Highway proximity associated with cardiovascular disease risk: the influence of individual-level confounders and exposure misclassification

Table 2 Regression models comparing hsCRP and IL-6 with distance from the highway

Highway distance	Unadjusted model		Exposure adjusted		Adjusted model
Highway distance	(N = 260)		(N = 252)		(N = 225)
hsCRP	%Diff	95%CI	%Diff	95%CI	%Diff	95%CI
hsCRP	Adj. R² = 0.05		Adj. R² = 0.14		Adj R² = 0.38
0-50 m	67%	(-8%,197%)	99%	(12%,254%)	43%	(-16%,141%)
50-150 m	-15%	(-48%,38%)	-24%	(-53%,22%)	7%	(-39%,45%)
150-250 m	75%	(9%,180%)	70%	(7%,169%)	54%	(-2%, 142%)
250-450 m	31%	(-20%,116%)	29%	(-27%,107%)	49%	(-4%,131%)
≥1000 m	ref	ref	ref
IL-6	Adj. R ² = 0.04		Adj. R ² = 0.17		Adj R ² = 0.29
0-50 m	51%	(4%,119%)	72%	(20%,146%)	49%	(6%,110%)
50-150 m	28%	(-6%,75%)	29%	(-4%,73%)	41%	(6%,86%)
150-250 m	54%	(13%,108%)	43%	(7%,90%)	18%	(-11%, 57%)
250-450 m	46%	(-5%,101%)	50%	(11%,101%)	42%	(6%,89%)
≥1000 m	ref	ref	ref

Values represent percent difference between distance category and urban background population.
Exposure adjusted models.
hsCRP adjusted for time spent at home, windows opened in winter and summer, smoking pack years and driving on highway.
IL 6 adjusted for time spent at home, windows opened in winter, work combustion exposures and air conditioner type.
Fully adjusted models.
hsCRP adjusted for age, smoking status, gender, income, BMI, born in the USA, vigorous physical activity, travel on highway, cooked with oil, non-workday time spent inside home, insulin medication, statin medication, heart attack.
IL6 adjusted for age, gender, smoking status, BMI, workday time spent at home, windows opened in winter and air conditioner type.

ISSN: 1476-069X