Functional and Symbolic Aspects of App Use for Improving Physical Activity: A Six-month Prospective Analysis

doi:10.21203/rs.3.rs-4670553/v1

Download PDF

Research Article

Functional and Symbolic Aspects of App Use for Improving Physical Activity: A Six-month Prospective Analysis

https://doi.org/10.21203/rs.3.rs-4670553/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background:

Mobile health technology plays an important role in improving physical activity (PA). However, commercial healthcare applications for smartphones (apps) have poor retention, and understanding how people adopt and integrate app use in daily life is critical. We investigated the use patterns of PA apps and explored the use styles that are predictive of (dis)continuation of use and changes in PA levels over time.

Methods:

We analyzed two-wave longitudinal survey data concerning commercial PA-app use, which included 4465 respondents (mean age = 50.7; 1932 women) identified as PA-app users at baseline. The participants completed a questionnaire regarding how and for what purpose they used the apps. A six-month follow-up survey was administered that asked participants about their current app use and PA levels.

Results:

At baseline, 2737 were identified as long-term users of a PA app (i.e., use for more than six months). Long-term users reported appreciating the lifestyle management aspects (e.g., constant accessibility to health information, tracking, and monitoring PA), whereas short-term users indicated that they appreciated their app’s distraction and building relationships (e.g., finding like-minded peers) aspects. Prospective analyses demonstrated that lifestyle management was associated with continuing to use the app and increased PA levels at the 6-month follow-up, whereas distraction predicted discontinuing the use of the app and decreased PA levels.

Conclusions:

These findings suggest that on-purpose use (i.e., using a PA app to improve one’s PA) is the key feature of being in an appropriation state, whereas off-purpose use may hinder app use, leading to less active lifestyles. The implications of appropriation theory and practice are also discussed.

mobile health

appropriation

mobile phone

smart phone

longitudinal

Mobile health has received increasing attention from researchers, practitioners, and other stakeholders as the ubiquity of mHealth tools allows the delivery of the right intervention to the right person at the right moment (e.g., [1, 2]). The mHealth approach is expected to play a particularly important role in promoting physical activity (PA), which is effectively and efficiently supported by smartphone applications (apps) and wearable activity trackers that implement various behavior change techniques [3–5]. Several controlled and uncontrolled trials have assessed and established the efficacy of mHealth interventions for PA, and umbrella reviews have concluded that most mHealth interventions are effective albeit documenting high heterogeneity [6–8].

Regardless of their clinical and economic impact [9, 10], mHealth interventions face important issues in implementation and distribution, namely poor retention rates for commercial app users. These challenges are a general issue for healthcare apps that is not limited to PA apps, as research indicates that only 4% of users who install mental health apps continue using the apps daily [11], and the median monthly usage time of mHealth apps, including both mental and physical health, is no more than 5 minutes [12]. Active user engagement has been suggested to be associated with the quality of product design [12]; for example, reward and personalization functions may contribute to a good retention rate. One longitudinal study of commercial PA apps equipped with these functions suggested that 60% of users maintained active app use for at least six months [13]. Interestingly, the retention rate in randomized controlled trials of mHealth interventions is estimated to be 91%, which is much higher than that of commercial apps on the market. Thus, the discontinuation of app use is a unique phenomenon that can be observed in a daily, free-living context where no external regulation is expected by healthcare professionals or researchers. Believing that a good product spreads spontaneously by word of mouth would be somewhat naive, given that an explosive number of products appear and disappear on the market annually. Knowing what factors are predictive of the continuation of app use is particularly important for stakeholders, as it can help build an effective strategy to facilitate and maintain app use, resulting in healthy lifestyle changes. Therefore, we aimed to explore how people continue using apps to support PA and exercise to clarify the technology appropriation processes.

Appropriation perspectives

Appropriation is the way people adapt, adopt, and integrate new technology into their daily lives [14–16]. Carrol et al. [15] distinguished between technology-as-designed (i.e., the way of use that developers and designers intended) and technology-in-use (i.e., how technology is currently used). Through an appropriation process, users implicitly or explicitly transform technology-as-designed into technology-in-use—that is, they “trial and evaluate a new technology, select and adapt some of its attributes and so take possession of its capabilities in order to satisfy their needs” (p. 4). New technology must go beyond how it is designed by developers to become part of users’ daily routines, and this process often involves users’ active adoption and integration (e.g., reshaping and customizing a mobile device).

An increasing number of studies have investigated mHealth adoption, most having a theoretical basis in the technology acceptance model (TAM) and unified theory of acceptance and use of technology (UTAUT and UTAUT2) [17]. The TAM highlights perceived usefulness and ease of use as fundamental determinants of user acceptance of information technology [18, 19]. Conversely, UTAUT highlights four constructs (i.e., performance expectancy, effort expectancy, social influence, and facilitating conditions) and three extensions (i.e., hedonic motivation, price value, experience, and habit) influencing the intention to use new technology. Empirical studies have shown that perceived usefulness and ease of use are significantly associated with continuance intention to use health apps on smartphones [20], and these perceptions can be explained by external variables such as health consciousness, subjective norms, and Internet health information use efficacy [21]. Similarly, the UTAUT constructs (e.g., performance expectancy, effort expectancy, and social influence) have been shown to have a significant impact on users’ behavioral intentions to adopt mHealth services, although some inconsistencies have been documented [22, 23].

The Mobile Phone Appropriation Model

Technology adoption models, including the TAM and UTAUT, typically target binary use intention, namely either the adoption or rejection (and use or non-use thereof), as the dependent variable. However, Wirth et al. [16] argued that appropriation is a more complex concept that cannot necessarily be boiled down to the adoption-rejection dichotomy, which ends in various usage and meaning patterns at the individual and social levels. The Mobile Phone Appropriation (MPA) model explicitly theorizes the multifaceted patterns of everyday integration of mobile/smartphones and individual apps [17, 24, 25]. This model assumes two aspects of usage—symbolic and functional. The former represents the goal for which a mobile phone or app is used, also known as the gratification dimensions [16, 26]. Stehr et al. [24] applied the MPA model to their analyses of nutrition app use and divided the functional aspects into the following three subdimensions: distraction (i.e., using an app for pastime), lifestyle management (i.e., continuous monitoring and tracking of users’ own health states and behaviors to fulfill informational needs), and building relationships (i.e., exchanging with like-minded peer users, seeking and receiving support from peers, and competing with other users). Symbolic aspects are subdivided into psychological and social dimensions pertaining to behaviors important to the users themselves and in relation to their social surroundings [24]. These aspects cover preference and suitability (e.g., how the user likes the app and how the app fits the user), as well as prestige and identity, using the app as a way of expressing the user’s sense of self in public, such as in a fashion statement [17, 27]. The MPA model places the functional and symbolic aspects of use in a cycle of appropriation, in which metacommunication (i.e., communication on how individuals use an app) and evaluations (e.g., prospects about future app use for functional and symbolic aspects, beliefs about social norms, and barriers hindering app use) dynamically interact with and influence app use behavior.

Evidence gap

Although the MPA model is a comprehensive and sophisticated framework for analyzing different app use patterns, empirical evidence is still lacking on how predictive the model is for the actual continued use of a healthcare app. The model correctly points to the importance of understanding user behavior with multifaceted aspects rather than the adoption-rejection dichotomy [17, 24, 25]. However, it is an important question (especially for stakeholders) how likely users are to continue using an app with a particular task and purpose – for example, whether people using a PA app for a pastime would be more likely to continue using the app for lifestyle management. Most appropriation process studies rely on qualitative or cross-sectional analysis. Longitudinal evidence is required to establish the predictive value of the aspects of use listed in the MPA model.

Another notable gap in the research is that studies on adoption and appropriation focus almost exclusively on IT use (or use intention) of information technology as the dependent variable. Analyses of these proximate outcomes are meaningful for designing and updating service products. However, when it comes to a healthcare app, distal outcomes are equally important; that is, how the functional and symbolic aspects of app use are associated with actual health outcomes such as engagement in PA and exercise (in the case of apps supporting PA and exercise).

Objectives

Therefore, the current study investigated how the functional and symbolic aspects of PA-app use would predict the (a) (dis) continued use of apps and (b) changes in PA levels over time. Questionnaire data from a longitudinal survey were analyzed. Participants reported how they used a PA app (for the functional and symbolic aspects) at baseline, and at the six-month follow-up, they completed a questionnaire regarding the current use (vs. non-use) of the app as well as their levels of PA.

Our analyses were conducted in a somewhat explorative manner, as we did not have a clear a priori hypothesis regarding which aspects of app use would be predictive of its continued use and participants’ PA levels at follow-up. However, the stage model of Benamar et al. [28], conceptualizing appropriation as a dynamic process of four stages (i.e., symbolic appropriation, exploration, use construction, and stabilization), may hint at how different app use patterns predict continuation. In the symbolic stage, users encounter a new app (a smartwatch in the analyses by Benamar et al. [28]) and imagine what they will do after its acquisition. Users then experience the app in terms of its sensory aspects and potential (the exploration stage). Then, users’ interactions with the app become more regular and functionalist by learning the functionality of the app (the use construction stage), implying that users have increased awareness of what they need and what they can do with the app. Simultaneously, they sort which functions to use, and the functions that were not viewed as useful are used less frequently through this sorting process, as those functions are less relevant to achieving a specific objective, such as improving PA. The lifestyle management dimension of the functional aspects of the MPA model is conceptually relevant for the use construction stage, whereas distraction and building relationships are less on purpose and may be deemed less functionalist. The stabilization stage, corresponding to the appropriation state, is characterized by a good knowledge of the app, affective attachment, and identity pertaining to the symbolic aspects of use in the MPA model.

Taken together, the lifestyle management dimension would be predictive of continued app use at the 6-month follow-up, as it may indicate that users are close to (or have already reached) the appropriation state; however, distraction and building relationships may be less predictive, as the sorting process is yet to be triggered. Symbolic aspects, which are thought to be the key characteristics of the appropriation state or stabilization stage, would also be predictive of continued app use at the follow-up.

Data

We analyzed a dataset of how people use commercial apps to support PA and exercise, some of which have been published elsewhere [29, 30]. The data contained questionnaire responses from 20,573 Japanese-speaking adults who were online panels registered in a sample pool database (see Oba et al. [29] for more details on the sampling procedure). The inclusion criteria were being aged > 18, having a good command of Japanese, and having residency in Japan. In this dataset, 5030 (24.4%) participants reported that they had used a PA-supporting app and wearable activity tracker, of which 4465 completed a questionnaire regarding the functional and symbolic aspects of the app use. These participants were invited to complete a 6-month follow-up survey, in which 3825 completed questionnaires reported on current app use and PA levels. The overarching study was approved by the Ethics Committee of the National Institute of Advanced Industrial Science and Technology (approval ID: 2022 − 1279).

Measures

Use of Apps for Supporting Physical Activity

Participants provided a binary response regarding their use of apps to support PA or exercise. Those who responded affirmatively provided further information on how they used the app. The questions included (a) the names of the apps in use, (b) how long they had been using the app that was used most frequently (less than a week to more than a year), and (c) how frequently they were using the app (less than once per month to multiple times per day). The complete description of the questionnaire is available in Oba et al. [29]. At the 6-month follow-up, participants reported on their current app use (vs. non-use) again, selecting one from the three response options (using an app, used but not using it anymore, or have never used it). The latter two responses were regarded as app non-use (i.e., discontinuation), although a response of never used would be inconsistent with the participants’ baseline response.

Functional and Symbolic Aspects of App Use

We adapted the MPA scale for nutritional apps [24] to assess the functional and symbolic aspects of PA app use. Participants indicated their attitudes toward and impressions of the PA application. Nine items assessed functional aspects, including distractions (three items), lifestyle management (three items), and building relationships (three items). Seven items assessed symbolic aspects, including psychological (four items) and social (three items) dimensions. We did not aggregate these items into factor scores because each item conveys slightly different concepts. For example, each social-dimension item represents showing off, a supportive environment, or identification. Participants rated the extent to which each item was applicable to situations or reasons why they used the indicated PA app on a 5-point scale (1 = not at all to 5 = very much).

Physical Activity

The International Physical Activity Questionnaire-Short Form (IPAQ-SF) [31, 32] was used to assess average weekly PA levels. Participants indicated the number of days and duration (in minutes) spent on three PA domains—(a) walking, (b) moderate-intensity, and (c) vigorous-intensity. The reported number of days and minutes were aggregated to represent the total PA time (min/week). We also converted this into metabolic equivalents (METs-hour) and assessed whether each participant met the PA level recommended by the Ministry of Health, Labor, and Welfare in Japan [33] (i.e., 23 METs-hour/week for adults aged < 65 years; 10 METs-hour/week for older adults ≥ 65 years).

Statistical Analyses

Cross-sectional and prospective analyses were conducted. The cross-sectional analyses focused on the duration of app use reported at baseline, examining differences in demographics and app use between long-term users (those using an app for ≥ 6 months) and relatively new short-term users (those using an app < 6 months). Given the large sample size (and power) of the current dataset, we interpreted the standardized mean differences (Cohen’s d) instead of basing our inferences on statistical hypothesis testing (i.e., p-values) for demographic and descriptive analyses. The app-use duration cutoff of six months was arbitrarily selected because we assumed that six months was long enough to indicate the maintained use of the app according to the stage-of-change theory [34, 35], considering individuals in the maintenance stage have maintained the desired behavior for six months.

The prospective analyses highlighted how the functional and symbolic aspects of app use predicted app use continuation at the 6-month follow-up. A logistic regression was conducted, in which users who reported their continued versus discontinued uses at follow-up were predicted by each app-use aspect assessed at baseline. A similar logistic regression analysis was conducted with the follow-up levels of PA as the binary dependent variable (i.e., adherence to the national PA guidelines), in which each app-use aspect was included as a predictor while controlling for the baseline levels of PA. For the prospective analyses, data from those who completed both the baseline and follow-up were used (n = 3825). Prior to these analyses, characteristics of the participants lost to the follow-up were explored.

Demographic characteristics and descriptions are presented in Table 1. Long-term users were older, had higher education levels and income, and were more likely to be men and to adhere to the PA guidelines, compared with short-term users. Notably, long-term users interacted with the apps more frequently (n = 2174 [79.4%] used them more than once per day) and were more likely to continue using the apps at the 6-month follow-up (n = 1609, 58.8%).

Table 1

Demographics and Descriptives at the Baseline and Follow-up
Variable	Short-term user (n = 1088)	Long-term user (n = 2737)	Difference
Baseline
Age, years: M (SD)	48.5 (17.8)	53.6 (16.0)	t (3823) = 8.599, p < .001, d = 0.31
Gender, women: n (%)	500 (46.0%)	1130 (41.3%)	χ² (1) = 6.753, p = .001
BMI, kg/m²: M (SD)	22.1 (3.4)	22.5 (3.8)	t (3823) = 2.703, p = .007, d = 0.10
Education, university or above: n (%)	537 (49.4%)	1453 (53.1%)	χ² (1) = 4.193, p = .041
Income, > 3 million JPY: n (%)	591 (54.3%)	1655 (60.5%)	χ² (1) = 11.887, p < .001
IPAQ-SF, total PA time, min/week: M (SD)	601.4 (809.6)	670.3 (824.1)	t (3823) = 2.346, p = .019, d = 0.08
IPAQ-SF, total PA, METs-hour/week: M (SD)	46.3 (64.1)	50.3 (66.3)	t (3823) = 1.681, p = .093, d = 0.06
Adherence to the PA guideline: n (%)	592 (54.5%)	1747 (63.8%)	chi-square (1) = 28.668, p < .001
Use app at least once per day: n (%)	633 (58.2%)	2174 (79.4%)	χ² (1) = 178.91, p < .001
Six-month follow-up
BMI, kg/m²: M (SD)	22.2 (3.9)	22.4 (3.5)	t (3823) = 1.486, p = .137, d = 0.05
IPAQ-SF, total PA time, min/week: M (SD)	535.8 (723.7)	584.1 (721.7)	t (3823) = 1.866, p = .062, d = 0.07
IPAQ-SF, total PA, METs-hour/week: M (SD)	42.1 (59.3)	44.2 (58.2)	t (3823) = 0.977, p = .329, d = 0.04
Adherence to the PA guideline: n (%)	663 (60.9%)	1841 (67.3%)	χ² (1) = 13.50, p < .001
Current app use, affirmative: n (%)	455 (41.8%)	1609 (58.8%)	χ² (1) = 89.53, p < .001
Note. BMI, body mass index; IPAQ-SF, International Physical Activity Questionnaire–Short From; 1 USD = 140 JPY. Guideline-recommended levels of PA = 23 METs hours/week for adults aged < 65 years; 10 METs hours/week for adults aged ≥ 65 years). Short-term or long-term users = those using apps for < 6 or ≥ 6 months.

We observed dropout of 640 (14%) participants to the follow-up. Characteristics of the dropout (vs. completer) samples were explored. Specifically, a logistic regression was estimated, with dropout (vs. completer) being predicted by the following seven baseline variables: age, gender (0 = men and 1 = women), BMI, education, income, PA adherence, and frequency of app use (i.e., use app at least once per day). Older participants (OR = 0.97, p < .001, 95%CI [0.96, 0.97]), men (OR = 1.28, p = .010, 95%CI [1.06, 1.54]), and frequent app users (OR = 0.78, p = .009, 95%CI = [0.65, 0.94]) were less likely to dropout from the study. These variables were included as covariates in the following regression analyses.

Table 2 illustrates the differences in the functional and symbolic aspects of app use between short- and long-term users at baseline. Lifestyle management items (Functional Aspect (FA) Items 4–6) were rated higher by long-term users (|ds| > 0.23), whereas short-term users appreciated distraction (FA Item 3, I'm using the app when I'm bored) and building relationships (FA items 7–8; |ds| > 0.23). Regarding the symbolic aspects, long-term users rated the psychological dimensions higher (except for Symbolic Aspect (SA) Item 3, “I’m using a cutting-edge app”) than short-term users. No substantial differences were found in the social dimensions (|ds| < 0.20, SA Items 5–7).

Table 2

Means (SDs) of the Functional and Symbolic Aspects of Short vs. Long-term Users at Baseline
Item	Short-term user (n = 1088)	Long-term user (n = 2737)	Cohen's d
Functional aspects
FA1, distraction: I'm using the app for diversion	2.79 (1.10)	2.74 (1.15)	-0.045
FA2, distraction: I'm using the app when there's nothing else to do	2.63 (1.13)	2.44 (1.13)	-0.174
FA3, distraction: I'm using the app when I'm bored	2.68 (1.17)	2.39 (1.15)	-0.256
FA4, lifestyle management: I'm using the app to be able to access health and exercise information at any time	3.03 (1.12)	3.34 (1.11)	0.279
FA5, lifestyle management: I'm using the app to lead a healthy life	3.41 (1.13)	3.77 (1.00)	0.348
FA6, lifestyle management: I'm using the app to keep track of my daily body condition or the quality of daily exercise/training	3.13 (1.14)	3.40 (1.13)	0.234
FA7, building relationships: I'm using the app to get to know people who share the same interests	2.32 (1.19)	2.04 (1.16)	-0.237
FA8, building relationships: I’m using the app to get support from others pursuing the same goal as me	2.28 (1.19)	1.99 (1.12)	-0.252
FA9, building relationships: I'm using the app to compete with other users of the app	2.08 (1.14)	1.80 (1.07)	-0.252
Symbolic aspects
SA1, psychological dimension: The app is a good fit for me	3.19 (1.04)	3.66 (0.92)	0.490
SA2, psychological dimension: I like using the app	3.24 (1.03)	3.60 (0.92)	0.386
SA3, psychological dimension: I'm using a cutting-edge app	2.85 (1.07)	2.96 (1.08)	0.105
SA4, psychological dimension: I can access my apps at any time	3.29 (1.17)	3.84 (0.98)	0.530
SA5, social dimension (showing-off): Sometimes, I brag with my usage of the app	2.33 (1.18)	2.17 (1.14)	-0.143
SA6, social dimension (supportive environment): The people close to me support my usage of the app	2.58 (1.18)	2.60 (1.18)	0.016
SA7, social dimension (identity): Who I am is also reflected in the way I use the app	2.68 (1.13)	2.73 (1.16)	0.047
Note. FA = Functional Aspect; SA = Symbolic Aspect. Short-term or long-term users = those using apps for < 6 or ≥ 6 months.

Table 3 shows the results of the logistic regression predicting PA-app use (vs. non-use) at the 6-month follow-up. FA Items 4–6 (i.e., accessing health information; leading a healthy life; keeping track of body conditions and exercise) were predictive of continuation of app use; however, FA2 and FA7 had significant negative associations, indicating that those using PA apps for distraction (FA2) or finding peers (FA7) were likely to discontinue app use. Similarly, SA2 and SA4 (liking the app; being able to access the app at any time) were positively associated with continued use.

Table 3

A Logistic Regression Predicting App Use (vs. Non-use) at the Six-Month Follow-Up (n = 3825)
Independent Variable	Estimate	SE	OR	95%CI	z	p
Age	0.003	0.002	1.003	[0.999, 1.008]	1.411	.158
Gender	-0.399	0.076	0.671	[0.578, 0.779]	-5.247	< .001
BMI	0.005	0.010	1.005	[0.986, 1.025]	0.496	.620
Duration	0.305	0.081	1.357	[1.157, 1.590]	3.764	< .001
Frequency	0.343	0.083	1.409	[1.197, 1.658]	4.124	< .001
FA1	-0.032	0.040	0.969	[0.896, 1.047]	-0.806	.420
FA2	-0.123	0.057	0.884	[0.791, 0.989]	-2.162	.031
FA3	-0.047	0.055	0.954	[0.856, 1.062]	-0.859	.390
FA4	0.126	0.041	1.135	[1.048, 1.229]	3.106	.002
FA5	0.131	0.047	1.140	[1.040, 1.249]	2.804	.005
FA6	0.242	0.040	1.274	[1.178, 1.379]	6.019	< .001
FA7	-0.110	0.054	0.896	[0.806, 0.996]	-2.028	.043
FA8	0.040	0.056	1.041	[0.932, 1.163]	0.713	.476
FA9	-0.049	0.050	0.952	[0.863, 1.050]	-0.987	.324
SA1	0.006	0.058	1.006	[0.898, 1.128]	0.110	.912
SA2	0.148	0.061	1.160	[1.029, 1.307]	2.426	.015
SA3	0.052	0.040	1.054	[0.974, 1.140]	1.299	.194
SA4	0.101	0.042	1.106	[1.018, 1.202]	2.385	.017
SA5	-0.052	0.044	0.949	[0.871, 1.035]	-1.179	.238
SA6	0.045	0.042	1.046	[0.964, 1.136]	1.073	.283
SA7	0.004	0.043	1.004	[0.922, 1.093]	0.096	.924
Note. Gender = 0 for men and 1 for women. BMI, body mass index (kg/m2). Duration was coded as 0 for app use < 6 months and 1 for app use ≥ 6 months. Frequency was coded as 0 for app use once in two or more days and 1 for app use at least once per day.

We conducted another logistic regression to predict adherence to the PA guidelines at follow-up after controlling for PA guideline adherence at baseline (Table 4). The results showed that FA2 (using apps for distraction) was negatively associated with adherence at follow-up, whereas FA6 (leading a healthy life) had a significant positive effect. None of the symbolic aspects were significantly associated with guideline adherence at follow-up.

Table 4

A Logistic Regression Predicting Adherence to the PA Guidelines at the Follow-up (n = 3825)
Independent Variable	Estimate	SE	OR	95%CI	z	p
Age	0.021	0.003	1.022	[1.016, 1.027]	8.069	< .001
Gender	-0.208	0.086	0.812	[0.686, 0.961]	-2.420	.016
BMI	-0.001	0.011	0.999	[0.977, 1.021]	-0.133	.894
PA adherence at baseline	2.170	0.082	8.755	[7.458, 10.298]	26.363	< .001
Duration	0.081	0.092	1.084	[0.905, 1.298]	0.878	.380
Frequency	0.281	0.093	1.325	[1.103, 1.590]	3.019	.003
FA1	0.033	0.045	1.033	[0.946, 1.128]	0.728	.467
FA2	-0.133	0.064	0.876	[0.772, 0.994]	-2.059	.039
FA3	0.000	0.062	1.000	[0.885, 1.129]	-0.007	.994
FA4	0.065	0.047	1.067	[0.973, 1.170]	1.386	.166
FA5	-0.016	0.053	0.985	[0.887, 1.092]	-0.294	.769
FA6	0.153	0.046	1.165	[1.065, 1.275]	3.324	.001
FA7	0.032	0.061	1.032	[0.917, 1.163]	0.521	.602
FA8	0.066	0.063	1.068	[0.944, 1.209]	1.039	.299
FA9	-0.065	0.057	0.937	[0.838, 1.047]	-1.148	.251
SA1	-0.063	0.066	0.939	[0.825, 1.068]	-0.963	.336
SA2	0.086	0.069	1.090	[0.952, 1.248]	1.242	.214
SA3	-0.004	0.045	0.996	[0.911, 1.089]	-0.083	.934
SA4	0.058	0.047	1.060	[0.966, 1.163]	1.225	.220
SA5	-0.043	0.050	0.958	[0.869, 1.057]	-0.855	.392
SA6	0.027	0.048	1.027	[0.935, 1.128]	0.560	.575
SA7	0.041	0.049	1.041	[0.946, 1.147]	0.829	.407
Note. FA = Functional Aspect; SA = Symbolic Aspect. See Table 2 for the item descriptions. Duration was coded as 0 for app use < 6 months and 1 for app use ≥ 6 months. Frequency was coded as 0 for app use once in two or more days and 1 for app use at least once per day. Guideline-recommended levels of PA = 23 METs-h/week for adults aged < 65 years and 10 METs-h/week for adults ≥ 65 years.

In this study, we analyzed two waves of longitudinal survey data on commercial app use from a community sample to evaluate the predictive value of the MPA model. Specifically, we aimed to identify the functional and symbolic aspects of app use that predict continued app use and PA levels.

Regarding functional aspects, cross-sectional analyses at baseline revealed that long-term users (those already using an app for more than six months) appreciated aspects related to lifestyle management (e.g., constant accessibility to health information, tracking, and monitoring PA), whereas short-term users rated distraction and building relationships as appreciated aspects. Similar patterns of results were observed in the prospective analysis—people appreciating lifestyle management aspects were more likely to continue using the app at the 6-month follow-up. In contrast, distraction and building relationships were associated with discontinuation. These lifestyle management findings, the core aspect of healthcare apps [36, 37], are the key dimensions that inform the appropriation state, while distraction and building relationships can be seen as behaviors signaling that appropriation is in progress. Users who appreciate these off-purpose dimensions of the app might be in the exploration or early use construction stages [28], as neither distraction nor building relationships are functionalist behaviors to achieve a specific objective, namely, to improve and maintain PA. Importantly, our results also showed that lifestyle management was positively associated with adherence to the national PA guidelines at the 6-month follow-up, whereas distraction was negatively associated with adherence. On-purpose (i.e., health-conscious [24]) use appears to be crucial for maintaining app use and achieving an actual health goal.

For symbolic aspects, we found that the psychological dimensions (e.g., I like using the apps; I can access my apps at any time) were positively associated with the continuation of app use, which may reflect intrinsic motivation driving continuous app engagement. Contrary to our hypothesis, social dimensions were not associated with the reported duration of app use at baseline or the reported continuation of app use at the six-month follow-up. We have no clear explanation to readily reconcile these unexpected null results (particularly for identity). However, one possibility is that users internalize the apps after they have integrated app use into their daily routines, suggesting that repeated or regular use of apps may foster affective attachment and a sense of identity (but not in the other way around). Studies suggest that smartphones are used at any time and in nearly every environment and that people are likely to develop an attachment (or even addiction) to their mobile devices through regular exposure and interactions in their daily lives [38, 39]. As this is a speculation, future research should test temporal order and causality.

Our findings should be interpreted considering important methodological limitations. First, we included in the analyses any PA apps that participants were using, as we wanted to maintain the generalizability of the results across different PA apps. Most participants identified iOS Healthcare and Google Fitness (see Oba et al. [29] for details), whereas others used different apps, such as those with a specific focus on fitness, training, or disease management (e.g., for hypertension and diabetes). Future research should investigate the differences in app-use aspects based on the types of apps and their implemented functions. Second, the analyses exclusively targeted self-reported data on app use and PA levels; however, apps often collect user-behavior data automatically (e.g., which functions are used, when they are used, and how frequently they are used). Analyses of in-app behavior may provide objective and behavioral phenotypes of the appropriation process, although this may have to be limited to particular applications (e.g., MyFitnessPal [40]). Third, generalizability is still a matter in the current study because of the nature of the sample—Japanese-speaking adults. We do not have a clear theory regarding the differences in user behaviors between the East and West, but some participants indicated that they were using apps that were available only in Japan. Some country-specific or cultural differences are likely bound to the apps on the market or even to the cultural norms and policies of countries and regions. Fourth, we exclusively assessed the functional and symbolic aspects of app use, although the MPA model highlights other components (i.e., metacommunications and evaluations) at play in appropriating mobile technology [17, 24, 25]. Future research should shed light on these components together with their use aspects, allowing researchers to clarify the dynamic cycle of the appropriation process for PA apps.

Despite these limitations, our findings contribute to the literature on the appropriation of healthcare apps. We based our analyses on the MPA model, which has been applied to research on nutrition and diabetes [17, 24]. Our results showed that the MPA model is also a meaningful theoretical basis for analyzing user behavior regarding PA apps. Both the functional (particularly lifestyle management) and symbolic aspects (psychological dimensions) were associated with long-term app use at baseline and continued use at the six-month follow-up. Simultaneously, our findings highlight the importance of considering the degree or stages of appropriation. Distraction and building relationships were associated with discontinued use, and furthermore, distraction led to poorer health outcomes (or lower adherence to the PA guidelines). On-purpose use of an app (i.e., using a PA app to improve PA) would be the key feature of the appropriation state, whereas off-purpose use was associated with discontinuation and could be sorted out in the process of acquiring a functionalist use style and integrating app use into one’s daily routine. One implication for practice is that developing solid on-purpose functions (e.g., strengthening tracking and monitoring features for a PA app) would be a better strategy to maintain users than equipping an app with off-purpose components, such as gamification [41, 42]. Our findings echo the importance of assessing multifaceted patterns of use beyond the adoption-rejection dichotomy but point to the possibility that some use aspects may not be regarded as active features of the appropriation state. Clarifying the process is important, including how appropriation progresses and ends and updating the theoretical assumptions to better understand how people adopt and appropriate new mobile technologies.

Ethics approval and consent to participate

All participants provided informed consent. This study was approved by the Ethics Committee of the National Institute of Advanced Industrial Science and Technology (approval ID: 2022-1279)

Consent for publication

Individual participants’ data were not reported in this study.

Availability of data and materials

The dataset analyzed during the current study is not publicly available because we did not obtain consent from participants for placing data on a public registry. However, the data are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

The study was supported by an internal fund of the AIST.

Authors' contributions

KT designed the work and drafted the manuscript; OT conducted the formal analyses; KK and KK curated the data. All authors read and approved the final manuscript.

Acknowledgements

Not applicable.

Mair JL, Hayes LD, Campbell AK, Buchan DS, Easton C, Sculthorpe N. A Personalized Smartphone-Delivered Just-in-time Adaptive Intervention (JitaBug) to Increase Physical Activity in Older Adults: Mixed Methods Feasibility Study. JMIR Form Res. 2022;6(4):e34662. doi:10.2196/34662
Mauch CE, Edney SM, Viana JNM, et al. Precision health in behaviour change interventions: A scoping review. Prev Med. 2022;163:107192. doi:10.1016/j.ypmed.2022.107192
Michie S, Richardson M, Johnston M, et al. The Behavior Change Technique Taxonomy (v1) of 93 Hierarchically Clustered Techniques: Building an International Consensus for the Reporting of Behavior Change Interventions. Ann Behav Med. 2013;46(1):81-95. doi:10.1007/s12160-013-9486-6
Mollee JS, Middelweerd A, Kurvers RL, Klein MCA. What technological features are used in smartphone apps that promote physical activity? A review and content analysis. Pers Ubiquitous Comput. 2017;21(4):633-643. doi:10.1007/s00779-017-1023-3
Middelweerd A, Mollee JS, van der Wal CN, Brug J, Te Velde SJ. Apps to promote physical activity among adults: a review and content analysis. Int J Behav Nutr Phys Act. 2014;11:97. doi:10.1186/s12966-014-0097-9
Fiedler J, Eckert T, Wunsch K, Woll A. Key facets to build up eHealth and mHealth interventions to enhance physical activity, sedentary behavior and nutrition in healthy subjects – an umbrella review. BMC Public Health. 2020;20(1):1605. doi:10.1186/s12889-020-09700-7
Ferguson T, Olds T, Curtis R, et al. Effectiveness of wearable activity trackers to increase physical activity and improve health: a systematic review of systematic reviews and meta-analyses. Lancet Digit Health. 2022;4(8):e615-e626. doi:10.1016/S2589-7500(22)00111-X
Mair JL, Salamanca-Sanabria A, Augsburger M, et al. Effective Behavior Change Techniques in Digital Health Interventions for the Prevention or Management of Noncommunicable Diseases: An Umbrella Review. Ann Behav Med. 2023;57(10):817-835. doi:10.1093/abm/kaad041
Iribarren SJ, Cato K, Falzon L, Stone PW. What is the economic evidence for mHealth? A systematic review of economic evaluations of mHealth solutions. PloS One. 2017;12(2):e0170581. doi:10.1371/journal.pone.0170581
Rinaldi G, Hijazi A, Haghparast-Bidgoli H. Cost and cost-effectiveness of mHealth interventions for the prevention and control of type 2 diabetes mellitus: A systematic review. Diabetes Res Clin Pract. 2020;162:108084. doi:10.1016/j.diabres.2020.108084
Baumel A, Muench F, Edan S, Kane JM. Objective User Engagement With Mental Health Apps: Systematic Search and Panel-Based Usage Analysis. J Med Internet Res. 2019;21(9):e14567. doi:10.2196/14567
Baumel A, Kane JM. Examining Predictors of Real-World User Engagement with Self-Guided eHealth Interventions: Analysis of Mobile Apps and Websites Using a Novel Dataset. J Med Internet Res. 2018;20(12):e11491. doi:10.2196/11491
Mitchell M, Lau E, White L, Faulkner G. Commercial app use linked with sustained physical activity in two Canadian provinces: a 12-month quasi-experimental study. Int J Behav Nutr Phys Act. 2020;17(1):24. doi:10.1186/s12966-020-00926-7
Maliwichi P, Chigona W, Sowon K. Appropriation of mHealth Interventions for Maternal Health Care in Sub-Saharan Africa: Hermeneutic Review. JMIR MHealth UHealth. 2021;9(10):e22653. doi:10.2196/22653
Carroll J, Howard S, Vetere F, Peck J, Murphy J. Identity, power and fragmentation in cyberspace: technology appropriation by young people. Working paper 01/IDG/2001, Department of Information Systems, University of Melbourne. Published 2001. Accessed December 5, 2023. https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.453.6847&rep=rep1&type=pdf
Wirth W, Von Pape T, Karnowski V. An Integrative Model of Mobile Phone Appropriation. J Comput-Mediat Commun. 2008;13(3):593-617. doi:10.1111/j.1083-6101.2008.00412.x
Rossmann C, Riesmeyer C, Brew-Sam N, et al. Appropriation of Mobile Health for Diabetes Self-Management: Lessons From Two Qualitative Studies. JMIR Diabetes. 2019;4(1):e10271. doi:10.2196/10271
Davis FD. Perceived Usefulness, Perceived Ease of Use, and User Acceptance of Information Technology. MIS Q. 1989;13(3):319. doi:10.2307/249008
Davis FD. A Technology Acceptance Model for Empirically Testing New End-User Information Systems : Theory and Results. Massachusetts Institute of Technology; 1986. Accessed January 5, 2024. http://hdl.handle.net/1721.1/15192
Cho J. The impact of post-adoption beliefs on the continued use of health apps. Int J Med Inf. 2016;87:75-83. doi:10.1016/j.ijmedinf.2015.12.016
Cho J, Quinlan MM, Park D, Noh GY. Determinants of Adoption of Smartphone Health Apps among College Students. Am J Health Behav. 2014;38(6):860-870. doi:10.5993/AJHB.38.6.8
Hoque R, Sorwar G. Understanding factors influencing the adoption of mHealth by the elderly: An extension of the UTAUT model. Int J Med Inf. 2017;101:75-84. doi:10.1016/j.ijmedinf.2017.02.002
Yuan S, Ma W, Kanthawala S, Peng W. Keep Using My Health Apps: Discover Users’ Perception of Health and Fitness Apps with the UTAUT2 Model. Telemed E-Health. 2015;21(9):735-741. doi:10.1089/tmj.2014.0148
Stehr P, Karnowski V, Rossmann C. The multi-faceted usage patterns of nutrition apps: a survey on the appropriation of nutrition apps among German-speaking users of MyFitnessPal. BMC Med Inform Decis Mak. 2020;20(1):279. doi:10.1186/s12911-020-01294-9
Lee SK, Karnowski V, Von Pape T, Cionea IA. An English scale for measuring mobile phone appropriation: Translation and assessment. Stud Commun Media. 2016;5(4):397-426. doi:10.5771/2192-4007-2016-4-397
Ruggiero TE. Uses and Gratifications Theory in the 21st Century. Mass Commun Soc. 2000;3(1):3-37. doi:10.1207/S15327825MCS0301_02
Katz JE, Sugiyama S. Mobile phones as fashion statements: evidence from student surveys in the US and Japan. New Media Soc. 2006;8(2):321-337. doi:10.1177/1461444806061950
Benamar L, Balagué C, Zhong Z. Internet of Things devices appropriation process: The Dynamic Interactions Value Appropriation (DIVA) framework. Technovation. 2020;89:102082. doi:10.1016/j.technovation.2019.06.001
Oba T, Takano K, Katahira K, Kimura K. Use Patterns of Smartphone Apps and Wearable Devices Supporting Physical Activity and Exercise: Large-Scale Cross-Sectional Survey. JMIR MHealth UHealth. 2023;11:e49148-e49148. doi:10.2196/49148
Oba T, Takano K, Katahira K, Kimura K. Revisiting the Transtheoretical Model for Physical Activity: A Large-Scale Cross-Sectional Study on Japanese-Speaking Adults. Ann Behav Med. Published online January 2, 2024:kaad069. doi:10.1093/abm/kaad069
Craig CL, Marshall AL, Sjöström M, et al. International Physical Activity Questionnaire: 12-Country Reliability and Validity: Med Sci Sports Exerc. 2003;35(8):1381-1395. doi:10.1249/01.MSS.0000078924.61453.FB
Murase, N, Katsumura, T, Ueda, C, Inoue, S, Shimomitsu, T. Validity and reliability of Japanese version of International Physical Activity Questionnaire. J Health Welf Stat. 2002;49(11):1-9.
Ministry of Health, Labor and Welfare. Physical Activity Standards for Health Promotion.; 2013. Accessed January 23, 2023. https://www.e-healthnet.mhlw.go.jp/information/policy/guidelines_2013.html
Marcus BH, Simkin LR. The stages of exercise behavior. J Sports Med Phys Fitness. 1993;33(1):83-88.
Prochaska JO, Velicer WF. The Transtheoretical Model of Health Behavior Change. Am J Health Promot. 1997;12(1):38-48. doi:10.4278/0890-1171-12.1.38
Brown III W, Yen PY, Rojas M, Schnall R. Assessment of the Health IT Usability Evaluation Model (Health-ITUEM) for evaluating mobile health (mHealth) technology. J Biomed Inform. 2013;46(6):1080-1087. doi:10.1016/j.jbi.2013.08.001
Dennison L, Morrison L, Conway G, Yardley L. Opportunities and challenges for smartphone applications in supporting health behavior change: qualitative study. J Med Internet Res. 2013;15(4):e86. doi:10.2196/jmir.2583
Vincent J. Emotional attachment and mobile phones. Knowl Technol Policy. 2006;19(1):39-44. doi:10.1007/s12130-006-1013-7
Sohn S, Karampournioti E, Wiedmann K, Fritz W. The sources of the many faces of consumer smartphone attachment: A value‐in‐use perspective. Int J Consum Stud. 2022;46(4):1399-1412. doi:10.1111/ijcs.12765
Gordon M, Althoff T, Leskovec J. Goal-setting And Achievement In Activity Tracking Apps: A Case Study Of MyFitnessPal. In: The World Wide Web Conference. ACM; 2019:571-582. doi:10.1145/3308558.3313432
Six SG, Byrne KA, Tibbett TP, Pericot-Valverde I. Examining the Effectiveness of Gamification in Mental Health Apps for Depression: Systematic Review and Meta-analysis. JMIR Ment Health. 2021;8(11):e32199. doi:10.2196/32199
Mazeas A, Duclos M, Pereira B, Chalabaev A. Evaluating the Effectiveness of Gamification on Physical Activity: Systematic Review and Meta-analysis of Randomized Controlled Trials. J Med Internet Res. 2022;24(1):e26779. doi:10.2196/26779

No competing interests reported.

STROBEchecklistv4combinedPlosMedicine1.docx

Download PDF

Editorial decision: Revision requested
01 Aug, 2024
Reviews received at journal
01 Aug, 2024
Reviews received at journal
29 Jul, 2024
Reviewers agreed at journal
28 Jul, 2024
Reviewers agreed at journal
19 Jul, 2024
Reviewers invited by journal
19 Jul, 2024
Editor invited by journal
08 Jul, 2024
Editor assigned by journal
03 Jul, 2024
Submission checks completed at journal
03 Jul, 2024
First submitted to journal
01 Jul, 2024

You are reading this latest preprint version

Functional and Symbolic Aspects of App Use for Improving Physical Activity: A Six-month Prospective Analysis

Status:

Version 1

Abstract

Introduction

Methods

Results

Discussion

Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1