Skip to main content

Table 4 Other Sensitivity Analyses

From: A common functional variant on the pro-inflammatory Interleukin-6 gene may modify the association between long-term PM10 exposure and diabetes

Sensitivity analysis

Genotype

Adjusted OR (95 % CI)

P-value*

P-value**

P-value***

P-value****

Diabetes defined as self-reported physician diagnosis and medication use [N(diabetes) = 196]

IL6 -572G > C

     

GG

1.41 (1.11, 1.79)

0.005

0.001

n.d.

0.004

GC + CC

0.63 (0.38, 1.04)

0.070

   

IL6 -174G > C

     

GG

1.23 (0.88, 1.71)

0.222

0.931

0.797

0.794

GC

1.33 (0.98, 1.81)

0.065

   

CC

1.17 (0.58, 2.35)

0.669

   

Diabetes defined as self-reported physician diagnosis only [N(diabetes) =193]

IL6 -572G > C

     

GG

1.43 (1.13, 1.83)

0.003

0.008

n.d.

0.003

GC + CC

0.63 (0.38, 1.04)

0.072

   

IL6 -174G > C

     

GG

1.23 (0.88, 1.71)

0.226

0.881

0.749

0.694

GC

1.38 (1.01, 1.88)

0.041

   

CC

1.16 (0.57, 2.33)

0.680

   

Diabetes defined as self-reported use of diabetes medication only [N(diabetes) =125]

IL6 -572G > C

     

GG

1.25 (0.95, 1.66)

0.113

0.008

n.d.

0.031

GC + CC

0.49 (0.22, 1.10)

0.085

   

IL6 -174G > C

     

GG

1.10 (0.72, 1.66)

0.666

0.954

0.600

0.803

GC

1.24 (0.86, 1.78)

0.246

   

CC

0.90 (0.36, 2.28)

0.829

   

Diabetes cases identified from blood tests only [N (diabetes) = 184]

IL6 -572G > C

     

GG

1.65 (1.27, 2.13)

<0.001

0.002

n.d.

0.006

GC + CC

0.70 (0.39, 1.25)

0.223

   

IL6 -174G > C

     

GG

1.46 (1.04, 2.06)

0.030

0.738

0.553

0.962

GC

1.41 (1.00, 1.99)

0.052

   

CC

1.83 (0.86, 3.90)

0.118

   

Model applying random slopes for study areas [N(diabetes) = 252]

IL6 -572G > C

     

GG

0.92 (0.44, 1.92)

0.819

0.004

n.d.

0.008

GC + CC

0.44 (0.19, 1.05)

0.065

   

IL6 -174G > C

     

GG

0.86 (0.41, 1.83)

0.704

0.814

0.928

0.687

GC

0.78 (0.36, 1.70)

0.540

   

CC

0.85 (0.33, 2.20)

0.735

   
  1. Adjusted models include age, sex, educational attainment, neighborhood-level socio-economic status, smoking status, pack-years of smoking, exposure to passive smoke and occupational dusts and fumes, dietary fibre intake, alcohol consumption, physical activity, body mass index (BMI), PM10. Study area was treated as random effects in all models except the model with random slopes for study area.OR: odds ratio; CI: confidence intervals; OR values represent percent increase in odds of diabetes per 10 μg/m3 increase in PM10 exposure. PM10: particulate matter <10 μm in diameter; n.d: not done due to very low sample size for CCgenotype. * P-value of genotype specific association between air pollutant and diabetes. **additive model (per C allele); ***dominant model (GG + GC vs. CC); ****recessive model (GC + CC vs.GG)