| ObjectiveThis study aims to explore the relationship between cognitive function, social network, and social capital among urban elderly and to explore the rationality of the Social Network-Social Capital-Cognitive Function path within a three-year follow-up survey. And this study could offer more suggestions for community health intervention policies on MCI elderly.MethodStudy PopulationThe target population is community-dwelling individuals aged 65 years or older in Wuhan. The sample was recruited from urban registered elderly residents or permanent residents in the baseline survey. In the follow-up study, the sample was recruited from the baseline sample and the residents who reached 65 years old during 2011-2014.The research contents are consisted of four parts:prevalence and prognosis of mild cognitive impairment (MCI); influence on cognitive outcome/decline from social capital; influence on cognitive outcome/decline from social networks; rationality of the Social Network-Social Capital-Cognitive Function path.Settings and SamplingShouyi road community and Honggang city road community were selected as the study site which consisted 23 neighborhood committees. The target neighborhood committees (6 out of 23) were selected randomly by using cluster sampling. Participants were recruited until the sample size was fulfilled. All the elders who meet the inclusion and exclusion criteria were under investigation.MeasurementThe baseline survey was conducted during the months of July to September in 2011. The followed-up survey was conducted during the months of July to September in 2014. The baseline survey includes the demography and cognitive function. The follow-up survey includes the cognitive function, social network, and social capital. Older adults diagnosed with mild cognitive impairment based on the Peterson criteria.Participants were categorized as normal cognitive (NC) or MCI by using the Montreal Cognitive Assessment (MoCA). Cognitive changes were defined as’normal to normal (MCI to normal)’’normal to MCI’’MCI to MCI’. The social network was measured by Name-Generate. The social capital was measured using PSCS-16. Statistical analysisDatabase was established by EPIDATA3.02. Data was analyzed using the SPSS 20.0verson and the SAS 9.2 version.ResultsThe prevalence of MCI and its transition. (1) Situation of epidemic:the sample size was 1108 in baseline survey, and 1218 in follow-up survey. The prevalence of MCI in the baseline and follow-up survey are 17.33 percent and 27.91 percent, respectively.There were 517subjects (50.90%) who have both baseline and follow-up record. The proportion of’normal to MCI’was 24.18%. (2) Influencing factor:there were statistical differences on age (t=-5.24; p<0.001), gender (x2=4.5184; p=0.0335), educational level (t=4.39;p＜0.001) and marital status (x2=9.3298;p=0.0023) among NC and MCI at baseline, and on age (t=-6.60,p＜0.001), educational level (t=-4.49.p＜0.001) and marital status (x2=5.4506, p=0.0196) among NC and MCI in the follow-up survey. Age (F=13.10, p＜0.0001) and educational level (F=3.23,p=0.0402) were influencing factors of cognitive decline.The reliability and validity of PSCS-16. The Cronbach a in both the bonding social capital scale and bridging social capital scale are higher than 0.8 and the Cronbach a is 0.9 in the full scale. This shows the scale has good internal consistency reliability. According to confirmatory factor analysis, good construct validity of PSCS-16 was shown among elderly population. The ICC of bonding and bridging social capital scale are both higher than 0.7, which shows the scale has good the test-retest reliability.The association between social capita and cognitive function.(1) The distribution characteristics and influencing factors:the means of bonding social capital and bridging social capital score were (21.50±6.24)> (14.94±5.99) with statistical significance between groups(t=34.63,p＜0.001). The bonding social capital was significantly associated with age (F=6.28,p=0.0003), living condition (t=-2.43, p=0.0155), educational level (F=23.32,p＜0.0001) and marital status (t=2.04,p=0.0420). The bridging social capital was significantly associated with age (F=3.16,p=0.0239).(2)The relationship between social capital and cognitive outcome/decline:in the logistic model that using ’cognitive outcome’ as outcome, bonding social capita was associated with MCI (OR=0.964,95% CI:0.942-0.986). The path coefficient between cognitive outcome and educational level/age/bonding social capital were 0.23,0.21 and -0.17, respectively. Neither bonding social capital (OR=0.980,95%CI:0.944-1.018) nor bridging social capital (OR=1.022,95%CI:0.988-1.057) was significantly associated with cognitive decline.The association between social network and cognitive function.(1) The distribution of social network parameters on cognitive outcomes:(a) Size. The proportion of ’10-19’ was 30.89% in NC which was highest and the proportion of ’0-4’ was 29.88% in MCI. The statistically significant size difference (Z=1.6982,p=0.0434) was found, (b) Relationship. The proportion of parent-child and spouse relationships was highest with 92.65% in NC and 90.86% in MCI. No statistically significant relationship difference (x2= 2.8918, p= 0.0890) was found. (c) Density. The mean of density was (0.89±0.26) in NC and (0.85±0.27) in MCI. No statistical significance difference (t=1.05, p=0.2960) was found between groups. (d) Strength. The mean of strength was (2.29±0.92) in NC and (2.74±1.15) in MCI. Statistical significance difference (t=-6.27, p<0.001) was found between groups.(e) Reciprocity. The mean of reciprocity was (1.16±0.35) in NC and (1.20±0.39) in MCI. Statistical significance difference (t=-1.96, p=0.0398) was found between groups.(2) The distribution of social network parameters on cognitive decline:(a) Size. The proportion of’10-19’was 33.43% in’normal to normal(MCI to normal)’which was highest and the proportion of’0-4’was 32.23%in’normal to MCI’ and the proportion of’5-9’ was 33.33%in ’MCI to MCF’. The statistically significant size difference{x= 4.2662, p= 0.0389) was found, (b) Relationship. The proportion of parent-child and spouse relationships was highest with 90.54% in’n-n’and 91.71% in ’n-m’ and 83.67% in’m-m’. No statistically significant relationship difference (y?= 5.6620,/?=0.0590) was found, (c) Density. The mean of density was (0.91±0.29) in’n-n’and (0.90±0.30) in ’n-m’ and (0.86±0.35) in ’m-m’. No statistical significance difference (F=0.46, p= 0.6317) was found between groups, (d) Strength. The mean of strength was (2.35±0.92) in’n-n’and (2.86±1.12) in’n-m’and (2.43±1.15) in’m-m’. Statistical significance difference (F=6.46,p<0.0001) was found between groups, (e) Reciprocity. The mean of reciprocity was (1.20±0.37) in’n-n’ and (1.26±0.40) in’n-m’and (1.21±0.51) in’m-m’. No Statistical significance difference (F= 1.16, p=0.3148) was found between groups.The analysis of the Social Network-Social Capital-Cognitive Function path.(1) networks-social capital-cognitive score:(a) Outcomes of Model Fitting.x2/Df=1.5899; p= 0.1894; GFI=0.9992; AGFI=0.9847; RMSEA=0.0228. (b) Model interpretation. Strength (r=-0.21,p<0.05), reciprocity (r=0.04,p0.05), size (r=0.13,p＜0.05) and bonding social capital (r=0.06,p＜0.05) were associated with cognitive score; a inverse correlation existed between strength/reciprocity and cognitive score and a positive correlation existed between size/bonding social capital and cognitive score.(2)network-cognitive decline:(a) Outcomes of Model Fitting. x2/Df=6.04; p=0.004; GFI=0.9908; AGFI=0.8896; RMSEA=0.1025. (b) Model interpretation. There was no association between bonding social capital and cognitive decline; The cognitive decline was mainly influenced by age (r=0.20, p<0.05)、 strength (r=0.19, p<0.05) and educational level (r=0.18, p<0.05).Conclusion1. PSCS-16 has good reliability in elderly.2. The elderly population in Chinese communities mainly rely on bonding social capital. The bonding social capital score is low among the elderly, living alone and low educational level population.3. Increase the frequency of interaction with the old man and increasing the social participation in the elderly will help maintain cognitive function in the elderly;4. The association between bonding social capital and cognitive function is weak. The mediate effect of this association in the Social Network-Social Capital-Cognitive Function path is also weak. Other mediators are possibly existed in this path. |
PDF Full Text Request