A comparison of self reported air pollution problems and GIS-modeled levels of air pollution in people with and without chronic diseases

Table 3 Individual associations between health variables and APP/NO₂ ¹.

	Logistic regression analysis Dependent variable: APP. OR (95% CI) and p-values.		OLS regression analysis Dependent variable: NO_2. Standardized Beta Coefficients (sig.).
Variables	a) Initial model²	b) Full model³	Full model
Asthma	1.51 (1.30 – 1.76) p = 0.000	1.31 (1.12 – 1.55) p = 0.001	-.012 (.154)
COPD	1.80 (1.48 – 2.21) p = 0.000	1.50 (1.21 – 1.85) p = 0.000	.007 (.392)
CHD	1.43 (1.20 – 1.70) p = 0.000	1.17 (0.98 – 1.41) p = 0.082	-.005 (.513)
Diabetes	1.01 (0.73 – 1.38) p = 0.940	1.03 (0.74 – 1.43) p = 0.831	-.005 (.508)
Fibromyalgia	1.45 (1.19 – 1.77) p = 0.000	1.24 (1.01 – 1.53) p = 0.038	-.015 (.068)
Osteoporosis	1.40 (1.08 – 1.82) p = 0.010	1.27 (0.97 – 1.68) p = 0.080	-.005 (.551)
Poor self rated health (SRH)	1.42 (1.26 – 1.59) p = 0.000	1.19 (1.04 – 1.35) p = 0.008	-.016 (.057)

¹Adjusted for all variables in table 2. Each health variable entered separately in the model. The associations reported in the model are not adjusted for the other health variables.
²The variables included in the initial model were: age, sex, employment status, education, smoking, area deprivation, season of examination, and NO₂.
³The variables included in the full model were: all variables in the initial model and self reported pollution due to wood or oil heating, factory etc., dwelling damp, draughts or cold, poor indoor climate and depression.

ISSN: 1476-069X