Skip to main content

Table 3 Final multivariable logistic regression models for six organ involvements showing associations with occupational and environmental exposures in patients with sarcoidosis

From: Associations between occupational and environmental exposures and organ involvement in sarcoidosis: a retrospective case-case analysis

Exposure

Pulmonary only (n = 164)

Liver involvement (n = 23)

Splenic involvement (n = 37)

Cardiac involvement (n = 17)

Eye involvement (n = 29)

Skin granulomas (n = 23)

OR

95% CI

p

OR

95% CI

p

OR

95% CI

p

OR

95% CI

p

OR

95% CI

p

OR

95% CI

p

Reactive chemicals (n = 26)

 

  

  

 

5.08*

1.28–19.2

0.016

 

  

 

Inorganic dust (n = 74)

2.11*

1.11–4.17

0.026

 

  

 

0.21*

0.04–0.76

0.029

 

  

 

Organic dust (n = 63)

1.86

0.95–3.82

0.079

 

  

 

0.22

0.03–0.91

0.066

 

  

 

Contact with livestock (n = 15)

0.23*

0.07–0.71

0.012

3.68*

0.91–12.7

0.047

4.94*

1.46–16.1

0.008

9.86*

1.95–49.0

0.004

0.00

 

> 0.99

 

 

Close human contact (n = 31)

 

 

4.33*

1.57–11.3

0.003

3.78*

1.47–9.46

0.005

 

  

  

 

Administrative work only (n = 42)

 

  

 

2.52*

0.99–6.16

0.046

 

  

  

 

Active smoker (n = 43)

 

  

  

  

 

3.26*

1.33–7.79

0.008

2.50

0.89–6.54

0.069

  1. For each organ involvement a separate multivariable logistic regression model was constructed. Odd ratios shown result from the best-fit multivariable logistic regression models, using Akaike information criterion (AIC) for model selection. OR = Odds Ratio for having been exposed (cases with a given organ involvement versus all other cases) while adjusting for other exposures and covariates retained in the final models (See “Materials and methods”). CI = Confidence Interval; p = p-value; *when statistically significant (p < 0.05)