Skip to main content

Table 4 Estimates of regression coefficient of multilevel regression analysis for practice and patient characteristics for mental components score (MCS-12)

From: The impact of health literacy and life style risk factors on health-related quality of life of Australian patients

Parameters (reference category) Regression coefficients (standard error)
Empty Model Model only with health literacy Main model with all patient and practice variables Model with interactions
Patient demographics
Intercept 50.97 54.36 52.66 54.76
Female patients (male)    0.19 (0.84) 0.19 (0.83)
Age, years
55–70 (40–54)    2.13 (0.83)* 0.73 (1.15)
Born in Australia (born in outside Australia)    −0.61 (0.85) −0.77 (0.84)
Owner-occupier (rented)    0.23 (1.31) –0.18 (1.30)
College/university (elementary/high School)    0.01 (0.77) 0.08 (0.77)
Employed patients (unemployed)    2.42 (1.13)* 0.17 (1.62)
Retired patients (unemployed)    2.87 (1.32)* 1.92 (1.76)
Patient risk factors
Insufficient health literacy (sufficient health literacy)   –7.02 (0.70) –6.43 (0.74)*** –8.38 (2.43)***
Smoker (non-smoker)    –1.33 (1.44) –1.17 (1.43)
Poor diet (having adequate fruits and vegetables intake)    –0.34 (0.91) –0.20 (0.90)
Alcohol risk (drinking ≤2 standards drinks in a day)    0.07 (1.59) 0.27 (1.58)
Insufficient physical activity (adequate physical activity)    –2.56 (0.73)*** –0.56 (0.97)
Overweight (having normal range BMI)    0.78 (0.76) 0.81 (0.76)
Patient interaction effects
Low health literacy × 55–70     2.70 (1.51)a*
Low health literacy × employed     3.95 (2.01)*
Low health literacy × retired     1.03 (2.60)
Low health literacy × Insufficient physical activity     −4.42 (1.44)**
Practice characteristics
Size 1–3 general practitioners (4 or more GPs)    −0.88 (0.87) −0.06 (0.91)
More than one nurses (one or no nurse)    −0.62 (0.85) −0.85 (0.88)
Random effects
Variance between general practices (standard errors) 1.50 (1.44) 0.25 (0.94) 0.00 (0.00) 0.29 (0.94)
% variance explainedb   83.5 100.0 80.7
Variance between patients (standard errors) 95.24 (5.22)*** 82.64 (4.57)*** 77.39 (4.40)*** 75.16 (4.36)***
% variance explainedb   13.2 18.7 21.1
  1. *P <0.05, **P <0.01, ***P <0.001
  2. aBorderlind significant (P = 0.07)
  3. bExplained variance using the variance in the empty model as reference