Skip to main content

Table 2 Regressing family income variables on the presence of a longstanding illness or health problem in 2012

From: Family income and health in Canada: a longitudinal study of stability and change

 

Women

Men

OR (95% CI)

PP1

PP2

RR

RD

OR (95% CI)

PP1

PP2

RR

RD

Model 1

 Average income

0.79 (0.72–0.86)

0.43

0.37

0.87

0.06

0.87 (0.79–0.97)

0.40

0.37

0.92

0.03

Model 2

 Years in bottom $ quartile

1.04 (1.02–1.07)

0.38

0.39

1.02

0.01

1.05 (1.02–1.08)

0.35

0.36

1.03

0.01

 Average income

0.86 (0.78–0.95)

    

0.95 (0.86–1.06)

    

Model 3

 Years in top $ quartile

1.02 (0.98–1.06)

0.40

0.41

1.01

0.00

0.99 (0.95–1.03)

0.39

0.38

0.99

0.00

 Average income

0.74 (0.62–0.90)

    

0.90 (0.75–1.08)

    

Model 4

 SD of incomes

0.97 (0.88–1.05)

0.41

0.41

0.98

0.01

1.04 (0.94–1.14)

0.38

0.39

1.02

0.01

 Average income

0.80 (0.72–0.89)

    

0.85 (0.75–0.97)

    

Model 5

 2011 $ – 2002 $

0.81 (0.74–0.88)

0.43

0.38

0.89

0.05

0.88 (0.78–1.00)

0.39

0.37

0.93

0.03

 2002 income

0.77 (0.71–0.84)

    

0.87 (0.78–1.74)

    

Model 6

 Downward trajectory of $

1.21 (0.98–1.48)

0.41

0.45

1.11

0.04

1.05 (0.82–1.34)

0.38

0.39

1.03

0.01

 Average income

0.76 (0.69–0.84)

    

0.87 (0.77–0.97)

    

Model 7

 Upward trajectory of $

1.04 (0.83–1.30)

0.41

0.42

1.02

0.01

1.04 (0.81–1.33)

0.38

0.39

1.02

0.01

Average income

0.78 (0.71–0.86)

    

0.87 (0.77–0.97)

    
  1. Note: For average income, SD of incomes and 2011 $ – 2002 $, PP1and PP2 are the predicted probabilities at 0.5 standard deviations below and above their means, respectively. For years in bottom quartile, years in top quartile, downward trajectory of $ and upward trajectory of $, PP1and PP2 are the predicted probabilities at 0 and 1, respectively. RR (risk ratio) is PP2 divided by PP1 and RD (risk difference) is the absolute value of PP1 subtract PP2. Each model controls for age in years, immigrant status, marital status and parental education. Person and replicate weights are applied to each model