Examining Social-Ecological Determinants of RPAamong Chinese Elderly
Xiangren Yi ,Shandong University, Jinan, China;ZanGao, University of Minnesota, Minneapolis, MN; Leon Chen, Delaware State University, Dover, DE.
Background/Purpose
With the most rapid expansion occurring over aged 60 in China, the current estimate of older adults will increase 17% of aging population in 2020 to 30% by 2040 (National Council on Aging (NCOA), 2006).China Report on Development of the Cause of Aging（2013）was published by NCOA that indicated that the number of aging population have reached from the 0.194 billion（14.3%）in 2012 to 0.202 billion（14.8%）in 2013. China Health Statistical Yearbook 2010 showed that 64.54% of over aged 65 have chronic diseases, which have seriously threatened their health and converted intothe substantial social and economic issues and become the significant challengesChina has to confront with. Scientists and policy makers in China have focused increasingly on identifying determinants that may have extensive positive impacts on the active and healthy aging process. To meet the huge challenge of chronic diseases, the Ministry of Health of China(MHC), with the support of WHO, and in cooperation with relevant sectors, has been developing the first medium and long-term high level national plan for chronic disease control and prevention (2005–2015)，which mandate an integrated and comprehensive approach to the control and prevention of cardiovascular disease, cancer, chronic respiratory disease, and diabetes. There will be priority actions in at least four areas: adult male smoking, hypertension, overweight and obesity, and capacity building for chronic disease control.
The modifiable health risk behaviors such as physical inactivity, unhealthy eating, tobacco use and excessive alcohol consumption leading to the chronic diseases are already common (MHST, 2004).Research has demonstrated that physical activity(PA) is directly associated with improvements and with lower mortality, morbidity and mortality rates from cardiovascular diseases(Kujala, Kaprio, Sarna, &Koskenvuo, 1998; Kaplan, Strawbridge, Cohen, &Hungerford, 1996;Wannamethee, Shaper, &Walker, 2000), including reductions in risks of stroke (Rodriguezet al, 2002) , diabetes mellitus (Gill& Cooper, 2008) , cancer (Giovannucci, Ascherio, Rimm, Colditz, Stampfer, &Willett, 1995), osteoporosis (Borer, 2005; Preisinger et al, 1996) , falling(Tinett,Baker & McAvay, 1994; Campbell, Borrie, Spears, 1989), and having a strong effect on longevity (Lee, Paffenbarger, &Hennekens, 1997). Furthermore, maintaining a physically active lifestyle facilitates independent functioning in activities of daily living, enhances quality of life, relieve of feeling of depression and promote psychological well-being (Fried et al,1998; Sigh, 2002).
Despite the benefits of regular physical activity (RPA) throughout the lifespan, the majority of adults in China do not achieve the recommended amount of PA and the proportion of old adults who do not participate in any leisure-time PA has remained close to 60% for the past decade (CSPAA, 2006). The research demonstrated that only 20.2% Chinese elderly participated in RPA in 30 minutes of
moderate intensity activity at least three days per week. An estimated over 70%
of Chinese are considered sedentary based on no RPA or inadequate levels of
physical activity (General Administration of Sport (GAS), 2008). Some researchers found in China that the determinates of RPA for old adult were lack of time and equipments ( Che, 2008; GAS, 2008; Wang & Li, 2010), social support system ( Wang, 2007) and motivation ( Xiao, 2005)). Studies showed that the environment, individual, cognitive, income, efficiency were associated with the behavior of PA
adherence of elderly (Gao& Wang, 2005; Wang & lu, 2004).
Social ecological model (SEM) of health promotion have increasingly been adopted to better understand the relations among the individual factors(IF), social environment (SE), physical environment (PE) and regular physical activity (RPA) in the U.S., but little is known regarding the application of SEM in predicting RPA among elderly in China. The SEM proposed in this study was a synthesis of the models that were formulated and specifically designed for elder’s physical activity in Chinese community. The purpose of this study is to validate the questionnaires assessing social-ecological determinants (IF, SE, and PE) and examine the relationships of the determinants and RPA among Chinese elderly.For the purpose of this study, a multidimensional SEM was formulated to identify the constructs that were used to determine perceived determinants of RPA for elderly. In this hypothesized model, overall perceived determinants were determined by old adult.
Method
Participant
In the first phase, two groups of participants were included in this study: (a) the participants for the initial scale construction and (b) the participants for the pilot study of the initial scale. The participants included in the initial scale-construction phase were varied and represented various personnel in the community. A total of 325 members (255 women and 75 men) agreed to participate in this study. Most of these members were between the ages of 55 and 75 years old (86%) and were married (95%). In the second phase, there are 30 investigators, graduate or undergraduate, who have trained formally and participated in indoor or outdoor interview. Two people formed a group for indoor interview to assist and check each other’ work. We contacted with 1896 participants, 33 people dropped out during the investigation. A total of 1863 elderly (male = 36.7%, female =63.3%;M age = 64.11±9.77 years) responded to the questionnaire assessing their RPA, IF (health knowledge, health behavior, sport skills, current physical ability/disability, and psychological factors), SE (social support, social security,organizational participation, social relationship, and social activity), and PE (facilities). Structural equation modeling was utilized to test relations among the determinants and physical activity
Instrument
The process of formulating the preliminary instrument included a literature review, observation, interviews with focus groups in the community, and a modifiedapplication of the Delphi technique.A preliminary questionnaire with 150 items based on SEM was formulated through literature review, field observations, interviews, the Delphi technique, and a pilot study.The final validated questionnaire comprised of 120-itemsvia exploratory factor analysis and reliability tests.
Procedure
Two methods were used to collect data. First involved indoor visit, accompanying with administrator or dormitory leaders, during which the old adults were briefed on the study, signed a consent form, and then investigator completed the questionnaire based on their interview. Other method is to put table on the street that administrator or dormitory leaders call for them from each home to participate in this survey.
After the questionnaire completed, the 4 graduates every day will check whether each questionnaire have been finished correctly. If the questionnaires have any problem, telephone interview will be followed next day.
Data analysis
Structural equation modeling using LISREL for Windows, version 8.7was used to test the fit of the hypothesized physical activity models to the data collected from participants. The confirmatory factor analysis was utilized to identify the factor structure of model constructs prior to evaluate any relationships among constructs. The structure model showed a good fit to the data with chi square, p< .05. The fit indices met the acceptable criteria: RMSEA = .43, NNFI = .94, CFI= .95, SRMR= 0.05.
Analysis/Results
As defined in the multivariate analysis literature, the content validity of a construct implies that all the items used to measure a construct should show high loadings on their respective constructs. Tables 1 showed that the factor loading were significant at the 0.05 level and ranged between 0.74-0.94 from health knowledge, current physical ability/disability, sport skill, psychological factors, social safety, social support, social activity, community, facilities. All the variables significantly loaded on their respective constructs and this confirmed that the measurement model used possesses the required content validity.
Table 1 Component Loadings and T-values
Construct / Loading / Standard Error / T-valueHealth Knowledge / 0.74 / 0.13 / 6.69
Psychological Factor / 0.86 / 0.03 / 2.74
Physical Ability / 0.76 / 0.04 / 2.75
Sport Skill / 0.91 / 0.05 / 16.89
Social Safety / 0.76 / 0.03 / 2.27
Social Support / 0.77 / 0.03 / 2.32
Social Activity / 0.80 / 0.03 / 3.87
Community / 1.00 / 0.03 / 3.67
Facility / 0.94 / 0.06 / 11.82
The convergent validity is defined to indicate the extent to which a set of items converges in measuring aparticular construct (Bagozzi & Yi, 1988; Hair et al., 2010), which can be confirmed by examining the item's reliability, composite reliability (CR) and the average variance extracted (AVE).That is, the items of each construct are highly loaded and statistically significant in measuring their respectiveconstructs with at least 0.7 factor loadings, CR of each construct is at least 0.75 exceedingthe cut off value of 0.7 and AVEfor each construct is at least 0.5. These resultsindicated that the measurement model has an adequate convergent validity.
Table 2 Convergent Validity
Construct / AVE / CR / Cronbach’s ɑ / CommunalityHealth Knowledge / 0.61 / 0.76 / 0.8 / 0.61
Psychological Factor / 0.57 / 0.80 / 0.62 / 0.57
Physical Ability / 0.55 / 0.83 / 0.74 / 0.55
Sport Skill / 0.61 / 0.90 / 0.87 / 0.61
Social Safety / 0.59 / 0.81 / 0.68 / 0.59
Social Support / 0.65 / 0.78 / 0.62 / 0.65
Social Activity / 0.61 / 0.75 / 0.71 / 0.61
community / 0.62 / 0.76 / 0.79 / 0.62
Facility / 0.58 / 0.81 / 0.71 / 0.58
The discriminant validity is defined to be the degree to which a set of items can differentiate a construct from other constructs. This means that variance shared among the items of each construct should be greater than the variance shared with other constructs (Compeau et al., 1999). the results in the correlation matrix illustrated in Table 3 ensure that the discriminant validity is confirmed.
Table 3 Correlations and DiscriminantValidity
Construct / V1 / V2 / V3 / V4 / V5 / V6 / V7 / V8 / V9Health Knowledge / 0.72
Psychological Factor / 0.67 / 0.75
Physical Ability / 0.57 / 0.61 / 0.73
Sport Skill / 0.66 / 0.66 / 0.6 / 0.78
Social Safety / 0.61 / 0.62 / 0.43 / 0.7 / 0.73
Social Support / 0.58 / 0.64 / 0.47 / 0.6 / 0.6 / 0.77
Social Activity / 0.47 / 0.54 / 0.57 / 0.57 / 0.64 / 0.54* / 0.74
community / 0.57 / 0.57 / 0.64 / 0.65 / 0.63 / 0.69 / 0.66 / 0.67
Facility / 0.6 / 0.59 / 0.59 / 0.52 / 0.54 / 0.7 / 0.61 / 0.51 / 0.74
Results indicated thatRPAwere positively associated with sport skills (= 0.492, t = 10.71), current physical ability/disability (= 0.291, t = 5.14) , psychological factors (=0.108, t = 2.54), social support(= 0.279, t = 6.74), social safety ( = 0.135, t = 2.57), social activity( = 0.137, t =3.42), facilities ( = 0.382, t = 6.315).Among them, facilities and sport skillshad stronger predictive strength on RPA for Chinese elderly.
Table 4 The Structural Model and Hypothesis Testing
Structural Model / Path Coefficient / Standard Error / T-valueHealth Knowledge / 0.001 / 0.04 / 1.03
Psychological Factor / -0.07 / 0.039 / 2.54**
Physical Ability / 0.111 / 0.048 / 5.14***
Sport Skill / 0.358 / 0.045 / 10.71***
Social Safety / 0.08 / 0.038 / 2.57**
Social Support / 0.064 / 0.035 / 6.74***
Social Activity / 0.116 / 0.032 / 3.42***
Community / 0.136 / 0.035 / 3.876***
Facility / 0.254 / 0.039 / 6.315***
***P<0.01, **P<0.05,
Conclusion
SEM provides an overarching framework, or set of theoretical principles, for understanding the interrelations among diverse personal and social environmental factors in human health and illness (Stokols, 1996).Understanding and addressing social and contextual correlates of physical activity among elderly in China is necessary to establish comprehensive intervention programs, particularly within community settings. This study was designed to identify the structure of determinants of impact of RPA on elderly in China under SEM, which is a multidimensional construct that requires multiple dimensionsto evaluate the perceptions of elderly in China, including three dimensions. Individual, social environment, physical environment.
The findingsrevealedthat facilities,sport skills and social supportemerged as major determinants of RPA. Psychological factors also played a mediating role in the model.Health knowledgewasnot significant associated with RPA. Future studies willassess IF, SE, and PE asinteractive contributors to physical activity behavior and to betterunderstand the pathways through which they influence RPA for Chinese elderly.