significant effects (H6: β=−.131, P=.16). Intention of continued use was positively related to performance expectancy (H7: β=.848, P<.001). Finally, the intention of continued use had a positive effect on habitual use (H9: β=.439, P<.001), thereby supporting H9. A mediation analysis was conducted to explore the direct and indirect effects of performance expectancy on habitual use (Table 6). Partial mediation was observed in the path between performance expectancy and habitual use, which was mediated by the intention of continued use. The indirect effect of performance expectancy on habit was significantly mediated by the intention of continued use (=.372, P=.03), while performance expectancy had significant direct (=.537, P<.001) and total (=.909, P<.001) effects on habitual use.

    We present Figure 2 for a more meaningful interpretation of the summary of the research results that conform to the research model. The research results were summarized in Figure 2. Variables from the HBM are colored in red, variables from the UTAUT2 are colored in green, and the variable colored in yellow demonstrates that it is an exogenous variable. The black lines represent the significance of the paths (P<.01), and the dotted line represents an insignificant (P>.01) path. The fitness indices of the structural model were ideal: chi-square or df (χ²/df)=2.096, goodness-of-fit index=0.871, adjusted goodness-of-fit index=0.841, comparative fit index=0.917, normed fit index=0.854, incremental fix index=0.918, Tucker Lewis index=0.906, and root-mean-square error of approximation=0.061. The model explained 86.6% of the variance in the habitual use of wearable health devices.

    Table 5. Regression coefficients of pathways in the model.
    Path (Hypothesis)EstimateSEResult
    βaP value
    H1: Perceived severity → Health motivation.243.008.076Supported
    H2: Perceived vulnerability → Health motivation.562<.001.121Supported
    H3: Health motivation → Performance expectancy.497<.001.070Supported
    H4: Risk perception → Performance expectancy.137.02.032Supported
    H5: Effort expectancy → Performance expectancy.558<.001.060Supported
    H6: Effort expectancy → Intention of continued use–.131.164.101Not supported
    H7: Performance expectancy → Intention of continued use.848<.001.144Supported
    H8: Performance expectancy → Habitual use.537<.001.157Supported
    H9: Intention of continued use → Habitual use.439<.001.112Supported

    aStandardized regression coefficient.

    Table 6. Direct and indirect effects of performance expectancy on habitual use.
    Classification ICa (PEb → HBc)Point estimate (β)Product of coefficientsBootstrapping (95% CI)
    Bias-correctedPercentile
    SEP valueLowerUpperLowerUpper
    Indirect effects.372.132.03.122.599.057.569
    Direct effects.537.153<.001.284.861.301.898
    Total effects.909.053<.001.787.998.791.999

    aIC: intention of continued use.

    bPE: performance expectancy.

    cHB: habitual use.

    ‎
    Figure 2. Empirical analysis results of the research model. EE: effort expectancy; HB: habitual use; HM: health motivation; IC: intention of continued use; PE: performance expectancy; PS: perceived severity; PV: perceived vulnerability; RP: risk perception. *P<.01, **P<.005, and ***P<.001.

    Principal Findings

    We evaluated the predictors and the structural relationships thereof related to the habitual use of wearable devices for self-health management in adults aged >40 years and having risk factors for MetS. Our results suggest that high health motivation and high expectations regarding the effectiveness and usefulness of wearable devices for health management (ie, performance expectancy) were associated with habitual use through the voluntary determination of continued use. Performance expectancy, a key component of this process, was influenced by the ease of use and risk perception related to wearable devices, while effort expectancy did not significantly influence the intention of continued use.

    Health motivation was influenced by perceived severity and vulnerability, that is, individuals with higher perceived vulnerability and severity had strong health motivation. In this study, users with higher vulnerability were more motivated to manage their health compared to those with greater severity of the disease. Kim et al [57] found that the awareness of the deterioration of their health condition motivated people to take action for self-health management. In a previous study of extended UTAUT2, perceived vulnerability and severity of disease were used as moderators [58]. The authors found that combined perceived usefulness and vulnerability had a greater effect on behavioral intention than combined perceived usefulness and severity of disease; these findings suggest that the recognition of nonspecific risk factors for CVD may more strongly affect the preventive health behaviors than the socioeconomic risks associated with the disease.

    Health motivation positively influences performance expectancy. This was the unique path of the research model that connected health motivation from HBM to performance expectancy from UTAUT2. This combined path explains the acceptance of wearable devices for self-health management to lower the risk of perceived vulnerability or disease severity, which is not adequately explained by technology innovation. A similar study from China [23] evaluated the factors that affect behavioral intention to use a smartphone fitness app. In that study, individuals with greater weight loss intentions were more likely to expect that the fitness app would help them lose weight; such individuals were also more likely than others to use the app. A study from South Korea [48] reported that health belief factors, including perceived vulnerability and severity, positively influenced the intention to accept mobile health care services, suggesting that users may seek appropriate tools for self-health management rather than using wearable devices for health care. Therefore, performance expectancy is a belief regarding the health benefits of using wearable devices for self-health management [48].

    Although health motivation is essential for building belief in performance expectancy, it is influenced by device-related factors of effort expectancy and risk perception. In this study, effort expectancy, as measured by ease of use, was positively influenced by performance expectancy. Several previous studies have reported that ease of use is closely related to performance expectancy and plays a critical role in adopting new technology [35,38]. However, many recent studies have found that older users are willing to learn when adopting new technologies, such as wearable health devices, and that they are more concerned about the usefulness of the device than its ease of use [28,59,60]. Our results revealed that effort expectancy was not related to the development of the intention to continue using wearable health devices. A previous study reported that older age was associated with more sustained use of technological devices [61]. A qualitative study reported that certain features of wearable health devices, such as alarms and monitoring of daily activities, were effective for developing habitual use of a tracker in older, long-term users [62]. Therefore, device development should focus on its performance rather than its ease of use.

    Risk perception (ie, perception of privacy insecurity) was positively associated with performance expectancy, which is inconsistent with some previous results [23,34]. Several studies have explored the paradoxical relationship between the demand for personalized health care through data sharing and the refusal to adopt the technology due to a perceived risk to privacy [63]. The paradox occurs when individuals willingly use health technology despite high privacy concerns [23,63]. For example, Wei et al [23] reported that risk perception was positively related to the behavioral intention to use a fitness mobile app and negatively related to performance expectancy. The authors suggested that some users might have overcome the perceived risks and adopted new technology [23]. Similarly, the privacy calculus theory posits that individuals compare the benefits and risks of behavior [63]. The findings of certain studies are aligned with those of our study as well as the privacy calculus theory. Certain study participants may have perceived the positive health effects of using wearable health devices with data sharing. Furthermore, our results suggest that disclosing information has a positive relationship with the cognitive evaluation of a device’s performance in the post-adoption phase. However, the perception of privacy insecurity may vary depending on the sociocultural context. Therefore, future studies should develop measurement items that are suitable for Korean culture.

    The perception of the usefulness and efficiency of health monitoring and management using wearable devices (ie, performance expectancy) had a significant positive effect on the intention of continued and habitual use. Additionally, the intention of continued use was significantly related to habitual use. Our results suggest that the positive recognition of self-health management using wearable devices developed through a conscious evaluation process and may lead to habitual self-health management. Lankton et al [64] explored information technology habits and determinants and found that habitual use predicts continued use of wearable devices. They concluded that the antecedents of habitual behaviors were users’ thoughts on personal relevance and importance of performing activities, satisfaction, prior use, and task complexity in the postadoption stage. Our results suggest that an evaluation of the importance and appropriateness of technology was needed before its automatic use. One study found that several functions of wearable health devices facilitate their habitual use, such as sports activity monitoring, heart rate feedback, step reminders, and activity statistics [65]. The most commonly used functions of wearable health devices by our study participants were monitoring heart rate, blood pressure, and physical activity. The development of devices should focus on advanced features that target the specific health needs of users to manage their risk factors for MetS. This recommendation is supported by our finding that performance expectancy is strongly related to health motivation, the intention of continued use, and habitual use.

    Practice Implications

    Our study results have several practical implications. First, perceived vulnerability was strongly related to health motivation and, subsequently, performance expectancy. Therefore, it is necessary to raise awareness about the advantages of the use of wearable devices that can continuously monitor health before CVD development. Second, individuals aged >40 years old with risk factors for MetS in South Korea reported positive perceptions about self-health management using wearable devices. Habitual health management through wearable health devices can be acquired when maintenance strategies are supported by identifying the specific health needs of users. Third, advanced features of wearable health devices should be developed to embed new healthy habits through habitual use of the devices, which can be integrated with self-monitoring. These combined functions may enhance health motivation and performance expectancy, enabling more personalized self-health management.

    Limitations

    Several study limitations should be considered. First, we enrolled only middle-aged Koreans with at least 1 risk factor for MetS. This was a conscious decision to reduce the possibility of confounders derived from demographic characteristics by ensuring that the study population was homogenous. Therefore, our results cannot be applied to other groups with different demographic characteristics. Second, more measurement items are needed for individuals with risk factors for MetS, even though they had high reliability and validity in this research. Finally, the proposed integrative model does not include the theoretical backgrounds and characteristics of the original models and instead interprets each construct from a new perspective that is appropriate for the study subject.

    Conclusions

    Although many studies have found the effects of wearable devices on health improvement, long-term adaptation and sustained use of wearable health care devices to maintain healthy behaviors remain the challenge for MetS prevention. We conducted an empirical analysis of the predictors of habitual use of wearable health devices by combining HBM, UTAUT2, and perceived risk. Our findings revealed that habitual use of wearable health devices was influenced by users’ expectations that the devices could help effectively manage MetS risk factors. This relationship was partially mediated by the willingness of users to continue the use of the device, and strong motivation was an important predictor of the expectancy of the device’s performance.

    The results suggest that developers and health care practitioners should find better ways to meet the performance expectations of the middle-aged population with MetS risk factors. This may involve identifying specific health indicators to be monitored, personal goal setting with detailed behavioral instructions, and reflecting on the context of existing daily routines to form new habits. This also requires implementing easier and more intuitive features that enhance the performance expectancy of wearable health devices, which ultimately influence the long-term and habitual use of the devices.

    Acknowledgments

    This research was supported by the MSIT (Ministry of Science and ICT), Korea, under the ITRC (Information Technology Research Center) support program (IITP-2020-0-01778) supervised by the IITP (Institute of Information and Communications Technology Planning and Evaluation).

    Data Availability

    The data sets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

    Conflicts of Interest

    None declared.

    1. Saklayen MG. The global epidemic of the metabolic syndrome. Curr Hypertens Rep 2018;20(2):12 [FREE Full text] [CrossRef] [Medline]
    2. Kim D, Yoon SJ, Gong YH, Kim YA, Seo HY, Yoon J, et al. The economic burden of cancers attributable to metabolic syndrome in Korea. J Prev Med Public Health 2015;48(4):180-187 [FREE Full text] [CrossRef] [Medline]
    3. Ranasinghe P, Mathangasinghe Y, Jayawardena R, Hills AP, Misra A. Prevalence and trends of metabolic syndrome among adults in the Asia-Pacific region: a systematic review. BMC Public Health 2017;17(1):101 [FREE Full text] [CrossRef] [Medline]
    4. Kim MH, Lee SH, Shin KS, Son DY, Kim SH, Joe H, et al. The change of metabolic syndrome prevalence and its risk factors in Korean adults for decade: Korea National Health and Nutrition Examination Survey for 2008–2017. Korean J Fam Pract 2020;10(1):44-52. [CrossRef]
    5. Lee HJ, Chang JS, Ahn JH, Kim MY, Park KS, Ahn YS, et al. Association between low muscle mass and non-alcoholic fatty liver disease diagnosed using ultrasonography, magnetic resonance imaging derived proton density fat fraction, and comprehensive NAFLD score in Korea. J Prev Med Public Health 2021;54(6):412-421 [FREE Full text] [CrossRef] [Medline]
    6. Kim HJ, Lee KH, Lee JH, Youk H, Lee HY. The effect of a mobile and wearable device intervention on increased physical activity to prevent metabolic syndrome: observational study. JMIR mHealth uHealth 2022;10(2):e34059 [FREE Full text] [CrossRef] [Medline]
    7. Status of metabolic syndrome risk factors by age and gender. Korean Statistical Information Service. 2022.   URL: https://kosis.kr/statHtml/statHtml.do?orgId=350&tblId=DT_35007_N139 [accessed 2022-04-22]
    8. Eckel RH, Grundy SM, Zimmet PZ. The metabolic syndrome. Lancet 2005;365(9468):1415-1428. [CrossRef] [Medline]
    9. Wannamethee SG, Shaper AG, Whincup PH. Modifiable lifestyle factors and the metabolic syndrome in older men: effects of lifestyle changes. J Am Geriatr Soc 2006;54(12):1909-1914. [CrossRef] [Medline]
    10. Kim G, Lee JS, Lee SK. A technology-mediated interventional approach to the prevention of metabolic syndrome: a systematic review and meta-analysis. Int J Environ Res Public Health 2021;18(2):512. [CrossRef] [Medline]
    11. Wong EML, Leung DYP, Tam HL, Wang Q, Yeung KW, Leung AYM. The effect of a lifestyle intervention program using a mobile application for adults with metabolic syndrome, versus the effect of a program using a booklet: a pilot randomized controlled trial. Clin Interv Aging 2021;16:633-644 [FREE Full text] [CrossRef] [Medline]
    12. Yen HY. Smart wearable devices as a psychological intervention for healthy lifestyle and quality of life: a randomized controlled trial. Qual Life Res 2021;30(3):791-802. [CrossRef] [Medline]
    13. Franssen WMA, Franssen GHLM, Spaas J, Solmi F, Eijnde BO. Can consumer wearable activity tracker-based interventions improve physical activity and cardiometabolic health in patients with chronic diseases? A systematic review and meta-analysis of randomised controlled trials. Int J Behav Nutr Phys Act 2020;17(1):57 [FREE Full text] [CrossRef] [Medline]
    14. Alturki MR. A systematic review on what features should be supported by fitness apps and wearables to help users overcome obesity. Int J Eng Res Technol 2016;5:197-206. [CrossRef]
    15. Shin G, Jarrahi MH, Fei Y, Karami A, Gafinowitz N, Byun A, et al. Wearable activity trackers, accuracy, adoption, acceptance and health impact: a systematic literature review. J Biomed Inform 2019;93:103153 [FREE Full text] [CrossRef] [Medline]
    16. Lin KY, Chien CF, Kerh R. UNISON framework of data-driven innovation for extracting user experience of product design of wearable devices. Comput Ind Eng 2016;99:487-502. [CrossRef]
    17. Davis FD. Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly 1989;13(3):319-340. [CrossRef]
    18. Venkatesh V, Morris MG, Davis GB, Davis FD. User acceptance of information technology: toward a unified view. MIS Quarterly 2003;27(3):425-478. [CrossRef]
    19. Salgado T, Tavares J, Oliveira T. Drivers of mobile health acceptance and use from the patient perspective: survey study and quantitative model development. JMIR mHealth uHealth 2020;8(7):e17588 [FREE Full text] [CrossRef] [Medline]
    20. Handarkho YD, Harjoseputro Y, Samodra JE, Irianto ABP. Understanding proximity mobile payment continuance usage in Indonesia from a habit perspective. J Asia Bus Stud 2021;15(3):420-440. [CrossRef]
    21. Tamilmani K, Rana NP, Dwivedi YK. Use of ‘habit’ is not a habit in understanding individual technology adoption: a review of UTAUT2 based empirical studies. : Springer; 2018 Presented at: IFIP WG 8.6 International Conference on Transfer and Diffusion of IT; June 25, 2018; Portsmouth, UK p. 277-294. [CrossRef]
    22. Ahadzadeh AS, Pahlevan Sharif S, Ong FS, Khong KW. Integrating health belief model and technology acceptance model: an investigation of health-related internet use. J Med Internet Res 2015;17(2):e45 [FREE Full text] [CrossRef] [Medline]
    23. Wei J, Vinnikova A, Lu L, Xu J. Understanding and predicting the adoption of fitness mobile apps: evidence from China. Health Commun 2021;36(8):950-961. [CrossRef] [Medline]
    24. Tam C, Santos D, Oliveira T. Exploring the influential factors of continuance intention to use mobile apps: extending the expectation confirmation model. Inf Syst Front 2020;22(1):243-257. [CrossRef]
    25. Wei J, Xu J, Zhao D. Public engagement with firms on social media in China. J Inf Sci 2015;41(5):624-639. [CrossRef]
    26. Sergueeva K, Shaw N, Lee SH. Understanding the barriers and factors associated with consumer adoption of wearable technology devices in managing personal health. Can J Adm Sci 2020;37(1):45-60 [FREE Full text] [CrossRef]
    27. Lee JS, Kang MA, Lee SK. Effects of the e-Motivate4Change program on metabolic syndrome in young adults using health apps and wearable devices: quasi-experimental study. J Med Internet Res 2020;22(7):e17031 [FREE Full text] [CrossRef] [Medline]
    28. Peng W, Li L, Kononova A, Cotten S, Kamp K, Bowen M. Habit formation in wearable activity tracker use among older adults: qualitative study. JMIR mHealth uHealth 2021;9(1):e22488 [FREE Full text] [CrossRef] [Medline]
    29. Li L, Peng W, Kononova A, Bowen M, Cotten SR. Factors associated with older adults' long-term use of wearable activity trackers. Telemed J E Health 2020;26(6):769-775. [CrossRef] [Medline]
    30. Taylor D, Bury M, Campling N, Carter S, Garfied S, Newbould J, et al. A review of the use of the Health Belief Model (HBM), the Theory of Reasoned Action (TRA), the Theory of Planned Behaviour (TPB) and the Trans-Theoretical Model (TTM) to study and predict health related behaviour change. National Institute for Health and Clinical Excellence. 2006.   URL: https://www.nice.org.uk/guidance/ph6/resources/behaviour-change-taylor-et-al-models-review2 [accessed 2023-03-13]
    31. Rosenstock IM. The health belief model and preventive health behavior. Health Educ Monogr 1974;2(4):354-386. [CrossRef]
    32. Xu W, Murphy F, Xu X, Xing W. Dynamic communication and perception of cyber risk: evidence from big data in media. Comput Hum Behav 2021;122:106851. [CrossRef]
    33. Bauer RA. Consumer behavior as risk taking. In: Cox DF, editor. Risk Taking and Information Handling in Consumer Behavior. Cambridge: Harvard University Press; 1960:389-398.
    34. Luo X, Li H, Zhang J, Shim JP. Examining multi-dimensional trust and multi-faceted risk in initial acceptance of emerging technologies: an empirical study of mobile banking services. Decis Support Syst 2010;49(2):222-234. [CrossRef]
    35. Mehta A, Morris NP, Swinnerton B, Homer M. The influence of values on E-learning adoption. Comput Educ 2019;141:103617. [CrossRef]
    36. Cocosila M, Archer N, Yuan Y. Early investigation of new information technology acceptance: a perceived risk - motivation model. Commun Assoc Inf Syst 2009;25(1):30. [CrossRef]
    37. Featherman MS, Pavlou PA. Predicting e-services adoption: a perceived risk facets perspective. Int J Hum Comput 2003;59(4):451-474. [CrossRef]
    38. Baptista G, Oliveira T. A weight and a meta-analysis on mobile banking acceptance research. Comput Hum Behav 2016;63:480-489. [CrossRef]
    39. Son HJ, Lee SW, Cho MH. Influential factors of college students' intention to use wearable device-an application of the UTAUT2 model. Korean journal of communication and information 2014;68:7-33.
    40. Indrawati, Putri DA. Analyzing factors influencing continuance intention of e-payment adoption using modified UTAUT 2 model. 2018 Presented at: 6th International Conference on Information and Communication Technology (ICoICT); May 3-5, 2018; Bandung, Indonesia p. 167-173. [CrossRef]
    41. Nikolopoulou K, Gialamas V, Lavidas K. Habit, hedonic motivation, performance expectancy and technological pedagogical knowledge affect teachers’ intention to use mobile internet. Comput Educ Open 2021;2:100041. [CrossRef]
    42. Park M, Lee G, Park JY. Changes in the convenience of providing inventory information affect the intention in omni-channel. Korea Assoc Bus Educ 2018;33(4):415-436. [CrossRef]
    43. Wang C, Harris J, Patterson P. The Roles of Habit, Self-Efficacy, and Satisfaction in Driving Continued Use of Self-Service Technologies. Journal of Service Research 2013 Jan 23;16(3):400-414. [CrossRef]
    44. Venkatesh V, Thong J, Xu X. Consumer acceptance and use of information technology: extending the Unified Theory of Acceptance and use of technology. MIS Quarterly 2012;36(1):157. [CrossRef]
    45. Moore G, Benbasat I. Development of an Instrument to Measure the Perceptions of Adopting an Information Technology Innovation. Information Systems Research 1991 Sep;2(3):192-222. [CrossRef]
    46. Karppinen P, Oinas-Kukkonen H, Alahäivälä T, Jokelainen T, Teeriniemi AM, Salonurmi T, et al. Opportunities and challenges of behavior change support systems for enhancing habit formation: A qualitative study. J Biomed Inform 2018 Aug;84:82-92 [FREE Full text] [CrossRef] [Medline]
    47. Lally P, Wardle J, Gardner B. Experiences of habit formation: a qualitative study. Psychol Health Med 2011 Aug;16(4):484-489. [CrossRef] [Medline]
    48. Cho SC, Han YJ. A study on the effect of health belief factors on the acceptance of mobile healthcare: focusing on mediating effectects of perceived usefulness. Reg Ind Res 2020;43:263-280. [CrossRef]
    49. An S, Kang H, Chung S. Older adults' adoption of health-realted mobile application: the role of empowerment. J Public Relat 2018;22(6):53-74. [CrossRef]
    50. Cha MH, Kim YK. Moderating effect of health motivation, health concern and food involvement on the relationship between consumption value and purchasing intentions of healthy functional food. J Korean Soc Food Sci Nutr 2008;37(11):1435-1442. [CrossRef]
    51. Malhotra N, Kim S, Agarwal J. Internet Users' Information Privacy Concerns (IUIPC): The Construct, the Scale, and a Causal Model. Information Systems Research 2004 Dec;15(4):336-355. [CrossRef]
    52. Sideridis G, Saddaawi A, Al-Harbi K. Internal consistency reliability in measurement: Aggregate and multilevel approaches. J. Mod. Appl. Stat. Methods 2018 Jun 26;17(1):15. [CrossRef]
    53. Lerdpornkulrat T, Poondej C, Koul R. Construct reliability and validity of the shortened version of the information-seeking behavior scale. Int J Inf Commun Technol Educ 2017;13(2):27-37. [CrossRef]
    54. Harun M, Ahmad S. Factor structure, reliability and validity of attitudes of biostatistics scale. BJMCS 2016 Jan 10;16(6):1-9. [CrossRef]
    55. Hamid MRA, Sami W, Sidek MHM. Discriminant validity assessment: use of Fornell and Larcker criterion versus HTMT criterion. J Phys Conf Ser 2017;890:012163. [CrossRef]
    56. Sujati H, Sajidan, Akhyar M, Gunarhadi. Testing the construct validity and reliability of curiosity scale using confirmatory factor analysis. J Educ Soc Res 2020;10(4):229. [CrossRef]
    57. Kim J, Park HA. Development of a health information technology acceptance model using consumers' health behavior intention. J Med Internet Res 2012 Oct 01;14(5):e133 [FREE Full text] [CrossRef] [Medline]
    58. Beh P, Ganesan Y, Iranmanesh M, Foroughi B. Using smartwatches for fitness and health monitoring: the UTAUT2 combined with threat appraisal as moderators. Behaviour & Information Technology 2019 Nov 19;40(3):282-299. [CrossRef]
    59. Kim T, Ho C. Validating the moderating role of age in multi-perspective acceptance model of wearable healthcare technology. Telematics and Informatics 2021 Aug;61:101603. [CrossRef]
    60. Yu-Huei C, Ja-Shen C, Ming-Chao W. Why Do Older Adults Use Wearable Devices: A Case Study Adopting the Senior Technology Acceptance Model (STAM). 2019 Presented at: Portland International Conference on Management of Engineering and Technology (PICMET); August 25-29, 2019; Portland, OR, USA p. 1-8. [CrossRef]
    61. Hermsen S, Moons J, Kerkhof P, Wiekens C, De Groot M. Determinants for Sustained Use of an Activity Tracker: Observational Study. JMIR Mhealth Uhealth 2017 Oct 30;5(10):e164 [FREE Full text] [CrossRef] [Medline]
    62. Kononova A, Li L, Kamp K, Bowen M, Rikard RV, Cotten S, et al. The Use of Wearable Activity Trackers Among Older Adults: Focus Group Study of Tracker Perceptions, Motivators, and Barriers in the Maintenance Stage of Behavior Change. JMIR Mhealth Uhealth 2019 Apr 05;7(4):e9832-e9299 [FREE Full text] [CrossRef] [Medline]
    63. Fox G. “To protect my health or to protect my health privacy?” A mixed‐methods investigation of the privacy paradox. Journal of the Association for Information Science and Technology 2020 May 20;71(9):1015-1029. [CrossRef]
    64. Lankton N, Wilson E, Mao E. Antecedents and determinants of information technology habit. Information & Management 2010 Aug;47(5-6):300-307. [CrossRef]
    65. Rieder A, Lehrer C, Jung R. Understanding the habitual use of wearable activity trackers. 2019 Presented at: 14th International Conference on Wirtschaftsinformatik; February 24-27, 2019; Siegen, Germany p. 24-27.

    ‎
    AVE: average variance extracted
    CVD: cardiovascular disease
    HBM: health belief model
    MetS: metabolic syndrome
    TAM: technology acceptance model
    UTAUT: Unified Theory of Acceptance and Use of Technology
    WAT: wearable activity tracker

    Edited by A Mavragani; submitted 22.08.22; peer-reviewed by D Tom-Aba, B Oh; comments to author 15.02.23; revised version received 27.02.23; accepted 28.02.23; published 06.04.23

    Copyright

    ©Jaeyoung Ha, Jungmi Park, Sangyi Lee, Jeong Lee, Jin-Young Choi, Junhyoung Kim, Sung-il Cho, Gyeong-Suk Jeon. Originally published in JMIR Formative Research (https://formative.jmir.org), 06.04.2023.

    This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on https://formative.jmir.org, as well as this copyright and license information must be included.