Skip to main content

Table 3 Multivariate weighted ordered logistics model used for sensitivity analysis on the association between blood log-BEX and hearing loss

From: Association between blood volatile organic aromatic compound concentrations and hearing loss in US adults

Characteristics

Model 1

Intercepts

Hearing loss

OR (95%CI)

Normal vs. Mild

Mild vs. Moderate to severe

 Ethylbenzene

1.24 [0.04,0.39]

1.99

4.16

 M-/p-Xylene

1.19 [0.02,0.34]

2.39

4.56

 O-Xylene

1.27 [0.03,0.44]

1.98

4.14

 Benzene

1.17 [0.01,0.03]

2.17

4.34

 BEX

1.24 [0.04,0.39]

2.43

4.60

SFHL

   

 Ethylbenzene

1.07 [-0.10,0.24]

3.82

5.39

 M-/p-Xylene

1.04 [-0.13,0.21]

3.97

5.54

 O-Xylene

1.07 [-0.12,0.25]

3.84

5.41

 Benzene

1.09 [-0.07,0.22]

3.75

5.32

 BEX

1.07 [-0.10,0.23]

3.86

5.53

HFHL

   

 Ethylbenzene

1.25 [0.05,0.40]

1.97

4.11

 M-/p-Xylene

1.21 [0.03,0.35]

2.37

4.52

 O-Xylene

1.28 [0.05,0.45]

1.95

4.10

 Benzene

1.18 [0.01,0.31]

2.17

4.32

 BEX

1.25 [0.05,0.40]

2.42

4.57

  1. Abbreviations: CI, confidence interval; OR, odds ratio. SFHL: Speech-frequency hearing loss; HFHL: High-frequency hearing loss
  2. Model 1: adjusted for age, sex, race, BMI, diabetes status, drinking status, hypertension, education level, smoking status, Marital status, PIR, work noise, recreational noise, firearm noise, and year cycle
  3. Log-BEX: log-transformed volatile organic aromatic compound values
  4. Participants were reclassified into three levels based on the average hearing threshold (PTA): normal (< 20 dB), mild (20–35 dB), and moderate to severe (> 35 dB)
  5. Intercepts: represent estimated intercepts for different outcome levels in the ordered logistic regression model