Associations with VRIA and visual impairment—controls as comparator: multinomial logistic regression
| Relative risk ratio (95% CI) compared with controls: univariate | Relative risk ratio (95% CI) compared with controls: multivariable* | |||
| Variable | Visual impairment but no VRIA (n=1671) | Visual impairment and any VRIA (n=419) | Visual impairment but no VRIA (n=1671) | Visual impairment and any VRIA (n=419) |
| Age (per decade)† | 1.02 (0.99 to 1.06) | 1.2 (1.05 to 1.37) | 1.03 (0.995 to 1.07) | 1.22 (1.07 to 1.40) |
| Sex (compared with male)† | ||||
| Female | 0.94 (0.90 to 0.98) | 1.3 (1.09 to 1.54) | 0.94 (0.90 to 0.98) | 1.28 (1.08 to 1.52) |
| Ethnicity (compared with White)† | ||||
| Asian or Asian British | 0.87 (0.75 to 1.02) | 1.62 (1.12 to 2.34) | 0.83 (0.71 to 0.97) | 1.60 (1.10 to 2.32) |
| Black or Black British | 0.89 (0.79 to 1.01) | 1.11 (0.75 to 1.64) | 0.79 (0.69 to 0.90) | 1.17 (0.79 to 1.74) |
| Other ethnic group | 1.16 (0.94 to 1.43) | 1.07 (0.56 to 2.02) | 1.04 (0.83, 1.29) | 1.10 (0.58 to 2.09) |
| Townsend deprivation index quintiles(compared with first quintile—least deprived)† | ||||
| Second | 1.09 (0.90 to 1.32) | 0.98 (0.69 to 1.40) | 1.09 (0.91 to 1.32) | 0.99 (0.69 to 1.41) |
| Third | 1.28 (1.07 to 1.54) | 1.20 (0.86 to 1.68) | 1.29 (1.07 to 1.55) | 1.21 (0.86 to 1.69) |
| Fourth | 1.23 (1.03 to 1.47) | 1.21 (0.88 to 1.67) | 1.24 (1.04 to 1.49) | 1.21 (0.87 to 1.68) |
| Fifth—most deprived | 1.56 (1.31 to 1.85) | 1.19 (0.86 to 1.65) | 1.59 (1.33 to 1.91) | 1.18 (0.85 to 1.64) |
| BP diastolic (per increase 1) | 0.998 (0.993 to 1.003) | 0.998 (0.989 to 1.007) | – | – |
| BP systolic (per increase 1) | 0.999 (0.997 to 1.002) | 1.001 (0.997 to 1.006) | – | – |
| BMI category (compared with normal/underweight) | ||||
| Overweight | 0.95 (0.84 to 1.07) | 0.94 (0.75 to 1.19) | – | – |
| Obese | 1.04 (0.90 to 1.2) | 1.07 (0.82, 1.38) | – | – |
| Whole body impedance (per increase 100)† | 1.04 (0.99 to 1.09) | 0.88 (0.79 to 0.98) | – | – |
| Waist–hip ratio (per increase 0.1)† | 1.03 (0.99 to 1.08) | 1.11 (1.001 to 1.23) | – | – |
| Pulse wave arterial stiffness index (per increase 1) | 0.99 (0.97 to 1.005) | 1.00 (0.996 to 1.01) | – | – |
| Current smoking status (compared with No)† | ||||
| Yes | 1.39 (1.14 to 1.69) | 1.23 (0.85 to 1.79) | – | – |
| Occasionally | 0.97 (0.70 to 1.36) | 1.20 (0.68 to 2.10) | – | – |
| Medications (compared with none) | ||||
| BP | 1.05 (0.93 to 1.19) | 1.13 (0.90 to 1.41) | – | – |
| Cholesterol† | 1.08 (0.96 to 1.22) | 1.32 (1.06 to 1.65) | – | – |
| Medical history (compared with none) | ||||
| Cancer | 0.99 (0.83 to 1.18) | 0.95 (0.69 to 1.32) | – | – |
| Diabetes mellitus† | 1.13 (0.91 to 1.39) | 1.61 (1.15 to 2.25) | – | – |
| Cardiovascular and/or cerebrovascular disease | 1.00 (0.89 to 1.13) | 1.07 (0.86 to 1.33) | – | – |
| Depression | 0.90 (0.70 to 1.15) | 0.80 (0.50 to 1.28) | – | – |
*Those variables identified as associated univariately (<5% level) were included in a multivariable model and remained in model if significantly associated at 5% level. As cases were matched by age, sex and ethnic group, these variables were included in the multivariable model irrespective of their association.
† Univariate association with visual impairment at 5% level—included in initial multivariable model.
BP, blood pressure; BMI, body mass index; VRIA, vitreoretinal interface abnormalities.