Table 4
Linear regression models of refraction, by education, occupation, and income, adjusted for ocular biometric components
| Linear regression coefficients of refraction (dioptres) | |||||
|---|---|---|---|---|---|
| Model 1 | Model 2 (+ axial length) | Model 3 (+ anterior chamber) | Model 4 (+ lens thickness) | Model 5 (+ vitreous chamber) | |
| Data are linear regression coefficients (95% confidence intervals) of refraction, with education, occupation, and income as independent variables. There were 29 people with missing data on occupation (unclassifiable occupation), 104 with missing data on income (including 95 who had retired) and 5 with missing data on housing type. | |||||
| Model 1 includes the following independent variables: age, sex, education, occupation, income, and housing. Linear regression coefficients for a particular variable (eg, education) are adjusted for other independent variables (ie, age, sex, occupation, income, and housing). For example, a 10 year difference in education is associated with a −1.50 dioptres (95% CI: −2.08 to −0.92) difference in refraction, adjusted for age, sex, occupation, income ,and housing. Model 2–5 includes specific biometric components added as additional independent variables to Model 1. | |||||
| *Refer to Table 1 for definitions of occupation, income, and housing type categories. | |||||
| Education, per 10 years | −1.50 (−2.08 to −0.92) | −0.68 (−1.14 to −0.21) | −1.42 (−2.00 to −0.85) | −1.49 (−2.07 to −0.91) | −0.71 (−1.19 to −0.24) |
| Occupation, near work v others* | −0.71 (−1.24 to −0.17) | −0.32 (−0.75 to 0.11) | −0.68 (−1.21 to −0.15) | −0.71 (−1.25 to −0.17) | −0.35 (−0.78 to 0.09) |
| Income, per category* | −0.25 (−0.52 to 0.03) | −0.05 (−0.27 to 0.17) | −0.22 (−0.49 to 0.05) | −0.25 (−0.53 to 0.02) | −0.08 (−0.29 to 0.15) |
