Table 3.

Predictive models of FA exposure class and levels in the PARIS cohort study by multivariate logistic and linear regression models.

Multivariate logistic regression model			Multivariate linear regression model
Predictive factors	OR (95% CI)	p	β (95% CI)	p
Home characteristics
Construction date
1976–1990 (vs. before 1975)		1.26 (0.41, 3.92)		0.686		0.04 (–0.17, 0.24)		0.722
After 1990 (vs. before 1975)		3.61 (1.09, 11.98)		0.036		0.09 (–0.09, 0.27)		0.334
Housing area
≥ 70 m² (vs. < 70 m²)		2.07 (0.94, 4.58)		0.070		0.10 (–0.03, 0.23)		0.141
No. of occupants
> 3 (vs. ≤ 3)		2.11 (0.96, 4.64)		0.064		0.12 (–0.02, 0.25)		0.086
Continuous sources
Wall coating (paint or fiber cloth)
Yes ≥ 1 year (vs. no)		5.34 (1.84, 15.46)		0.002		0.16 (–0.01, 0.33)		0.053
Yes < 1 year (vs. no)		5.14 (1.76, 15.03)		0.003		0.17 (–0.01, 0.34)		0.052
Wood-pressed products for flooring or varnished parquet floor
Yes ≥ 1 year (vs. no)		1.98 (0.87, 4.51)		0.103		0.15 (0.01, 0.29)		0.037
Yes < 1 year (vs. no)		3.70 (1.06, 12.86)		0.040		0.23 (0.03, 0.44)		0.023
Particle board furniture
Yes ≥ 1 year (vs. no)		4.20 (1.16, 15.17)		0.028		0.37 (0.16, 0.58)		0.001
Yes < 1 year (vs. no)		4.34 (1.30, 14.53)		0.017		0.42 (0.23, 0.61)		< 0.001
Aeration/air stuffiness
Mechanical ventilation
Yes (vs. no)		1.74 (0.72, 4.21)		0.221		0.08 (–0.06, 0.23)		0.266
Double glazing
Yes (vs. no)		2.76 (1.22, 6.28)		0.015		0.28 (0.14, 0.42)		< 0.001
Duration of windows opened (> 1 hr)a		0.89 (0.81, 0.99)		0.028		–0.02 (–0.03, 0.01)		0.061
β, partial regression coefficient. aAdjusted for season of declaration.