Table 3.

Unconditional logistic regression analysis of the DNA Repair score^*

Variable	Score value	No. of control individuals (%)	No. of case patients (%)	Adjusted OR† (95% CI)	P	Cross-validation Adjusted OR‡ (95% CI)	P
Score per 1 SD decrease§	0.98 U	140 (100.0)	149 (100.0)	2.7 (1.9 to 3.7)	<.001	2.5 (1.8 to 3.4)	<.001
Score by tertiles‖
Highest tertile	≥4.52	47 (33.6)	6 (4.0)	1.0 (Referent)	—	1.0 (Referent)	—
Middle tertile	3.68–4.51	46 (32.9)	22 (14.8)	3.0 (1.0 to 8.7)	.048	1.6 (0.6 to 4.2)	—
Lowest tertile	≤3.68	47 (33.6)	121 (81.2)	11.5 (4.3 to 31.2)	<.001	7.2 (3.0 to 17.5)	—

^*DNA Repair score definition: 0.00621 × APE1 – 0.047 × OGG1 – 0.0223 × MPG. APE1 = apurinic/apyrimidinic endonuclease 1; CI = confidence interval; MPG = methylpurine DNA glycosylase; OGG1 = 8-oxoguanine DNA glycosylase; OR = odds ratio.

^†Unconditional logistic regression adjusted for age, sex, and smoking status (smoker, past smoker, never smoker).

^‡Cross-validated odds ratios calculated according to the SDs and tertiles of the cross-validated DNA Repair scores.

^§Score was fitted in the unconditional logistic regression model as a continuous variable. 0.98 U is the 1 SD in the control group for the DNA Repair scores. For each model, the estimated odds ratios for smoking were slightly different and the range was as follows: current vs never = 57.1–59.4; current v ex = 1.27–1.29.

^‖Tertiles of the control individuals’ values. The upper tertile was chosen as the referent.