Table 3

 Final logistic regression models for radiographic abnormalities

Logit coefficient (SE)Odds ratio (95% CI)
*Compared to Group I.
The following variables were considered for model inclusion: age (>50 years), group of exposure, obstructive and restrictive impairments, cumulative exposure (⩾10 fibres/y/cc), latency (⩾20 years), and smoking (current or past).
Parenchymal abnormalities
    Age >50 years1.86 (0.25)6.47 (3.89 to 10.75)
    Cumulative exposure ⩾10 fibres/y/cc1.10 (0.28)3.02 (1.72 to 5.31)
    Current or past smoking0.39 (0.27)1.48 (0.86 to 2.56)
    Constant−5.30 (0.29)
Pleural abnormalities
    Age >50 years1.24 (0.24)3.48 (2.16 to 5.60)
    Cumulative exposure ⩾10 fibres/y/cc0.36 (0.24)1.44 (0.89 to 2.33)
    Current or past smoking0.31 (0.21)1.36 (0.90 to 2.06)
    Group II*−0.70 (0.28)0.49 (0.28 to 0.86)
    Group III*−1.10 (0.36)0.33 (0.16 to 0.67)
    Constant−3.24 (0.38)
Pleural and/or parenchymal abnormalities
    Age >50 years1.37 (0.20)3.95 (2.66 to 5.87)
    Group II* and cumulative exposure ⩾10 fibres/y/cc0.75 (0.44)2.12 (0.88 to 5.10)
    Group III* and current or past smoking0.73 (0.37)2.07 (0.99 to 4.30)
    Cumulative exposure ⩾10 fibres/y/cc0.19 (0.31)1.21 (0.66 to 2.22)
    Current or past smoking0.09 (0.21)1.10 (0.72 to 1.68)
    Group II*−1.24 (0.45)0.29 (0.12 to 0.69)
    Group III*−1.66 (0.42)0.19 (0.08 to 0.45)
    Constant−2.62 (0.39)