Institution-based Home-dwelling
Senior Apartment
95% CI Means of Physical Domain
16.00
15.50
15.00
14.50
14.00
13.50
13.00
Senior apartment Home-dwelling Institution-based Living Arrangements
- 53 -
Living Arrangements
institution based home-dwelling
Senior Apartment
95% CI Means of Psychological Domain
15.50
15.00
14.50
14.00
13.50
13.00
12.50
Figure 4-1 95%CI means of psychological domain of QOL by different living arrangements (continued)
…
Senior apartment Home-dwelling Institution-based Living Arrangements
- 54 -
Living Arrangements
Institution-based Home-dwelling
Senior Apartment
95% CI Means of Social Relationship Domain
16.50
16.00
15.50
15.00
14.50
14.00
13.50
Figure 4-1 95%CI means of social relationships domain of QOL by different living arrangements (continued)
Senior apartment Home-dwelling Institution-based
Living Arrangements
- 55 -
Living Arrangements
Institution based Home-dwelling
Senior Apartment
95% CI Means of Environment Domain
16.00
15.50
15.00
14.50
14.00
13.50
13.00
Figure 4-1 95%CI means of environment domain of QOL by different living arrangements (continued)
Senior apartment Home-dwelling Institution-based
Living Arrangements
- 56 -
4.4▓Contribution of the four QOL domains
The results of utilization of the STATA program to analyze the predictors of four QOL domains are shown in Table 4-4. Dependent variables included WHOQOL-BREF Taiwan version physical domain, psychological domain, social relationship domain and environment domain. While the independent variables included age, living arrangement, IADL, ADL, TDQ, health status (perception of their health status whether it was very poor, poor, neither poor nor good, good, and very good), life conditions (perceptions of their life whether it was very unhappy, unhappy, neither happy nor unhappy, happy, very happy), gender, education, religion, marriage, disease (the physical illness), health status, individual income, family income, level of dependence (ADL, IADL), depression symptoms (TDQ), and living places arrangement.
With the STATA program, these variables were entered to find out the best predictor for our outcome. For physical domain, the variables which were the age, and health status, showed the significant level of affecting the satisfaction of their physical health (p <.05). The health status was the major determinant of physical domain (beta coefficienthealth status=
2.96(poor), 3.35(neither poor nor good), 4.22 (good), 4.70 (very good). Secondly, the age showed the negative effect where beta coefficient= -0.05). That mean as the people get older the less satisfaction on their physical domain.
There were two independent variables contributing to the psychological domain. Our model, which included depression symptoms, explained 50% where adjusted R2=0.50 of the variance in psychological domain. Of these two variables, health status made the largest unique contribution to psychological domain (beta coefficient= 2.60 (good), 2.49(very good), p<.05). The depression symptoms contributed significantly (beta coefficient= -0.78, p<.05) to the outcome of psychological domain where the higher the score of their depression symptoms the less satisfaction on their emotional state or psychological domain.
- 57 -
For social relationship domain, health status (beta coefficient= 1.97, p< .05) were the dominant contribution to the quality of life. The satisfaction of the social relationship or participation would be attenuated greatly by the health status. This meant that if the health status was good, they might participate actively in the social activities. Secondly, the gender predictor contributed to the social relationship with beta coefficient= 0.85, p<.05, where the female was better in social relationship compare to male. Adjusted R2=0.21, which mean the variance of this model is 21.0%.
For the environment domain, there was 42.0% of the variance of the model (adjusted R2= 0.42). The two different variables were the same as psychological domain contributed to the quality of life in environment domain. The variables included depressions symptoms (beta coefficient=-0.76, p<.05), and health status (beta coefficient= 1.48, p<.05).
Scores on the four instruments were compared among the groups to determine which environment was more conducive to supporting a senior’s quality of life. Overall, the scores of the quality of life of the elderly people staying at home were the highest and youngest among this three types living places arrangement. However, our findings showed that living arrangement does not contribute significantly (p> .05) to the outcome of the quality of life in these four domains. The perceived health status was the major determinants of quality of life.
Secondly, the age and gender also contributed significantly to the outcome of physical domain and social relationship domain respectively.
- 58 -
Table 4-4▓Scores of predictors for contributions of the four QOL domains
Unstandardized Coefficients (SD) t p 95% Confidence Interval
Physical domain (TW) (adjusted R2=0.44)
*Age
*Health Status (poor)
*Health Status (neither poor nor good )
*Health Status (good)
*Health Status (very good)
Living Arrangement (home-dwelling)
*Health Status (good)
*Health Status (very good)
Living Arrangement (home-dwelling)
- 59 -
Table 4-4▓Scores of predictors for contributions of the four QOL domains (continued)
Unstandardized Coefficients (SD) t p 95% Confidence Interval
Social relationship domain(TW) (adjusted R2=0.21)
*Gender
*Health Status (Good)
Living Arrangement (home-dwelling)
*Health Status (Good)
Living Arrangement (home-dwelling)
- 60 -
4.5 Analysis of covariance
STATA program was conducted to examine between subjects differences, with different living arrangement (senior apartment coded 1, home-dwelling coded 2, institution-based coded 3) as the fixed factor; the three set measures were adjusted as the covariate. The dependent variables were the physical domain, psychological domain, social relationship domain, and environment domain. Model 1, there were no covariate added. Model 2 were controlling the first set covariates including demographic characteristic. Model 3 were controlling the second set covariates including disease, depression symptoms, perceived life conditions and health status and lastly model 4 controlled all covariates including demographic characteristic, disease depression symptoms, perceived life conditions and health status respectively. Table 4-5 summarized the analysis of covariates in the main effects of different living arrangement on quality of life.
For Model 1, the STATA program demonstrated of significant main effect of different living arrangement at home-dwelling and senior apartment on the quality of life with four domains (physical, psychological, social relationship, environment) with p< .05 (beta coefficientphysical=1.33, beta coefficientpsychological=1.13, beta coefficientsocial relationship=1.07, beta coefficientenvironment=1.00 ) without any covariate added. However, qualities of life with four domains of were not significant differences between institution-based and senior apartment (beta coefficientphysical=0.76, beta coefficientpsychological=0.30, beta coefficientsocial relationship=0.20, beta coefficientenvironment=0.34 ) with p>.05.
For Model 2, the first set measure of demographic characteristic were added as covariate, the STATA solutions showed that there were no significant effects for the association of living arrangement senior apartment and institution-based with quality of life (beta coefficientphysical=0.23, beta coefficientpsychological=0.20, beta coefficientsocial relationship=0.13, beta coefficientenvironment=-0.05 ) with p>.05. However, the significant effect also found in the four domains for living arrangement of senior apartment and home -dwelling (beta coefficientphysical=1.09, beta coefficientpsychological=1.53, beta coefficientsocial relationship=1.16, beta coefficientenvironment=1.05) with p<.05.
- 61 -
In Model 3, with the second and third set measures of covariate were added. The STATA demonstrated a non-significant different living arrangement on the quality of life giving the condition of controlling the depression and health status for the physical, psychological and environment domain of quality of life. The depression state assessed by TDQ, and health status showed a significant effect on the contribution to the physical domain (beta coefficientdepression= -0.77, p<.01, =.04), psychological domain, and environment domain (beta coefficient=4.67, p< .05, =.032). The main effects were no longer significant with p>.05.
All covariates were added to the model 4. The health status (beta coefficient= -0.77, p<.01, =.04) showed significant covariate effect to the association of living arrangement and quality of life of physical (p<.000), psychological (p<.05) and environment (p<.05) domains.
The age and gender also showed the significant effects to the physical domain (beta coefficient= -0.05, SD 0.02, p= .04) and social relationship domain (beta coefficient= 0.85, SD 0.36, p= .02) respectively but the results did not show significant main effects to the association of living arrangement and quality of life (p>.05).
In conclusion, after controlling for the covariate age, gender, health status, life conditions of the three groups, (home-dwelling, senior apartment, institution-based) do not differ significantly in their quality of life. The health status and mental health are the major determinant to the quality of life.
- 62 -
Table 4-5: Effect of living arrangements on QOL (Physical Domain) outcomes, Kaohsiung, Taiwan (N=151)
Model 1
F(2, 148)=5.57 Adj R2=0.057
β(SD) p 95%CI
Model 2 F(8,138)=2.37 Adj R2=0.12
β(SD) p 95%CI
Model 3
F(12,138)=9.76 Adj R2=0.41
β(SD) p 95%CI
Model 4
F(18,128)=7.35 Adj R2=0.44
β(SD) p 95%CI
Age 0.04 (0.03) .029 -0.09 0.02 -0.05(0.02) .043 -0.10 0.00
Gender -0.66 (0.41) .109 -1.48 0.15 -0.52(0.33) .111 -1.17 0.12
Education -0.35 (0.68) .604 -1.70 0.99 -0.08(0.55) .885 -1.18 1.02
Marriage -0.21 (0.91) .815 -2.01 1.59 -0.68(0.74) .362 -2.14 0.79
Individual Income -0.42 (0.68) .540 -1.76 0.93 -0.34(0.55) .536 -1.43 0.75 Family Income -0.38 (1.03) .715 -2.40 1.65 0.02(0.85) .982 -1.66 1.70
Disease -0.52(0.42) .214 -1.34 0.30 -2.20(1.99) .273 -6.14 1.75
Depression score -0.77(0.36) .035 -1.49 -0.06 -0.74(0.38) .056 -1.50 0.02
Life condition 2 -1.92(1.97) .332 -5.81 1.98 -2.20(1.99) .273 -6.14 1.75 Life condition 3 -2.51(1.84) .174 -6.14 1.12 -2.73(1.86) .145 -6.42 0.95 Life condition 4 -1.89(1.87) .313 -5.59 1.81 -2.05(1.90) .284 -5.81 1.72 Life condition 5 -0.30(1.96) .879 -4.17 3.58 -0.42(2.00) .835 -4.38 3.54
Health status 2 2.70(0.87) .002 0.98 4.41 2.96(0.86) .001 1.25 4.67
Health status 3 3.17(0.79) .000 1.61 4.73 3.35(0.79) .000 1.78 4.92
Health status 4 4.29(0.85) .000 2.61 5.97 4.22(0.85) .000 2.53 5.90
Health status 5 4.48(1.03) .000 2.44 6.52 4.70(1.03) .000 2.66 6.74
Home-dwelling 1.33 (0.41) .001 0.53 2.13 1.09 (0.46) .020 0.17 2.01 -0.04(0.37) .915 -0.77 0.69 -0.29(0.41) .484 -1.09 0.52 Institution-based 0.76 (0.50) .130 -0.23 1.75 0.23 (0.55) .674 -0.86 1.33 -0.11(0.44) .799 -0.99 0.76 -0.56(0.48) .245 -1.51 0.39 Constant 13.69(0.24) .000 13.20 14.17 19.01(2.61) .000 13.84 24.17 13.87(2.22) .000 9.47 18.27 19.68(3.04) .000 13.66 25.71
- 63 -
Table 4-5: Effect of living arrangements on QOL (Psychological Domain) outcomes, Kaohsiung, Taiwan (N=151) (continued)
Model 1
F(2,148)=3.36 Adj R2=0.03
β(SD) p 95%CI
Model 2 F(8, 138)=1.39 Adj R2=0.07
β(SD) p 95%CI
Model 3
F(12,138)=12.95 Adj R2=0.49
β(SD) p 95%CI
Model 4
F(18,128)=9.14 Adj R2=0.50
β(SD) p 95%CI
Age 0.05(0.03) .164 -0.02 0.11 0.03(0.02) .170 -0.01 0.08
Gender -0.28(0.45) .538 -1.17 0.61 0.001(0.33) .996 -0.65 0.65
Education -0.15(0.75) .840 -1.63 1.32 0.02(0.56) .971 -1.09 1.13
Marriage -0.58(1.00) .562 -2.56 1.40 -0.77(0.75) .302 - 2.25 0.70
Individual Income -0.22(0.75) .770 -1.70 1.26 -0.21(0.55) .705 - 1.31 0.89
Family Income 1.35(1.13) .234 -0.88 3.57 1.59(0.86) .065 - 0.10 3.29
Disease 0.27(0.41) .513 -0.54 1.08 0.37(0.43) .383 - 0.47 1.22
Depression score -1.16(0.36) .002 -1.86 -0.45 -0.78(0.39) .045 - 1.55 -0.02
Life condition 2 -2.91(1.94) .136 -6.76 0.93 -2.60(2.01) .199 - 6.59 1.38 Life condition 3 -1.15(1.81) .526 -4.74 2.43 -0.96(1.88) .611 - 4.68 2.76 Life condition 4 0.34(1.84) .852 -3.31 3.99 0.61(1.92) .752 - 3.19 4.41 Life condition 5 1.89(1.93) .331 -1.94 5.71 2.28(2.02) .261 - 1.72 6.28 Health status 2 1.20(0.86) .163 -0.49 2.89 1.38(0.88) .117 - 0.35 3.10 Health status 3 0.96(0.78) .220 -0.58 2.50 1.24(0.80) .123 - 0.34 2.82
Health status 4 2.20(0.84) .010 0.54 3.85 2.60(0.86) .003 0.90 4.30
Health status 5 1.93(1.02) .060 -0.08 3.94 2.49(1.04) .018 0.44 4.55 Home-dwelling 1.13(0.44) .011 0.27 1.99 1.53(0.51) .003 0.52 2.54 -0.04(0.37) .924 -0.76 0.69 0.29(0.41) .484 -0.52 1.10 Institution-based 0.30(0.54) .574 - 0.76 1.37 0.20(0.61) .748 -1.01 1.40 -0.52(0.44) .237 -1.38 0.35 -0.45(0.48) .354 -1.41 0.51 Constant 13.48(0.26) .000 12.96 14.00 9.87(2.87) .001 4.21 15.54 13.60(2.19) .000 9.26 17.94 9.36(3.07) .003 3.28 15.44
- 64 -
Table 4-5: Effect of living arrangements on QOL(Social Relationship Domain) outcomes, Kaohsiung, Taiwan (N=151) (continued)
Model 1
F(2,148)=4.27 Adj R2=0.04
β(SD) p 95%CI
Model 2 F(8,138)=1.67 Adj R2=0.04
β(SD) p 95%CI
Model 3
F(12,138)=4.23 Adj R2=0.27
β(SD) p 95%CI
Model 4
F(18,128)=3.18 Adj R2=0.21
β(SD) p 95%CI
Age 0.001(0.03) .977 -0.05 0.06 -0.01(0.03) .842 -0.06 0.05
Gender 0.70(0.38) .070 - 0.06 1.46 0.85(0.36) .018 0.15 1.55
Education 0.44(0.64) .495 - 0.82 1.69 0.41(0.60) .494 - 0.78 1.61
Marriage -0.17(0.85) .846 -1.85 1.52 -0.51(0.81) .525 - 2.11 1.08
Individual Income 0.45(0.64) .482 -0.81 1.71 0.53(0.60) .374 - 0.65 1.72
Family Income 0.50(0.96) .602 -1.40 2.40 0.80(0.92) .386 - 1.02 2.63
Disease -0.13(0.44) .760 -1.00 0.73 -0.14(0.46) .758 - 1.05 0.77
Depression score -0.21(0.38) .589 -0.97 0.55 -0.07(0.42) .862 - 0.90 0.75 Life condition 2 0.32(2.08) .879 -3.79 4.43 0.39(2.17) .855 - 3.90 4.70 Life condition 3 1.67(1.94) .391 -2.17 5.50 1.57(2.03) .440 - 2.44 5.58 Life condition 4 2.54(1.97) .201 -1.37 6.44 2.45(2.07) .239 - 1.65 6.55 Life condition 5 3.39(2.07) .103 -0.70 7.48 3.27(2.18) .136 - 1.04 7.58 Health status 2 1.24(0.92) .177 -0.57 3.05 1.35(0.94) .154 - 0.51 3.21 Health status 3 1.17(0.83) .163 -0.48 2.82 1.23(0.86) .158 - 0.48 2.94
Health status 4 1.54(0.89) .087 -0.22 3.31 1.97(0.93) .036 0.13 3.81
Health status 5 1.94(1.09) .077 -0.22 4.09 2.11(1.12) .063 - 0.11 4.33 Home-dwelling 1.07(0.37) .004 0.34 1.81 1.16(0.43) .008 0.31 2.02 0.48(0.39) .218 -0.29 1.26 0.45(0.44) .310 - 0.42 1.33 Institution-based 0.20(0.46) .668 - 0.71 1.11 0.13(0.52) .803 - 0.89 1.15 -0.03(0.47) .951 -0.95 0.89 -0.12(0.52) .819 - 1.15 0.91 Constant 14.36(0.22) .000 13.92 14.81 11.94( 2.44) .000 7.11 16.76 11.41(2.35) .000 6.77 16.05 9.04(3.32) .007 2.47 15.60
- 65 -
Table 4-5: Effect of living arrangements on QOL (Environment Domain) outcomes, Kaohsiung, Taiwan (N=151) (continued)
Model 1
F(18,148)=4.55 Adj R2=0.046
β(SD) p 95%CI
Model 2
F(18,128)=2.04 Adj R2=0.05
β(SD) p 95%CI
Model 3
F(18,128)=6.95 Adj R2=0.42
β(SD) p 95%CI
Model 4
F(18,128)=6.95 Adj R2=0.42
β(SD) p 95%CI
Age -0.01(0.02) .738 -0.06 0.04 -0.02(0.02) .396 -0.05 0.02
Gender -0.19(0.34) .570 -0.86 -0.48 -0.08(0.27) .774 -0.61 0.46
Education 0.78(0.56) .166 -0.33 -1.88 0.75(0.46) .105 -0.16 1.66
Marriage 0.45(0.75) .545 -1.93 1.03 -0.50(0.61) .417 -1.71 0.71
Individual Income 0.84(0.56) .137 -0.27 1.94 0.85(0.45) .063 -0.05 1.75 Family Income -0.35(0.84) .675 -2.02 1.31 -0.44(0.70) .534 -1.83 0.95
Disease 0.26(0.34) .442 -0.41 0.93 0.10(0.38) .793 -0.60 0.78
Depression state -0.80(0.30) .008 -1.38 -0.21 -0.76(0.32) .018 -1.39 -0.13
Life condition 2 -0.80(1.60) .619 -3.96 2.37 -1.82(1.65) .271 -5.09 1.44 Life condition 3 0.93(1.50) .533 -2.02 3.88 0.01(1.54) .994 -3.03 3.06 Life condition 4 1.56(1.52) .306 -1.44 4.57 0.59(1.57) .707 -2.52 3.70
Life condition 5 3.27(1.59) .042 0.12 6.41 2.20(1.65) .187 -1.08 5.47
Health status 2 1.28(0.70) .071 -0.11 2.67 1.24(0.71) .085 -0.17 2.65
Health status 3 0.70(0.64) .276 -0.57 1.97 0.66(0.66) .315 -0.64 1.96
Health status 4 1.46(0.69) .036 0.10 2.82 1.48(0.70) .038 0.08 2.87
Health status 5 1.50(0.84) .076 -0.16 3.15 1.52(0.85) .077 -0.17 3.20
Home-Dwelling 1.00(0.33) .003 0.35 1.66 1.05(0.38) .007 0.29 1.80 0.25(0.30) .407 -0.35 0.85 0.18(0.34) .599 0.49 0.84 Institution-Based 0.34(0.41) .413 - 0.48 1.15 -0.05(0.46) .904 -0.96 0.85 -0.06(0.36) .867 -0.77 0.65 -0.44(0.40) .273 -1.22 0.35 Constant 13.87(0.20) .000 13.47 14.26 14.13(2.15) .000 9.88 18.37 12.25(1.81) .000 8.67 15.82 14.23( 2.52) .000 9.25 19.21
- 66 -