This item has the following downloads:
AGE RELATED DIFFERENCES IN SELF EFFICACY AND THE USE OF E HEALTH SUPPORTS FOR EXERCISE BEHAVIOR IN ADULTS By ASHLEY L. REYNOLDS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 3
2 201 3 Ashley L. Reynolds
3 I de dicate this dissertation to my h usband, mother and father, family friends, colleagues and mentors.
4 ACKNOWLEDGEMENTS I want to thank Drs. Roberts, Chen, Weber, Sutton and Garva n for their unwavering support. They were instrumental to my success and I would not have finished without their expert guidance, confidence, feedback and mentoring. I would also like to thank Martin Bates, without whom, my research would not have gone forward. Next I would like to thank my family for their un derstanding and support during this process. Finally and most importantly, I want to thank my husband for his support, understanding and encouragement. I could not have done this without him.
5 TABLE OF CONTENTS page ACKNOWLEDGEMENTS ................................ ................................ ............................... 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 AGE RELATED DIFFERENCES IN SELF EFFICACY AND USE OF E HEALTH SUPPORTS FOR EXERCIS E BEHAVIOR IN ADULTS ................................ ......... 12 Problem and Significance ................................ ................................ ....................... 12 Conceptual Framework ................................ ................................ ........................... 15 Self Efficacy ................................ ................................ ................................ ..... 15 Mastery Experiences ................................ ................................ ........................ 15 Vicarious Learning ................................ ................................ ............................ 17 Verbal Persuasion ................................ ................................ ............................ 17 Physical and Emotional States ................................ ................................ ......... 18 Outcome Expectations ................................ ................................ ..................... 19 Facilitation ................................ ................................ ................................ ........ 20 Self Regulation ................................ ................................ ................................ 20 Background ................................ ................................ ................................ ............. 21 Purpose of the Study ................................ ................................ .............................. 23 Specific Aim s ................................ ................................ ................................ .......... 23 Specific Aim 1 ................................ ................................ ................................ ... 23 Specific Aim 2 ................................ ................................ ................................ ... 23 Specific Aim 3 ................................ ................................ ................................ ... 23 Summary ................................ ................................ ................................ ................ 24 Definition of Key Terms ................................ ................................ ........................... 25 2 REVIEW OF RELATED LI TERATURE ................................ ................................ ... 28 Social Cognitive Concepts and Exercise ................................ ................................ 28 E Health Resources and Age related Differences ................................ .................. 31 Age Cohorts ................................ ................................ ................................ ..... 31 Age related Differences Affecting Technology Use ................................ .......... 32 Physical ................................ ................................ ................................ ............ 33 Environmental ................................ ................................ ................................ .. 34 Social ................................ ................................ ................................ ................ 35 Patterns of Adoption by Ad ults ................................ ................................ ......... 37 Internet and Information ................................ ................................ ................... 38 Age related Differences in Online Resource Utilization ................................ .... 39
6 Online Resources and Exercise ................................ ................................ .............. 40 So cial Cognitive Concepts and Online Resources ................................ ........... 41 Self Monitoring & Exercise ................................ ................................ ............... 41 Social Networking in Health & Exercise ................................ ........................... 43 Age related Differences in Social Network Use ................................ ................ 44 Social Cognitive Concepts and Social Networks ................................ .............. 45 Electronic Games ................................ ................................ ................................ ... 47 Age related Differences in Video Game Use ................................ .................... 49 Experiential Timing ................................ ................................ .................... 49 Physical Changes ................................ ................................ ...................... 50 Exergames, Health and Exercise ................................ ................................ ..... 51 Social Cognitive Concepts and Exertive Gaming ................................ ............. 54 Exerga mes and Barriers to Exercise ................................ ................................ 54 Exergames and Self efficacy ................................ ................................ ............ 55 Summary ................................ ................................ ................................ ................ 56 3 RESEARCH DESIGN AND ME THODOLOGY ................................ ....................... 58 Design ................................ ................................ ................................ ..................... 58 Sample ................................ ................................ ................................ .................... 58 Power Analysis ................................ ................................ ................................ 59 Recruitment ................................ ................................ ................................ ...... 60 Measurement ................................ ................................ ................................ .......... 60 Characterization of participants ................................ ................................ ........ 61 Use of e Health Supports ................................ ................................ ................. 62 Self efficacy for e Health Support Use ................................ ............................. 62 Self Efficacy for Exercise ................................ ................................ .................. 63 Outcome Expectations for Exercise ................................ ................................ 64 Barriers to Exercise ................................ ................................ .......................... 65 Exercise Participation ................................ ................................ ....................... 65 Procedures ................................ ................................ ................................ ............. 66 Data Analysis ................................ ................................ ................................ .......... 6 7 Human Subjects ................................ ................................ ............................... 69 4 RESULTS ................................ ................................ ................................ ............... 71 Sample ................................ ................................ ................................ .................... 71 Millennials ................................ ................................ ................................ ......... 72 Generation X ................................ ................................ ................................ .... 72 Baby Boomers ................................ ................................ ................................ .. 73 Silent Generation ................................ ................................ .............................. 74 Preliminary Analyses ................................ ................................ .............................. 75 Differences in e Health Support Use (Research Question 1) ........................... 77 Relationship of e Health Use to Self efficacy (Research Question 2) .............. 77 Relationship of e Health Use for Exercise and Social Cognitive Exercise Variables (Research Question 3) ................................ ................................ .. 77
7 5 DISCUSSION ................................ ................................ ................................ ......... 87 Individual e Health Tool Differences ................................ ................................ ....... 88 Internet ................................ ................................ ................................ ............. 88 Social Networks ................................ ................................ ................................ 89 Video Games ................................ ................................ ................................ .... 90 Self Monitoring ................................ ................................ ................................ 90 Patterns of Use ................................ ................................ ................................ ....... 91 Overall e Health Differences Among Generations ................................ .................. 92 Differences in Social Cognitive Factors Among Cohorts ................................ ........ 93 Self Efficacy for e Health ................................ ................................ .................. 93 Exercise ................................ ................................ ................................ ............ 94 Relationships Between Social Cognitive Factors and e Health ........................ 94 Limitations ................................ ................................ ................................ ........ 95 Clinical Implications ................................ ................................ ................................ 97 Research I mplications ................................ ................................ ........................... 101 Conclusion ................................ ................................ ................................ ............ 102 APPENDI X A SURVEY ................................ ................................ ................................ ............... 105 B RECRUITMENT LETTER ................................ ................................ ..................... 115 REFERENCES ................................ ................................ ................................ ............ 116 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 129
8 LIST OF TABLES Table page 4 1 Age Cohort Demographics ................................ ................................ ................. 79 4 2 e Health tool Use by Cohort ................................ ................................ ............... 81 4 3 Social Cognitive Factors in e Health and Exercise ................................ ............. 83 4 4 Age Differences in e Health Use ................................ ................................ ........ 85
9 LIST OF FIGURES Figur e page 1 1 Theoretical Framework for Study ................................ ................................ ........ 27 4 1 Survey flowchart ................................ ................................ ................................ 86
10 Abstract of Dissertation Presented t o the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy AGE RELATED DIFFERENCES IN SELF EFFICACY AND THE USE OF E HEALTH SUPPORTS FOR EXERCISE BEHAVIOR IN ADULTS By Ashley L. Reynolds August 2013 Chair: Beverly L Roberts Major: Nursing Science s E Health tools show great promise to promote health because of their broad reach, consistent interventions and reduced cost. L ittle is still known about the ways in which the public uses these tools, or their perceptions and beliefs about their use. T h e purpose of this cross sectional study was to describe the electronic health supports used by different age groups, compare efficacy beliefs regarding these tools and t o explore the relationship betwee n e Health support use and social cognitive concepts. A mailed survey ( n = 1,868) was used to gather data from a random sample of participants in a national employer sponsored wellness program. The age cohorts were Millennials, Generation X, Baby Boomers and Silent Generation. The total weekly minutes of all e Health tools and e Health tools to support exercise were significantly different among the cohorts ( X 2 (3) = 44.49, p < .0001) O verall as age increased the use of e Health tools declined Post hoc comparisons among the cohorts revealed that Millennials utilized e Health tools more frequently than their older counterparts Health tools was significantly higher than their older counterparts except for Generation X. In addition, the greater number of weekly minutes using e Health tools was significantly related to greater self efficacy for e
11 Health tool use ( r s = .50, p = .01 ) w hile greater e Health tool use for exercise was significantly related to greater exercise self efficacy ( r s = .21, p = .0003) and outcome expectations for exercise ( r s = .28, p < .0001) it was not significantly associated with perceived barriers for exercise ( r s = .05). These finding support the existence of a ge related differences in the use of e Health tools to support exercise and the relationship between social cognitive concepts and e Health tools C onsideration of these differences can gu ide e Health tool use in different aged groups. E Health tools should incorporate efficacy building mechanisms and p ractitioners should assess e Health self efficacy prior to using e Health intervention s Future e Health research should explore experiential timing as a factor in e Health tool adoption, video game use by older adults, tool use for behaviors other than exercise and the effect of multi component e Health tools on health behavior.
12 CHAPTER 1 AGE RELATED DIFFERENCES IN SELF EFFICACY AND USE OF E HEALTH SUPPORTS FOR EXERCISE BEHAVIOR IN ADULTS Problem and Significance The United States is on the verge of a health care catastrophe. With healthcare costs spiraling out of control, increased demand and lack of access have put a heavy strain on an already overburdened system. Genetic predisposition and environmental factor s notwithstanding, most leading causes of death in the United States posses a lifestyle or behavioral component (Centers for Disease Control [CDC] 2012). Chronic conditions such as obesity, heart disease, hypertension, stroke, some cancers and diabetes a re directly related to a sedentary lifestyle (Centers for Disease Control [CDC], 2009; Chodzko Zajko, Proctor, Singh, Minson, Nigg, & Salem, 2009; Millen & Bray, 2009 ; Nelson et al., 2007). According to one study, half of all annual deaths in the United States can be attributed to preventable causes that are facilitated by poor health be haviors (Mokdad et al., 2004). Implementation of primary preventive measures such as routine exercise can dramatically impact the development of diseases and even improve glucose control in diabetics ( Resnick, Luisi, Vogel, & Junaleepa, 2004 ; Song, Ahn, Roberts, Lee, & Ahn, 2009). In addition to impacting disease development and progression, the CDC (2006) reported that medical costs decrease dramatically a s physical activ ity increases. The American College of Sports Medicine (ACSM) states that lack of exercise in adults contributes to poor health, including the development of preventable illness and functional decline (Chodzko Zajko et al., 2009). In addition, the CDC (200 6) reported that over time, participation in regular physical activity declines with age. Recognizing the importance of exercise participation to personal health, both the ACSM and CDC
13 have issued guidelines for exercise in an effort to decrease mortality and reduce the development of preventable chronic conditions like diabetes and heart disease (CDC, 2006; Chodzko Zajko et al., 2009 ). Technology is ubiquitous and pervasive and the numbers are staggering. For example, the number of mo bile devices will outnumber global population by the end of 2012 (Institute of Engineering and Technology, 2011). As of May 2012, t here were 900 million users of the online social network Facebook with numbers expected to exceed 1 billion by the end of the year (Mashable, 2 012). Furthermore, of the 46% of adults who used the site to memorialize someone who suffered from a condition and 7% of them have obtained health information there (Fox 2013 ). Self monitoring tool use is also increasing, with 27% of adult Internet users tracking their health data online. Finally, there are currently 40,000 mobile health applications available and 247 million individuals have downloaded at least one to the ir mobile device (Jahns & Gair, 2012). Electronic Health (e health ) tools are proposed to extend the reach and efficacy of behavior change interventions across large populations with emerging data supporting their effectiveness ( Fukuoka, Kamitani, Bonnet & Lindgren, 2011; Noar & Harrington, 2012 ). The term e delineates a burgeoning field of study encompassing multiple disciplines (Office of Disease Prevention and Health Promotion, 2006, p.xi) Moreover, e Health supports are the specific digital tools used to support health behavior (U.S. Department of Health and Human Services, 2012). Further study is needed to provid e theoretical support for the
14 efficacy of e H ealth support use among different aged populations as well as expanding current knowledge of the ideal ways in which these tools should be constructed, tested and deployed. E Health tools offer various ways by which an individual can initiate, engage, track and maintain healthy behaviors, including exercise. Interest is mounting to assess the utility and efficacy of electronic devices and supports as means to curb diseases both nationally and ab road (World Health Organization, 2011). Although the use of e H ealth is expanding, little evidence exists supporting the efficacy of these tools in driving sustained behavior change. Furthermore, little is known about the patterns of adoption for e H ealt h supports among different age groups, the reasons why different groups use these supports and whether certain supports are preferred by specific ages. Finally, currently nothing is known about efficacy beliefs for the use of these tools among different a ged adults. Exercise participation and social cogn itive constructs have been well studied among different age groups; though few of these studies include the use of electronic tools to support healthy exercise behaviors. E Health support use may be effic acy building and warrant further study to determine impact on exercise participation. Considering the rapidity with which these emerging technologies are being adopted, and the inherent benefits of exercise participation, understanding how these supports a re used by different aged participants can aid in development e Health tools and interventions to improve health through exercise. E Health tools are being used more frequently to manage disease, which impacts older adults more frequently, thus
15 understandi ng age related differences will aid in delivering these tools appropriately (Topol, 2012) Conceptual Framework Electronic health tools offer ways in which an individual can interact with and receive various types of health related information, thereby he lping form beliefs about specific types of health behaviors. Social Cognitive Theory (SCT) is deeply rooted in the study of human learning and behavior. A central concept of the SCT is reciprocal determinism. H umans are in a constant state of interaction with their environment and are affected by it, while at the same time affecting the en vironment around them (Bandura, 1997; Glanz, Rimer, & Viswanath, 2008 ; Pajares, 2002) Using e H ealth supports can help build positive self maintain healthy exercise behaviors. Self Efficacy Bandura (1997) defines self ceived self efficacy is concerned not with the number of skills that you have, but with what you believe you can do with what you have a person is in their ability to complete a tas k, given a variety of stressors. Based on SCT, self efficacy can be influenced in different ways through the use e Health supports. Mastery Experiences Of the four mechanisms to increase self efficacy mastery experiences provide the greatest influence on personal efficacy beliefs (Bandura, 1997) As some experience continued successes with a task, their confidence in their ability to perform increases. Conversely, repeated failures can decrease personal efficacy and will be
16 detrimental to positive perfo rmance. If failures are experienced early in the behavior change process, they will have a greater negative impact than if the individual e xperiences a setback after some successes. Since mastery experiences provide capabilities, they are the most influential on self efficacy beliefs. Mastery experiences affect performance through cognitive appraisal of that experience (Bandura, 1997). Thus, perception of the effort required completing the task, along with judgme nts about the ease or complexity of the task and related circumstances would greatly influence how self efficacy will be affected. For example, success with a task that a person judges to be easy in nature will have less of a positive influence on efficac y belief than success with a complicated or difficult behavior. Moreover, the amount of effort expended or assistance received in performance of the task can influence the degree to which efficacy beliefs are enhanced. Electronic health supports are uniq uely positioned to provide mastery experiences for exercise. For example, many gaming platforms utilize progressive learning techniques that allow the user to move through increasingly difficult stages. As the user becomes more proficient and complete s e arlier levels, they are presented with more challenging tasks. In this way, electronic games provide opportunities to master tasks, thereby building efficacy beliefs. In another example, websites where users track progress are able to show the user how f ar they have progressed towards a goal. As they continue the healthy activity, in this case exercise, they are able to visualize their successes on the screen, which reinforces their beliefs that they are capable of continued success.
17 Vicarious Learning Another key component of the SCT is the concept of vicarious learning. and are an effective way to influence self efficacy. H umans tend to emulate those who appear most like us. The more a person observes another similar person completing a task or behaving a certain way, the more likely they are to attempt to repea t that behavior (Bandura, 1986, 1997). With the advent of online social networks, individuals can now interac t with others more easily than ever before. For example, Facebook has over 1 b illion users so far and hundreds of different groups related to health and exercise. In addition, many health specific websites have incorporated social networking as a way to allow like minded individuals the opportunity to share and interact with one another while attempting a common goal. The Livestrong website allows members to join different lose failures as well as provide each other support and encouragement. Members can also share their progress with others, providing support for the efficacy beliefs of other members who are working towards the same goal. Verbal Persuasion Feedback in the form of verbal communications also has an impact on self efficacy beliefs (Bandura, 1997). Affirmative statements given to individuals regarding their capabilities support their beliefs that they possess the characteristics needed to be successful in their endeavors. Individuals receiving verbal persuasions regarding their capabilities evaluate the persuasory statement against their beliefs and prior
18 experiences with the tar get action. Therefore the effectiveness of verbal persuasions is linked to whether the individual believes the person making the statements and whether they believe that they can actually produce the required outcome. In addition to social networks impac ting efficacy beliefs though vicarious learning, they can also be the source for statements of encouragement from other users. As an individual interacts with others in social networks who share the same goal of improving exercise, they can receive positi ve supporting statement s from other participants. In successful. These positive supportive statements can improve efficacy beliefs related to exercise participation. P hysical and Emotional States The fourth and final way to influence self efficacy beliefs is through the cognitive appraisal of somatic and affective arousal ( Bandura, 1997). These physical and ributions concerning the sources of the arousal, perceptions of the arousal and the efficaciousness of the individual experiencing the arousal. The influences of somatic indicators are particularly robust when related to health related tasks, coping strate gies and physical activities (Bandura, 1997). Social networks provide opportunities for ongoing support when individuals are experiencing somatic complaints or emotional states. Other members within the group provide positive supportive statements as w ell as share their own experiences and how they disregarded negative states in order to achieve success. Furthermore, many exercise based games today provide positive supporting statements while completing the exercise, especially during the most challeng ing portions.
19 Outcome Expectations One of the principal psychological determinants of health behavior is outcome expectations. Expecting a positive result from physical ac tivity has been shown to relate to exercise participation in many studies. Outcome expectations describe how a person feels or believe about a particular consequence of a behavior and the value the individual places on that outcome (Bandura, 1997; Glanz et al., 2008 ; Pajares, 2002). In addition to individual value judgments about the influenced by the belief that others will view the behavior in a certain way (Bandura, 1997 2006a ; Glanz et al. 2008). Finally, the individual also can be influenced by the expected sense of personal satisfaction ach ieved when successfully completing the i ntended behavior (Bandura, 1997, 2006a; Glanz et al. 2008). more likely they will engage in and be successf ul completing it (Bandura, 19 86, 1997). For example, if a person believes that their social support view s their participation in a particular behavior in a positive way, then the individual will be more likely to engage in it. Also, one may believe that maintaining physical activity levels does not provide any real benefit; therefore they would be less likely to engage in it routinely. Use of e H several different ways. First, as they interact with others who shar e common values about exercise within in a social networking environment, they are engaging others who may view exercise as important, thereby positively influencing their outcome expectation for exercise. Secondly, they are able to see positive results of exercise from other participants within their social group, thereby influencing their own expectations about results they will receive as well.
20 Facilitation The environmental determinants of behavior include incentive motivations and facilitation ( Bandur a, 1986, 1997 ; Glanz et al. 2008 ). Incentive motivations are external rewards or punishments used to direct behavior (Bandura, 1997) For example, providing punishment in the form of laws can have unintended consequences, such as increased smoking in teen s rather than a less punitive method such as increasing cigarette costs, which were effective at reducing smoking in teens (Glanz et al. 2008). In addition, facilitation affects behavior through the increase in available resources needed to change behavi or (Glanz et al. 2008). For example, facilitation can be achieved through effective identification and removal of barriers that prevent one from making a change (Glanz et al. 2008). As perceived barriers decrease, the likelihood of engaging in a behavi or increases. Self Regulation Within the SCT, a person can direct changes in behavior based on the rewards and changes to the environment they provide for themselves (Glanz et al. 2008). This is called self regulation and is achieved through the followi ng six methods: Self monitoring, goal setting, feedback, self reward, self instruction and enlisting social supports (Bandura, 1997; 1986; Glanz et al. 2008). E Health support use provides ways for the individual to utilize various self regulatory mechanisms to improve health. Many games, mobile devices and directive websites provide opportunities to set health related goals as well as providing ongoing feedbac k on goal attainment. In addition, these sites provide rewards in the form of badges or additional functionality for achieving certain milestones. Also, health related websites often provide self directed learning opportunities to improve health literacy As
21 explained earlier, social networks provide support for individuals who are seeking to change behaviors. Background The relationship between exercise and health outcomes has been well studied with similar results pointing to increased exercise and improved health across adult age groups (Anderson Bill, Winett, Wojcik & Winett, 2011 b ; Conn, Hafdahl, Brown & Brown, 2008; Dunlap & Barry, 1999; Elavsky, McAuley, Motl, Konopack, Marquez et al 2005; Telama, Yang, Viikari, Valimaki, Wanne et al 2005). Also well established is how a ge group exercise patterns vary. F or instance, exercise participation declines with age, even though evidence suggests continued participation improves quality of life, strength, balance, and cognition while promoting indepe ndence ( Elavsky et al., 2008 ; Gerling, Schild & Masuch, 2010; Telama et al., 1997 ). Social cognitive theory has been utilized frequently as a framework to explain various health behaviors, including exercise behavior in adults ( McAuley, Jerome, Marquez, Elavsky & Blissmer, 2003; McAuley, Morris, Motl, Hu, Konopack et al., 2007; Rejeski, Tian, Liao & McDermott, 2008 Resnick & Nigg, 2003). While self efficacy is most often mentioned as a significant predictor of exercise behavior (McAuley et al 2003 ; Mill en & Bray, 2009; Neupert, Lachman & Whitbourne, 2009), other variables such as outcome expectations ( Anderson Bill, Winett & Wojcik, 2011 a ; Hortz & Petosa, 2008; Millen & Bray, 2009; Resnick, Luisi, Vogel & Junaleepa, 2004) and perceived barriers to exerci se ( Daskapan, Tuzun & Eker, 2006; Dunlap & Barry, 1999; Lees, Clark, Nigg & Newman, 2005; Leveille, Cohen Mansfield & Guralnik, 2003) have also been demonstrated to be influential across age groups Whether self efficacy influences exercise to a greater d egree than other social cognitive variables is a matter
22 of some debate Resnick et al. (2004) found that outcome expectations were more influential than self efficacy in predicting exercise behavior. While evidence of self r has been well documented, no such relationship has been studied between self efficacy and the use of e H ealth supports. There are a ge related differences in the use of various types of technology such as: Internet use (Chu, Mastel Smith, 2010; Cohall, Nye, Moon Howard, Kukafka, Dye et al., 2011; McInnes, Gifford, Kazis, & Wagner, 2010; McMillan & Macias, 2008), online health programs (Bickmore, Caruso, Clough Gorr & Heeren, 2005), social networks (Fukuoka, Bonnet, Lindgren, 2011; Nahm, Resnick & Gaines, 2004), video games ( Gerling, Schild & Masuch, 2010 ; Nitz, Kuys, Isles & Fu, 2010; Pearce, 2008) and self monitoring devices ( Beaudin, Intille & Morris, 2006 ; Svensson & Lagerros, 2010). While knowledge exists regarding the use of these tools for health related activities by different age groups, too little is known about their use to support exercise behaviors and whether a ge related preferences exist in their utilization. Some suggest how e health support use may influence social cognitive variables such as self efficacy ( Anderson Bill et al 2011 a; Bandura, 2004; So ng, Peng & Lee, 2011 ), outco me expectations (Anderson Bill et al 20 1 1 a ; Band ura, 2004) and barriers to exercise ( Boschman, 2010 ; Center for Technology and Aging, 2011 ). Little is known about how these technologies support social cognitive influences of exercise behavior among adults. Furthermore, a significant gap exists in the ex ploration of a ge related differences between the influence of e health support use and social cognitive influencers of exercise.
23 Purpose of the Study There is little empirical data to support understanding efficacy beliefs regarding the use of e H ealth to ols by different age groups and whether they are used to guide exercise participation. This research was designed to build on existing knowledge about the use of e health tools, self efficacy for their use and exercise behavior across age groups. This purpose of this study was to describe the various electronic health supports used by different age groups, compare ef ficacy beliefs regarding these tools and to explore the relationship betwee n e Health support use and social cognitive concepts Figure 1 1 represents the theoretical framework for this study with accompanying relationships depicted. Specific Aims Specifi c Aim 1 1. To describe age related differences in the frequency of e health support use. a. H ypothesis 1: E health support users of different ages will demonstrate differences in the frequency of e health supports used. Specific Aim 2 1. To determine if self effi cacy for e health support use is related to increased utilization of these resources. a. Hypothesis 2: A higher level of self efficacy for e health support use is related to higher levels of utilization Specific Aim 3 1. To determine if e health support use for exercise increases self efficacy for exercise, outcomes expecta tions for exercise and decreases perceived barriers for exercise when used by adults? a. Hypothesis 3: E health support use will be related to greater self efficacy for exercise b. Hypothesis 4: E health support use will be related to greater outcome expectations for exercise
24 c. Hypothesis 5: E health support use will be related to decreased perceived barriers for exercise. Summary The proliferation of various diseases largely attributable to l ifestyle influences such as lack of exercise remains an important problem in the United States. With two thirds of the entire U.S. population overweight or obese (CDC, 2011), it is becoming imperative to find solutions to address obesity and its associate d sequelae. Behaviors such as poor eating habits and inactivity contribute to the epidemic of obesity and subsequent chronic disease and can often be treated or mitigated through healthy lifestyle changes. Simply maintaining a routine exercise program can drastically improve morbidity and mortality. Unfortunately, many do not maintain activity levels needed to prevent illness and promote health with significant decreases in exercise as we age. The benefits associated with routine exercise have been wide ly studied across populations with several theoretical perspectives demonstrating associations with sustained engagement in exercise behavior Self efficacy has also been examined using various theoretical perspectives and is overwhelmingly influential in driving actual behavior. The more efficacious the individual, the more likely they are to engage in the activity about which they feel most confident. In addition, perceived barriers, goals and expected outcomes have been repeatedly shown to influence e xercise. Electronic media use has grown with the ubiquity of home computer s gaming consoles, broadband and wireless Internet connections as well as mobile devices. Unfortunately the problem of sedentary lifestyles has often been attributed to various fo rms of media consumption since they historically were torpid activities. Now with social exercise networks, exercise gaming, online fitness programs and self monitoring
25 tools, the use of electronic media may be transformed into a health benefit rather than a risk. T his research is intended to expand current understanding of self efficacy as described in social cognitive theory in relation to e health support use, how different age groups use these technologies and how their use relates to known social cognitive influencers of exercise behavior. Definition of Key Terms Exercise P hysical activity is any skeletal muscle driven movement that inc reases energy expenditure By contrast exercise is defined as planned purposeful movements conducted with the intent of improving or maintaining physical fitness. Moderate activity is defined as purposeful movement with increased effort and can include ac tivities such as brisk walking, swimming or light cycling while vigorous exercise includes jogging, cycling at a faster pace and cardio respiratory fitness ( Thompson, Buchner, Pina, Balady & Williams et al 2003). Self Efficacy for Exercise A personal be engage in routine exercise (Bandura, 2004; White et al. 2011). Outcome Expectations for Exercise participating in exercise and its relative importance (Millen & B ray, 2009; Resnick & Nigg, 2003; Umstattd, & Hallam, 2007). Barriers for Exercise Exercise barriers are defined as perceived feelings, environments, conditions or situations that inhibit participation in regular exercise ( Resnick et al., 2004 ). e Health S upports. Forms of electronic media that provide information, directions, connections with others or entertainment and are categorized into the following groups based on how they are used support exercise.
26 Internet programs. Structured programs with specific instructions to follow in order to achieve a fitness goal and accessed through the web or mobile device. Social Networks Websites, programs or mobile applications where individuals can communicate, share information and connect with others. Self Monitoring. Methods for tracking various exercise behaviors through the use of pedometers, GPS mobile devices and fitness applications which gather exercise information for display and/or storage by the user. Exergaming /Exertive games Mobile, compute r or console based electronic games where users must use exertive interfaces resulting in increased energy expenditure above normal resting energy expenditure. Self Efficacy for e Health Supports Self efficacy, in general, is a personal belief or confide apabilities to complete a particular behavior (Bandura, 1997). Thus, self efficacy for e utilize these tools to support exercise behavior.
27 Figure 1 1 Theoretical Fram ework for Study
28 CHAPTER 2 REVIEW OF RELATED LITERATURE Social Cognitive Concepts and Exercise The literature review begins with an overview of how social cognitive concepts influence various health promoting behaviors, including exercise behavior. Next, the focus turn s to e health support patterns of adoption among different age groups and how they can influence behavior. Fin ally, e H ealth supports are explored for clues on their potential influence on self efficacy. Social cognitive concepts and their relationship have been well studied in the context of exercise, with most focusing frequently on exercise self efficacy E xercise self efficacy significantly increases exercise behaviors in adults of all ages (McAuley et al., 2007; McAuley et al 2003; Rejeski et al 2008; Resnick & Nigg, 2003). In persons with peripheral artery disea s e exercise was explained by exercise self efficacy and was increased when pain acceptance, satisfaction with function, perceived control and desire for functional competence were added (Rejeski et al 2008 ). In a randomized trial of strength training Neupert, Lachmnan and Whitbour n e (2009) found that exercise self efficacy beliefs were significantly associated with continued participation in strength training for 9 to 12 months post intervention. Thus, increased exercise self efficacy positively influences exercise behavior. The influence of self efficacy on exercise is not without some debate. Hortz and Petosa (2008) examined a self efficacy the mediating effects of S ocial C ognitive T heory (SCT) variables on moderate exercise participation. Social cognitive factors included self regulation, social situation, outcome expectancy values and self efficacy. The self efficacy intervention included instruction on identification and
29 overcoming of barriers to exercise but did not include interventions based on mastery experiences, social modeling, improving physical/emotion al states and verbal persuasion, which are efficacy increasing components in Social Cognitive Theory ( Bandura, 1997 ; Glanz, Rimer, & Viswanath, 2008). While the intervention did not improve self eff icacy and exercise behavior, improvement s in self regulation and social situations were significant and were in included in subsequent meditational analysis. So bel and Goodman pathway tests revealed that social situation and self regulation mediated the ef fects of the intervention on moderate intensity exercise In addition to self efficacy, outcome expectations for exercise have been linked to exercise participation A related to maintenance of exercise behaviors though its influence is mixed (Anderson Bill et al 2011a; Marks & Allegrante, 2005; McAuley et al., 2003 2007 ; Resnick et al., 2007; Wh ite, Wojcicki & McAuley, 2011) For instance, in a cross sectional study, structural equation mode ling revealed significant but small negative direct effects of outcome expectations on physical activity ( direct = .14, p = .02), w hich was somewhat offset by a small, positive indirect effect ( indirect = .03, p = .06) and significantly partially mediated ( total = .11, p = .007) through outcome expectations on self regulation ( Ander son Bill et al 2011 a ). B y comparison, significant relationships between physical outcome expectations and physical activity ( = .23, p < .05) have been found. In addition more efficacious individuals have significantly higher physical outcome expectations ( = .22), s elf evaluative outcome expecta tions ( = .26), social outcome expectations ( = .19), goals ( = .27), fewer disability limitati ons ( = .26), and par ticipate in greater levels of physical activity ( = .27) ( White et al 2011 ).
30 Barriers to exercise have been identified as influencing exercise behavior in young adults (Daskapan, Tuzun & Eker, 2006), adults (Murray, Rodgers & Fras er, 2011) and older adults (Dunlap & Barry, 1999; Resnick, Luisi & Vogel; 2008; Zizzi et al., 2006). Perceived exercise barriers can take different forms and are often dependent on age, cognitive ability, disease burden and environment. Exercise barriers i n y ounger adults include a lack of energy, motivation, self confidence, resources, soc ial support and time (Daskapan et al 2006) while pain, fear of injury, social isolation, weather and safe environment are barriers for older adults (Leveille, Cohen Man sfield & Guralnik, 2003). According to Neupert, Lachman and Whitbourne (2009 ), environmental barriers and exercise related barriers are important considerations in older population s because age related changes like limitations in mobility or cognition can impact the ability to exercise safely In one cross sectional study of Canadian adults personal constraints and exercise scheduling self efficacy mediated the relationship between exercise behavior and subjective socioeconomic status and income ( Murra y, Rodgers & Fraser, 2011) In addition, while not havin g enough time to exercise is a significant barrier in other studies (Daskapan et al 2006) it was insignificant in Appalachian older adults ( Zizzi et al., 2006 ). This inconsistency may be accounte d for due to the differences in the ages of the study populations and that fact that one study included college students who may perceive greater demands on their time while engaged in an academic program
31 E Health Resources and Age related Differences Age Cohorts The lexicon of social science commonly includes the def inition of different age cohorts as generational groups Labeling groups with similar characteristics such as age is common in soc ial research, especially when attempting to discern differences or make comparisons though care does. Using the term age cohort or age related group communicates th e logical grouping by age without ambiguity. Birth cohorts possess characteristics unique to their group, which are often a product of shared experiences, or social and environmental lar to others within their group and that these experiences shape how a particular cohort views their world. It is often these cluster characteristics that are of interest to science. Cohort size can impact individual perspectives (Ryder, 1965). For examp le, baby boomers outnumbered previous cohorts so they had different experiences finding housing, employment and class sizes. Unfortunately, classifying age groups with discrete age ranges oversimplifies the heterogeneity of any cohort. While larger socia l or cultural experiences influence the beliefs and perspectives of a cohort, individual differentiations between gender, faith, education experience, income and status cannot be overlooked. Indeed, social researchers agree that cohort effects along with life cycle effects and period effects all influence how individual beliefs and perspectives are shaped within a cohort (Pew Research Center, 2010). In addition, while there may be very little difference between a 30 year old Generation X and a 29 year old Millennial some age cutoff is necessary in order to group by age. As long these individual variant
32 influences are understood in any cohort, valuable information may still be gained by using age groups in research. Despite several limitations using discrete age groups to compare cohorts in social science research, there are numerous benefits that make these classifications worthwhile. Using commonly accepted nomenclature to group and define age cohorts allows for co mparisons across differing studies, thereby creating a more comprehensive acceptance of naming and grouping of different cohorts can add to the broader understanding of thes e groups beyond the scientific community. For purposes of this Age related Differences Affecting Technology Use The use of e health tools represents an opportunity to support the health behaviors of older adults, thereby improving quality of life and perhaps extend ing periods of independence, though little has been done to empirically test this. Older adults are using e Hea lth tools with increasing frequency ( Entertainment Software Association, 2011; Fox, 2010; Gerling et al 2010 ; Pew Research Center, 2010 ) and over half adults aged 65 or older would be willing to use wireless health monitoring devices in their home ( Barre tt 2011). Nevertheless, even though older consumer demand exists developers must consider age related changes in the creation of e Health tools so they can be more easily adopted. The ensuing discussion of age related changes will focus on those normal a lterations likely to affect interaction with e Health tools and is not meant to imply that they are a n obligatory component of aging Indeed,
33 distinction should be made betwee n normal aging patterns and pathological ones. Physical Physical frailty in aging is associated with increased weakness, loss of endurance, mobility limitations and loss of balance (Ham, Sloane, Warshaw, Bernard & Flaherty, 2007). As physical activity declines and muscle disuse sets in, significant losses of strength occur. Loss of range of motion coupled with decreased bone mass and loss of balance increase both the frequency and severity of falls in older adult s. In fact, fear of falling is a primary concern in older adults and moderates feelings of self efficacy and activity levels in this population (Deshpande et al., 2008). Musculoskeletal changes also include osteoarthritic changes limiting range of motion and increasing pain upon activity. Routine physical activity declines as we age (Nelson et al., 2007), though p hysical limitations can be improved with routine exercise thereby reducing the risk of falls through improved gait, strength and stability (Ha m et al 2007) Age related changes in vision impact how older adults interact with their environment. These changes are of great importance to e Health research since many of these tools are visually based. As we age, the lens of the eye loses its abil ity to accommodate or change shape so that moving a focal point from far to near or vice versa takes longer or requires corrective lenses ( Ham et al 2007) Age related ocular structural changes (changes in lens shape and size of the iris) limit the amoun t of light entering the eye thereby decreasing visual acuity. In addition, retinal neural attrition decreases the amount of visual information received and processed by the brain (Morrell, 2001). These changes, including the loss of the ability to discern details or contrast (Crassini, Brown & Bowman, 1988) and color discrimination ( Haegerstrom
34 Portnoy, Schneck, & Brabyn, 1999 ), coupled with increased sensitivity to glare, makes it difficult for older adults to read and visualize certain types of informatio n (Echt, Morrell & Park, 1998). Changes in cognition associated with aging results in decreased ability to perform mental operations such as simultaneously remembering and incorporating new information, multitasking and comprehending text ( Craik & Salthous e, 2000). These cognitive changes can begin in early adulthood (Salthouse, 2004). Evidence points to a typical linear cognitive decline in how new information is stored (encoding), how quickly information is processed and spatial awareness ( Hedden & Gabri eli, 2004 ). Other age related physiological changes impacting cognition include: reduced cerebral grey matter and neural demyelination resulting in slowed motor response and reaction time and should be considered when designing user interfaces for older ad ults (Kochunov et al., 2007). Environmental Numerous environmental factors that influence technology adoption are not limited to older adults, though they may experience them more frequently. Environmental access to technology is a function of infrastructure (availability) and resources (affordability). The digital divide refers to the limitations of some populations to receive digital services through reduced access or affordability ( Kreps & Neuhauser, 2010 ; Noar & Harrington, 2012). As mentio ned previously, many older adults live on fixed incomes and therefore may not be able to afford computers, Internet access, video games or mobile devices. According to Zickuhr and Smith (2012), one in five adults do not have access to the Internet with old er adults, those of lower socioeconomic status, Spanish speakers and less educated adults leading this group.
35 Social Older adults tend to be educated with about 34% having graduated high school, while by 2030 that number is expected to reach 86% (American Psychological Association [APA], 2012). Many older adults live on fixed incomes derived primarily from social security benefits though 13% live in poverty, compared to 15% aged less than 65. A lack of financial independence can affect the personality a nd individual dignity of older adults and can restrict autonomous decisions thereby forcing compromises on commonplace activities and precipitating startling role reversals between themselves and their children (Sijuwade, 2009). Social interaction often occurs in religious meetings with 50% of older adults participating in weekly services (APA, 2012). Also, older adults tend to remain politically active when compared to their younger counterparts. T he desire to maintain independence s ocial interaction s a nd family connections is important to older adults and they turn to technology for this purpose over any other (Barrett, 2011; Burnett, Mitzner, Charness & Rogers, 2011 ). Two thirds of adults aged 65 and over use computers to maintain communication with fa mily and friends, representing an increase of 36% since 2007. While both Millienials (24%) and Generation X (12%) viewed their use of technology as the defining characteristic of their respective generations, Baby Boomers cited work ethic (17%) and the Si lent Generation cited the depression and World War II (14%) as theirs (Pew Research Center, 2010) These viewpoints are not surprising when one considers the emergence of various technologies and the differences when they were first experienced and adopted
36 would not see or interact with these machines until much later. Prior to 1976, computers were often large, bulky devices that could not be used in the home (Computer History M useum, 2006). It wasn't until the advent of the personal computer in 1977 with the Apple II, the first widely adopted personal computer, that exposure to computers became widespread. From that point forward, working environments changed as businesses bega n to install and use these devices on a large scale. Baby boomers and Silent G eneration would have already entered the workforce, which for many, would be their first exposure to these devices. By 1982, Commodore woul d sell 22 million units of its C ommodor e 64 home computers, thereby extending the reach of exposure beyond the business environment to millions of homes across the U.S. Currently, the availability of computer techn ology and the expansion of the I nternet has allowed unfettered access to online c omputer programs, information and social networking capabilities. (Computer History Museum, 2006). Prior to that period, most video games were not commercially available to the ge neral public. Atari created the first widely adopted home video game system with Pong, a tennis like game played on a television with hand held controllers. Generation X would have been exposed to home video game systems early in their childhood if their Baby Boom er parents provided them. Most Baby Boomers and Silent G eneration members would have purchased these games as gifts for their Generation X children and Millennial grandchildren, thereby shaping their early thoughts about these systems. With the association of video games as a pleasant, social experience was made by many youth.
37 systems allowed home video game playing to ov ertake arcade gaming. Patterns of exposure to the Internet by different generations are s imilar to the exposure to computers in general. First exposure to the Internet for Baby Boomers and Generation X would hav e likely been in the work environment, whil e at the same time the Silent Generation and Millennial s would have likely used the Internet first at home. become known as the Internet didn't emerge until 1992 (Smithsonian, 2011). From that point forward, companies like Microsoft and America Online opened the door to online information through the creation of Internet browsers and information portals. It wasn't until web browsers and modem connections appeared in personal computers that Internet access in the home became common As the number of websites grew, it became necessary to access that information in a user friendly way. This was accomplished through the creation of search engines, which allowed users to search f or information using key terms and phrases. Patterns of Adoption by Adults Adults in each age group use e health tools on a routine basis, though age related differences exist in how often and which types are preferred. While younger adults view themselves as being defined by technology, older adults do not (Pew Research Center, 2010). Nearly 90% of younger adults use the Internet, compared to 79% of middle aged adults and 40% of older adults. Younger adults use social networking sites at much higher rates than their older counterparts with nearly three fourths accessing them com pared to 50% or less in older adults. In addition, t he average age of a video game player was 37 with a 12 year history of game playing.
38 Finally, 29% of those who play video games are older than 50 and the numbers are steadily increasing (Entertainment So ftware Association, 2011; Gerling, Schild & Mausch, 2010) Despite lagging behind younger adults in technology use in most categories, older adults do utilize technology more frequently than their younger counterparts in at least one area. According to th e Center for Technology and Aging 11% of adults over the age of 50 use mobile devices to track health metrics such as blood glucose, physical activity levels or blood pressure (2011). Other researchers also found that older adults health related use of t Rogers & Charness, 2011). Finally, use of online e health tools by older adults peaks before decreasing as age reach es the oldest old (McInnes, Gifford, Kazis & Wagner, 2010; McMillan & Macias, 2008). Inte rnet and Information The Internet has been accessible and used by large portions of the population for many years. In addition, the ways in which the Internet has been accessed have changed rapidly as advances in connectivity have evolved from early dial up connections to the high health and health promotion professionals began to recognize that the Internet programs could potentially impact large portions of the population and increase preventi ve and health promoting behaviors (Cassell, Jackson & Cheuvront, 1998). In addition to information on prescription medications, diet and diseases such as cancer and arthritis, exercise is one of the most popular topics of interest online (McMillian & Maci as, 2008).
39 The use of online resources to support healthy behaviors is much more developed than other forms of e health tools. Even though electronic games, social networks and self monitoring tools have existed for many years, their use to support heal th related activities is still relatively recent. The Internet has been rapidly adopted as a primary channel for seeking health related information for large segments of the population (Kreps & Neuhauser, 2010). Regardless of its reach, gaps still exist in its lower socioeconomic status or increased age (Kreps & Neuhauser, 2010) Despite its limitations, the I nternet can be a successful channel for delivering effectiv e behavioral interventions that rival and sometimes exceed the effectiveness of standard lifestyle interventions for smoking cessation, self efficacy and exe rcise and may even be the preferred method for receiving exercise information in certain groups ( Ma rshall, Hunt & Jenkins, 2008) Furthermore, self efficacy for the use of online resources can influence how frequently older adults use it (Chu & Mastel Smith, 2010). Age related Differences in Online Resource Utilization While there is inconsistent data about the degree to which individuals use the I nternet for health related activities (Kreps & Neuhauser, 2010; McInnes, Gifford, Kazis & Wagner, 2010), the I nternet is often used as a primary conduit for health information among many different groups (Atkinson, Saperstein & Pleis, 2009; Brouwer et al., 2010; Cohall et al., 2011; Center for Technology and Aging, 2011; Koch Weser, Bradshaw, Gualtieri & Gallagher, 2010). Y ounger adults, while significant users of the Internet in general, don't seek online health related support as often as their older counterparts, since very old adults use the I nternet less than a ll other adult groups (McInnes et al 2010; McMillan & Macias, 2008). The possible explanation
40 As a group, younger adults are typically healthier and are therefore not as inclined to seek online health information though this has not been empirically tes ted. Historically, lower prevalence rates of technology adoption have been associated with increased age, though incidence of technology adoption is increasing. Olde r adults have begun utilizing various forms of Internet support for health with some expressing a preference to use the Internet to communicate with their providers (Singh, Fox, Petersen, Shethia & Street, 2009). In a qualitative survey of older adults, ol der adults desire online access to several different types of health information, such as up to date treatment information, information so assist in choosing health providers, general health topics and information on diet and exercise ( Xie, 2009). Among a dults aged 55 and older, the methods most used for accessing online health information were search engines, commercial health sites and web portals like Yahoo ( McMillian & Macias, 2008). Online R esou rces and E xercise Some longitudinal studies have demonstrated that certain Internet programs when acting as a relational agent and providing information and feedback to users, can produce desired behavior changes in older adults ( Bickmore, Caruso, Clough Gorr & Heeren, 2005a; Bickmore, Gruber & Picard, 2005b). In addition, online programs can provide feedback and act as a virtual coach to im prove physic al activity (Watson, Bickmore, Cange, Kulshreshtha, & Kvedar 2012) or weight loss ( Hunter et al., 2008; Svensson & Lagerros, 2010) by providing supportiv e feedback on progress and a comparison to historical performance
41 Social Cognitive Concepts and Online Resources In addition to being a health information tool, the Internet can provide an avenue for participation in structured programs that can influenc e self efficacy and behavior (Bandura, 2004) Theory based online interventions would be most successful on social cognitive concepts. For example, a longitudinal study examining the relationship between an intervention and changes in physical activity a nd eating habits found that online interventions can significantly improve steps per day of activity, and healthy food consumption ( Anderson Bill et al 2011 b ). The web based online intervention consisted of tracking physical activity, eating habits and receiving goals and general feedback over a 52 week period These results supported earlier findings that online interactive programs are effective at improving self efficacy (Kreps & Neuhauser, 2010). Self Monitoring and Exercise Self monitoring tools can support self regulatory mechanisms such as feedback and goal setting. Using instruments that track data, contextual activities or feedback on data points provides the user with information regarding health behaviors. Significant effects of pedometers or other self monitoring devices on various health behaviors such as disease self management and exercise have been found (Heesch, Masse, Dunn, Frankowski & Mullen, 2003; Hollis et al., 2008). Even though little evidence exists about how different age groups use these devices, increasing numbers of adults are adopting self monitoring health tools, especially with the introduction of mobile technology. According to the Center for Technology and Aging (2011) 11% of adults over the age of 50 use mobile devices to track health metrics such as blood glucose, physical activity levels or blood pressure.
42 Though not empirically tested, self monitoring theoretically should enhance feelings of control, thereby boosting self efficacy and influencing behavior (Bandura, 1 997). The act of monitoring may influence behavior by raising awareness of the behavior that person is attempting to monitor. Self monitoring tools provide feedback and can improve motivation, self awareness and user satisfaction as well as increase the l ikelihood for sustained engagement with a weight loss program (Sevensson & Lagerros 2010) or exercise program (Heesch, Masse, Dunn, Frankowski & Mullen, 2003). In a randomized controlled trial comparing various methods for weight loss in obese adults, se lf monitoring through food tracking, when included as part of a comprehensive intervention program, added significantly to a model of factors predicting short term weight loss (Hollis et al., 2008). Further analysis in this study revealed that the online food tracking component, in conjunction with exercise, predicted greater weight loss in African Americans when compared to those who didn't track their food intake. monitori ng tools, identified benefits of self monitoring include d : behavior change support, making behavioral patterns more evident, health monitoring data, challenging or validating beliefs, recording events, and supporting social interaction (Beudin, Intille & M orris, 2006 ) In addition, potential limitations included: tracking certain behaviors may or valuable information, threatens self image, leads to social conflict, force s too much structure, and can be too comp licated, error prone or disruptive. Fogg and
43 of tracking using mobile devices that can persuade behavior change through social influences, providing feedback and contextual messaging. Social Networking in Health & Exercise All age groups are rapidly adopting social networking, though evidence points to differences in how quickly this is occurri ng. Facebook has grown to over 1 b illion users within just a few years. Individuals are now able to connect wit h each other, share experiences, offer encouragement and provide guidance and feedback using these new tools. They are becoming so popular in fact, that many health related websites now have some social networking aspect woven into them. Social networkin g can provide support to those with chronic disease and 55% of Americans can get information about a therapy or condition online (Deloitte Center for Health Solutions, 2010). In addition, websites such as Patients Like Me provide a milieu whereby individua ls can connect with others who share similar experiences, including ways in which they were able to overcome difficult circumstances and improve their own health (Deloitte Center for Health Solutions, 2010; Hwang et al., 2010). Even though social media use as a tool for health interventions shows promise, there are a few limitations. For example, the veracity of health information disseminated via social networks is sometimes questionable and using social networks as a vehicle for health interventions may exclude those socially disadvantaged, a group most often in need of health information (Chou, Hunt, Beckjord, Moser & Hesse, 2009). Many factors influence how social networks can sway behaviors for various groups of individuals. Social networks increase in terpersonal communications, thereby facilitating the transference of ideas, supportive statements, barrier identification and thought sharing among others (Bandura, 2004). Social networks can also be used to
44 facilitate awareness of health related topics. I ndeed, social networks can increase the power and reach of health communications campaigns thereby increasing their effectivene ss at changing behaviors (Chou et al 2009; Kreps & Neuhauser, 2010). While there is still some debate over whether open soci al networks such as Facebook rather than health specific ones will be used by individuals for health related functions, evidence is mounting that many prefer to use some form of social network in pursuit of better health (Fukuoka et al., 2011; Greene, Cho udhry, Kilabuk & Shrank, 2011). For instance, the ability to form social networks in which to share barriers, experiences and feelings was indicated as an important characteristic of a mobile diabetes community (Fukuoka et al., 2011). These findings suppo rted efficacy building assertions by Bandura (1997) which indicated that social modeling and verbal persuasory interactions positively influence health behaviors Age related Differences in Social Network Use Significant a ge related differences exist in t he adoption and use of social networking technology. As would be expected, Millennials use social networking technology more frequently than older generations. For example, 75% of Millennials have created a social networking profile compared to 50% of Ge neration X, 30% of Baby Boomers and ju st 6% of members of the Silent G eneration (Pew Research Center, 2010). Even though they comprise the largest group of social networking users, social network technology adoption is not limited to younger generations. In fact, while the use of social networking by 50 to 64 year olds was only 7% in 2005, by 2010 that percentage had increased to 42% (Center for Technology and Aging, 2011). In addition, just 2% of the adults aged 65 and older used social networking in 2005 and by 2010
45 that number had increased to 6%. Utilization of social media can support older adults in the following ways: managing chronic conditions, promoting primary preventive measures and eliciting peer support while decreasing social isolation (Cent er for Technology and Aging, 2011; Gracia & Herrero, 2009; Patrick, Griswold, Raab & Intille, 2008). Very few studies have been conducted testing the differences in the adoption of social networks by different age groups. However, limited evidence exists indicating that age significantly influences the adoption of social networks to support health (Chou et al 2009). Cross sectional survey results revealed that age was significantly associated with Internet and social networking use. Further analysis usi ng age stratified logistic regressions supported the significance of age as an influencer of social network use with increasing age being negatively related to social networking use for health. Social Cognitive Concepts and Social Networks While there is emerging evidence to support social networks as an efficacy building tool, more work needs to be done to test these effects Social networks can influence health behaviors by overcoming various barriers (Maitland, 2011). Personal or environmental constrai nts such as lack of experience, unfamiliar exercise environments, social or familial limitations, or lack of exercise role modeling can be countered through the application of a social network. For instance, modeling of exercise behavior through social net work exposure to exercisers who physically appear like the individual or that highlights alternative exercise methods and environments and provides affirmative reinforcement can influence exercise behavior. These strategies are congruent with efficacy buil ding through social modeling and verbal persuasion.
46 S ocial networks can change behaviors through socially mediated paths of influence. In addition, s ocial networks must be efficacy building in order to be effective and that sustained success in behavior c hange can be realized ( Bandura, 2004). Social networks provide guidance, support and incentives relevant to the individual, thereby exerting influence and promoting behavior change. S ocial networks aimed at health behavior change must be s tructured to mot ivate, promote self management and address health habits in order to be effective. Empirical testing of social networking influences on health behavior adoption is in its infancy. However, in a novel approach to testing social network interventions, res earchers conducted an experimental study to test the effects of homophily on the adoption of a health behavior within an online social network (Centola, 2011). Homophily is the tendency of individuals to construct social networks that are similar in charac teristics to themselves. The sample included 710 participants of an online fitness website Two groups were formed with one consisting of clustered groups by age, body mass index and gender and another, that was not grouped by any demographic variable. W ithin each group a single person displayed their use of an online nutrition journal to their social network. The researchers then measured how quickly a health behavior, nutrition journaling, was adopted by social network as evidenced by ini tiating the online health journal and sending a notification that they began the journal behavior to their group Those in the homogenous group significant ly adopted the journaling behavior nearly three times faster than as many individuals ( p < 0.01) tha n the control group. Furthermore, obese individuals adopted the behavior in the homophilous group twice as quickly as those obese individuals in the heterogeneous
47 group These findings were congruent with suggestions by Bandura that social modeling influ ences behavior (1997, 2004). In a meta analysis of interventions to increase exercise behavior in chronically ill adults, Conn Hafdahl, Brown and Brow n (2008) also found that adoption of healthy behaviors could be facilitated through social networks. Hen ce, social network influences impact s exercise participation in many different types of studies Electronic Games Human beings have engaged in game playing since the beginning of history with evidence of games found preserved in ancient burial sites thousands of years old ( Baranowski, Buday, Thompson & Baranowski, 2008). Playing games satisfies emotional and psychological needs by promoting feelings of competence, achievement, and satisfac tion, and in solitary games, autonomy ( Ryan, Rigby, Przybylski, 2006). While the playing of games is not new, the ways in which games are played has evolved considerably. New technologies offer significant opportunities to engage people in activities aim ed at improving health that are both fun and motivating. The use of games as health interventions is being explored more frequently although significant gaps in knowledge still exist. Digital games are being played on an ever increasing basis and that th ey can be effective at improving physical activity as well as confidence in activity participation. In addition, games are no longer restricted to home use and can be played by many newer mobile devices (Li & Counts 2007). Using electronic games is becoming increasingly popular. In fact, the number of people who play electronic games in the U.S. numbers in the millions (Entertainment Software Association, 2011). In 2010, gamers spent 25.1 billion dollars on electronic
48 game s and hardware with 72% of American households playing electronic games of some type. Gamers are also diverse, with female gamers (42% overall) over the age of 18 (37%) outnumbering male gamers 17 or younger (13%). As of 2011, the age of the average game player was 37 with a 12 year history of game playing. Finally, older gamers over the age of 50 represent ed about 29% of the American gaming population and the numbers are climbing (Entertainment Software Association, 2011; Gerling, Schild & Mausch, 2010) Electronic games are able to provide fully immersive environments in which players can complete tasks, solve problems, and assume different identities. Historically, electronic games played at a computer or using a console were considered a health risk since interacting required little physical exertion by the gamer (Kautiainen, Koivusilta, Lintonen, Virtanen & Rimpela, 2005). However, gaming mechanics have changed from simple joystick controllers to ones augmented by accelerometers, infrared cameras or other motion sensing technology. Players can now control games using body movements, with or without the aid of handheld controllers. With the advent of motion based game controls, some forms of gaming have transitioned from a sedentary health risk to a tool to that can improve health by increasing physical activity. Electronic games are capable of reaching large, diverse audiences and can be experienced for extended periods of time while maintaining the attention of the player. Thus, making electronic games ideally suited as vehicles for behavior change ( Baranowski et al 2008).
49 Age related Differences in Video Game Use Experiential T iming Electronic games have been a source of recreation for children and young adults for many years and are often used to socialize and share experiences with others in their age group. As games became commercially available in the late 70s and early '80 s, individuals were able to engage in electronic game play at home as well as social venues like video arcades. Meeti ng at the arcade was often a social event in which young game players were able to interact with each other and play. Many individuals who were older at that time did not participate in gaming the way children did, therefore, their experience possibly sha ped their idea that certain electronic games were for children. Although not empirically tested, several investigators have suggested that timing of exposure to technology may influence the use of electronic games (Boschman, 2010; Ijsselsteijn, Nap, de K ort & Poels, 2007; Pearce, 2008 ). Age related factors such as prior exposure to technology, cost perception and desire for instructions can influence how different games are adopted and enjoyed by different age groups. In a pilot study to explore patterns of adoption for exergames in novice users over 40 years of age Boschman (2010), compared two exergames with exertive interfaces while assessing how instruction clarity impac ts perception of the experience. In a qualitative interview participants regarded certain types of gaming such as personal computer based games as appropriate for older gamers rather than console based games, which were seen primarily for children (Boschman, 2010). Nearly all of the study participants stated that they played electronic games on a personal computer (PC) rather than a gaming console such as a Nintendo, a finding that was also observed by Pearce (2008).
50 Adults over 65 may possess characteristics that influenced their level of comfort using electronic games ( Ijsselsteijn e t al 2007) For instance, they may have retired without using a computer at work and possibly learned computer games differently due to this lack of experience. Also, o lder adults tend to be fearful that they will damage the game and will consult manuals more frequently than younger players Furthermore, Pearce conducted a mixed methods study of 300 Baby Boomers born between 1946 and 1964, and their use of computer games (2008). Almost half of the participants did not play their first game until they were 20 years old or older; with 16% not playing initially until after age 40. In addition, 76% received their first gaming system after age 20 and 16% after the age of 40. Further qualitative analysis revealed that they were fearful of breaking the games. Unfortunately the survey was delivered via a website, potentially biasing the sample towards those more comfortable using technology, though this was not addressed. In summary, timing of exposure to technology may influence its subsequent adoption. Physical C hanges While experiential timing seems to influence how electronic games are adopted and in what format, other cohort characteristics may also play a role (Dunning 2008 b ). Age related changes affected the game interface for elderly players (Gerling et al., 2010). For instance, cognitive changes, sensory changes or presence of illness may Additionally, physical limitations may limit the range of motion or precision needed to successfully control the game. Electronic exercise games may be a useful tool to increase physical activity in older adults (Dunning 2008a; Gerling et al., 2010). H owever, physical declination
51 normally associated with the aging process can raise challenges for older game players who wish to use exertive interfaces to exercise (Gerling et al., 2010). Game players who are not successful due to physical limitations may not find the experience enjoyable, thereby decreasing the likelihood of continued interaction. The physical limitation s of older adults or a ge related differences should be taken into account when designing and providing older adults with video games. For instance, games should be flexible enough to allow various styles of game play to accommodate the physical limitations such as difficulty standing or maintaining balance. As in some dance games with team based play, social interaction and teamwork are facilitated and many older adults find these enjoyable (Boschman, 2010; Dunning, 2008b). Also older adults preferred games with simplicity, adjustable displays, immediate performance feedback and fewer required steps (Gerling et al., 2010). Including the se design Electronic games can be enjoyable for all age groups. In a cross sectional study ( Graves, Ridgers, Williams, Stratto n, Atkinson & Cable 2010 ) 14 adolescents aged 11 to 17 years o ld were compared to 15 young adults aged 21 to 38 and 13 older adults aged 45 to 70. How much a person enjoyed an activity was the primary determinant of the likelihood of continued participation Not only were the games enjoyable, but they also generated significant increases in oxygen consumption. While older adults preferred exergame interfaces using the balance board, all groups preferred exergames to traditional exercise. Exergames, Health and Exercise While some controversy exists regarding whether exercise games as a group can promote health, the evidence is clear that certain exertive interface video games
52 are beneficial and effective at generating energy expenditures needed to qualify as moderate exe rcise. In some studies, exergames did not produce energy expenditure s needed to coun t as moderate physical activity, which is found to have health promotion benefits (Graves, Stratton, Ridgers & Cable, 2007; Graves, Ridgers, Stratton, 2008; Lanningham Fost er, Foster, McCrady, Jensen, Mitre & Levine, 2009). Miyachi, Yamamoto, Ohkawara & Tanaka (20 10 ) countered that the measurement of energy expenditures was flawed in these studies. O ften arm and leg movements were discounted thereby underestimating the amou nt of energy expended in exergames. Instead of measuring energy expenditure wearing a facemask as others have done Miyachi and associates (20 10 ) used a metabolic chamber that produc ed a far more accurate reading, than facemask measurements. They found tha t exergames do generate energy expenditures needed to reach moderate levels of physical activity. In addition to generating energy expenditures needed to promote health, exercise games can also improve athletic performance as well as other health measures in adults. Warburton and colleagues (2007) found that interactive video game exercise significantly impr oved maximal oxygen uptake, leg power, and vertical jump height and reduced resting systolic pressure, when compared to control group performing stand ard exercise interventions Interestingly, they found that adherence to exercise training was significantly higher in their intervention group and mediated the relationship between training and health outcomes. Therefore, participants using video games ex ercised more frequently, likely accounting for the improvement in health outcomes rather than the intervention itself. Similar significant findings were reported by others who found that electronic games were effective at increasing physical activity,
53 ener gy expenditure, oxygen consumption, and weight loss (Inzitari, Greenlee, Hess, Perera & Studenski, 2009), cardiorespiratory benefits (Siegel, Haddock, Dubois & Wilkin, 2009) and balance and leg strength in older adults (Nitz, Kuys, Isles & Fu, 2010). An expert panel at the proceedings for the American Heart Association described several benefits of exergame participation, such as: i mproved confidence, social for other exer cise experiences ( Le iberman, Chamberlin, Medina & Franklin 2011). For example, o therwise sedentary individuals would begin an exercise routine using the exergame then progress to other forms of exercise. P articipants extended their exercise routines to i nclude other traditional forms of exercise and also suggested that exergames make exercise more appealing across different age groups. This is of particular interest from a health behavior standpoint since it points to the relationship between exergames an d increased exercise behaviors. The panel cited study results from the American Heart Association which included 58% who stated they began other fitness pursuits such as tennis or walking after beginning an exergame based fitness program. In addition, 82% of those surveyed stated they played more with family and friends, thus increasing social connectivity Women were more likely than men to use exergames, to try new activities, stay active or cha llenge their physical limits. It is likely that through the se self efficacy improvements, participants were confident to engage in more traditional activities, wh ich is consistent with Bandura (1997) that highly efficacious individuals are more likely engage in increasingly challenging tasks and remain resilient w hen embarking on a new activity. In addition to behavioral benefits, the authors noted several physiological impacts such as: improved energy expenditure,
54 vertical jump height heart rate and blood pressure. Another physiological benefit was distraction f rom discomfort. P articipants tended to participate regardless of painful perceptions, especially if social or collaborative aspects were included (Lieberman et al., 2011). Distraction from discomfort may counteract pain as a barrier to exercise in older adults. Social Cognitive Concepts and Exertive Gaming E xercise games are used by all age groups and are capable of producing numerous physical benefits and can produce health behavior changes. In a meta analysis of games as behavior change tools, resear chers found two mechanisms by which behavior can be changed by games that are consisten t with Social Cognitive Theory (Bandura, 1997, 2004) 1) using behavior change techniques such using goal setting as a tool to play the game, or 2) using a story during the game to teach specific behaviors (Baranowski et al 2008). Authors provided the example of learning self regulatory goal setting in a story based game aimed at increasing fruit consumption among children as evidence how a theory based cha nge concept can be incorporated into a game so that as goals are set and achieved, healthy behaviors are learned and improved. Other game components found to be important to change behavior include : game immersion, interactivity and fantasy. However, games as health behavior change vehicles are in their infancy and require further study on their effectiveness at motivating different types of players. Exergames and Barriers to E xercise In addition to providing exercise and changing behaviors, exergames can help overcome barriers to exercise. Inzitari, Greenlee, Hess, Perera & Studenski (2009) conducted a qualitative study of post menopausal women aged 45 75 years old and
55 assessed their attitudes towards electronic dance game as tool for exercise Women tend ed to adhere to dance based exercise programs and had fewer injuries when compared to traditional exercise. The investigators speculate d that electronic dance games might promote exerc ise adherence through its system of rewards and feedback, which is consistent with self regulatory motivators (Bandura 1997 ). Focus groups identified barriers such as ti me commitment and weather In addition, advantages of electronic dance games were: fun increased social i nteraction physical and mental benefits while reducing perceived barriers to exercise through convenience, not dependent on weathe r, easy to do in group or alone. Furthermore p otential disadvantages were: difficulty in learning, long term adherence, technical aspects, cost, may not like music o r dance style in game. Exergames and S elf efficacy In addition to influencing behavior, promoting social interaction, improving various physiological measures and increasing enjoyment, video game s are also efficacy building In one example, Bandura (2004) describes how in children video games build self efficacy for self care behaviors in chronic diseases su ch as asthma (Lieberman, 1997) and diabetes (Brown et al., 1997) The children engaged in various tasks which facilitated learning of diabetes management techniques such as insulin dosing and nutrition management. Follow up at 6 months demonstrated a significantly higher rate of diabetes control than those children wh o played non health related games. In addition, self efficacy for diabetes care was significantly higher as well as an increased ability to discuss various self care techniques. The findings of other s support a positive relationship between video game use and self efficacy ( Lieberman, Chamberlin, Medina & Franklin, 2011)
56 Thomas, Cahill and Santilli (1997) conducted an Intervention study with inner city adolescents and young adults ( n = 324) using a computer based electronic game to increase self efficacy for negotiating safe sex. Ages of participants ranged from 12 to 22 and were racially diverse. The game was designed to repeat various negotiating behaviors and reinforce simulated avatars completing the task. Intervention was specifically designed to b oost self efficacy through mastery experiences by repeating negotiating tasks in different computerized scenarios, and by social modeling through viewing ones digital self and others in similar situations successfully negotiation condom or dental dam use p rior to an implied sexual encounter. Pre and post intervention knowledge regarding safe sex practices revealed a significant increase in knowledge within seven safe sex knowledge categories and safe sex negotiation Participants with the lowest initial sel f efficacy scores demonstrated the largest increase when compared to those who were already highly efficacious. V ideo games may boost self efficacy and more work needs to be done in order to test whether exertive video games promote exercise self efficacy. Summary Age related differences exist in the utilization and perception of electronic resources. Older adults view electronic resources differently than their younger counterparts in that they tend to be more fearful in the use and are slower to adopt ne w technologies. Also, younger generations feel that technology is a defining characteristic of their group, a feeling that is not shared by older adults. Despite a ge related differences, electronic resources can provide a mechanism to improve physical functioning, social interaction and confidence in behavior changes among all groups
57 wh ile reducing perceived barriers and theoretically improving self efficacy. E H ealth supp orts show promise in improving health through exercise but work must be done to understand which supports are effective at changing exercise behaviors. E H ealth supports can fill a basic human need for social interaction, promote autonomy and might even bu ild self efficacy, but little is known about how their impact on this important concept. Furthermore, while a ge associated characteristics might affect patterns of adoption for e Health support use, little has been done to empirically test this. As obesi ty rates climb and lifestyle related disease proliferate, it is becoming imperative to find ways to engage individuals in their own health through personal action and controlling those risk factors on which they can have an impact. Many individuals can dr astically improve their health risks and quality of life by engaging in behaviors like routine exercise. Unfortunately, many do not exercise as often as needed to prevent illness and promote health. Ways must be found to sustain individual engagement in health promoting behaviors so that the healthy period of life on the health/illness continuum is expanded, while illness and disease is compressed. Significant technological advances in electronic health supports have flooded the population with too ls promising to aid users in being healthy Unfortunately, little work has been done to understand exactly how different a ge s utilize these tools and whether they are effective at influencing known components of health behavior. This research will add to current knowledge about self efficacy and exercise among different age cohorts as well as support current understanding of how electronic health support tools are used by these groups.
58 CHAPTER 3 RESEARCH DESIGN AND METHODOLOGY Design This cross sectiona l study explore d the use of e health resources and their effects on self efficacy and known factors that influence exercise behavior A systematic random sample was obtained from participants that are stratified by different a ge related cohorts. A survey was mailed to them to assess outcomes. Sample A stratified random sample was selected from potential participants in a nationwide wellness program, The Prevention Plan This plan was an employer or health plan sponsored program that included employees and retirees of major corporations and state agencies. The Prevention Plan was a voluntary wellness program delivered online, telephonically or through paper based learning modules based on preferences of participants Stratification based on ag e cohort increased the probability that the study groups represented those in the Prevention Plan. Hence, comparisons of the cohorts in e H ealth resource utilization, efficacy beliefs and social cognitive variables could be generalized to the population of persons in wellness programs There were a total of 96,287 potential participants in the wellness program. Of that number, 58.8% were female ( n = 56,599) and 56.4% ( n = 54,261) were employed either part time or full time. The overwhelming majority of participants were Caucasian (53.5%), followed by African American (4.7%), Asian (2.5%) and Hispanic (3.4%), with the remaining individuals declining to state their race. Prevention plan member a ges range d from 3 years old to 110 with a mean age of 47 years though only adults were
59 included in this study Age distributions within each age category were as follows: Millennial (8.1%), Generation X (39.1%), Baby Boomers (47.8%) and Silent Ge neration (5%). Inclusion criteria were men and women ages 18 or older Wellness program participation or access to a computer, video game system self monitoring tool or mobile phone was not used to exclude participants from the study. Participants were stratified into age categories based on their age at the time of the study. Potential p articipants were grouped into the following four categories by age: Millennial: born since 1983 and aged 29 or younger ; Generation X: born between 1965 and 1982 aged 30 to 47 years old ; Baby Boomers: born between 1946 and 1964 aged 48 to 66 years old ; and finally, the Silent Generation: born earlier than 1946 and aged 67 or older. Power Analysis Power analysis was conducte d a priori with alpha power of 80. In similar studies to the proposed, where relationships between social cognitive variables and exercise were tested, effect sizes of 0.30 (Hortz & Petosa, 2008) and 0.41 ( Conn, Hafdahl, Br own & Brown, 2008 ) were found. According to Portney and Watkins (2009), correlational relationships between 0.25 0.5 indicate a small relationship, while correlations 0.5 and higher indicate moderate or higher relationships. A conservative effect size of 0 .20 was selected as the minimum effect to be detected within the study. Using G*Power version 3.1.2 for Mac (Heinrich Heine University) with alpha set at 0.05 and a power of 80, the sample size needed to achieve the required power in this study is 280 T here was a total of 70 needed in each age group.
60 Recruitment Based on prior survey responses with the study population, a 15 % response rate was estimated and used to determine the initial number of surveys to be sent. Using the required number to achieve p ower in the study ( n = 280), divided by the response rate (15 % ) a total of 1,868 surveys was mailed, with 467 mailed to each age group. R ecruitment continue d until completed data was received on 70 participants in each group. Thirty days after the initial surveys were mailed, analysis of returned surveys reveal ed if further recruitment was necessary to reach statistical power for each group. If a group remained underrepresented having less that 70 an analysis to determine the n umber of new surveys to be sent commence d utilizing the following formula: ( 70 number of surveys received)/(0.15 ) Reassessment of responses occurred 30 days after the second mailing. If recruitment was ultimately unsuccessful to reach required numbers options to move forward with the study include d grouping the two oldest groups together as one group and grouping the two youngest groups together. According to the Pew survey (2010), older aged cohorts (Silent Generation and Baby Boomers) did not defin e themselves using technology, while younger age groups (Generation X and Millennial s) named technology as their defining characteristic, thus indicating logical clusters within the four groups. Measurement Study participants were mailed a questionnaire (see Appendix A ) along with a cover letter (see Appendix B) introducing the survey, instructions and a postage paid return envelope. The anonymous questionnaire was include d the following assessments: basic demographic information, exercise participation, exercise self
61 efficacy, outcomes expectations for exercise, perceived barriers to exercise, e H ealth support use and self efficacy for e H ealth support use. Survey headers were color coded by age group (Silent Generation =Red, Baby Boomers=Blue, Gen eration X=Green, Millennial s=Orange) so that if age and year of birth was not answered the survey data could still be used in the correct category. The survey was written using plain language and organized so that common concepts were addressed together (Dillman, Smyth & Christian, 200 9) For example, questions related to physical activity participation and social cognitive factors related to physical activity were grouped while questions related to each e health resource were grouped as well. Instructions for completing each section were included and ambiguously worded questions were avoided. Arial 12 point font was used to increase readability especially for the older cohorts It was anticipated that the su rvey would take no longer than 15 minutes, thereby increasing the perception that the survey would not be onerous to complete. There were no incentive s to complete this survey. Content validity for all developed questions wa s verified using an expert physician scientist with experience using survey instrumentation for health promotion. A Cronbach's alpha exceeding .70 was considered an acceptable level of internal consistency (Nunnaly & Bernstein 1994 ). Corrected item correlations between .3 and .7 were considered evidence of a good scale (Ferketich, 1991). Characterization of participants Participant age, gender, marital status, educational level, geographic location, employment status and estimated number of years us ing a computer, mobile phone, video games and social networks were assessed.
62 Use of e H ealth Supports The taxonomy of e H ealth supports described in the literature review was used to develop questions addressing each type of resource Each e H ealth cat egory was defined in the questionnaire immediately preceding the questions to which they relate d These were : Internet use self monitoring, social networking and exertive games. Respondents were asked first how many minutes per week they engage d in the us e of the particular resource and how many minutes each week they use each resource to support exercise behavior. If a respondent d id no t use a particular resource, the score was zero for use and frequency for that resource. There were 3 scores: 1) total we ekly minutes using each type of e H ealth resource, 2) the total number of weekly combined minutes using all e H ealth resources, and 3) The combined number of weekly minutes using all e H ealth tools to support exercise behavior also by health resource Self efficacy for e H ealth Support Use No current instrument existed to measure self efficacy for e H ealth support use though guidance exists on their construction (Bandura, 1997; 2006). Self efficacy scales follow a similar pattern in which the responden t rates their level of confidence for a particular behavior. Efficacy scales should be unipolar, beginning at 0 with higher self efficacy ratings coinciding with higher numbers on the scale. Also, scales should range from 0 10 or 0 100 and allow the user t o choose between at least 11 intervals. Scales with fewer intervals lose differentiation in efficacy beliefs and are therefore discouraged. Finally, efficacy scales should assess current beliefs about operative capabilities rather than prospective or fu ture based assessments. Accordingly, four questions were developed regarding confidence in using each type of e health support. Respondents were asked rate their current level of confidence
63 for using each type of e health resource on a Likert scale from 0 to 10 with higher levels indicating higher self efficacy for e health support tool use. The ratings were summed and the total score used to indicate self efficacy for e health support use. The same process was used to construct four corresponding questions rating self efficacy for each e Health tool used to support exercise. The internal consistency of the self efficacy for e Health tool use overall scale was very good ( Cronbach alpha = .80) and inter item correla tions between the four e Health sub scale items ranged between ( r = .43) and ( r = .58), indicating a strong relationship between the items. Also, corrected item correlations, which is the correlation between a single item and the rest of the items ranged from ( r = .58) to ( r = .68), indicating a strong single item relationship with the remaining combined scale items. Finally, no item, if deleted, improved overall Cronbach alpha, thereby indicating that all sub items should be included for the e Hea lth self efficacy overall scale. The reliability of the self efficacy for e Health tool use for exercise scale was also very good ( Cronbach alpha = .82) and inter item correlations between the four e Health for exercise sub scale items ranged between ( r = .4 0 ) and ( r = .65), indicating a strong relationship between the items. Corrected item correlations ranged between ( r = .57) and ( r = .72), indicating a strong single item relationship to the remaining combined scale items. Lastly, no scale item if del eted from the analysis, improved the overall calculated alpha, thereby indicating that all the sub items for the self efficacy for exercise e Health tool use should be included. Self Efficacy for Exercise Self efficacy for exercise was assessed using the Self Efficacy for Exercise (SEE) (R esnick, 2004). This was a nine question survey that participants rank their
64 exercise confidence on a scale from 0 to 10 Internal consistency for the scale was sufficient ( Cronba ch alpha = 0.70). R esponses were summed and efficacy for exercise. Evidence of validity was obtained through confirmatory factor analysis and hypothesis testing. Factor loadings were greate r than .50 at baseline and follow up testing. All factors extracted were related to perceived confidence in the ability to exercise despite perceived barriers. For instance, one factor was being confident to exercise when the weather is not pleasant or i n the presence of pain. These factors are consistent with theoretical expectations that highly efficacious individuals are resilient and engage in exercise despite barriers. Construct validity was also supported through hypothesis testing by statistically significant differences in exercise behaviors for those who had higher SEE scores than those who did not, which was consistent with theoretical expectations. For the current study, internal consistency was very strong for self efficacy for exercise ( Cro nbach alpha = .94). Outcome Expectations for Exercise Ou tcome expectations for exercise were assessed using a 9 item Outcome E xpectations for E xercise tool (Resnick, 2004 ). Respondents rate d expected mental and physical outcomes of participation in exercise E vidence for construct validity was based on hypothesis testi ng and confir matory factor analysis and through structural equation modeling (Resnick et al., 2001 ). Construct validity was supported by statistically significant differences in exercise behaviors for those who had higher outcome expectancy scores than those who did not which was consistent with theoretical expectations Results of structural equation modeling for the nine items included significant path coefficients ranging from .69 to .87
65 with outcomes expectations for exercise as the outcome and evi dence of model fit ( X 2 of 69, df = 27, p < .05, NFI = .98, RMSEA = .07). The ratings were summed with higher scores indicating greater outcome expectations for exercise. Internal consistency for the scale was very good in prior research ( Cronbach alpha = 0.90) and was ( Cronbach alpha = .87) in the present study. Barriers to Exercise The Barr iers for Exercise s cale was used to assess perceived barriers to exercise ( Sechrist, Walker & Pender, 1987) The respondents rate d barriers from 4 The ratings were summed and scores range between 14 and 56. Test r etest reliability was .77. Construct validity was established using confirmatory factor analysis. Factor analysis yielded a nine factor solution initially which an explained variance of 65.2%. The 14 item Barriers Scale has strong internal consistency with Cronbach alpha in prior studies of .8 7 and .88 in this study Exercise Participation Exercise participation was assessed using International Physical Activity Questionnaire (IPAQ) that identifies the frequency and intensity of physical activity comple ted on a weekly basis. The IPAQ measures physical activity for a wide ran ge of adults (15 69 years old) and has been tested in numerous countries (Craig et al., 2003) Instrument validity was tested through corroboration of accelerometer based measures of physical activity with measured estimates. Test retest reliability correlations between instrument items were very good, ranging from .32 to .88 when repeated a week apart. Estimates of physical activity are provide d in minutes over the past 7 days and i nclude descriptions of the different categories of physical activity The total number of minutes in each category were summed, representing the amount of physical activity for
66 each category a respondent engages in weekly. Physical activity was considered when one of the following criteria is met : Vigorous intensity activity on at least 3 days or 7 or more days of any combination of walking, moderate intensity or vigorous intensity activities Physical activity was ollowing conditions: 3 or more days of vigorous activity of at least 20 minutes per day 5 or more days of moderate intensity activity or walking of at least 30 minutes per day or 5 or more days of any combination of walking, moderate intensity or vigorou s intensity activities. Physical activity that does not meet the criteria for either high or moderate activity was Procedures The Prevention Plan donate d administrative support to provide the identification, stratification, postage, mailing and data entry of survey results. The vice president (VP) for health intelligence generate d the eligible list. Since age related characteristics were of interest in this study and to ensure adequate representation from ea ch group, all program members were loaded into a SQL database and a random number generator used to assign a unique identifier to each potential participant. Potential p articipants were further grouped into the followi ng four categories: Millennial, Gener ation X, Baby Boomers and Silent Generation. Once grouped by age, the health intelligence VP sort ed each group by the randomly assigned member number in ascending order. The total number needed in each group was determined using the total sample size ( n = 280 ) divided by the number of groups so that there were an even number in each group ( n = 70 ). Onc e sorted numerically, the first participants in each age group were selected to r eceive the mailed questionnaire All potential participants were stored on a table for future use if needed. An analysis was conducted four weeks after the initial
67 mailing to determine response rate. In the event that response is less than 70 in each group the stored tables were use d in the same manner so that additi onal participants could be identified and surveys mailed until the required number in each category was reached Throughout the process, the researcher was blinded to the sampling procedure and identification of study participants. T he researcher provided the survey list to a secure mail house provider to mail merge the participant list with the surveys, stuff envelopes, apply postage and send surveys to participants. Surveys were returned using the self addressed stamped envelope. The researcher open ed e ach survey and enter ed survey data into a Microsoft spreadsheet. A second administrative person conduct ed a data entry integrity verification to check for keying errors. Response d ata may differ on the returned surveys for some questions For instance, in questions where the respondent is asked to estimate time in years or minutes, responses may be alpha or numeric. In addition, data may include fractions or decimals denoting partial years or minutes. Numerical data that is spelled out were converted to corresponding numerical values prior to entering into the spreadsheet. Fraction or decimal responses were rounded to the nearest whole number prior to entering into the spreadsheet. Initial administrative support training include d how to handle data of these types. Data Analysis All data analysis was conducted using Predictive Analytics Software (PASW) version 18.0.3 (International Business Machines [IBM]) Descriptive univariate statistics was used to describe the population. Pr ior to statistical testing, all data was examined to determine whether it me t the assumptions for parametric tests. In the event that the
68 data did not meet parametric assumptions, non parametric methods were used for each specific aim if applicable Logar ithmic transformations were attempted in the even t the data were not normally distributed. Missing data were identified. I f age related questions were not answered the color coded headers were used to group the survey into the appropriate age category. Mean replacements were used for questions where less than 10% of the data is missing. Surveys where greater than 10% of the data is missing were rejected. For specific aim 1, total w eekly minutes of e health support use was tested for normal distribution using Shapiro Wilk. If minutes of e health support use not normally distributed, t he Kruskal Wallis test was used instead of analysis of variance. Homogeneity of variance was f variances. Cases must meet independence assumptions and the number in each age group should be approximately equal. If assumptions were met, a nalysis of v ariance was used to explore the differences between the inde pendent variable of age categories (Cate gorical) and total weekly minutes of e health support use (Ratio). In the event that a significant group difference was found Wilcoxon pairwise comparisons were used to conduct post hoc testing. For specific aim 2, total weekly minutes of e health suppor t use was tested for normal distribution using Shapiro Wilk. In the event that the total weekly minutes were not normally distributed, the Spearman Rho rank order correlation coefficient was used ( Burns & Grove, 2005 ; Portney & Watkins, 2009). Homogeneity of variance was tested Scatterplot diagrams were used to determine linearity of the relationships. Careful attention was paid to outliers since
69 I nfluential outliers were examined and excluded from further analysis If assumptions are met, then c orrelation was used to explore the relationship between the independent variable of self efficacy for e health support use ( ratio ) and total weekl y minutes of e health support use ( ratio ). For specific aim 3, total weekly minutes of e health support use for exercise was tested for normal distribution using Shapiro Wilk. In the event that the total weekly minutes were not normally distributed, the S pearman Rho rank order correlation coefficient was used ( Burns & Grove, 2005 ; Portney & Watkins, 2009). Homogeneity of variance was diagrams were used to determine linearity of the r elationships. Careful attention was were examined and excluded from further analysis. If assumptions were met, then correlation were used to explore the relat ionship between the independent variable of total weekly minutes of e health support use for exercise ( ratio ) and SEE, OEE and BE ( ratio ). A priori adjustment of alpha to .01 was used to correct for family wise error. Human Subjects Institutional Review Board approval was obtained prior to init iating any research activities. Data integrity and security was protected using double layer password protection for all computers. Paper files were maintained in a locked file cabinet within a locked room. All information was kept in accordance with Health Insurance Portability Accountability Act ( HIPAA ) security requirements. Individually identifiable information was removed and destroyed once it was no longer needed unless otherwise requir ed by University policy File destruction for paper materials will be conducted through a licensed, bonded and insured medical information destruction service. In addition, all
70 computerized files, including flash drives and other portable storage media w ere password protected/encrypted and will be degaussed if replaced. The researcher maintain ed only the minimum necessary information to complete the study. Study data were maintained in accordance with University policies and procedures.
71 CHAPTER 4 RES ULTS This purpose of this study was to describe the various electronic health supports used by different age cohorts compare efficacy beliefs regarding these tools and to explore the relationship between e Health support use and social cognitive concepts The results of the data analysis include sample description, survey response rates, testing of assumptions and results of testing research questions. Data were analyzed using Predictive Analytics Software (PASW) version 18.0.3. Sample A total of 1,868 s urveys were mailed initially, with 467 sent to each cohort. Figure 4 1 describes the survey mailing process over the data collection period. After the first six weeks, the surveys returned were: M illennial ( n = 45, 10%), Generation X ( n = 57, 12%), Baby Boomer ( n = 84, 18%), S ilent Generation ( n = 66, 14%). Using the actual response rates for the first mailing, the number of surveys for a new mailing was calculated for each c ohort. For example, using the M illennial response rate and th e number still needed to reach 70, the following calculation was used: (70 45)/(.10) = 250 new surveys were mailed. The same process was applied to the remaining groups that did not reach the minimum number of surveys needed. The final number of surveys m ailed was 2,174 with a response rate of 13% overall. Overall, the Millennials had greater weekly minutes of technology use than other generations. Millennials engaged in the use of social media about three times more minutes each week than even Genera tion X and nearly 5 times more minutes than Silent Generation Greater detail regarding e Health technology use by age cohorts is in Table 4 2
72 Millennials The millennial group was primarily female (67.1%) single (58.9%) and Caucasian (80.8%) The mean a ge was 25 years They were well educated with 80.8% having some or more college. Most were employed either part time or full time (93.2%) and 61.6% reported annual household incomes greater than $35,000. Also, a majority (84.9%) reported moderate or hig h physical activity. (See Table 4 1 ) Compared to the older cohorts M illennials had the least years experience using computers ( M = 14.3), mobile phones ( M = 9.5), the Internet ( M = 12.2) or self monitoring tools ( M = 0.91) By contrast, M illennials had the most experience using socia l networks ( M = 6.7) and were second only to G eneration X in their years experience playing video games ( M = 8.5) ( Table 4 2 ). Millenials spent the most week ly minutes of any group using the Internet ( M = 945.3), social networks ( M = 395.7) and playing video games ( M = 102.2). Weekly minutes of self monitoring tool use among M illennials was the second lowest ( M = 21) of all the groups. (See Table 4 2) Millennials had the highest self efficacy for general technology use in each category as well as in support of exercise but with two exceptions T hey were second to Generation X in their confidence using video games to exercise ( M = 5.5) and self monitorin g tools for exercise ( M = 6.3). By comparison, there was far less distinction in social cognitive exercise variables between the groups Millennials were moderately confident in their ability to exercise ( M = 54.4), had low perceived exercise barriers ( M = 24) and had high outcome expectations for exercise ( M = 37.5). ( See Table 4 3 ) Generation X The G eneration X group was primarily female (64.7%), married (62.7%) and Caucasian (80%) with a mean age of 38 years. Generation X respondents were very
73 well educated with 93.3% reporting completing some or more college. Most were employed full or part time (90.7%) and were high earners with 66.7% reporting household incomes above $50,000. Nearly 85% reported moderate or higher physical activity. (See Ta ble 4 1 ) Compared to other groups, Generation X had the greatest experience with video games ( M = 11.9) and were second highest in years experience with the Internet ( M = 13.6), self monitoring tools ( M = 1.6), mobile phones ( M = 12.8) and social network s ( M = 3.8). Overall, Generation X was also heavy users of technology, though not to the same extent as Millennials. Generation X had the second highest weekly minutes use of the Internet ( M = 518.0), social networks ( M = 130.3), video games ( M = 54.1) and self monitoring tools ( M = 57.3). They were also second in their use of e Health tools to support exercise, but first in their use of video games to exercise ( M = 8.8). (See Table 4 2 ) Generation X was confident in their ability to use e Health tools in general as well as for exercise. They were second to Millennials in their high confidence for Internet use ( M = 8.96) as well as moderate confidence using social networks ( M = 7.25), video games ( M = 6.32) and self monitoring tools ( M = 6.36). Interestin gly, G eneration X had the highest confidence using video games ( M = 5.67) and self monitoring tools ( M = 6.39) to exercise than the other groups. The Generation X group was moderately confident in their ability to exercise ( M = 50.81) and perceived barriers ( M = 25.04) and had high outcome expectations for exercise ( M = 37.54). (S ee Table 4 3) Baby Boomers The baby boomer group was primarily female (59.3%), married (70.9%) and Caucasian (88.4%) with a mean age of 58.5 years. Ba by Boomer respondents were
74 very well educated with 90.7% reporting completing some or more college. Most were employed full or part time (65.1%) or were retired (31.4%) and were the highest earners of any group with 68.6% reporting household incomes above $50,000. Similar to Generation X, nearly 85% reported moderate or higher levels of physical activity. (see Table 4 1 ) Compared to other cohorts, Baby Boomers had the highest number of years experience with computers ( M = 21.1), mobile phones ( M = 13.3) a nd the Internet ( M = 14.2). By comparison, they were next to last in their experience with social networks ( M = 1.6). Baby Boomers were not heavy users of e Health tools with one exception, they were first in their weekly use of self monitoring tools over all ( M = 97.9) and to support exercise ( M = 106.0). (S ee Table 4 2) Baby Boomers were confident in their ability to use the Internet overall ( M = 7.95) and in support of exercise ( M = 7.77), but had low confidence in the use of all other e Health tools. Baby Boomers were moderately confident in their ability to exercise ( M = 54.94), had moderate perceived barriers ( M = 23.17) and high outcome expectations for exercise ( M = 37.67). (S ee Table 4 3) Silent G eneration The Silent Generation group was primarily male (58.3%), married (65.3%) and Caucasian (90.3%) with a mean age of 70.6 years. Silent respondents were very well educated with 93.1% reporting completing some or more college. Most were retired (58.3%) or employed full or part time (36.1%) and were high earners with 68.1% reporting household incomes above $50,000. The Silent Generation was the least active of any group with 68.1% reporting moderate or higher levels of physical activity. (See Table 4 1 )
75 When compared to the other gr oups, the Silents had the greatest number of years experience using self monitoring devices ( M = 2.3), second in years using a computer ( M = 20.8), but were last in their experience with social networks ( M = 1.2) and video games ( M = 2.6). The Silent group were not heavy users of e Health tools and were last in every category of technology use except weekly social network use ( M = 75.8), video games ( M = 36.5) and game use for exercise ( M = 4.4) where they were second to last. (See Table 4 2 ) The Silent Generation was moderately confident in their ability to use the Internet both in general ( M = 6.52), and in support of exercise ( M = 6.10). However, they had the lowest confidence of all the groups in their ability to use all other e Health tools. Finally the Silent Generation group was moderately confident in their ability to exercise ( M = 51.17), had the highest perceived barriers ( M = 21.12) and moderate outcome expectations for exercise ( M = 34.76), which was the lowest of the four groups. (see Table 4 3 ) Preliminary Analyses Thirteen cases had missing data and were excluded from the analysis. The cases with missing data represented a very small portion of the overall sample (4.3%) and in some cases had greater than 10% of the overall data missing fro m the results. Recreating data through mean replacements would have yielded very limited benefit Therefore all cases were excluded. S ample sizes, means ranges and standard deviations for the outcome variables can be found on Table 4 4 To assess the assumption of a normal distribution required for parametric statistical tests proposed for this study, skewness coefficients and Shapiro Wilk tests were analyzed. S kewness coefficients (SC) less than 1 or greater than +1 were
76 considered highly skewed ( Bul mer 1979). A significant Shapiro Wilk test indicates a non parametric distribution of data ( Field 2009). The t otal weekly minutes using eHealth resources across all groups was highly positively skewed ( SC = +2.71) and had a significant Shapiro Wilk ( p < .0001). In addition, the total weekly minutes of using eHealth tool use for exercise was highly skewed (+9.56) and demonstrated a significant Shapiro Wilk ( p < 0001). In addition, eHealth self efficacy was moderately negatively skewed ( .43) with a sig nificant Shapiro Wilk ( p < 0001). Exercise self efficacy, perceived barriers and outcome expectations were also skewed and were not normally distributed. Since all outcomes for technology use were significantly different than a normal distribution and transformation of the data di d not resolve this issue, non parametric methods were used. In situations where analysis of variance assumptions are violated, a comparable non parametric test can be used. The Kruskal Wallace statistic tests for differences in the distribution between gro ups and is based on ranked data (Field, 2009). As with the analysis of variance test, Kruskal Wallace only determines if a significant difference exists between groups, however, it does not provide detail about which groups are different. Therefore, post hoc tests were done to determine differences between the groups. The Wilcoxon rank sum test can be used for pairwise comparisons to determine significance. These multiple comparisons increase the likelihood of a type I error. The Bonferroni correction w as used as a conservative measure to control for family wise error. For six comparisons, an alpha of .0008 ( alpha .05 / 6 comparisons ) was used to assess significance of statistical tests.
77 Differences in e Health Support Use (Research Question 1) The dif ference between the generational cohorts for weekly minutes of e Health support was significant ( X 2 (3) = 44.49, p < .0001). Millennial's weekly minutes of overall e Health tool use was significantly higher than G eneration X, B aby Boomers and Silent Generation Likewise, Generation X's weekly minutes of e Health use was significantly higher than the Silent Generation B aby B oomers were not significantly different in weekly minutes of overall e Health use than G eneration X and the S ilent Generation ( S ee Table 4 4) Similar significant age related difference s were found in the weekly use of e Health tools to support exercise ( X 2 (3) = 31.75, p < .0001). Millennial's weekly minutes using e Health tools to support exercise was significantly higher than Baby Boomers and S ilent generation Generation X's use was significantly higher than Baby B oomers and S ilent Generation Finally, M illennial s use of eHealth tools for exercise was not significantly different than G eneration X and B aby B oomer s were not s ignificantly different than S ilent Generation (See Table 4 4) Relationship of e Health Use to Self efficacy (Research Question 2) A significant moderately positive relationship was found between the number of weekly minutes using e Health tools and self efficacy for e Health tool use ( r s = .50, p = .01). Relationship of e Health Use for Exercise and Social Cognitive Exercise Variables (Research Question 3 ) Weekly minutes of e Health tool use for exercise was positively related to self efficacy for exercise ( r s = .21, p = .0003) and outcome expectations for exercise ( r s = .28, p = <.0001). By contrast, weekly minutes of e Health tool use was not significantly
78 related to perceived exercise barriers ( r s = .05). C orrelations betwee n individual weekly minutes of e Health tools and perceived exercise barriers were not significant : Internet ( r s = .024) social networking ( r s = .014) video games ( r s = .083) and self monitoring ( r s = .114)
79 Table 4 1. Age Cohort Demographics Millennials Generation X Baby Boomers Silent Total ( n =73) ( n =75) ( n =86) ( n =72) ( n =306) n % n % n % n % n % Gender Male 24 32.9 26 34.7 31 36.0 42 58.3 123 40.2 Female 49 67.1 48 64.0 51 59.3 26 36.1 174 56.9 No answer 0 0.0 1 1.3 4 4.7 4 5.6 9 2.9 Marital Status Single 43 58.9 17 22.7 10 11.6 9 12.5 79 25.8 Married 27 37.0 47 62.7 61 70.9 47 65.3 182 59.5 Partnered 2 2.7 5 6.7 3 3.5 1 1.4 11 3.6 Divorced 1 1.4 5 6.7 9 10.5 9 12.5 24 7.8 Widowed 0 0.0 1 1.3 3 3.5 6 8.3 10 3.3 Ethnic Background African American 6 8.2 6 8.0 4 4.7 3 4.2 19 6.2 Asian 1 1.4 3 4.0 2 2.3 2 2.8 8 2.6 Caucasian 59 80.8 60 80.0 76 88.4 65 90.3 260 85.0 Hispanic 4 5.5 2 2.7 1 1.2 0 0 7 2.3 Native American 0 0.0 0 0.0 1 1.2 0 0 1 0.3 Pacific Islander 3 4.1 3 4.0 1 1.2 1 1.4 8 2.6 Other 0 0.0 0 0.0 1 1.2 0 0 1 0.3 No answer 0 0.0 1 1.3 0 0.0 1 1.4 2 0.7 Education Level 8 th Grade or less 1 1.4 0 0.0 0 0.0 0 0 1 0.3 Some high school 2 2.7 1 1.3 0 0.0 1 1.4 4 1.3 High School Graduate / GED 11 15.1 3 4.0 8 9.3 4 5.6 26 8.5 Some College or 2 year degree 21 28.8 25 33.3 26 30.2 11 15.3 83 27.1 4 year college graduate 21 28.8 21 28.0 23 26.7 11 15.3 76 24.8 More than 4 year degree 17 23.3 24 32.0 29 33.7 45 62.5 115 37.6 No answer 0 0.0 1 1.3 0 0.0 0 0 1 0.3
80 Table 4 1. Continued Millennials Generation X Baby Boomers Silent Total ( n =73) ( n =75) ( n =86) ( n =72) ( n =306) n % n % n % n % n % Employment Full time 53 72.6 61 81.3 47 54.7 21 29.2 182 59.5 Part time 15 20.5 7 9.3 9 10.5 5 6.9 36 11.8 Unemployed 5 6.8 5 6.7 1 1.2 0 0 11 3.6 Retired 0 0.0 0 0.0 27 31.4 42 58.3 69 22.5 Volunteering 0 0.0 0 0.0 1 1.2 2 2.8 3 1.0 No Answer 0 0.0 2 2.7 1 1.2 2 2.8 5 1.6 Annual household Income < 10,000 4 5.5 2 2.7 0 0.0 0 0 6 2.0 10,001 25,000 9 12.3 5 6.7 2 2.3 1 1.4 17 5.6 25,001 35,000 10 13.7 3 4.0 7 8.1 9 12.5 29 9.5 35,501 50,000 15 20.5 8 10.7 10 11.6 8 11.1 41 13.4 50,001 75,000 12 16.4 15 20.0 19 22.1 22 30.6 68 22.2 > 75,000 18 24.7 35 46.7 40 46.5 27 37.5 120 39.2 No Answer 5 6.8 7 9.3 8 9.3 5 6.9 25 8.2 Physical Activity Low 6 8.2 13 17.3 11 12.8 19 26.4 49 16.0 Moderate 25 34.2 31 41.3 40 46.5 30 41.7 126 41.2 High 37 50.7 29 38.7 33 38.4 19 26.4 118 38.6 No Answer 5 6.8 2 2.7 2 2.3 4 5.6 13 4.2
81 Table 4 2. e Health tool Use by Cohort Millenials Generation X Baby Boomers Silent M (SD) M (SD) M (SD) M (SD) Median Median Median Median [Range] [Range] [Range] [Range] Years using Computer 14.3 (4.3) 19.3 (6.5) 21.1 (9.1) 20.8 (14.0) 15.0 20.0 20.0 20.0 [4 24] [0 33] [0 40] [0 50] Mobile Phone 9.5 (2.7) 12.8 (4.5) 13.3 (6.6) 10.2 (8.2) 10.0 14.0 15.0 10.0 [2 16] [0 22] [0 32] [0 35] Internet 12.2 (3.0) 13.6 (5.3) 14.2 (6.2) 13.2 (8.0) 12.0 15.0 15.0 15.0 [4 20] [0 23] [0 30] [0 30] Social Networks 6.7 (2.8) 3.8 (4.2) 1.6 (2.0) 1.2 (3.7) 7.0 3.0 0.0 0.0 [0 14] [0 16] [0 7] [0 26] Video Games 8.5 (8.5) 11.9 (13.5) 2.9 (7.1) 2.6 (6.3) 0.0 0.0 0.0 0.0 [0 28] [0 35] [0 30] [0 25] Self Monitoring Devices 0.9 (1.8) 1.6 (2.7) 1.8 (4.5) 2.3 (5.8) 0.0 0.0 0.0 0.0 [0 10] [0 10] [0 25] [0 32] Weekly minutes of use Internet 945.3 (960.4) 518.0 (615.1) 510.5 (758.5) 349.6 (488.4) 600.0 360.0 200.0 210.0 [12 4000] [0 2600] [0 5040] [0 2400]
82 Table 4 2. Continued Millenials Generation X Baby Boomers Silent M (SD) M (SD) M (SD) M (SD) Median Median Median Median [Range] [Range] [Range] [Range] Social networks 395.7 (529.8) 130.3 (227.4) 44.1 (99.3) 75.8 (324.1) 200.0 45.0 10.0 0.0 [0 2520] [0 1500] [0 700] [0 2400] Video games 102.2 (323.5) 54.1 (125.3) 17.3 (63.0) 36.5 (114.5) 0.0 0.0 0.0 0.0 [0 2500] [0 840] [0 420] [0 600] Self monitoring tools 21.0 (46.8) 57.3 (193.0) 97.9 (741.5) 13.5 (44.1) 0.0 0.0 0.0 0.0 [0 30] [0 1440] [0 6720] [0 270] Weekly Minutes for exercise Internet 98.1 (343.1) 17.9 (31.5) 15.1 (32.8) 7.3 (23.4) 20.0 0.0 0.0 0.0 [0 2500] [0 180] [0 180] [0 180] Social networks 7.9 (19.5) 4.2 (12.4) 2.4 (12.9) 1.7 (14.3) 0.0 0.0 0.0 0.0 [0 120] [0 70] [0 60] [0 120] Video games 7.0 (22.7) 8.8 (28.7) 2.9 (17.8) 4.4 (27.3) 0.0 0.0 0.0 0.0 [0 120] [0 210] [0 150] [0 200] Self monitoring tools 54.9 (28.4) 88.1 (423.5) 106.0 (743.5) 51.3 (172.0) 0.0 0.0 0.0 0.0 [0 2400] [0 3600] [0 6720] [0 1080]
83 Figure 4 3. Social Cognitive Factors in e Health and Exercise Millennials Generation X Baby Boomers Silent M SD M SD M SD M SD Median Median Median Median [Range] [Range] [Range] [Range] Self efficacy Internet 9. 3 1. 2 9.0 1.4 8 0 2. 4 6.5 3.0 10.0 9.5 9.0 7.0 [5 10] [3 10] [0 10] [0 10] Internet for exercise 8.9 1. 7 8.3 2. 2 7. 8 2.7 6.1 3. 5 10.0 9.0 8.5 8.0 [2 10] [1 10] [0 10] [0 10] Social network 9.1 1.4 7. 3 3. 2 4. 6 3. 6 3. 7 3.3 10.0 9.0 5.0 3.0 [5 10] [0 10] [0 10] [0 10] Social network for exercise 6. 9 2.9 5. 9 3. 4 3. 9 3.6 2.9 3 0 7.0 6.0 3.0 2.0 [0 10] [0 10] [0 10] [0 10] Video game 7. 1 3.3 6.3 3. 3 3. 9 3.3 3.5 3.4 8.0 7.0 3.0 3.0 [0 10] [0 10] [0 10] [0 10] Video game for exercise 5.5 3.4 5. 7 3.3 3. 7 3.3 2.5 3.0 5.0 6.0 3.0 1.0 [0 10] [0 10] [0 10] [0 10] Self monitoring 6. 6 2. 9 6. 4 3.1 4. 8 3.3 4. 8 3.4 7.0 7.0 5.0 5.0 [0 10] [0 10] [0 10] [0 10] Self monitoring for exercise 6. 3 2. 9 6. 4 3.2 4. 8 3.6 4. 3 3.6 6.0 7.0 5.0 4.0 [0 10] [0 10] [0 10] [0 10] e Health tool use overall 32.0 6.2 28.9 8.3 21.2 9.9 18.2 10.2 34.0 29.0 21.0 19.0 [13 40] [5 40] [2 40] [0 40]
84 Figure 4 3. Continued Millennials Generation X Baby Boomers Silent M SD M SD M SD M SD Median Median Median Median [Range] [Range] [Range] [Range] e Health tool use for exercise 27. 6 8.0 26.2 9.4 20.1 10. 7 15.5 10.6 28.0 27.5 20.0 14.0 [2 40] [4 40] [0 40] [0 40] Exercise overall 54. 4 20.7 50.8 24.6 54.9 22.6 51. 2 28. 2 53.0 49.5 53.0 55.0 [0 90] [0 90] [0 90] [0 90] Exercise Barriers 2 4 0 6. 3 25.0 6. 9 23. 2 7.5 21.1 7.3 24.5 27.0 22.0 20.0 [14 37] [14 39] [14 39] [14 32] Outcome Expectations 34. 5 5.8 37.5 5. 6 37. 7 5. 8 34. 8 7. 9 38.0 37.0 37.0 37.0 [23 45] [22 45] [21 45] [13 45]
85 Table 4 4. Age Differences in e Health Use Millennials Generation X Baby Boomers Silent (N=73) (N=75) (N=86) (N=72) n n n n Chi Square Differences a Mean(SD) Mean(SD) Mean(SD) Mean(SD) P value Among Groups Median Median Median Median [Range] [Range] Range Range Total weekly minutes using e Health resources n=71 n=75 n=78 n=69 X 2 (3)=44.49 a > b, c, d 1480.69(1409.80) 759.73(843.04) 692(1142) 476.52(748.14) ( p <.0001) b > c 1020 515 285 300 [55 6795] [0 4440] [0 7710] [0 5070] Total weekly minutes using e Health resources for exercise n=71 n=74 n=78 n=69 X 2 (3)=31.75 a > c, d 171.52(447.39) 120.53(434.63) 128.06(765.92) 58.43(183.11) ( p <.0001) b > c, d 36 30 0 0 [0 2520] [0 3670] [0 6750] [0 1110] a Significant differences between the groups using Wilcoxon pairwise comparison with Bonferroni adjustment ( p < .008). a (Millennial), b (Generation X), c (Baby Boomers) and d (Silent Generation)
86 Figure 4 1. Survey flowchart
87 CHAPTER 5 DISCUSSION Technological advances are changing the way in which health interventions are being delivered. From innovative and engaging web based virtual coaches, to mobile wireless devices, smart phones and trackers, people are bein g exposed to new tools to improve their health. In fact, technology mediated health behavior and self monitoring is expected to grow to a multi billion dollar industry by 2015 (Jackson, 2011). For example, projected revenue from just one component of e Hea lth, mobile health applications, is expected to quadruple to 400 million annually by 2016 (Empson, 2011). Another component of e Health, disease specific wireless self monitoring, is expected to grow as health care providers begin prescribing these tools to their patients (Topol, 2012). Exercise is a well known behavior that influences health and well being and e Health tools are being used more frequently to support it. In addition, evidence for the relationships between social cognitive concepts and e xercise has been well documented. Some of the most downloaded mobile health applications today are related to fitness and frequently video games are being marketed for exercise. Social cognitive theory has been applied in multiple situations with varying effects, however a paucity of social cognitive research is focused on the relationship with e Health tools. This research study was designed to explore age related differences in e Health tool use, identify whether their use was related to e Health tool self efficacy, and finally, to assess whether using these tools to support exercise was related to known social cognitive exercise concepts of self efficacy, perceived barriers and outcome expectations.
88 A mailed survey was used to gather data from a ran dom sample of participants in a national employer sponsored wellness program. Not surprisingly, the younger cohorts were less responsive to a traditional mailed survey than the Baby Boomers and Silent Generation. Using only a mailed survey was an attempt to prevent biasing the sample towards those individuals comfortable with Internet use, which was a potential barrier to older respondents and was a variable of interest in this study. The sample was primarily well educated, married, Caucasian females, which was consistent with the workplace wellness population from which the subjects were drawn and other findings which demonstrated that this demographic is more likely to participate in worksite wellness programs (Person, Colby, Bulove & Eubanks, 2010). There were two notable exceptions to the sample demographic trends noted. First, Millennials were primarily single w hich is consistent with the 15% marriage rate for this group in the US population (Madland & Teixeira, 2009). Second, unlike the proporti on of males in the United States population over 65 years of age, the Silent Generation cohort was primarily male. According to the US Census, there are 1.4 times more women over the age of 65 than men, thus the ratio of men to women in this study was not consistent with the overall population thereby limiting the generalizability of these findings to the U.S. population (U.S. Census, 2010). Individual e Health Tool Differences While increasing age was associated with decreased use of e Health tools in this study overall distinctions were seen in the technology preferred by different cohorts. Internet The Internet was by far the most popular e Health tool used by all groups in general as well as to support exercise and was negatively related to age. Previous
89 studies found a similar positive relationship between using the Internet for health related activities and age (McInnes, Gifford, Kazis & Wagner, 2010; McMillan & Macias, 2008). By comparison, lower Internet use both overall and to support exerci se was related to greater age in the present study. Possibly the definition of use of the Internet to support exercise rather than more broadly related to health information seeking in the present study may have masked the positive relationship seen by o thers. A broader definition of health activities would likely have included activities more relevant to older individuals such as disease management or illness prevention, a concern for these groups since disease burden increases with age. Social Networks was much higher than the older age groups. In this sample, 97.3% of Millennials report ed social network use that was 1.3 times higher than Generation X (76%), 1.8 times higher than B aby Boomers (54.8%) and 3.4 times higher than Silent Generation (28.6%). By comparison, Pew researchers found that Millennials (75%) used social networking 1.5 times more frequently than Generation X (50%), 2.5 times higher than Baby Boomers (30%) and 12. 5 times higher than Silent Generation (6%) (2010). In the three years since the Pew study was conducted, social network use has continued to climb, and thereby may account for the higher rates of use seen in the present study. Regardless of the differences in the proportion of the generations using social networks, the negative relationship between age and social network use was evident and consistent with the declining relationship in general social network use seen in the Pew study. Overall, social network use to support exercise was low (12.7%) across all cohorts. Compared to social network use in general, its use as a tool to support
90 exercise was limited. In addition, social network use for exe rcise was negatively related to age, which supported earlier studies where increasing age was associated with lower social network use for health (Chou et al., 2009). It is important to note that while this negative relationship between social network use and health was seen in previous studies the health focus of the present study was exercise, thus the comparability of these findings is limited. Video Games In the present study, Generation X had the greatest number of years experience with video games and a greater proportion of them used video games to support exercise than any other group This finding was not surprising considering they were the first age cohort group to have access to video games at an early age. Timing of exposure to video games may influence their use (Boschman, 2010; Ijsselsteijn et al 2007; Pearce, 2008). While using video games to support exercise would appear to be a likely intervention for Generation X further study should be conducted to ascertain whether Millennials, wh o were exposed to video games even earlier than Generation X, use video games to exercise with increasing frequency over time. Finally, the upward trend in recreational and exercise video game use by Silent Generation seen in the present study should be e xplored further as it may hint to a growing acceptance of video game use by this age group. Self Monitoring Self monitoring was the most popular e Health tool used to support exercise in all but the Millennial group. Baby Boomers were not heavy users of e Health tools with one exception T hey were first in their weekly use of self monitoring tools overall and to support exercise. The rate of self monitoring tool use by Baby Boomers and Silent
91 Generation in this study was 29.5%, which was considerably higher than the rate of self monitoring tool use (11%) found by the Center for Technology and Aging for adults older than 50 (2011) and higher than the rate (27%) for all adults found by Fox (2010) The study sample, drawn from a group of worksite wellness participants, may have been more motivated to engage in health activities, possibly accounting for the difference in findings of the studies Self monitoring tools may be attractive to Baby Boomers and Silent Generation due to the relative simplicity of their design and use Pedometers have been used to track walking activity for many years and the user interface on these devices usually consists of few buttons and little extraneous information. These types of user interfaces are preferred by older ad ults and are important considerations when using an e Health tool for this group (Gerling et al., 2010). While not assessed in this study, testing for differences in Smartphone application based self monitoring tools versus more traditional pedometers in older adults would guide understanding about future development of self monitoring tools for Baby Boomers Possibly, age related increases in chronic disease and physical limitations make walking an attractive exercise option, which is the primary exercise tracked by pedometers. Finally, older adults tend to have more free time since they often do not work full time or are retired. Hence they can devote more time to walking and would be more likely to be interested in using these tools. Pattern s of Use Several patterns emerged in individual e Health tool use that were of interest. The weekly minutes of e Health tool use declined consistently with increased age : general Internet, Internet for exercise and social networking use for exercise. Both general and exercise self monitoring tool use increased with age, peaking with
92 Generation X and Baby Boomers before declining in the Silent Generation Finally, general social network use, general game use and games used for exercise declined with age until reaching the Silent Generation where they began to increase. This uptick in some e Health tool use at the older end of the age spectrum suggest s that older adults are in fact using these tools with increasing frequency that is consistent with earlier research with similar findings ( Entertainment Software Association, 201; Fox, 2010; Gerling, Schild & Mausch, 2010). In the case of social network use by the Silent Generation this increase may reflect the importance older adults place on maintaining social connections that have been observed by others (Barrett, 2011; Burnett, Mitzner, Charness & Rogers, 2011). Also, with advanced age, great er empha sis is placed on health, which may influence increased use of these tools. Overall e Health Differences Among Generations Different types of e Health tools may be used simultaneously, and significant differences among the cohorts were found for total wee kly minutes of all e Health tools. Post hoc comparisons among the cohorts revealed that Millennials utilized e Health tools more frequently than their older counterparts. Unexpectedly, Generation X's general e Health tool use was not significantly higher than Baby Boomers. In contrast, the Pew study (2010) revealed that Generation X was more similar to Millenials in using technology than their older counterparts The exact reasons why these differences exist is still unclear, however reasonable inferenc es based on the characteristics of each cohort may point to several possibilities. For instance, the age or context in which technology was first introduced may influence beliefs about the usefulness or necessity of technology use in general. Introduc tio n to technology at a younger age, while
93 formative processes are still active, may influence persistence of technology use in later life. A similar significant difference was observed in e Health tool use for exercise and subsequent comparisons were more consistent with expectations. Again, Millennials used e Health tools to support exercise significantly more than Baby Boomers and the Silent Generation ; nonetheless they were not significantly higher than Generation X that was suggested by the Pew study (2010). However, these findings conflict with Olsen, O'Brien, Roger & Charness's (2011) work that found that older adults used technology to support health more than younger adults. The difference may be related to the reason for e Health tool use. In the present study e Health tool use was focused on exercise, which is a specific subset of health behavior s rather than health in general. Study results might have been consistent with Olsen and associates if the target behavior was focused on disease mana gement as disease burden increases with age. Differences in Social Cognitive Factors Among Cohorts Sel f E fficacy for e Health Self efficacy for e Health tool use both overall and for exercise followed a consistent pattern that was negatively related t o age. Younger cohorts were generally more confident in their ability to use e Health tools in both contexts This pattern was also consistent with the age related declines in actual use of e Health tools described in above sections Specifically, the m inutes spent using e Health tools significantly decreased with greater age. Baby Boomers surpassed other groups in their years experience using computers, mobile phones and the Internet. Even though they had the most years experience with these technol ogies, their confidence and frequency of use was lower
94 than Millenials and Generation X. While social cognitive theory would suggest that years experience would increase confidence thereby prompting use, the pervasiveness of technology use by younger grou ps, the age that it was first experienced and frequency of use may be more influential on e Health efficacy beliefs (Bandura, 1997). Exercise Perceived exercise barriers were significantly lower for Millennials and Generation X than Baby Boomers and the Silent Generation Perceived barriers were not significantly higher between Generation X and Millennials and were not significantly higher between Baby Boomers and the Silent Generation Perceived exercise barriers can take different forms and are often d ependent on age, cognitive ability, disease burden and the environment. Thus, this was not an unexpected finding because perceived exercise barriers increase with age and declining physical abilities (Deshpande et al., 2008; Ham, Sloane, Warshaw, Bernard & Flaherty, 2007). By contrast, exercise self efficacy and outcome expectations were not significantly different among the age groups. This was somewhat surprising because exercise self efficacy would likely be higher in younger individuals who have not y et experienced age related changes and health concerns typically affecting the exercise self efficacy beliefs of older cohorts (Deshpande et al., 2008; Ham, Sloane, Warshaw, Bernard & Flaherty, 2007). Relationships Between Social Cognitive Factors and e H ealth Weekly minutes using e Health tools were significantly and positively related to the self efficacy for e Health tool use. This relationship is consistent with social cognitive theory that describes a positive relationship between self efficacy and the desired behavior (Bandura, 1997). Greater weekly minutes of e Health tool use overall
95 and specifically for exercise were associated with greater exercise self efficacy and outcome expectations f or exercise. As Bandura posited self efficacy is speci fic to a particular action or behavior (1997). This lends support to findings by others who suggest that using e Health tools can build self efficacy (Fukuoka et al., 2011; Greene, Choudhry, Kilabuk & Shrank, 2011). Increased outcome expectations for exer cise were significantly related to increased e Health tool use for exercise. Therefore, those who used e Health tools to support exercise had more positive expectations of the benefits of exercise overall. Even though no research has been done exploring t he relationship between outcome expectations for exercise and e Health tool use, these findings are theoretically consistent with relationships between outcome expectations and behavior (Bandura, 1997). The small negative relationship between weekly minut es of e Health tool use for exercise and perceived barriers was unexpectedly not significant. This was surprising since e Health tools such as exercise video games (Inzitari et al. 2009) and social networks (Maitland, 2011) have been reported to lower perceived barriers to exercise. The expected relationship was not seen even when assessing the relationship of perceived barriers with video games use for exercise, and social networ k use for exercise. It may be that the technological nature of these tools is in some way a barrier themselves though this has not been tested. Limitations Using a mailed only survey design increased the cost of the study and may account for the low res ponse rate. S ince the survey was in English non response bias may be a factor, as potential participants who were more comfortable using another
96 language may not have responded. Age related cohort differences in technology use identified in this study m ay account for the slow response by the younger cohorts Since the younger cohorts define themselves by their use of technology, using a n Internet survey methodology would be prudent and more likely to increase the response of these cohorts (Pew, 2010). M ixed method approaches that utilize both mailed and Internet surveys are reliable and should be considered if age based comparisons are of interest, though they must be tested for equivalence prior to deployment (Dillman et al., 2009). The survey respon se rate was low among all the cohorts and the sample may hence not be representative of the population. Also, the sample was fairly homogenous, composed primarily of Caucasian female English speakers, thereby limiting the generalizability of these finding s to other demographic and ethnic/racial groups. In addition, the sample was highly educated with higher incomes that may have increased their use of technology (Noar & Harrington, 2012). Those less educated or below the poverty level may not have access or understand how to use some types of technology may not be inclined to respond to surveys about technology use, however there was not enough data in each cohort to determine this. Comparisons with the Pew Center survey (2010) are not possible because d emographics of the responders and response rate were not reported. The sample demographics were consistent with the workplace wellness population from which the subjects were drawn and other findings which demonstrated that this demographic is more likely to participate in worksite wellness programs (Person, Colby, Bulove & Eubanks, 2010 ). Thus the sample was already likely to be engaged in wellness activities For example, unlike their age cohorts in the general
97 population, the participants were o verall moderately or highly physically active potentially biasing the results though this could not be confirmed due to the anonymous nature of the survey and the researcher being blinded to the sampling Being active already may have mitigated their perc eived exercise barriers or meant they were already efficacious related to exercise. The age distribution of members of the wellness program were not consistent with the U.S. population since this program was delivered primarily as an employer sponsored pr ogram. This study was focused on the health behavior of exercise and did not address other health related functions e Health tools could support. While these findings contribute to understanding e Health tool use overall and to support exercise, other pot ential health behavior contexts were not explored. For example, e Health tools can be used to monitor and manage chronic conditions such as diabetes while the present study only focus ed on exercise While the findings support a relationship between effica cy beliefs and e Health tools specific to exercise behavior, causality between e Health tool use and improved exercise self efficacy cannot be inferred in the present study Clinical Implications Often health promotion programs have limitations in size, scope or funding so targeting a health intervention using an e Health tool preferred by a particular age group would be useful. As these tools are increasingly being used to support health behavior, it is important to consider age differences in e Health tools to target them to a population with the greatest chance of being utilized. For example, Generation X and Millennials use social media and the I nternet more frequently than other older cohorts in the present study. Thus, the younger cohorts may be more responsive to these
98 technologies than the older ones Also, in this study, Generation X use d vi deo games more frequently and would be more likely respond more to these types of tools for exercise and health than older age groups. Even though older a dults are using technology more frequently, providing these tools to an older patient may present a barrier if age related differences such as self confidence using technology or technology preferences are not considered. This information may be especially useful in targeting interventions such as Internet campaigns for older adults or delivering an exercise intervention via a video game to younger groups. E Health technology is being used increasingly as a tool for disease support and self management. The overburdened healthcare system is bracing for the silver tsunami of Baby Boomer retirees who will begin to require health care services as they age. This confluence of technolog y and demand for health care provides opportunities and challenges to develop effective e Health strategies for greater self management and promote long term independence. Care must be taken to ensure that e Health tools used with ol der adults be considerate of their differences. Baby Boomers and the Silent Generation may perceive certain e Health technologies such as video games as a barrier rather than a helpful tool. In addition, while social networks to promote health may appeal to younger groups, older adults may not find them useful. As a group, Baby Boomers will force the healthcare system to change as there will not be capacity to meet their needs using current resources. Conversely, Baby Boomers and Silent Generation will n eed to explore personal health related technology use as it will continue to be deployed in clinical settings and extend into the home. In home self management of disease and health promotion will become critical to delivering
99 healthcare in the future so that inpatient, long term, skilled and assisted care will be minimized. Providers will continue to use e Health tools more frequently to manage and monitor their patients remotely. In addition these tools will provide new levels of connectivity to the pro vider. Designing e Health tools for specific populations should include consideration for the users age and e Health tool preferences. Creating an e Health tool for an older user should account for the age related visual and cognitive changes so that the experience can be less overwhelming and more enjoyable. Keeping the user interface clean, without extraneous details and simple, concise instructions will aid usability for older adults. In addition, older adults physical limitations such and mobili ty and balance may make some e Health tools difficult or uncomfortable to use. Having ergonomically designed controllers or games in which the player is able to set the difficulty or controller characteristics to suit their physical limitations would be b eneficial. While many e Health tools such as mobile applications are free and are being provided as part of existing condition support programs at no additional cost to the user, other technologies may be expensive, thereby reducing their appeal to those w ith limited incomes. I ncreasing self efficacy for e Health tool use may increase the likelihood the tool will be used. Social cognitive theory provides guidance on ways to improve self efficacy that can be adapted to improve e Health tool self efficac y. Health care professionals should familiarize themselves with the various types of e Health tools available, how they are used and which ones are effective. Providing a tool or prescribing one to be used without consideration for an individual's level of confidence would impact the efficacy of the e Health tool in supporting the intended behavior. A user's level of
100 confidence should be assessed prior to using an e Health tool as an intervention. If e Health self efficacy is low, then specific instruct ions, verbal encouragement or simpler initial tasks, which are easily mastered, will build confidence and familiarize persons with the tool and how it should be used. Therefore adequate training and social support are essential In the not too distant f uture, data from e Health tools will be integrated into the medical record and practitioners will need to understand the implications. It will become common to have patients who are using wireless devices, connected to mobile applications and managing con ditions at home Their data more than likely will be wirelessly connected to their provider, disease management nurse or health coach. Therefore these providers will be expected to support these new data streams and provide guidance to e Health tool user s in their practice. E H ealth tools should be designed so that they address the theoretical underpinnings of health behavior. E Health tools should include components that will improve user self efficacy through mastery experiences, vicarious learning, verbal persuasion and overcoming physical and emotional states. In addition, they should be constructed to reduce perceived barriers to their use and improve outcome expectations. They should include features suc h as task progression so that early mastery can support initiation of more challenging tasks and improve efficacy beliefs. For example, rather than an e Health tool immediately focusing on a complete dietary change, it should start with smaller tasks such as simple food journaling before progressing to more difficult changes. E Health tools should include self regulatory mechanisms so that the user can see and track their progress towards a goal.
101 Research Implications Using a mixed methods or Internet only survey approach es may increase response rates and mitigate non response bias found in the age groups in the present study If a technology based survey is used, then other potential confounds such as lack of Internet access can lead to response bias The digital divide, while shrinking, still impacts those whose characteristics may be of interest in e Health research, such as the poor or uneducated. Further research sampling random zip codes or populations such as local community residents not engaged in wellness activities might yield results more representative of the U.S. population. Despite the contribution to the evidence the findings of this research provide s further st udy of e Health tools is needed. W hile significant age related differences were found in use of e Health tools t here is not yet compelling evidence that experiential timing has an impact on the use of these tools. For example, older adults had greater years experience with many of these tools, however se lf efficacy and actual weekly use was much lower than younger groups. Future research is needed to explore the relationship between the age at which a person was first exposed to e Health tools and current usage patterns. Also, the expected relationship between e Health tool use for exercise and perceived exercise barriers was not seen. E xercise barriers such as an unsafe environment or lack of time should logically be mediated by an exercise intervention that could be done in the safety of the home. The observed upward trend in video game use by the Silent Generation seen in this study may hint to a growing acceptance of video game use by this group E Health tools such as video games are often marketed to be
102 a safe and convenient alternative to traditi onal exercise and research is needed whether these games reduce barriers to exercise There is limited data comparing the effectiveness of e Health tools constructed using a theoretical framework to those currently in the marketplace. Without research based evidence assessing the efficacy of e Health tools, consumers are left to the advertisements of manufacturers to determine whether a tool is safe and effective to use. Randomized controlled trials are also needed to determine the effectiveness of e Health tools. These will provide greater clarity about their efficacy and will begin to show that they do what they are advertised to do. Finally, very little work has been done about long term engagement with e Health tools. If these tools are used init ially to support health, do users continue to engage with them over time? Also, if e Health tools are effective, is this impact on behavior a long term benefit or do the effects lessen? In addition e Health tool use for exercise, additional research should be conducted to determine whether certain types of e Health tools are more effective for some behaviors over others. For example, is using a video game to impact exercise behavior as effective if it's used to control diabetes? Are self monitoring tools more effective for dietary management or medication adherence? Finally, this research has focused primarily on the independent use of e Health tools; however, as these technologies evolve and incorporate aspects from different e Health tools, does t his generate an additive effect on the target health behavior? Conclusion E Health tools show great promise such as broad reach, consistent interventions and reduced cost, however there are many unanswered questions. With such a rapid proliferation of available tools and the relatively short period since their application to
1 03 health, little is still known about the ways in which the public uses these tools, perceptions and beliefs about their use and whether or not they are effective. With over 40,000 he alth related applications available, it is easy to see how poor design can infiltrate consumer products, which at minimum won't work and at their worst, can injure (Cohn, 2012). As Noar and Harrington (2012) point out, this rapid expansion has not allowed sufficient time for scientific inquiry to determine which approaches are safe and effective. Demand for e Health tools will continue to increase as the U.S. population ages and the need for cost effective alternatives to manage health and disease are ne eded. In home wireless disease management monitoring, exercise gaming, tracking and integrated Smartphone applications are being used to improve health at a lower cost than traditional methods of self management. E Health tools should be used to improve h ealth though they should not be used indiscriminately. Some tools may be adopted and utilized with greater ease if appropriately applied. Furthermore, care must be taken to include behavioral science and clinical evidence in the development of these tool s. If these tools are not constructed in a way that supports health behavior theories, then their effectiveness may be hindered. Using a technology simply because it's in vogue or is "cool" is not reason enough to use it as a health intervention. As Ban dura points out, repeated failures can erode self efficacy (1997). If these tools don't work due to poor design or without consideration for the target population, the user may interpret their lack of progress as personal failure thereby impacting their s elf confidence. As this research shows, social cognitive theoretical concepts continue to be evident, even when using emerging
104 technologies. Thus, e Health tools that are intended to change or support health behavior should incorporate efficacy building mechanisms to enhance effectiveness.
105 APPENDIX A SURVEY
115 APPENDIX B RECRUITMENT LETTER Dear Sir or Madam: The Prevention Plan is pleased to assist Ashley Reynolds in a study of how adults use different types of technology for health. Mr. Reynolds is a doctoral candidate at the University of Florida and Senior Vice President for Health Services at The Prevention Plan. We are supporting this study by sending out a survey. The enclosed survey has questions about the types of technology you may use, how confident you are about using technology and how you exercise. There are no right or wrong answers. You will not have to answer any question you do not wish to answer There are questions on the front and back of each page. It should take about 15 minutes to complete. Please complete return the enclosed survey in the enclosed envelope. By completing and returning this survey you give permission to take part in the study. The survey is completely anonymous. There is no information that can link your responses to you. U.S. Preventive Medicine, your employer or Mr. Reynolds have no way of knowing whether you have returned the survey. U.S. Preventive Medicine and your employer or insurer will not h ave access to the study data. Participation will not affect your health insurance or other benefits. If you do not want to take part in the study, do not return the survey. It is completely voluntary. We appreciate your time and support this study. If you have any questions about this study please contact Ashley Reynolds at xxx xxx xxxx or his faculty supervisor, Dr. Beverly Roberts, who is a professor in the College of Nursing at xxx xxx xxxx Questions or concerns about your rights as a research partici pant rights may be directed to the IRB02 office, University of Florida, Box 112250, Gainesville, FL 32611; ( xxx ) xxx xxxx Paul Risner EVP U.S. Preventive Medicine The Prevention Plan The Prevention Plan XXXXX Gran Bay Parkway, Suite 2400 Jacksonville, FL 32258 Ph XXX XXX XXX www.ThePreventionPlan.com
116 REFERENCES American Psychological Association (2012). Older adults' h ealth and age related c hanges. Retrieved 6/25/12 from http://www.apa.org/pi/aging/resources/guides/older.aspx# Anderson Bill, E., Winett, R., Wojcik, J. & Winett, S. (2011 a ). Web Based guide to health: relationship of theoretical variables to change in physical activity, nutrition and weight at 16 months Journal Medical Internet Research 13 (1). doi: 10.2196/jmir.1614 Anderson Bill, S. E., Winett, A. R., & Wojcik, R. J. (2011 b ). Social cognitive determinants of nutrition and physical activity among web health users enrol ling in an online intervention: the influence of social support, self efficacy, outcome expectations, and self regulation Journal Medical Internet Research 13 (1), e28. Atki nson, L. N., Saperstein, L. S. & Pleis, J. (2009). Using the internet for healt h related activities: findings from a national probability sample Journal Medical Internet Research 11 (1), e4. Bandura, A. (1986). Social foundations of t hought and action: a social cognitive t heory Englewood Cliffs, NJ: Prentice Hall. Bandura, A. (1997 ). Self e fficacy, the e xercise of c ontrol New York: W. H. Freeman and Company. Bandura, A. (2004). Health promotion by social cognitive means Health Education and Behavior, 31 (2), 143 164. doi: 10.1177/1090198104263660 Baranowski, T. Buday, R., T hompson, D.I. & Baranowski, J. (2008). Playing for r eal: v ideo g ames and s tories for h ealth r elated b ehavior c hange American Journal of Preventive Medicine, 34 (1), 74 82. Barrett, L. (2011). Healthy @ Home 2.0: Full report. Retrieved 6/25/12 from http://assets.aarp.org/rgcenter/health/healthy home 11.pdf Bass, S. B., Ruzek, S. B., Gordon, T. F., Fleisher, L., McKeown Conn, N. & Moore, D. (2006). Relationship of Internet h ealt h i nformation u se w ith p atient b ehavior and s elf e fficacy: e xperiences of n ewly d iagnosed c ancer p atients w ho c ontact the n ational c ancer i nstitute's c ancer i nformation s ervice Journal of Health Communication, 11 (2), 219 236. doi: 10.1080/10810730500526794 Beaudin, S. J., Intille, S. S. & Morris, E. M. (2006). To t rack or n ot to t rack: u ser r eactions to c oncepts in l ongitudinal h ealth m onitoring. Journal Medical Internet Research 8 (4), e29.
117 Bickmore, T. W., Caruso, L., Clough Go rr, K., & Heeren, T. (2005 a ). It's just like you talk relational agents for older adults. Interacting with Computers, 17 (6), 711 735. doi: 10.1016/j.intcom.2005.09.002 Bickmore, T., Grub er, A. & Picard, R. (200 5 b ). Establishing the computer patient working alliance in automated health behavior change interventions. Patient Education and Counseling, 59 (1), 21 30. doi: 10.1016/j.pec.2004.09.008 Boschman, L. R. (2010). Exergames for adult users: a preliminary pilot study. Paper presented at the Proceedings of the International Academic Conference on the Future of Game Design and Technology, Vancouver, British Columbia, Canada. Brouwer, W., Oenema, A., Raat, H., Crutzen, R., de Nooijer, J., de Vries, N. K., & Brug, J. (2010). Characteristics of visitors and revisitors to an Internet delivered computer tailored lifestyle intervention implemented for use by the general public. Health Education Research, 25 (4), 585 595. doi: 10.1093/her/cyp063 Brown, Lieberman, Gemeny, Fan, Wilson & Pasta., (1997) Educational video game for juvenile diabetes care: Results of a controlled trial. Med ical Inform atics, 22 77 89 Bulmer, M., (1979). Principles of S tatistics Mineola, NY : Dover p.45 68. Burnett, J.S., Mitzner, T.L., Charness, N., & Rogers, W.A. (2011). Understanding Predictors of Computer Communication Technology Use by Older Adults. Proceedings of the Human Factors and Ergonomics Society 55th Annual Meeting. Las Vegas, NV: Human Factors and Ergonomics Society. Cassell, M., Jackson C ., & Cheuvront, B. (1998). Health c ommunication on the Internet: a n e ffective c hannel for h ealth b ehavior c hange? Journal of Health Communication, 3 (1), 71 79. doi: 10.1080/108107398127517 Ce nters for Disease Control. (2011 ). Physical activity and heal th: the benefits of physical activity. Retrieved from the web http://www.cdc.gov/physicalactivity/everyone/health/index.html Center for Technology and Aging (2011). mHealth technologies: applications to benefit older adults (pp. 1 38). Centola, D. (2011). An e xperimental s tudy of h omophily in the a doption of h ealth b ehavior. Science 334 1269 1272. Chodzko Zajko, Proctor, Singh, Minson, Nigg Salem et al. (2009, July). American college of sports medicine position stand. exercise and physical activity for older adults Medicine and Science in Sports and Exercise, 41 (7), 1510 1530. Retrieved September 8, 2009, from CINAHL Plus with Full Text database
118 Chou, S. Hunt M., Beckjord, B., Moser, P. & Hesse, B. (2009). Social m edia u se in the United States: i mplications for h ealth c ommunication. Journal Medical Internet Research 11 (4), e48. Chu, A. & Mastel Smith, B. (2010). The o utcomes of a nxiety, c on fidence, and s elf efficacy w ith Internet h ealth i nformation r etrieval in o lder a dults: a p ilot s tudy. CIN: Computers, Informatics, Nursing, 28 (4). Cohall, A. T., Nye, A., Moon Howard, J., Kukafka, R., Dye, B., Vaughan, R. D., & Northridge, M. E. (2011). Computer u se, Internet a ccess, and o nline h ealth s earching a mong H arlem a dults. American Journal of Health Promotion, 25 (5), 325 333. Cohn, M., (2012). Hopkins researchers aim to uncover which mobile health applications work Baltimore Sun. Retrieved 4 10 2013 from http://articles.baltimoresun.com/2012 03 14/health/bs hs mobile health apps 20120314_1_health apps app works mobile health Computer History Museum (2006). Timeline of c omputer h istory Retrieved 3/25/2012, from http://www.computerhistory.org/timeline/?category=cmptr Conn, V. S., Hafdahl, A. R., Brown, S. A., & Brown, L. M. (2008). Meta analysis of patient education interventions to increase physical activity among chronically ill adults. Patient Education and C ounseling, 70 (2), 157 172. doi: 10.1016/j.pec.2007.10.004 Craig, C., Marshall, A., Sjostrom, M., Bauman, A., Booth, M., Ainsworth, B., . Oja, P. (2003). International p hysical a ctivity q uestionnaire: 12 Country r eliability and v alidity 8. Retrieved 80 05433, mg8, 35, from http://ovidsp.ovid.com/ovidweb.cgi?T=JS&PAGE=referen ce&D=ovftf&NEWS=N&AN=00005768 200308000 00020 Craik, F & Salthouse, T. A. ( Ed ) (2000). Handbook of a ging and c ognition (2nd Ed.). Hillsdale, N.J.: Lawrence Erlbaum Associates. Daskapan, A., Tuzun, E., & Eker, L. (2006). P erceived b arriers to p hysical a ctivity in u niversity s tudents Journal of Sports Science and Medicine 5 615 620. De Cocker, K. A., De Bourdeaudhuij, I. M., & Cardon, G. M. (2008). The effect of pedometer use in combination with cognitive and behavioral support materials to promote physical activity. Patient Education and Counseling, 70 (2), 209 214. doi: 10.1016/j.pec.2007.10.008 Deloitte Center for Health Solutions. (2010). Social Networks in health care: communication collaboration and insights.
119 Der Ananian, C., Wilcox, S., Watkins, K., Saunders, R., & Evans, A. (2008). Factors a ssociated w ith e xercise p articipation in a dults w ith a rthritis. Journal of Aging and Physical Activity, 16 125 143. Deshpande, N., Metter, E., Lauretani, F., Bandinelli, S., Guralnik, J., & Ferrucci, L. (2008). Activity restriction induced by fear of falling and object ive and subjective measures of physical function: A prospective cohort study. Journal of the American Geriatrics Society, 56, 615 620. DOI: 10.1111/j.1532 5415.2007.01639 Dillman, D., Smyth, J. & Christian, L. (2009 ). Internet, m ail and m ixed m ode s urveys: the t ailored d esign m ethod (3 rd ed). Hoboken, New Jersey: John Wiley & Sons. Dunlap, J., & Barry, H. C. (1999). Overcoming exercise barriers in older adults. Physician and Sports M edicine, 27 (11), 69. Dunning, T. (2008 a ). Aging, activities, and the internet Activities, Adaptation and Aging 32 (2), 158 159. doi: 10.1080/01924780802256196 Dunning, T. (2008 b ). Dance, dance revolution (DDR). Activities, Adaptation and Aging 32 (3 4), 238 239. doi: 10.1080/01924780802563203 Echt, K.W., Morrell, R.W. & Park, D.C. (1998). The effects of age and training formats on basic computer skill acquisition in older adults. Educational Gerontology, 24 3 25. Elavsky, S., McAuley, E., Motl, R., Konopack, J., Marquez, D., Hu, L., . Die ner, E. (2005). Physical activity enhances long term quality of life in older adults: efficacy esteem, and affective influences. Annals of Behavioral Medicine, 30 (2), 138 145. doi: 10.1207/s15324796abm3002_6 Empson, R. (2011). Market For m obile h ealth a p ps p rojected t o q uadruple t o $400 m illion b y 2016. TechCrunch retrieved 4 1 13 from http://techcrunch.com/2011/11/29/market for mobile health apps projected to quadruple to 400 million by 2016/ Entertainment Software Association. (2011). Essential facts about the computer and video game industry. Retrieved 6/15/2012 from http://www.theesa.com/facts/pdfs/ESA_EF_2011.pdf Ferketich, S. (1991). Focus on psychometrics: aspects of item analysis. Research in Nursing and Health 14 165 168. Fogg, B., Eckles, D., (2007). Mobile persuasion : 20 perspectives on the future of behavior change Palo Alto: Stanford Captology Media.
120 Fox, S. (2012). What's the future for self tracking? Retrieved 2/5/2013 from http://www.pewinternet.org/presentations/2012/sep/medicine x.aspx Friedman, H., Martin, L., Tucker, J., Criqui, M., Kern, M. & Reynolds, C. (2008). Stability of p hysical a ctivi ty across the l ifespan. Journal of Health Psychology, 13 1092 1104. doi: DOI: 10.1177/1359105308095963 Fukuoka Y, K amitani E., Bonnet K, Lindgren T. (2011). Real time social support through a mobile virtual community to improve healthy behavior in overwe ight and sedentary adults: a focus group analysis Journal Medical Internet Research 13 (3). doi: 10.2196/jmir.1770 Glanz, K., Rimer, B., & Viswanath, K. (2008). Health behavior and health education, theory, research and practice (4 th ed). San Francisco, CA: Jossey Bass Gerling, K. M., Schild, J., & Masuch, M. (2010). Exergame design for elderly users: the case study of SilverBalance. Paper presented at the Proceedings of the 7th International Conference on Advances in Computer Enter tainment Technology, Taipei, Taiwan. Gracia, E. & Herrero, J. (2009). Internet u se and s elf r ated h ealth a mong o lder p eople: A n ational s urvey. Journal Medical Internet Research 11 (4), e49. Gr aves L Ridgers N. & Stratton G. (2008). The contribution of upper limb and total body movement to adolescents' energy exp endi ture whilst playing Nintendo Wii. European Journal of Applied Physiology, 104 : 617 23. Graves, L., Ridgers, N., Williams, K., Stratton, G., Atkinson, G., & Cable, N. T. (2010). The p hysiological c ost and e njoyment of Wii Fit in a dolescents, y oung a dults, and o lder a dults. Journal of Physical Activity and Health, 7 (3), 393 401. Graves L S tratton G Ridgers N & Cable N. (2007). Comparison of energy expenditure in ado lescents when playing new generation and sedentary computer games: cross sectional study. B ritish M edical J ournal 335 :1282 4. Gre ene J Choudhry N Kilabuk E. & Shrank W. (2011). Online social networking by patients with diabetes: a qualitative evaluation of communication with Facebook. Journal of General Internal Medicine 26 (3):287 292. [doi: 10.1007/s11606 010 1526 3] Haegerstrom Portnoy, G., Schneck, M. E., & Brabyn, J. A. (1999). Seeing into old age: Vision function beyond acuity. Optometry and Vision Science, 76 (3), 141 158.
121 Hall, K., Crowley, G., McConnell, E., Bosworth, H., Sloane, R., Ekelund, C., & Morey, M. (2010). Change in g oal r atings as a m ediatin g v ariable b etween s elf efficacy and p hysical a ctivity in o lder m en. Annals of Behavioral Medicine, 39 (3), 267 273. doi: 10.1007/s12160 010 9177 5 Ham, Sloane, Warshaw, Bernard & Flaherty. (2007). Primary care geriatrics: a case based approach Philadelphia, PA: Mosby. Hedden, T. & Gabrieli, J. (2004). Insights into the ageing mind: a view from cognitive neuroscience. National Review of Neuroscience, 5 (2), 87 96. Heesch, K. C., Masse, L. C., Dunn, A. L., Frankowski, R. F., & Mullen, P. D. (2003 ). Does a dherence to a l ifestyle p hysical a ctivity i ntervention p redict c hanges in p hysical a ctivity? [Article]. Journal of Behavioral Medicine, 26 (4), 333. Hollis, J. F., Gullion, C. M., Stevens, V. J., Brantley, P. J., Appel, L. J., Ard, J. D., . Sv etkey, L. P. (2008). Weight l oss d uring the i ntensive i ntervention p hase of the w eight l oss m aintenance t rial. American Journal of Preventive Medicine, 35 (2), 118 126. doi: 10.1016/j.amepre.2008.04.013 Hortz, B. & Petosa, R. (2008). Social c ognitive t heory v ariables m ediation of m oderate e xercise. American Journal of Health Behavior, 32 (3), 305 314. Hunter, C. M., Peterson, A. L., Alvarez, L. M., Poston, W. C., Brundige, A. R., Haddock, C. K., . Foreyt, J. P. (2008). Weight m anagement u sing the I nternet: a r andomized c ontrolled t rial. American Journal of Preventive Medicine, 34 (2), 119 126. doi: 10.1016/j.amepre.2007.09.026 Hwang, K. O., Ottenbacher, A. J., Lucke, J. F., Etchegaray, J. M., Graham, A. L., Thomas, E. J., & Bernstam, E. V. (2010). M easuring s ocial s upport for w eight l oss in an Internet w eight l oss c ommunity. Journal of Health Communication, 16 (2), 198 211. doi: 10.1080/10810730.2010.535106 Inzitari, M., Greenlee, A., Hess, R., Perera, S., & Studenski, S. A. (2009). Attitudes of p ost menopausal w omen toward i nteractive v ideo d ance for e xercise. Journal of Women's Health, 18 (8), 1239 1243. doi: doi:10.1089/jwh.2008.1176 IJsselsteijn, W.A., Nap, H., de Kort, Y.A.W., & Poels, K. (2007). Digital g ame d esign for e lderly u sers, Proceedings of Futureplay 17 22. Jackson, S. (2011). Digital h ealth market pegged to reach 5.7 b illion by 2015. Fierce Mobile Healthcare Retrieved 4 10 2013 from http://www.fiercemobilehealthcare.com/story/digital health market pegged reach 57b 2015/2011 02 08
122 J ahns, R. & Gair, G. (2012). US$ 1.3 billion: The market for mHealth applications in 2012. Mobile Health Market Report, 2012 2016. Retrieved 6/26/12 from http://www.research2guidance.com/us 1.3 billion the market for mhealth applications in 2012/ Kautiain en, S., Koivusilta, L., Lintonen, T., Virtanen, S. M., & Rimpel, A. (2005). Use of information and communication technology and prevalence of overweight and obesity among adolescents International Journal of Obesity, 29 (8), 925 933. doi: 10.1038/sj.ijo.080 2994 Kochunov, P., Thompson, P., Lancaster, J., Bartzokis, G., Smith, S., Coyle, T., Royall, D., Laird A. & Foxa P (2007). Relationship between white matter fractional anisotropy and other indices of cerebral health in normal aging: tract based spatial statistics study of aging. NeuroImage, 35 478 487. Koch Weser, S., Bradshaw, Y. S., Gualtieri, L., & Gallagher, S. S. (2010). The Internet as a h ealth i nformation s ource: FINDINGS from the 2007 h ealth i nformation n ational t rends s urvey and i mplications for h ealth c ommunication. Journal of Health Communication, 15 (sup3), 279 293. doi: 10.1080/10810730.2010.522700 Kreps, G. L., & Neuhauser, L. (2010). New directions in eHealth communication: opportunities and challenges. Patient Edu cation and Counseling, 78 (3), 329 336. doi: 10.1016/j.pec.2010.01.013 Lanningham Foster L Foster R C McCrady S K Jensen T B Mitre N Levine JA. (2009). Activity promoting video games and in creased energy expenditure. J ournal of Pediatr ics, 154 819 23. Lees, F., Clark, P., Nigg, C., & Newman, P. (2005). Barriers to exercise behavior among older ad ults: a focus group study. Journal of Aging and Physical Activity, 13 (1), 23 33. Leveille, S. G., Cohen Mansfield, J., & Guralnik, J. M. (2003). Th e impact of chronic musculoskeletal pain on exercise attitudes, self efficacy, and physical activity. Journal of Aging and Physical Activity, 11 (2), 275 283. Li, K. & Counts, S. (2007). Exploring social interactions and attributes of casual multiplayer m obile gaming. Proc eedings of the 4th Intl. Conf erence on Mobile Technology, Applications and Systems 696 703. Lieberman, D., (1997). Interactive video games for health promotio n: effects on knowledge, self efficacy, social support, and health I n Stree t RL, Gold WR, Manning T ( E ds .): Health promotion and interactive technology: theoretical applications and future directions Hillsdale, NJ, Lawrence Erlbaum, pp. 103 120.
123 Lieberman, D. A., Chamberlin, B., Medina, E., Franklin, B. A., Sanner, B. M., & Vafiadis, D. K. (2011). The Power of p lay: Innovations in g etting a ctive s ummit 2011. Circulation, 123 (21), 2507 2516. doi: 10.1161/CIR.0b013e318219661d Maitland, J. (2011). Towards negotiation as a framework for health promoting technology. SIGHIT Rec or d 1 (1), 10 19. doi: 10.1145/1971706.1971710 Marks, R., & Allegrante, J. P. (2005). Chronic o steoarthritis and a dherence to e xercise: A r eview of the l iterature. Journal of Aging and Physical Activity, 13 (4), 434. Marshall, A., Hunt, J., & Jenkins, D. ( 2008). Knowledge of and preferred sources of assistance for physical activity in a sample of urban indigenous Australians. International Journal of Behavioral Nutrition and Physical Activity, 5 (1), 22. Mashable, (2012). Facebook hits one billion active u sers. Retrieved 11/1/2012 from http://mashable.com/2012/10/04/facebook one billion/ McAuley, E., Jerome, G., Marquez, D., Elavsky, S., & Blissmer, B. (2003). Exercise self efficacy in older adults: Social, affective, and behavioral influences. Annals of Behavioral Medicine, 25 (1), 1 7. doi: 10.1207/s15324796abm2501_01 McAuley, E., Jerome, G. J., Elavsky, S., Marquez, D. X., & Ramsey, S. N. (2003). Predicting long term maintenance of physical activity in older adults. Preventive Medicine, 37 (2), 110 118. doi: 10.1016/s0091 7435(03)00089 6 McAuley, E., Morris, K. S., Motl, R. W., Hu, L., Konopack, J. F., & Elavsky, S. (2007). Long term follow up of physical activity behavior in older adults. Health Psychology, 26 (3), 375 380. doi: 10.1037/0278 6184.108.40.206 5 McInnes, D. K., Gifford, A. L., Kazis, L. E., & Wagner, T. H. (2010). Disparities in health related internet use by US veterans: results from a national survey. Informatics in Primary Care, 18 (1), 59 68. McMillan, S. J., & Macias, W. (2008). Strengthening the s afety n et for o nline s eniors: Factors i nfluencing d ifferences in h ealth i nformation s eeking a mong o lder Internet u sers. Journal of Health Communication, 13 (8), 778 792. doi: 10.1080/10810730802487448 Millen, J., & Br ay S. (2009). Promoting s elf efficacy and o utcome e xpectations to e nable a dherence to r esistance t raining a fter c ardiac r ehabilitation. Journal of Cardiovascular Nursing, 24 (4), 316 327. Miyachi, M., Yamamoto, K., & Ohkawara, K. (2010). METs in a dults w hile p laying a ctive v ideo g ames: a m etabolic c hamber s tudy Medicine and Science in Sports and Exercise, 42 (6), 1149 1153. doi: 10.1249/MSS.0b013e3181c51c78
124 Mokdad, A., Marks, J., Stroup, D. & Gerberding, J. (2004). Actual causes of death in the United States, 2000. Journal of the American Medical Association, 291 1238 1245. doi:10.1001/jama.291.10.1238. Morrell, R. (2002). Older a dults, h ealth i nformation, and the w orld w ide w eb Mahwah, NJ: Lawrence Erlbaume Associates. Murray, T., Rodgers, W., & Fraser, S. (2011). Exploring the relationship between socioeconomic status, control beliefs and exercise behavior: a multiple mediator model. Journal of Behavioral Medicine 1 11. doi: 10.1007/s108 65 011 9327 7 Nahm, E., Resnick, B. & Gaines, J. (2004). Testing the r eliability and v alidity of c omputer mediated s ocial s upport m easures a mong o lder a dults: A p ilot s tudy. CIN: Computers, Informatics, Nursing, 22 (4), 211 219. Nelson, M., Rejeski, J., Blair, S., Duncan, P., Judge, J., King, A., Macera, C., Castaneda Sceppa, C. (2007). Physical a ctivity and p ublic h ealth in o lder a dults: recommendation from the A merican College of Sports Medicine and the American Heart Associatio n. Medicine and Science in Sports and Exercise 1435 1445. doi: 10.1249/mss.0b013e3180616aa2 Neupert, S. D., Lachman, M. E., & Whitbourne, S. B. (2009). Exercise s elf e fficacy and co ntrol b eliefs: Effects on e xercise b ehavior a fter an e xercise i nterventio n for o lder a dults. Journal of Aging and Physical Activity, 17 (1), 1 16. Noar, S. & Harrington, N. (2012). eHealth a pplications: Promising s trategies for b ehavior c hange New York, NY:Routledge. Nitz, J. C., Kuys, S., Isles, R., & F u, S. (2010). Is the Wii Fit a new generation tool for improving balance, health and well being? A pilot study. Climacteric, 13 (5), 487 491. doi: 10.3109/13697130903395193 Nunnaly, J. C., & Bernstein, I. H. (1994). Psychometric theory Sydney: McGraw Hill Office of Disease Prevention and Health Promotion, (2006). Expanding the reach and impact of consumer e health tools. Retrieved 2/5/2013 from http://www.health.gov/communication/ehealth/ehealthTools/pdf/ehealthreport.pdf Rogers, W. A., & Charness, N. (2011). Diffusion of technology: Frequency of use for younger and older adults. Ageing International, 36 (1), 123 145. Pajares, F. (2002). Overview of s ocial c ognitive t heory and s elf efficacy. Retrieved 2/28/2010 from http://www.des.emory.edu/mfp/eff.html
125 Patrick, K., Griswold, W. G., Raab, F., & Intille, S. S. (2008). Health and the m obile p hone. American Journal of Preventive Medicine, 35 (2), 177 181. doi: 10.1016/j.amepre.2008.05.001 Pearce, C. (2008). The truth about baby boomer gamers: a study of over forty computer game players Games and Culture, 3 (142). doi: DOI=10.1177/1555412008314132 Person, A., Colby, S., Bulova, J. & Eubanks, J., (2010). Barrier s to participation in a worksite wellness program. Nutrition Research and Practice, 4 (2). 149 154. P ew Research Center (2010). MILLENNIALS: confident. connected. open to change Retrieved 3/29/2012, from http://pewsocialtrends.org/files/2010/10/millennials confident connected open to change.pdf Portney, L. & Watkins, M. (2009). Foundations of c linical r esearch: Applications to p ractice (3 rd E d ). Upper Saddle River, NJ: Pearson/Prentice Hall. Rejeski, W., Tian, L., Liao, Y., & McDermott, M. (2008). Social c ognitive c onstructs and the p romotion of p hysical a ctivity in p atients w ith p eripheral a rtery d isease. Journal of Cardiopulmonary Rehabilitatio n and Prevention, 28 65 72. Resnick, B., Zimmerman, S., Orwig, D., Furstenberg, A., and Magazine, J. (2001). Model t esting for r eliability and v alidity of the o utcome e xpectations for e xercise s cale. Nursing Research, 50 (5), 293 299. Resnick, B. & Nigg, C (2003). Testing a t heoretical m odel of e xercise b ehavior for o lder a dults. Nursing Research, 52 (2), 80 88. Resnick, B., Luisi, D., Vogel, A. & Junaleepa, P. (2004). Reliability and v alidity of the s elf e fficacy for e xercise and o utcome e xpectations for e xercise s cales with m inority o lder a dults. Journal of Nursing Measurement, 12 235 248. doi: 10.1891/jnum.12.3.235 Resnick, B., Luisi, D., & Vogel, A. (2008). Testing the s enior e xercise s elf efficacy p roject (SESEP) for u se with u rban d w elling m inority o lder a dults. Public Health Nursing, 25 (3), 221 234. doi: 10.1111/j.1525 1446.2008.00699.x Resnick, B., Orwig, D., Yu Yahiro, J., Hawkes, W., Shardell, M., Hebel, J. R., . Magaziner, J. (2007). Testing the e ffectiveness of the e xercise p lus p rogram in o lder w omen p ost h ip f racture. Annals of Behavioral Medicine, 34 (1), 67 76. doi: 10.1080/08836610701495367 Ryan, R., Rigby, C., & Przybylski, A. (2006). The m otivational p ull of v ideo g ames: A s elf d etermination t heory t pproach Motivation and Emotion, 30 (4), 344 360. doi: 10.1007/s11031 006 9051 8
126 Ryder, N. (1965). The cohort as a concept in the study of social change. American Sociological Review, 30 (6), pp. 843 861 Salthouse, T. A. (2004). What and when of cognitive aging. Cu rrent Directions in Psychological Science, 13 140 144. Sechrist, K. R., Walker, S. N., & Pender, N. J. (1987). Development and psychometric evaluation of the exercise benefits/barriers scale. Research in Nursing & Health, 10 (6), 357 365. doi: 10.1002/nur .4770100603 Siegel, S., Haddock, B., Dubois, A. & Wilkin, L. (2009). Active v ideo/ a rcade g ames ( e xergaming) and e nergy e xpenditure in c ollege s tudents. International Journal of Exercise Science, 2 (3), 165 174. Sijuwade, P. (2009). Attitudes towards old age: a study of the self image of aged. Studies of Home Community Science, 3 (1), 1 5. Singh, H., Fox, A. S., Petersen, J. N., Shethia, A., & Street, L. R. (2009). Older p e nthusiasm to u se e lectronic m ail to c ommunicate w ith t heir p hysicians: cross s ectional s urvey. Journal Medical Internet Research 11 (2), e18. Song, R., Ahn, S., Roberts, B., Lee, E., & Ahn, Y. (2009). Adhering to a t'ai chi program to improve glucose control and quality of life for individuals with type 2 diabetes. Journal of Alternative and Complementary Medicine, 15 (6), 627 632. Retrieved from CINAHL database. Song, H., Peng, W., & Lee, K. M. (2011). Promoting e xercise s elf e fficacy w ith an e xergame. Journal of Health Communication, 16 (2), 148 162. doi: 10.1080/10810730.2010.535107 Svensson, M., & Lagerros, Y. T. (2010). Motivational techno logies to promote weight From Internet to gadgets. Patient Education and Counseling, 79 (3), 356 360. doi: 10.1016/j.pec.2010.03.004 Telama, R., Yang, X., Laakso L., & Viikari, J. (1997). Physical activity in childhood and adolescence as predictor of physical activity in young adulthood. American Journal of Preventive Medicine, 13 (4), 317 323. Telam a, R., Yang, X., Viikari, J., Valima ki, I., Wanne, O., & Raitak ari, O. (2005). Physical activity from childhood to adulthood: a 21 year tracking study. American Journal of Preventive Medicine, 28 (3), 267 273. doi: 10.1016/j.amepre.2004.12.003
127 Thalheim er, W., & Cook, S. (2002 ). How to calculate effect sizes from publi shed research articles: a simplified methodology. Retrieved February 5, 201 2 from http://work learning.com/effect_sizes.htm. Thomas, R., Cahill, J., & Santilli, L. (1997). Using an i nteractive c omputer g ame to i ncrease s kill and s elf e fficacy r egarding s a fer s ex n egotiation: Field t est r esults. Health Education and Behavior, 24 (1), 71 86. doi: 10.1177/109019819702400108 Thompson, P. D., Buchner, D. Pin a, I. L., Balady, G. J., Williams, M. A., Marcus, B. H., . Wenger, N. K. (2003). Exercise and p hysical a ctivity in the p revention and t reatment of a therosclerotic c ardiovascular d isease. Circulation, 107 (24), 3109 3116. doi: 10.1161/01.cir.0000075572.40158.77 Topol, E., (2012). The creative destruction of medicine: how the digital revolution will c reate better healthcare Perseus: NY, NY. Umstattd, M., & Hallam, J. (2007). Older a e xercise b ehavior: roles of s elected c onstructs of s ocial c ognitive t heory. Journal of Aging and Physical Activity, 15 206 218. U.S. Census (2010). U.S. p opulation by age, July 2010. Retrieved 4 10 13 from http://www.aoa.gov/AoARoot/Aging_Statistics/Census_Population/census2010/In dex.aspx U.S. Department of Health and Human Services. (2012). What is e Health? Retrieved 6/26/2012 from http://www.healthit.gov/patients families/e health Warburton, D. R., Bredin, S. D., Horita, L. L., Zbogar, D., Scott, J. M., Esch, B. A., & Rhodes, R. E. (2007). The health benefits of interactive video game exe rcise. Applied Physiology, Nutrition and Metabolism, 32 (4), 655 663. doi: doi:10.1139/H07 038 Watson, A., Bickmore, T., Cange, A., Kulshreshtha, A., & Kvedar, J. (2012). An Internet b ased v irtual c oach to p romote p hysical a ctivity a dherence in o verweight a dults: r andomized c ontrolled t rial. Journal Medical Internet Research 14 (1), e1. Webb, L. T., Joseph, J., Yardley, L., & Michie, S. (2010). Using the Internet to p romote h ealth b ehavior c hange: A s ystematic r eview and m eta analysis of the i mpact of t heoretical b asis, u se of b ehavior c hange t echniques, and m ode of d elivery on e fficacy. Journal Medical Internet Research 12 (1), e4. White, S. M., Wojcicki, T. R., & McAuley, E. (2011). Social c ognitive i nfluences on p hysical a ctivity b ehavior in m iddle a ged and o lder a dults. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences doi: 10.1093/geronb/gbr064
128 World Health Organization. (2011). mHealth New horizons for health through mobile technologies. Retrieved 6/15/2012 from ht tp://www.who.int/goe/publications/goe_mhealth_web.pdf Xie, B. (2009). Older a dults' h ealth i nformation w ants in the i nternet a ge: Implications for p atient p rovider r elationships. Journal of Health Communication, 14 (6), 510 524. doi: 10.1080/10810730903089614 Zickuhr, K. & Smith, A. (2012). Digital differences. Retrieved 5/12/2012 from http://www.pewinternet.org/Reports/2012/Digital differences.aspx Zizzi, S., Goodrich, D., Wu, Y., Parker, L., Rye, S., Pawar, V., . Tessaro, I. (2 006). Correlates of p hysical a ctivity in a c ommunity s ample of o lder a dults in a ppalachia. Journal of Aging and Physical Activity, 14 (4), 423 438.
129 BIOGRAPHICAL SKETCH Ashley Reynolds is a Florida native who currently resides in Port St. Lucie Florida wi th his husband and three dogs. He is a registered nurse and has been in the healthcare industry for twenty years. He is employed as a senior vice president for health services at Sensei Health He obtained his associate s degree in nursing at Florida Co mmunity College at Jacksonville and his bachelor s and master s degrees in nursing from Jacksonville University.