Table 4. Demographic and clinical characteristics associated with differences in composite outcome score.
| Bivariate models | Multivariate modela | |||
|---|---|---|---|---|
| Δ Outcome (95% CI)b | P value | Δ Outcome ( 95% CI) b | P value | |
| Female | -7.3 (-13.3 to -1.3) | 0.02 | -13.8 (-18.9 to -8.7) | <0.001 |
| Age, years | 0.5 (0.1 to 0.9) | 0.02 | 0.0 (-0.5 to 0.5) | 0.99 |
| Ethnicity: European | reference | — | reference | — |
| Indigenous | -0.8 (-10.9 to 9.3) | 0.87 | -4.9 (-12.2 to 2.5) | 0.20 |
| Other | 7.1 (-9.4 to 23.7) | 0.40 | 6.1 (-7.0 to 19.1) | 0.36 |
| Citizenship: Canadian | reference | — | reference | — |
| Non-Canadian | 10.9 (2.4 to 19.4) | 0.01 | 5.3 (-5.3 to 15.9) | 0.33 |
| Education: ≤High school | reference | — | reference | — |
| Some college/university | 12.1 (3.7 to 20.5) | 0.005 | 13.7 (0.3 to 27.0) | 0.04 |
| Graduate degree | 18.0 (11.6 to 24.4) | <0.001 | 21.6 (6.1 to 37.0) | 0.006 |
| Employment: Fully employed | reference | — | reference | — |
| Retired | 2.8 (-2.5 to 8.2) | 0.30 | -3.7 (-11.0 to 3.5) | 0.31 |
| Other | 0.1 (-5.9 to 6.0) | 0.98 | -2.1 (-9.9 to 5.6) | 0.59 |
| Relationship: Partnered | reference | — | reference | — |
| Non-partnered | 3.0 (-3.8 to 9.8) | 0.39 | -2.2 (-9.1 to 4.6) | 0.52 |
| Income: $0–$60 000 | reference | — | reference | — |
| $60 000–$100 000 | -2.3 (-9.0 to 4.5) | 0.52 | -7.0 (-17.0 to 2.9) | 0.16 |
| ≥$100 000 | -6.1 (-12.4 to 0.1) | 0.06 | -11.5 (-20.7to -2.4) | 0.01 |
| Smoking status: Non-smoker | reference | — | reference | — |
| Smoker | -16.7 (-23.7 to -9.7) | <0.001 | -8.8 (-17.5 to -0.1) | 0.05 |
| Alcohol consumption: Never | reference | — | reference | — |
| Less than weekly | 9.6 (1.9 to 17.4) | 0.02 | 3.4 (-3.5 to 10.3) | 0.34 |
| Weekly or more | -1.2 (-6.5 to 4.2) | 0.68 | -11.5 (-22.3 to -0.6) | 0.04 |
| Exercise: <150 minutes/week | reference | — | reference | — |
| >150 minutes/week | -1.3 (-8.5 to 5.9) | 0.73 | -1.2 (-7.8 to 5.5) | 0.73 |
CI = confidence intervals. aMultivariate models contain all listed variables, plus variables for study site. b Data are difference in composite outcome relative to the reference condition for the variable (for example, women achieved a mean of 7.3% less than men on the unadjusted analysis, and 13.8% less than men on the composite outcome after adjustment for the effect of other covariates)