Table 4.
Multivariate linear regression models 1 predicting EQ VAS scores and utility scores for the entire sample
| Model 1 | Model 2 | Model 3 | |
|---|---|---|---|
| β (95% CI) | β (95% CI) | β (95% CI) | |
| EQ VAS scores | |||
| Low HL2 (ref. high HL) | −3.27 (−5.8, −0.73) | −0.70 (−3.99, 2.60) | −2.57 (−5.34, 0.21) |
| With CD2 (ref. without CD) | N/A | N/A | −14.47 (−16.73, −12.22) |
| F value (P) | 8.26 (P < 0.001) | 6.24 (P < 0.001) | 24.21 (P < 0.001) |
| R-square | 0.05 | 0.05 | 0.20 |
| Adjusted R-square | 0.05 | 0.05 | 0.19 |
| Utility scores | |||
| Low HL2 (ref. high HL) | −0.023 (−0.044, −0.002) | −0.017 (−0.044, 0.009) | −0.017 (−0.041, 0.008) |
| With CD2 (ref. without CD) | N/A | N/A | −0.117 (−0.137, −0.097) |
| F value (P) | 5.48 (P < 0.001) | 3.83 (P < 0.001) | 18.65 (P < 0.001) |
| R-square | 0.03 | 0.03 | 0.16 |
| Adjusted R-square | 0.03 | 0.03 | 0.15 |
1: Linear regression models, for the entire sample, model 1 adjusting for age, ethnicity, income, geographical location; model 2 further adjusting for education based on model 1; model 3 further adjusting for the presence of chronic disease based on model 2. Significant regression coefficients (β) and 95% CI were shown in bold at P <0.05.
2: CD = chronic disease HL = health literacy.