In recent years, occupational health promotion researchers and practitioners have placed increased attention to the concept of facilitating organizational cultures that support employee health as opposed to approaches based on participation in individual programs. A workplace culture that supports health entails a broader population-based approach to health promotion that seeks to impact attitudes, thoughts, and behaviors related to health independent of participating in specific well-being programs. A culture of health includes having a leadership vision of employee health and well-being as an organizational priority that is supported by policies and supervisors. Health-related behaviors are also practiced and encouraged among coworkers.
Well-being programs and participation contribute to a culture of health. However, beyond the wellness programs, health and healthy behaviors are also part of the larger social fabric of the organization. Thus, a culture of health supports the entire population not just those participating in specific well-being programs. This broader approach to health promotion combining population-based strategies with those that leverage aspects of the culture is a relatively new approach for improving employee health in an occupational health promotion field. Such a population-based approach that leverages culture and the social ecological model1 has received much attention from a number of organizations that were not satisfied with the efficiency of health care cost and low participation in individualized wellness programs. Several studies in occupational health promotion describe that a population-based strategy plays a significant role in improving employee health.2–4 However, the empirical evidence on the association between a worksite culture of health and employees’ health is still lacking. Others have examined the effect of specific strategies (eg, the link between stair use promotion and physical activity)5–6 on outcomes of interest. To our best knowledge, there has been no empirical study to date that shows whether an overall worksite culture influences employee health, especially with consideration of social networks and well-being initiatives designed to enhance employee health.
Although the level of support in the workplace culture for health has been understood as being fundamental for well-being programs and individual success, there has been a paucity of research in this area.7–8 One reason has been the lack of validated tools in peer-reviewed literature to measure support for health in the culture.9 This study utilizes the Workplace Culture of Health (CoH) scale which has been shown to have satisfactory reliability and validity characteristics and has been the source of previous research in peer-reviewed literature. Workplace CoH scale scores have been directly associated with job satisfaction and inversely related to intention to leave10 and body mass index.11 Unpublished work shows higher Workplace CoH scores associated with higher work engagement and lower stress. Although currently not widely established in the occupational health field, there are several projects underway at multiple academic institutions utilizing the tool.
Workplace health promotion proponents have long purported that well-being initiatives improve social ties, sense of support from employer, and social capital, yet few studies have researched this specifically. To unpack the underlying dynamics and understand associations between social capital, workplace culture supporting health, and employee health, we draw the concept of social capital from the sociology literature. Social capital is broadly defined as “an asset that inheres in social relations and networks.”12 That is, individuals can access resources gathered through relationships among people. Such concept of social capital applies well to an organizational context.13 Employees enjoy benefits of social capital when the organization facilitates development of stronger social ties and networks. Specific to this study, the relationships fostered through sports participation contribute to strengthening social capital14 and health behaviors.15 Kawachi and Berkman reintroduced the concept of social capital in public health.16–17 The authors illustrated that higher levels of social support buffer psychological distress and mitigate damaging effects on mental health.17 Consistent with those findings, Oksanen et al showed that a low level of social capital is detrimental to employees’ health.18
Occupational stress is known to impact physical health. Chen et al reported that lower levels of stressors are associated with better self-rated health.19 In related studies, the direction of job level as a moderating effect is unclear. For example, some studies find that employees with high occupational position tend to have greater job control20 and good mental health status.21 The higher level position is occupied by the ones who conduct work precisely and rapidly. Such higher level employees are more resistant to stress than lower ones and manage more work.22 On the contrary, others find that employees in managerial levels often have a variety of work-relate demands, which could cause stronger stress.23–24
Kawachi and Berkman suggested the need to examine various avenues that yield social capital to better understand the relationship between social capital and health.17 Thus, the present study considers the worksite context as a source of social capital and focuses on the relationship between a worksite cultural support for health and employee health with three main objectives. First, as discussed earlier, a population-based strategy in occupational health promotion is a substantively important topic. The addition of metrics pertaining to workplace culture of health and the effects of CoH on employee health should be of great interest to scholars as well as employers in organizational health management. We extend the work of Kawachi and Berkman's with empirical evidence that efforts to establish a workplace culture of health enhance employee health directly though instrumental support and indirectly by facilitating higher levels of social capital and lower levels of occupational stress. Third, we further extend the literature by exploring moderating effects of sex on the social capital–self-rated health relationship and of job level (managerial employees vs nonmanagerial employees) on the occupational perceived stress and self-rated health relationship. Previous work using the workplace CoH scale showed that job level was a mediator in the relationships between perceived culture of health, job satisfaction, and intention to leave.10 Our findings will aid worksite health promotion program directors in understanding the influence of social capital associated with participation of sports events on self-rated health and the impact of culture of health and individual factors on the proposed relationships.
Setting and Well-Being Initiatives
The study was conducted at one of the largest life insurance companies in South Korea with approximately 4100 employees. The company initiated well-being in 2014 with goals of improving employee health and the workplace culture for supporting health. The programs included but were not limited to no-smoking policies, healthy foods in the cafeteria, and signs to promote stair use. As a part of the initiative, the company also encouraged its employees to participate voluntarily in sports (eg, grassroots sports events such as local running and walking races). Events were monthly and the cost was partially subsidized. In addition, the company awarded prizes (eg, laptop, individual PTO days, sports wearables) based on participating. Goals of the initiative were multifold. First was to improve employee health across the whole population (not just people participating in specific programs). Second was to create a workplace culture that supports health. Beyond traditional well-being initiatives of no-smoking policies, encouragement of physical activity, and healthy eating, the sports programs facilitated meeting and socializing with coworkers and others during events, thus influencing the larger culture.
We collected data via a survey tool during the first quarter of 2017. The study data were limited to employees in its headquarter and branches in Seoul. The survey was self-administered online, and a random drawing incentive (gift cards) was offered to encourage participation. The e-mail addresses of all employees were not accessible. Authors first sent the link to the questionnaire to managers, and then they forwarded it to employees. The respondents were informed that the survey would remain completely anonymous to encourage honest responses. The survey was sent to 902 full-time employees working in the Seoul area; there were 816 total survey respondents, of which 725 had complete data for the study (88.8%). The final study sample represented approximately 80.4% of the eligible employee population.
Measures and Instruments
Workplace CoH, participation in sports activity, social capital, perceived stress, and self-rated health were measured in this study. The workplace CoH survey was employed as an exogenous variable, whereas participation in sports activity, social capital, perceived stress, and self-rated health were employed as endogenous variables. An endogenous variable can be both independent and dependent variables. As self-rated health was only used for the final outcome endogenous variable, the other three endogenous variables are referred to mediating variables from this time forward to help to better understand our proposed relationships.
Workplace cultural support for health was measured with the Workplace CoH survey, a 36-item scale previously utilized in other studies10 and with established reliability and validity.2 It includes five major areas of the environment and culture for supporting health at the worksite. The scale is a composite score calculated from the 36 items on the Workplace CoH survey. Each item utilized a seven-point Likert-type scale (1 = strongly disagree; 7 = strongly agree). The five factors as validated by previous factor analyses (ie, EFA and CFA) are the following: (1) senior leadership support for health and policies supporting health; (2) well-being programs and rewards; (3) quality assurance presence of evaluation processes; (4) supervisor support for health and well-being programs; and (5) coworker support for healthy practices, concern for each others’ health and norms around health-related behaviors. The previous study showed strong internal consistency (eg, Cronbach's α = 0.91 to 0.97). The overall score resulting from the Workplace CoH survey is a summative index of the items and yields an indication of the overall culture of health at the worksite. The specific CoH survey items are not available for publication. For more detailed information as to the specific items in the survey and its utilization, parties are encouraged to contact the authors.
Social capital was measured using an eight-item scale developed by Kouvonen et al.25 The authors designed to assess social capital specifically in a worksite setting. Thus, this scale indicates whether employees feel that they are respected, valued, and treated as equals at workplace: (1) our supervisor treats us with kindness and consideration; (2) we have a “we are together” attitude; (3) our supervisor shows concern for our rights as an employee; (4) people feel understood and accepted by each other; (5) people keep each other information about work-related issues in the work unit; (6) we can trust our supervisor; (7) Do members of the work unit build on each other's ideas to achieve the best possible outcome?; and (8) people in the work unit cooperate in order to help develop and apply new ideas. The response format for the eight items was a seven-point Likert-type scale anchored by 1 = strongly disagree to 7 = strongly agree. Participation of sports activities was measured by the format: never, 1 to 3 times per year, 4 to 6 times per year, 7 to 9 times per year, 10 to 12 times per year. Occupational stress was measured with Williams et al's26 four-item scale. This scale was originally adapted from Motowidlo et al27 and consists of 4 items (eg, “I feel a great deal of stress because of my job,” “Very few stressful things happen to me at work (reverse scored),” “My job is extremely stressful,” and “I almost never feel stressed at work (reverse scored).” The items were scaled using a seven-point Likert-type scale (1 = strongly disagree; 7 = strongly agree). Higher scores on the Likert scale indicate greater job stress.
We selected self-rated health as a health outcome (ie, the end outcome variable in this study) which is the most commonly used validated indicator of health.28 In most cases, self-rated health was measured on a five-point Likert-type scale with questions on respondents’ perceived health—for instance, “How would you describe your overall state of health these days?” “In general, would you say that your health is: very good, good, fair, poor, very poor.”25 Higher scores represented better health.
Moderating and Control Variables
Sex (ie, male vs women) and job level (ie, managerial employees vs nonmanagerial employees) were used for moderating variables on the social capital and self-rated health relationship along with the occupational perceived stress–self-rated health relationship, respectively. Socioeconomic status (ie, income and age) was utilized as a control variable to control for the final outcome endogenous variable of self-rated health.29
Data analysis was conducted in three steps. First, a confirmatory factor analysis (CFA) using AMOS version 21.0 (Chicago, IL) tested to assess the dimensionality of latent variables (ie, CoH score, social capital, and occupational stress) in the present study. This was performed before estimating and testing the hypothesized structural paths (ie, full structural model) to detect any potential measurement issues, as recommended by Marsh et al.30 In the first step, we also performed validity and reliability tests during the CFA. Second, a structural equation modeling (SEM) using AMOS version 21.0 (Chicago, IL) was employed on the basic hypotheses to examine the relationships among perceived cultural support for health, the participation of sports events, social capital, occupational stress, and self-rated health. Following the procedure outlined by Becker,31 we added a control variable (ie, socioeconomic status: income and age) to the hypothesized model and then compared SEM results both with and without the control variable. Finally, overall χ2 difference tests for the moderator variables (ie, sex and job level), following the recommendations of Dabholkar and Bagozzi,32 were conducted.
The study sample (N = 725) was composed of 56.1% male and had an average age of 38.2 ± 7.5 years. More detailed demographic information for the study sample is shown in Table 1.
The CFA was performed with latent variables (ie, workplace Culture of Health, social capital, and occupational stress) to test the measurement model. All the constructs on latent variables in the model met the recommended level, thereby providing evidence of good psychometric properties of the scales (Table 2). Measurement validity of each latent variable construct appearing in the structural model was performed in two phases before testing the structural paths.33 In the first phase, convergent validity of the scale was examined by the composite reliability (CR) and average variance extracted (AVE) values. As shown in Table 2, each CR score of the latent variables was greater than its AVE score, and the AVE scores were greater than 0.50 (eg, CoH: CR = 0.91 > AVE = 0.67). In the second phase, the discriminant validity between the three latent constructs was evaluated by testing that the square root of AVE score was greater than the rest of the interconstruct correlations.34 All the latent constructs met the recommended level, thereby providing the evidence of convergent and discriminant validity.
To test reliability, the CR score of each construct was estimated because it is a more suitable indicator of reliability than Cronbach's coefficient alpha.35–37 The CR score of all of the latent variables was greater than 0.70. In addition, the authors employed MaxR(H) measures, Maximal Reliability, on each latent construct. MaxR(H) measure is argued to be more robust than construct reliability. According to Hancock and Mueller, “Coefficient H describes the relation between the latent construct and its measured indicators… coefficient H is unaffected by the sign of indicators’ loadings, drawing information from all indicators in a manner commensurate with their ability to reflect the construct.”38 Each maximal reliability score for the three latent constructs was greater than 0.80, which is indicating adequate to excellent reliability.38
SEM was utilized to test relationships among study constructs. The model was estimated before and after the addition of the control variable (ie, socioeconomic status: income and age). Both models, with and without the control variable, provided similar results in terms of standardized regression coefficients and t values. As shown in Table 3, all basic hypotheses were statistically significant (P
< 0.001) in both models. The fit of the model without the control variable (χ2/df = 2.978; CFI = 0.939; SRMR = 0.058; RMSEA = 0.052) was less satisfactory than the fit indices of the model with the control variable included (χ2/df = 2.890; CFI = 0.948; SRMR = 0.053; RMSEA = 0.048). Considering these fit indices and the suggestions of Becker,31 the model with the control variable was selected for the next step, and revealed that all hypothesized relationships were significant and in the theorized direction (Table 3): the CoH–participation of sports events relationship (β = 0.452, t = 11.79), the participation of sports events–social capital (β = 0.410, t = 11.23), the social capital–stress relationship (β = −0.145, t = −3.74), the social capital–self-rated health relationship (β = 0.226, t = 6.96), the CoH–stress relationship (β = −0.281, t = −6.91), and the stress–self-rated health (β = −0.498, t = −14.80). Finally, socioeconomic status, a control variable, had significant and positive effects on self-rated health (income: β = 0.159, t = 2.58, age: β = −0.152, t = −2.47).
In the final step, once support for the main effects was found, the proposed moderator variables, sex (male vs female) and job level (managerial employees vs nonmanagerial employees), were included in the model. Following the approach of Dabholkar and Baggozi,32 an overall χ2 difference test for the moderator variables was first conducted for each variable. As shown in Table 4, the χ2 difference between the fully constrained model and the unconstrained model on sex was 8.688, with a P value of 0.192. This indicates that there is not a significant moderating effect of sex somewhere in the parameters of the research model.39 Thus, the moderating effect of sex on the social capital–self-rated health relationship was not tested. However, the χ2 difference was 18.689 (P
< 0.01) for job level, which indicates that there was a moderating effect of job level somewhere in the parameters of the model. After obtaining this evidence, the specific moderating effects that we proposed of occupational perceived stress were further examined. The relationship between occupational stress and self-rated health was statistically stronger for employees (β = −0.589, P
< 0.001) than for managers (β = −0.481, P
The aim of the present study was to examine the associations between worksite CoH and self-rated health of employees with the mediating effects of social capital after implementing environmental design strategies and sports programs (to encourage social ties and networking). Our study makes several theoretical contributions. First, although prior research has suggested that support for health in the workplace culture helps to improve employee's health, the present study is, to our best knowledge, the first attempt to provide empirical evidences on the association between workplace culture of health with employees’ overall self-rated health, especially in the mechanism how social capital acts between a worksite CoH scores and health. Jia et al presented the positive association between workplace health culture and health-related outcomes, including self-rated health, mental health, and happiness40; however, their study did not contribute to understanding how a worksite culture of health leads to employees’ health, by not including a compelling theory related to the link between culture and health. Previous studies in social capital and health have shown that individuals exposed to community-based health promotion programs tend to have healthy behaviors and even affect other individuals’ health-related behaviors.41–42 In a similar vein, the exposure to population-based programs at the worksite tends to make employees more interested in becoming actively participated in sports, which affects social capital (and thus self-rated health) at the organizational context.
Second, our findings, in line with previous studies on social capital and health,17,43 show that social capital associated with sports participation and stronger cultural support for health positively influences health directly, and indirectly through stress. Unlike our findings, social support of organizations may have a negative effect on employees’ health by increasing stress due to excessive pressure to participate in sports.44 However, culture programs of the targeting company encouraged their employees to voluntarily participate in sports with prizes, which appears that employees build positive social capital from sports participation without excessive stress and finally have positive self-rated health. In addition, as noted earlier, this finding extends the social capital literature by exploring the relationship between social capital and health at a new structure level (ie, the organization implementing the culture of health programs).17
Third, empirical research supports the notion that many individual factors are associated with the relationship between social ties and health outcomes (stress, self-rated health status).45 Likewise, we empirically show that job classification (ie, managerial employees vs nonmanagerial employees) moderated the relationship between stress and self-rated health. Results from our moderating analyses showed that the link between stress and self-rated health is stronger for nonmanagerial employees as compared with managerial employees. This is in line with previous studies that found that the positions at higher hierarchical levels are often occupied by the individuals who better handle and cope with pressures arising from threatening workload under excessive time pressure.10,20–21 Our findings highlight that the impact of the stress associated with social capital and perceived cultural support for health on self-rated health depends on the job level in which employees are in the workplace. The job level and stress construct may also vary by organization and/or industry. As such, researchers may find job level as an important boundary condition that can stimulate the stress–self-rated health relationship. With regard to the moderating effects of sex on the social capital–self-rated health relationship, sex differences were not observed. This is consistent with Kawachi and Berkman's view that sex difference is based on how much social support is given to the group (eg, organization in this study) to which the person belongs.17 In light of this, the authors, in a community setting, reported that women tend to have lower self-rated health status than men due to low social network involvement. However, employees, both women and men, of the target company were given social support equally.
The following limitations need to be discussed. First, the present study used the overall score of the CoH measure. It would be interesting to see if particular elements (ie, leadership, policies, programs, quality assurance, supervisor support, and coworker support) have a differential impact on self-rated health through social capital. Second, the CoH survey used in the present study has been the source of previous studies in peer-reviewed literature. However, few such measures exist and as inherent in any one tool, our survey scale may not be a complete representation of the worksite culture of health. Third, the survey was administered via managers sending a link to employees. Although employees were assured their responses were anonymous and confidential, the invitation coming from the manager could have created reticence in answering candidly. Fourth, this study is the first attempt to draw social capital in understanding the relationship between workplace culture supporting health and employee health. That is, the study provides a “big picture” view of the importance of social capital in an occupational health promotion field. Thus, it would be interesting if future research focuses more on the nature of social capital at the worksite level. For example, social capital includes several types (eg, bonding, bridging, and linking) as the CoH scale has particular elements. Future research might consider the differential impacts of social capital types on employees’ self-rate health. A fifth limitation is that the impact of the CoH was tested at one point in time in the present study. However, if the cultural change tests at several points in time, it would be more meaningful in representing the effects of perceived cultural support. A final limitation is related to the understanding of the relationship between a population-based program and health. Although some existing studies involved testing the effects of workplace cultural support on health, more theory-based research is needed (other than social capital) in future to extend our understanding of the importance of a population-based program in worksite health promotion.
This study showed that social capital plays a significant role in the dynamic of workplace culture supporting health and self-rated health. Well-being strategies focusing on social capital and encouraging healthy behaviors are likely to have increased effectiveness. The moderating effects of job classification in the model revealed that the relationship between occupational perceived stress as connected with a workplace cultural support for health, social capital, and self-rated health is stronger for employees at lower hierarchical levels. Future studies should not only examine how particular elements of a culture of health (leadership, policies, manager, and/or peer support) are related to self-rated health, but how the social capital types may play a role between a worksite's culture of health and self-rated health.
1. Stokols D. Establishing and maintaining healthy environments: toward a social ecology of health promotion
. Am Psychol
2. Kwon Y, Marzec M, Edington DW. Development and validity of a scale to measure workplace culture of health
. J Occup Environ Med
3. O’Donnell MP. Health Promotion
in the Workplace. Boston, MA: Cengage Learning; 2002.
4. Pronk NP, Kottke TE. Physical activity promotion as a strategic corporate priority to improve worker health and business performance. Prev Med
5. Ball K, Bauman A, Leslie E, Owen N. Perceived environmental aesthetics and convenience and company are associated with walking for exercise among australian adults. Prev Med
6. Engbers LH, Van Poppel MNM, Chin A, Paw MJM, van Mechelen W. Worksite health promotion
programs with environmental changes: a systematic review. Am J Prev Med
7. Goetzel RZ, Henke RM, Tabrizi M, et al. Do workplace health promotion
(wellness) programs work? J Occup Environ Med
8. Pronk N. A workplace culture of health
. ACSM's Health Fit J
9. Aldana SG, Anderson DR, Adams TB, et al. A review of the knowledge base on healthy worksite culture. J Occup Environ Med
10. Kwon Y, Marzec ML. Does worksite culture of health
(CoH) matter to employees? Empirical evidence using job-related metrics. J Occup Environ Med
11. Marzec M. Measuring subcultures and leveraging data to optimize wellbeing initiatives. Am J Health Promot
12. Leana CR, Van Buren HJ. Organizational social capital
and employment practices. Acad Manage Rev
13. Bolino MC, Turnley WH, Bloodgood JM. Citizenship behavior and the creation of social capital
in organizations. Acad Manage Rev
14. Perks T. Does sport foster social capital
? The contribution of sport to a lifestyle of community participation. Sociol Sport J
15. Lindström M, Hanson BS, Östergren PO. Socioeconomic differences in leisure-time physical activity: the role of social participation and social capital
in shaping health related behaviour. Soc Sci Med
16. Kawachi I, Berkman LF. Berkman LF, Kawachi I. Social cohesion, social capital
, and health. Social Epidemiology
. New York: Oxford University Press; 2000. 174–190.
17. Kawachi I, Berkman LF. Social ties and mental health. J Urban Health
18. Oksanen T, Kouvonen A, Kivimäki M, et al. Social capital
at work as a predictor of employee health
: multilevel evidence from work units in Finland. Soc Sci Med
19. Chen DR, Lin YY, Chung KP. Career stress
and female managers’ health in Taiwan's hospitals: a multilevel model approach. Health Care Manage Rev
20. Kawakami N, Haratani T, Kobayashi F, et al. Occupational class and exposure to job stressors among employed men and women in Japan. Int J Epidemiol
21. Iwata N, Suzuki K, Saito K, Abe K. Type A personality, work stress
and psychological distress in Japanese adult employees. Stress Med
22. Beehr TA, Drexler JA. Social support, autonomy, and hierarchical level as moderators of the role characteristics-outcome relationship. J Organ Behav
23. Schieman S, Whitestone YK, Van Gundy K. The nature of work and the stress
of higher status. J Health Soc Behav
24. Schuler RS. Definition and conceptualization of stress
in organizations. Org Behav Hum Decis Process
25. Kouvonen A, Kivimäki M, Vahtera J, et al. Psychometric evaluation of a short measure of social capital
at work. BMC Public Health
26. Williams ES, Manwell LB, Konrad TR, Linzer M. The relationship of organizational culture, stress
, satisfaction, and burnout with physician-reported error and suboptimal patient care: results from the MEMO study. Health Care Manage Rev
27. Motowidlo SJ, Packard JS, Manning MR. Occupational stress
: its causes and consequences for job performance. J Appl Psychol
28. Idler EL, Benyamini Y. Self-rated health
and mortality: a review of twenty-seven community studies. J Health Soc Behav
29. Kennedy BP, Kawachi I, Glass R, Prothrow-Stith D. Income distribution, socioeconomic status, and self-rated health
in the United States: multilevel analysis. BMJ
30. Marsh HW, Byrne BM, Yeung AS. Causal ordering of academic self-concept and achievement: reanalysis of a pioneering study and revised recommendations. Educ Psychol
31. Becker TE. Potential problems in the statistical control of variables in organizational research: a qualitative analysis with recommendations. Organ Res Methods
32. Dabholkar PA, Bagozzi RP. An attitudinal model of technology-based self-service: moderating effects of consumer traits and situational factors. J Acad Mark Sci
33. Anderson JC, Gerbing DW. Assumptions and comparative strengths of the two-step approach: comment on Fornell and Yi. Sociol Method Res
34. Fornell C, Larcker DF. Structural equation models with unobservable variables and measurement error: algebra and statistics. J Mark Res
35. Lin HF, Lee GG. Impact of organizational learning and knowledge management factors on e-business adoption. Manage Decis
36. Molina ML, Lloréns-Montes J, Ruiz-Moreno A. Relationship between quality management practices and knowledge transfer. J Oper Manag
37. Raza SA, Qazi W, Umer A. Facebook is a source of social capital
building among university students’ evidence from a developing country. J Educ Comput Res
38. Hancock GR, Mueller RO. Cudeck R, Du Toit S, Sörbom D. Rethinking construct reliability within latent variable systems. Structural Equation Modeling: Present and Future
. Lincolnwood: Scientific Software International; 2001; 195-216.
39. Maiyaki AA. Moderating effect of individualism/collectivism on the association between service quality, corporate reputation, perceived value and consumer behavioural intention. J Mark Manag
40. Jia Y, Gao J, Dai J, Zheng P, Fu H. Associations between health culture, health behaviors, and health-related outcomes: a cross-sectional study. PLoS One
41. Kawachi I, Subramanian SV, Kim D. Kawachi I, Subramanian SV, Kim D. Social capital
and health. Social Capital and Health
. New York: Springer; 2008; 1-26.
42. Merzel C, D’Afflitti J. Reconsidering community-based health promotion
: promise, performance, and potential. Am J Public Health
43. Bartley M. Health Inequality: An Introduction to Theories, Concepts and Methods. Cambridge, MA: Polity Press; 2004.
44. Eriksson M. Social capital
and health—implications for health promotion
. Glob Health Action
45. Lindström M, Moghaddassi M, Merlo J. Individual self-reported health, social participation and neighbourhood: a multilevel analysis in Malmö, Sweden. Prev Med
46. Walsh G, Evanschitzky H, Wunderlich M. Identification and analysis of moderator variables: investigating the customer satisfaction-loyalty link. Eur J Marketing