Table 2

Longitudinal regression of joint damage and TNFA -308G > A genotype (n = 208)

Effect

Estimate

Standard error (estimate)

P-value


Intercept

5.8

1.18

< 0.0001

Genotype GA+AA

-5.0

1.62

0.0020

Year

1.6

0.38

< 0.0001

Year2

-0.13

0.020

< 0.0001

Genotype GA+AA × year

0.17

0.34

0.61


Shown are the variables in the longitudinal regression analysis over 9 years, with joint erosion score as the dependent variable and TNFA -308G > A or TNFA -308G > A × year as primary dependent variables. Year2 allows for a non-linear course of joint progression score over time. The analysis was corrected for gender, and time-averaged 28-joint disease activity score (DAS28), rheumatoid factor and age were not confounders.

Toonen et al. Arthritis Research & Therapy 2012 14:R264   doi:10.1186/ar4110

Open Data