Sage Journals: Discover world-class research

Abstract

Since the COVID-19 pandemic, the use of mobile health apps has been increasing significantly. Many have started using mobile health apps to manage their health and well-being. People need to develop healthy habits that will increase their job productivity, life expectancy and improve their social relationships. This study tests the Health Belief Model with adults in China in influencing health motivation and in turn improving their health behavior. A mobile and online survey was developed, and 120 usable responses were collected. A structural equation modeling was used to test the relationships between perceived benefits, perceived susceptibility, perceived barriers, health motivation, and health behavior. We found perceived self-efficacy, perceived benefits, and perceived susceptibility predicts health motivation, and health motivation predicts health behavior. We did not find that perceived barrier is a predictor for health motivation. We conclude that to encourage behavioral change, it is important to enhance self-efficacy and address the benefits; in the meantime, educating people on susceptibility is also important.

Keywords

Health belief model self-efficacy health motivation health behaviors health promotion

Introduction

The mobile app market grew 65.7% compared to pre-Covid (Fortune Business Insight 2020).¹ China's revenue in the Digital Health market is projected to reach US$46.00bn in 2022 with an annual growth rate of 12.98% till 2027. IQVIA reports that over 90,000 digital health apps were marketed last year, with an average of 250 apps per day (2021). Health mobile apps play an important role in health promotion and motivation. Harvard Medical Publishing (2015)² claims that downloadable smartphone apps can coach people to check and improve their health. Current evidence shows that individuals and families practicing self-management behaviors improve their health outcomes.³ Mobile health apps can be used to educate patients, and the public about health and even serve as interventions for disease treatment and recovery. Increase in demand for healthcare services during covid, inability of people to visit health care facilities due to social distancing measures, different regions within China having different COVID-19 policies highlighting the need for communication, increased awareness of maintaining good general health have all heightened the need to study mobile health apps adoption in China. It was interesting to find that the apps were mostly used for “accessing information on COVID-19.” Chinese respondents overwhelmingly identified “information” as the most valuable feature of COVID-19 apps.⁴ In addition, it is believed that the pandemic has accelerated the adoption of health apps in China⁷ and this trend may continue thus having a significant long-term impact. That is evidence that mHealth apps are an effective health care approach gaining in popularity among the Chinese population following the COVID-19 outbreak.⁵ The next step after recognizing the importance mobile applications is to explore the factors that initiates or improves adoption of mobile apps in China using established theoretical models. Such an understanding can not only improve the adoption of mobile apps by suggesting ways to improve both usage and adoption.

Theory-guided research is needed to understand health beliefs and attitudes toward technological interventions such as mobile health apps. The Health Belief Model (HBM) is widely accepted and used in explaining people's health beliefs and behaviors. It has been tested extensively in different countries and cultures.⁶ The classic Health Belief Model contains six constructs perceived seriousness, perceived threat, perceived susceptibility, health motivation, cue to action, and the likelihood of behavior change. Over the years, many researchers have modified the health belief model by adding moderators and mediators.⁷

The Pandemic has affected every people’s lives in unprecedented ways. Many people's health behaviors have changed. According to Statistica, China is ranked second in the mobile health app user share globally.⁸ The COVID-19 pandemic pushed patients and providers to adopt digital health tools as care moved outside the doctor's office. Despite the pandemic triggering an explosion of new digital health apps, its user acceptance is still an area for further investigation.

It's projected that the mobile platform will deliver more health services. A study compares age differences in users' adoption of mobile health services in China and finds that perceived value, attitude, perceived behavior control, and resistance to change predict the intention to use mobile health services for the middle-aged group. Compared to old users, perceived value, attitude, perceived behavior control, technology anxiety, and self-actualization need positively affected the behavior intention of older users.⁹ Though the limited studies in China on mobile apps provide some useful results, in general, studies on mobile health apps in China are limited.¹⁰ More studies are needed to further validate HBM in China contexts and people's beliefs and behavior change over time.

The aims of this study are to (1) apply the health belief model to mobile health apps in China context, (2) examine important predictors of health behavior, and (3) discuss implications. We intend to explore the users' adoption of mobile health apps and their attitudes toward mobile health apps, health beliefs, and health behavior. The remainder of this paper is organized as follows. We first provide an overview of mobile health apps and conduct a literature review on the health belief model. The Methods section describes instrument development, translations, and implementation. The analysis section explains analysis methods and model testing. The final two sections report the results of our analysis and a discussion of our findings.

Health belief model

Mobile health apps

Mobile health has a promising future in the healthcare system in patient education, treatment access, and recovery support. Harvard Medical Publishing (2015)² claims that downloadable smartphone apps can coach people to check and improve their health . Mobile messaging is a feasible addition to church-based programs to improve dietary and physical activity behaviors.⁷ Medical professionals often find it challenging to alter patients’ habits¹¹ for a variety of reasons. As an example, physicians do not want to engage in a weight-loss conversation for fear of damaging patient-physician relationships. Patients' self-awareness, self-education, self-regulation, and self-motivation are the starting points for better health.

Research shows that counselling services with frequent contact with patients provided by healthcare professionals can make patients change their health-related behavior and lifestyle.¹² Chisholm et al.¹¹ suggest that the barriers to behavioral change are context factors, motivational stance, and awareness of the need to change. Mobile health apps can provide counselling and frequent contact at the most convenient time and location. Whenever a patient is available, they can use health apps to self-manage their health-related needs.

China has made significant investments in its mobile technology infrastructure in the past decade. Mobile health apps hold the promise of patient education, disease prevention, disease management, and improved communication with providers. In general, mobile health app users in China are younger, better educated, and living in urban areas.¹³ A descriptive study of China's mobile health market concludes that clinician-facing apps feature education and peer-reviewed papers. In contrast, consumer-facing apps contain features like scheduling appointments, exercise, diet, or disease-specific apps.¹⁴

The use of mobile health apps could improve patient experience, such as accessing health information, making physician-patient communication more convenient, ensuring transparency in medical charges, and ameliorating short-term outcomes.¹³ A recent study on ankylosing spondylitis health apps finds that most apps lacked comprehensive functions for ankylosing spondylitis management. This calls for the room for improvement for disease-specific apps. A fully-functional and patient-centered apps need concerted effort among healthcare providers, patients, and app developers to develop high-quality, evidence-based apps.¹⁵

Health belief theory

Health belief theory states that individuals' beliefs determine health behavior about disease and strategies available to decrease its occurrence.¹⁶ It postulates that an individual's likelihood of engaging in a health-related behavior is determined by his/her perception of perceived susceptibility, perceived severity, perceived benefits and risks, and cues to action.¹⁷ Later, variables from Social Determination Theory (e.g. motivation)¹⁸ and Social Cognitive Theory (self-efficacy) were added to HBM to increase the predictive power of HBM.¹⁹ The Health Belief Model (HBM) is one of the most popular models used to prevent diseases with social, psychological, and behavioral underpinnings. HBM is the most widely used model explaining health beliefs and behavior validated in many cultures and contexts.^20–22

Health Belief Model has a great impact on health education, designing both long and short-term health behavior interventions and health promotion.^23,24 Although this model is useful in general, prior findings report the model only explains 20% of variances in behavior change.²⁵ Recently, a systematic review of COVID-19-related preventive behavior shows a 25% predictive validity, and the explained variance for HBM ranged from 6.5% to 90.1% from the global studies of HBM.²⁰ However, studies have used a varied number of HBM constructs. Some did not include perceived benefits, whereas others did not incorporate self-efficacy into their model. In general, perceived benefits and self-efficacy are the two most frequently used constructs in all HBM studies.²⁰

In our study, we intend to survey adults' attitudes and perceptions towards mobile health apps, we deem perceived barriers, perceived benefits, perceived susceptibility, and self-efficacy are important predictors to motivate individuals to engage in healthy behaviors. We did not include perceived severity in our model as prior studies show no significant difference was observed between the two groups regarding perceived severity.⁶ Harrison et al.²⁶ suggested severity might not vary as much as some of the other variables because few people consider outcomes like breast cancer as anything other than extremely severe.²⁷

Figure 1 shows the constructs and relationship between constructs in the Health Belief Model.

Figure 1.

Research model.

Research model

Self-efficacy refers to the belief that a person can control challenging environmental demands by taking adaptive action.²⁸ It is perceived to be domain specific. In the past, many self-efficacy scales have been developed, for example, computer self-efficacy²⁹ breast cancer self-efficacy,³⁰ and generalized self-efficacy scale.^31,32 revealed that self-efficacy was the most important predictor in the HBM constructs to predict breast self-examination behavior in female students. In a quasi-experimental study of mobile health app intervention on women's health beliefs on breast self-examination, self-efficacy, and health motivation were significantly higher in the intervention group compared to the control group.⁶ The relationship between self-efficacy and motivation has also been established in the literature. For example, a recent study on a university gym evaluation program explores how perceived changes in health and motivation are related to changes in self-efficacy by age.³³ Motivation could be intrinsic and extrinsic. Intrinsic motivation refers to "people's spontaneous tendencies to be curious and interested, to seek out challenges and to exercise and develop their skills and knowledge, even in the absence of operationally separable rewards".¹⁸ Drawing from studies on computer usage, researchers have theorized that higher self-efficacy leads to easier use and lower resistance to adoption.³⁴ Although intuitively it appears that such a theorization is plausible, results from other mobile app studies are conflicting. Alharbi et al.³⁵ found a strong support perceived benefits and self-efficacy to be the strong predictors of mobile health applications. However, Vancouver et al.,³⁶ describe other studies where the evidence is conflicting. To further test the role of self-efficacy we posit

H1: Self-efficacy is positively associated with health motivation.

Perceived benefits refer to individuals' opinions or values of a new behavior’s benefits.³⁷ Perceived benefits play an important role in adopting preventive behaviors.²⁷ Graham³⁸ found that women who perceived the benefits of breast self-examination had higher odds of performing the screening than those who did not. Interventions based on perceived benefits are recommended to increase women's regular breast cancer screening.³² People who perceive the benefits are more motivated to engage in certain health behavior. While there are some studies that support the influence of perceived benefits on motivating behavioral change,³⁹ there is some doubt cast by others whether the motivation will lead to lasting change in behavior Borland et al., 2010).⁴⁰ To test this relationship further, we include perceived benefits in the model and posit that

H2: Perceived benefits are positively associated with health motivation.

Perceived susceptibility/risk refers to the likelihood of an individual developing a health condition. People will be more motivated to seek healthy behavior if they believe they are at risk for developing a specific negative health outcome.⁴¹ Usually, the higher the perceived risk is, the higher the likelihood of an individual engaging in behavioral changes to reduce the risk. Increased belief in risks will motivate people to take action.⁴² For example, Mullens et al.⁴³ report behavior changes in colon cancer survivors. We therefore posit:

H3: Perceived susceptibility is positively associated with health motivation.

Perceived barriers mean the obstacles an individual perceives to adopting a new behavior. It is the most significant⁴⁴ and strong predictor⁴⁵ in behavior change. When individuals feel the barriers are high, they are less motivated to participate in behavior-change activities. We therefore posit:

H4: Perceived barriers are negatively associated with health motivation.

Motivation is the general desire to do something. Motivation is driven by our needs, and affects how we think, feel, and act. Each of us is unique in our needs, wants, and desires which drive us to achieve better health, and well-being, live a purposeful life, or grow professionally. Becker⁴⁶ proposed that overall motivation to pursue healthy behavior should be included in HBM. Motivation helps us understand human nature as it explains why we set goals, strive for achievement and power, and why experience emotions like fear, anger, and compassion. Health motivation is important in changing behavior, as numerous studies show intention and motivation drive behavior.⁴⁷

H5: Health motivations are positively associated with health behaviors.

Research methods

The authors’ Institutional Review Board approved this study. To examine the relationships among constructs in the Health Belief Model, we used a survey method. Data were collected in a fitness club near a university in Hangzhou, China. Most of the club members are young adults. We recruited five college students to collect data for us. The five students went to the fitness club asking for consent for participation in this survey. Two survey delivery methods were used to provide convenience to the study participants: a mobile survey and an online survey. The same survey was used in both the delivery methods. Once the consent was obtained, a bar code containing a link to the survey was texted to WeChat on their mobile phone and emailed to the study participants. The data collection lasted about 2 months between Oct. 1.2021 – Nov. 30, 2021. A total of 122 surveys were collected, 120 were usable surveys.

Measures

All measures were adapted from existing literature. The self-Efficacy scale was adapted and based on the theoretical perspectives proposed by Bandura.²⁸ Perceived benefits, perceived seriousness, perceived susceptibility, health motivation, and barriers were adapted from.⁴⁸ For health behavior, we use three items eating, exercise, and body weight index, to measure behaviors.

An English version of health belief model questions specifically targeting mobile health apps was developed. Questions were measured on a seven-point Likert scale (with one indicating strongly agree and seven indicating strongly disagree). The questionnaire contains four questions related to susceptibility, four about seriousness, six about benefits, seven about barriers, three about self-efficacy, three about health behavior, and six about health motivation. Low scores indicated high levels of perceived dimensions. Two researchers translated the English version into the Chinese version. A back translation was also conducted⁴⁹ from Chinese to English to ensure equivalency and reliability between the English and Chinese versions. Participants then completed a Chinese version of the HBM questionnaire. Appendix 1 contains our survey.

Results

Descriptive statistics

Table 1 shows the descriptive statistics and demographic characteristics of our participants. The majority of our respondents are females (62.5%). The males consist of 37.5% of the total sample. With regards to age distribution, 34.17% are 20 years and under; 59.17% are between 21 and 30, and 6.33% are over 30. Our sample contains more young adults; 45% of the respondents have used mobile health apps for less than 1 year; 28.33% have used mobile health apps for two to 3 years; 17.5% have used four to 5 years, and 9.17% for more than 5 years.

Table 1.

Sample descriptive statistics.

Characteristics	N	Percent (%)
Gender
Male	45	37.50
Female	75	62.50
Age
under 20	41	34.17
21-30	71	59.17
31-40	1	0.83
41-50	3	2.50
>50	4	3
Years of mobile health app experience
=<1	54	45.00
2-3	34	28.33
4-5	21	17.50
>5	11	9.17

Analysis

A confirmatory factor analysis using Stata 17 was used to test the measurement model.

We first validated our measurement model for construct validity. Three items were removed due to their poor loadings. To establish construct validity, we must validate reliability, convergent and discriminant validity. Table 2 shows descriptive statistics of constructs, mean, standard deviation, Cronbach alpha, composite reliability, and average variance extracted. Our values of Cronbach alpha range from 0.74 to 0.91 exceeding the recommended level of 0.7.⁵⁰ Composite reliability for all constructs is between 0.8 and 0.96 and the average variable extracted is between 0.51 and 0.64 exceeding the suggested threshold of 0.5.⁵¹

Table 2.

Descriptive statistics of constructs.

Constructs	Mean	Standard Deviation	Cronbach alpha	Composite reliability	Average variance extracted
Health behavior	2.43	1.16	0.84	0.96	0.64
Perceived susceptibility	3.99	1.72	0.75	0.93	0.58
Self-efficacy	2.46	1.18	0.85	0.86	0.67
Perceived benefits	2.65	1.23	0.86	0.8	0.63
Health motivation	2.36	0.91	0.74	0.94	0.51
Perceived barriers	4.06	1.52	0.91	0.87	0.55

Table 3 shows factor loadings. According to Hair et al.,⁵² factor loadings greater than 0.5 are considered acceptable. All the factor loadings in Table 3 are in the range of 0.68-0.89. Thus convergent validity was established with the combined evidence provided in both Table 2 and Table 3.

Table 3.

Factor loadings.

	Measure	Loadings
Self-efficacy	SEQ1	0.68
	SEQ2	0.82
	SEQ3	0.93
Benefits	PBQ1	0.88
	PBQ2	0.73
	PBQ3	0.76
Perceived barriers	PEB1	0.85
	PEB2	0.85
	PEB3	0.89
	PEB4	0.82
Health motivation	HM3	0.72
	HM4	0.70
	HM5	0.73
Susceptibility	PSU1	0.85
Susceptibility	PSU2	0.67
Health behavior	Eat	0.84
	Exercise	0.84
	BMI	0.71

To test discriminant validity, we used two well-established methods: item-to-construct correlation,⁵³ and Fornell-Larcker's method on correlation and AVE, which requires the square root of the AVEs of each latent variable must be greater than its correlations with any other constructs.⁵¹ Appendix 2 shows item-to-construct correlations, which provide evidence that all items correlated highly with their intended constructs.

Table 4 provides correlations between constructs and the square root of the average variance extracted. The bold numbers are the square root of the average variance extracted. As shown in Table 4, all the bold numbers are greater than the correlations of the construct with other constructs. Based on the statistical analysis of validity criteria depicted in Tables 2–4, we, therefore, conclude the establishment of construct validity.

Table 4.

Discriminant validity.

Constructs	Se	PE	HM	PS	BH	BR
Self-efficacy (SE)	0.82
Benefits (PE)	0.63	0.79
Motivation (PM)	0.37	0.44	0.71
Susceptibility(PS)	0.22	0.22	0.29	0.76
Behavior (BH)	0.31	0.34	0.51	0.27	0.80
Barrier (BR)	0.03	0.14	0.13	0.51	0.11	0.85

Structural model

After validating the factor structure of the constructs, we proceed to test the structural relationships using a confirmatory approach. Stata 17 was used to conduct a CFI on our research model. Stata provides CFI, RMSEA, and TLI for model evaluation. Our model shows CFI=0.92, RMSEA 0.07, TLI =0.90. These fit indices satisfy the criteria for a good model fit (Bentler, 1990).⁵⁴ Our RMSEA is 0.07, which is an acceptable fit level (Browne and Cudeck, 1992).⁵⁵ The diagram below shows significant paths in the model. As shown in the diagram, when four variables are used to predict health promotion, only three are significant: self-efficacy perceived benefits and perceived susceptibility. The perceived barrier is not a significant predictor. Collectively the three constructs perceived efficacy, perceived benefits, and perceived susceptibility, explain 34.5% of the variance. The path coefficient from perceived efficacy to health motivation is 0.19 (p < 0.01). The path coefficient from perceived benefits to health motivation is 0.28 (p < 0.01). The path coefficient from perceived susceptibility to health motivation is 0.29 (p < 0.01). The path coefficient from health motivation to health behavior is 0.62 (p < 0.001) and health motivation explains 39.5% of the variance in health behavior.

When conducting multiple regression tests, it is important to check multicollinearity. As shown in Table 5, all the values of variance inflation factor (VIF) are less than 2, thus, multicollinearity is not of concern (O'brien, 2007)⁵⁶

Table 5.

Multicollinearity variance inflation factor (VIF).

	VIF	VIF
	Health promotion	Health behaviour
Health promotion		1
Perceived barrier	1.54
Perceived susceptibility	1.57
Self-efficacy	1.96
Perceived benefit	1.94

Table 6 presents the results of the hypotheses testing. Among five hypotheses, four were confirmed except one. We did not find a significant relationship between perceived barriers and health motivation.

Table 6.

Hypotheses testing.

Hypotheses	Standardized regression coefficients	p-value	Conclusion
H1: Self efficacy--> health motivation	0.19	0.04	Supported
H2: Perceived benefits--> health motivation	0.28	0.03	Supported
H3: Perceived susceptibility--> health motivation	0.29	0.04	Supported
H4: Perceived barrier--> health motivation	0.11	0.12	Rejected
H5: Health motivation--> health behavior	0.62	0.00	Supported

Analysis

Mediation test

According to Williams and MacKinnon,⁵⁷ bias-corrected bootstrapped confidence intervals (CIs) are considered more favorable compared to traditional theorem estimates. Using Stata, we applied bias-corrected bootstrapping techniques (1,000 bootstrap samples) and tested for biased-corrected two-tailed significance to confirm that the indirect effect was present Figure 2.

Figure 2.

Path results.

The table below shows the results of mediation analysis. As the results of estimating specific indirect effects, all the estimates of specific indirect effects were positive and significant (p < 0.05) and well within their bias-corrected bootstrapped 95% CIs. Hence, we conclude that health motivation mediates the path self efficacy à health behavior; health motivation mediates the path of perceived benefits à health behavior; and health motivation mediates the path of perceived susceptibility à health behavior. The results of direct effects analysis show a non-significant effect exists between self efficacy and health behavior, perceived benefits and health behavior, and perceived susceptibility and health behavior. Therefore, the type of mediation is full for all three paths and there is no mediation for Perceived barriers-Health behaviour relationship Table 7.

Table 7.

Direct and Indirect effects of Mediation.

			Bias-corrected 95% CI
Specific indirect effect	Point estimate	Bootstrap std err	Lower	Upper
Self efficacy--> health motivation--> health behavior	0.12*	0.16	0.17	0.48
Perceived benefits--> health motivation--> health behavior	0.17*	0.1	0.05	0.36
Perceived susceptibility--> health motivation--> health behavior	0.18*	0.08	0.01	0.32
Direct effect
Self-efficacy--> health behavior	0.09 ns	0.14	–0.26	0.3
Perceived benefits--> health behavior	0.08 ns	0.15	–0.21	0.38
Perceived susceptibility--> health behavior	0.12 ns	0.1	–0.08	0.32
Perceived barrier--> health behavior	0.03 ns	0.11	–0.19	0.26

Note: ns: non-significant *=p ≤ 0.05.

Discussion

China's Ministry of Education reports that about 30% of Chinese college students failed to reach the “qualified” level of physical health as of 2020.⁵⁸ This calls for urgency in educating young people about health. The health of people determines their productivity. The Pandemic has made more people more conscious about their health, resulting in more extensive adoption of mobile health apps. Mobile apps promote health and wellness.⁵⁹ The health belief model has been validated in many contexts. Mobile health (m-health) apps can bring health prevention and promotion to the general population.⁶⁰ This study assesses adults' health perceptions, self-reported health attitudes toward using mobile health apps, and their behaviors. Built on Health Belief Model and Self Determination theory, this study tests a structural equation model of user perceptions of mobile health apps with motivation as a mediator which predicts health behaviors.

Several key findings are developed. First, this study finds that self-efficacy positively influences health motivation. This finding is consistent with what the literature reports.^32,21,61 For example, people who are more knowledgeable about breast cancer self-examination are more motivated to conduct the self-examination³² Another study on the obese population also reports that education is a significant predictor of the performance of health promotion. When applying this finding to practice, education and training is the key. When people gain knowledge and skills, they are more confident and motivated to change their behavior.

Second, this study also finds the positive influence of perceived benefits on health motivation. Intuitively, this relationship is expected. When the benefits are tangible or can be visualized, people are motivated to change their behavior.⁴²

A recent systematic review of the COVID-19 study²⁰ concludes that perceived benefits are the most significant predictor of behavior. Thus, any education materials for disease prevention or treatment should start with potential benefits. Letting people know the benefits would motivate them to engage in desired health behavior.

Third, this study confirms the positive relationship between perceived susceptibility and motivation. When people are worried about something or feel vulnerable to health-related issues, they are more likely to want to change their behavior. The worry or perceived risk is sufficient for people to change their behavior.⁴² In a recent COVID-19 study on nursing students,²¹ increased susceptibility will increase nursing students' intention to change their behavioral intention to prevent the spread of COVID-19.

Lastly, unlike prior studies,²¹ this study did not find perceived barriers a predictor for motivation. This could be due to our study participants being familiar with mobile apps and how to use them, and they do not feel any extra effort or risks in using mobile apps. Furthermore, we confirm motivation is a very strong predictor of health behaviors. Thoughts and beliefs drive one's actions; in essence, changing people's beliefs will lead to changes in behaviors.⁴² Access to smartphone applications improves one's beliefs of perceived susceptibility, self-efficacy, and health motivation.⁶

Conclusion

Health has been a focused concern since the Pandemic. Mobile health apps have become an important platform for healthcare and gained popularity during the Pandemic. This study applied Health Belief Model to assess users' acceptance and attitudes toward using mobile health apps for health interventions. Our study focuses on the consumer-facing mobile health apps, which could be a useful health intervention for people to change their behavior. Healthcare mobile apps can satisfy our health-related needs and empower us to manage our health any time anywhere with an intuitive interface that is easy to learn and use for most consumers. Moreover, this study further confirms the importance of self-efficacy and perceived benefits in changing consumers' behavior. In practice, mobile health app vendors should target the benefits when promoting their health apps; it is also essential to tackle self-efficacy issues when addressing behavior change. Yet, long-term sustainability, user behavior, and engagement with mobile health apps continue to be important research areas. Our future studies will address the health belief model in different cultures and contexts.

Limitations

We recognized a few limitations with our research. First, we collected data in China to test our model. The literature documented evidence of cultural differences in people's attitudes and behavior²⁸ Thus, generalization should be applied to similar contexts. Future studies could test our model in a different cultural context. Second, we did not use a random sampling strategy. Though our sampling method is not random, the selection of the study participants suits the purpose of this study. Future studies will use the random sample to test the same model to examine the predictive validity of the model. It should be borne in mind that the survey results are not generalizable across the entirety of the Chinese population. Third, our questionnaire on mobile health apps did not target a particular disease category. There are many disease-specific mobile apps on the market. It will be helpful to expand our study to disease-specific apps and mobile health apps that are recommended by clinicians. Cross-validation of the results is useful in knowledge transfer and generalization. Future studies may tailor questions to disease types to further validate the health belief model.

Footnotes

Acknowledgments

The authors would like to thank all people who participated in this study.

Author contributions

The first and second authors conceptualized the paper and developed the research model. The second author developed the survey instrument. The first author performed the analysis and drafted the paper. The third author collected data.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Zhejiang National Science Foundation (PI: Xueyan Yang, No. LY 20G02001).

Ethical Statement

The institution Review Board at Northern Kentucky University approved this study (IRB#344). Zhejiang National Science Foundation (LY 20G02001) in China approved this study. Online consent was obtained prior to data collection.

ORCID iD

Xiaoni Zhang

Appendix 1

Table A1.

Survey questions.

Questions
SEQ1 I am able to use mobile apps to achieve most of the goals that I have set for myself
SEQ2 when facing difficult tasks in mobile devices, I am certain that I will accomplish them
SEQ3 I am able to successfully overcome many challenges in using mobile technology
PBQ1 Getting a fitness app will prevent me from worsening my health
PBQ2 Getting a fitness app will help my household from getting sick
PBQ3 Getting a fitness-app will prevent me from being absent from work
PEB1 Downloading a health-related app into my mobile phone is NOT convenient for me
PEB2 In order to get a health-related app into my mobile device, I would have to give up quite bit such as space in my phone
PEB3 Downloading or using a fitness-related app is a drain on my time
PEB4 Downloading a fitness-related app interferes with my daily activities
PEB5 mobile devices have been associated with the risk of cancer
PEB6 mobile devices influence nervous system negatively
PEB7 smartphone harbor viruses and bacteria (e.g. Coronavirus, E.Coli) that can make you sick
PES2 if my health deteriorates, my future career plans would be in danger
PES3 if my health deteriorates, it would disrupt my family
PES4 if I fall sick, it would make my daily activities more difficult
HM3 I frequently do things on my own to improve my health
HM4 I search for new information related to my health
HM5 I have the recommended yearly physical exams in addition to visits related to my health
PSU1 Working with multiple people each day increases my chances health risks
PSU2 only people over 65 years of age have greater health risks
PSU4 I feel my chance of health risks in the future are high
HPB1 I eat a balanced meal
HPB2 I exercise regularly each week
HPB3 I maintain a healthy body weight

Appendix 2

Table A2.

Item-to-construct correlation: discriminant validity.

Constructs	Barrier	Self efficacy	Perceived benefits	Health motivation	Susceptibility	Health behavior	Seriousness
SEQ1	0.04	0.80	0.61	0.33	0.21	0.24	0.27
SEQ2	0.03	0.90	0.46	0.27	0.22	0.27	0.15
SEQ3	0.02	0.92	0.59	0.39	0.16	0.30	0.12
PBQ1	0.11	0.64	0.89	0.33	0.16	0.23	0.23
PBQ2	0.08	0.49	0.86	0.39	0.18	0.36	0.23
PBQ3	0.19	0.51	0.86	0.42	0.23	0.30	0.20
PEB1	0.90	0.10	0.16	0.10	0.50	0.14	0.23
PEB2	0.90	0.06	0.16	0.16	0.40	0.19	0.19
PEB3	0.92	-0.05	0.06	0.08	0.42	0.02	0.18
PEB4	0.88	0.01	0.14	0.14	0.50	0.05	0.22
HM3	0.03	0.28	0.28	0.81	0.10	0.44	0.17
HM4	0.01	0.33	0.36	0.78	0.14	0.41	0.10
HM5	0.22	0.32	0.42	0.88	0.38	0.41	0.10
PSU1	0.36	0.16	0.14	0.20	0.88	0.19	0.27
PSU2	0.53	0.24	0.25	0.31	0.91	0.28	0.26
PES1	0.36	0.07	0.15	0.20	0.39	0.27	0.61
PES3	0.15	0.16	0.19	0.03	0.24	0.20	0.88
PES4	0.16	0.19	0.27	0.08	0.19	0.32	0.87
Eat	0.06	0.24	0.28	0.43	0.16	0.88	0.26
Exercise	0.16	0.27	0.34	0.43	0.29	0.90	0.27
BMI	0.04	0.28	0.26	0.46	0.24	0.83	0.19

References

mHealth apps market size, growth | analysis report, 2021-2028. [online] Available from: https://www.fortunebusinessinsights.com/mhealth-apps-market-102020. (2020, accessed

11 October 2023)

Better health smartphoneHarvard

apps.Health.

2015. [online] Available from: https://www.health.harvard.edu/staying-healthy/better-health-with-smartphone-apps. (2015, accessed 11 October 2023).

Ryan

Sawin

. The individual and family self-management theory: Background and perspectives on context, process, and outcomes. Nurs Outlook 2009; 57(4): 217–225.

Zhou

Jia

Skinner

, et al. Lessons on mobile apps for COVID-19 from China. J of Saf Sci and Resilience 2021; 2(2): 40–49. DOI: 10.1016/j.jnlssr.2021.04.002.

Yang

, et al. Changes in Mobile Health Apps Usage Before and After the COVID-19 Outbreak in China: semi longitudinal Survey. JMIR Public Health and Surveillance 2023; 9(1): e40552.

Shakery

Mehrabi

Khademian

. The effect of a smartphone application on women’s performance and health beliefs about breast self-examination: a quasi-experimental study. BMC Med Inform and Decis Mak 2021; 21(1): 1–10.

Whitley

Payán

Flórez

, et al. Feasibility and acceptability of a mobile messaging program within a church-based healthy living intervention for African Americans and Latinos. Health Inform J 2020; 26(2): 880–896.

Statistica . Share of health app users in selected countries as of 2022. https://www.statista.com/forecasts/1181436/share-of-health-app-users-by-country. (accessed 11 October 2023)

Deng

Liu

. Comparison of the middle-aged and older users’ adoption of mobile health services in China. International J of Med Inform 2014; 83(3): 210–224.

10.

Yang

Kovarik

. A systematic review of mobile health interventions in China: Identifying gaps in care. J Telemed Telecare 2021; 27(1): 3–22. DOI: 10.1177/1357633X19856746

11.

Chisholm

Hart

Lam

, et al. Current challenges of behavior change talk for medical professionals and trainees. Patient Educ and Counselling 2012; 87(3): 389–394.

12.

Walsh

Swangard

Davis

, et al. Exercise counseling by primary care physicians in the era of managed care. American J of PreV Med 1999; 16(4): 307–313.

13.

Xie

, et al. The Use of Mobile Health Applications to Improve Patient Experience: Cross-Sectional Study in Chinese Public Hospitals. JMIR mhealth and uhealth 2018; 6(5): e126.

14.

Hsu

, et al. The Top Chinese Mobile Health Apps: A Systematic Investigation. J Med Internet Res 2016; 18(8).

15.

Song

Chen

others . Evaluating Chinese mobile health apps for ankylosingspondylitis management: systematic app search. JMIR mHealth and uHealth 2021; 9(7): e27234.

16.

Hochbaum

. Public participation in medical screening programs: a sociopsychological study. 1958. Washington, DC: U. S. Government Printing Office.

17.

Rosenstock

Strecher

Becker

. Social learning theory and the healthbelief model. Health Educ Q 1988; 15(2): 175–183.

18.

Ryan

Deci

. Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychol 2000; 55(1): 68.

19.

Rosenstock

. The health belief model and preventive health behavior. Health Education Monog 1974; 2(4): 354–386.

20.

Zewdie

Mose

Sahle

, et al. The health belief model's ability to predict COVID-19 preventive behavior: A systematic review. Sage Open Med 2022; 10: 20503121221113668. DOI: 10.1177/20503121221113668.

21.

Tsai

Chen

, et al. Using the health belief model to explore nursing students’ relationships between COVID-19 knowledge, health beliefs, cues to action, self-efficacy, and behavioral intention: a cross-sectional survey study. Med. 2021; 100.

22.

National Cancer Institute . Constructs and measures for health behavior. https://cancercontrol.cancer.gov/brp/research/constructs (accessed 20 October 2022).

23.

Glanz

Bishop

. The role of behavioral science theory in the development and implementation of public health interventions. Annu Rev of Public Health 2012; 21: 299–418. DOI: 10.1146/annurev.publhealth.012809.103604

24.

Jones

Smith

Llewellyn

. Evaluating the effectiveness of health belief model interventions in improving adherence: a systematic review. Health Psych Review 2014; 8(3): 253–269.

25.

Orji

Vassileva

Mandryk

. Towards an effective health interventions design: an extension of the health belief model. Online J of Public Health Inform 2012; 4(3).

26.

Harrison

Mullen

Green

. A meta-analysis of studies of the health belief model with adults. Health Educ Res 1992; 7: 107–116.

27.

Carpenter

. A meta-analysis of the effectiveness of health belief model variables in predicting behavior. Health Comms 2010; 25(8): 661–669. DOI: 10.1080/10410236.2010.521906

28.

Bandura

. Self-efficacy: toward a unifying theory of behavioral change. Psychol Review 1977; 84(2): 191.

29.

Compeau

Higgins

. Computer self-efficacy: Development of a measure and initial test. MIS Q. 1995; 189–211.

30.

Champion

Ziner

Monahan

, et al. Development and psychometric testing of a breast cancer survivor self-efficacy scale. Oncology Nurs Forum 2013; 40: E403.

31.

Schwarzer

Jerusalem

. Generalized self-efficacy scale. In: Weinman

Wright

Johnston

(eds). Measures in health psychology: A user’s portfolio causal and control beliefs, Windsor: NFER-NELSON, 1995, p. 35-37.

32.

Didarloo

Nabilou

Khalkhali

. Psychosocial predictors of breast self-examination behavior among female students: an application of the health belief model using logistic regression. BMC Publ Health 2017; 17(1): 1–8.

33.

Heinrich

Cosgrove

Goodman

, et al. Age differences for relationships between perceived health, exercise motivation and self-efficacy factors after HIFT: 438 Board #254 May 27 10:30 AM - 12:00 PM. Med & Sci in Sports and Exercise 2020; 52(7S): 113–113.

34.

Balapour

Reychav

Sabherwal

, et al. Mobile technology identity and self-efficacy: Implications for the adoption of clinically supported mobile health apps. International J of Inform Mgmt 2019; 49: 58–68.

35.

Alharbi

Alghanmi

Fahlevi

. Adoption of Health Mobile Apps during the COVID-19 Lockdown: A Health Belief Model Approach. IJERPH 2022; 19.

36.

Vancouver

Thompson

Tischner

, et al. Two studies examining the negative effect of self-efficacy on performance. J of Applied Psychol 2002; 87(3): 506.

37.

Champion

. Revised susceptibility, benefits, and barriers scale for mammography screening. Research in Nurs & Health 1999; 22(4): 341–348.

38.

Graham

. Health beliefs and self breast examination in black women. J of Cultur Diversity 2002; 9(2): 49.

39.

Sallis

Owen

Fisher

. Ecological models of health behavior. Health Educ Behav 2008; 4: 465–486.

40.

Borland

H-H

, et al. Motivational factors predict quit attempts but not maintenance of smoking cessation: findings from the International Tobacco Control four country project. Nicotine and Tobacco Res, 2010.

41.

Haefner

Kirscht

. Motivational and behavioral effects of modifying health beliefs. Publ Health Rep 1970; 85(6): 478.

42.

Mullens

McCaul

Erickson

, et al. Coping after cancer: risk perceptions, worry, and health behaviors among colorectal cancer survivors. Psychooncology 2004; 13(6): 367–376.

43.

Janz

Becker

. The health belief model: A decade later. Health Educ Q 1984; 11(1): 1–47.

44.

Jones

Jensen

Scherr

, et al. The health belief model as an explanatory framework in communication research: exploring parallel, serial, and moderated mediation. Health Comm 2015; 30(6): 566–576.

45.

Becker

. The health belief model and sick role behavior. Health Educ Monogr 1974; 2(4): 409–419.

46.

Turan

Koç

. Health information technology adoption and acceptance of Turkish physicians-A model proposal and empirical assessment. Health Inform J 2022; 28(2): 14604582221096040.

47.

Fall

Izaute

Chakroun-Baggioni

. How can the health belief model and self-determination theory predict both influenza vaccination and vaccination intention? A longitudinal study among university students. Psych & Health 2018; 33(6): 746–764.

48.

Brislin

. Back-translation for cross-cultural research. J of Cross-cultur Psych 1970; 1(3): 185–216.

49.

Nunnally

Bernstein

. Psychometric theory. 3rd ed. New York: MacGraw-Hill, 1994.

50.

Fornell

Larcker

. Evaluating structural equation models with unobservable variables and measurement error. J of Mark Res 1981; 18(1): 39–50.

51.

Hair

Jr Babin barry

Anderson

, et al. Multivariate data analysis. 8th ed. Cengage Learning EMEA., 2018.

52.

Campbell

Fiske

. Convergent and discriminant validation by the multitrait-multimethod matrix. Psychol Bulletin 1959; 56(2): 81.

53.

Bentler

. Comparative fit indexes in structural models. Psychol Bulletin 1990; 107(2): 238.

54.

Browne

Cudeck

. Alternative ways of assessing model fit. Soc Methods & Res 1992; 21(2): 230–258.

55.

O’brien

. A caution regarding rules of thumb for variance inflation factors. Quality & Qty 2007; 41(5): 673–690.

56.

Williams

MacKinnon

. Resampling and distribution of the product methods for testing indirect effects in complex models. Struct Equ Model 2008; 15(1): 23–51.

57.

30% Chinese college students ‘unqualified’ in physical health: official data - global Times. https://www.globaltimes.cn/page/202104/1221944.shtml. (2021, accessed 11 October 2023)

58.

Kim

10 Mobile Apps That Promote Health and Wellness. CIO. Available from: https://www.cio.com/article/300264/10-mobile-apps-that-promote-health-and-wellness.html. (2012, accessed

11 October 2023)

59.

García-Gómez

de La Torre-Díez

Vicente

, et al. Analysis of mobile health applications for a broad spectrum of consumers: a user experience approach. Health Inform J 2014; 20(1): 74–84.

60.

Noman

Shahar

Abdul Rahman

, et al. The effectiveness of educational interventions on breast cancer screening uptake, knowledge, and beliefs among women: a systematic review. International J of Envir Res and Public Health 2021; 18(1): 263.

61.

Fisher

Kridli

SAO

. The role of motivation and self-efficacy on the practice of health promotion behaviours in the overweight and obese middle-aged American women. International J of Nurs Pract 2014; 20(3): 327–335.

Health perceptions and attitudes on mobile health apps in China

Abstract

Keywords

Introduction

Health belief model

Mobile health apps

Health belief theory

Research model

Research methods

Measures

Results

Descriptive statistics

Analysis

Structural model

Analysis

Mediation test

Discussion

Conclusion

Limitations

Footnotes

Acknowledgments

Author contributions

Declaration of conflicting interests

Funding

Ethical Statement

ORCID iD

Appendix 1

Appendix 2

References