Table 3.
Associations between ABO blood types determined by 6 ABO SNPs and risk of gastric precancerous lesions, in comparison with normal epithelium or non-atrophic gastritis
| Normal/non- atrophic gastritis |
Atrophic gastritis | Intestinal metaplasia | Dysplasia | Intestinal metaplasia + Dysplasia |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Hp cagA status |
Blood genotypes |
No. | No | ORa | (95% CI) | No. | ORa | (95% CI) | No. | ORa | (95% CI) | No. | ORa | (95% CI) |
| Negative | O | 285 | 59 | 1 | 76 | 1 | 16 | 1 | 92 | 1 | ||||
| B/AB | 44 | 16 | 1.69 | (0.88-3.26) | 14 | 1.18 | (0.60-2.33) | 1 | - | 15 | 1.05 | (0.54-2.02) | ||
| A | 172 | 30 | 0.83 | (0.51-1.36) | 33 | 0.68 | (0.43-1.08) | 2 | - | 35 | 0.60 | (0.38-0.94) | ||
| AO | 155 | 27 | 0.83 | (0.50-1.38) | 31 | 0.70 | (0.43-1.12) | 1 | - | 32 | 0.60 | (0.38-0.95) | ||
| AA | 17 | 3 | 0.89 | (0.24-3.20) | 2 | 0.48 | (0.11-2.17) | 1 | - | 3 | 0.63 | (0.18-2.26) | ||
| A1b | 132 | 24 | 0.84 | (0.49-1.42) | 25 | 0.65 | (0.39-1.07) | 1 | - | 26 | 0.56 | (0.34-0.91) | ||
| A2c | 40 | 6 | 0.81 | (0.32-2.04) | 8 | 0.80 | (0.33-1.82) | 1 | - | 9 | 0.78 | (0.36-1.71) | ||
|
| ||||||||||||||
| Positive | O | 348 | 127 | 1 | 244 | 1 | 47 | 1 | 291 | 1 | ||||
| B/AB | 46 | 19 | 1.12 | (0.63-2.00) | 26 | 0.88 | (0.52-1.48) | 4 | 0.69 | (0.23-2.04) | 30 | 0.85 | (0.51-1.41) | |
| A | 172 | 72 | 1.14 | (0.80-1.61) | 167 | 1.36 | (1.03-1.79) | 42 | 1.78 | (1.11-2.85) | 209 | 1.42 | (1.09-1.86) | |
| AO | 156 | 63 | 1.11 | (0.77-1.59) | 152 | 1.38 | (1.04-1.84) | 38 | 1.78 | (1.09-2.89) | 190 | 1.44 | (1.10-1.90) | |
| AA | 16 | 9 | 1.38 | (0.59-3.26) | 15 | 1.16 | (0.55-2.45) | 4 | 1.77 | (0.55-5.76) | 19 | 1.25 | (0.62-2.54) | |
| A1b | 122 | 46 | 1.02 | (0.69-1.53) | 118 | 1.34 | (0.98-1.83) | 32 | 1.88 | (1.13-3.15) | 150 | 1.42 | (1.06-1.92) | |
| A2c | 50 | 26 | 1.40 | (0.83-2.36) | 49 | 1.40 | (0.90-2.17) | 10 | 1.50 | (0.69-3.23) | 59 | 1.42 | (0.93-2.17) | |
| Pinteractiond | 0.189 | 0.006 | 0.0006 | |||||||||||
Odd ratios were adjusted for age, gender, family history of gastric cancer, smoking status, length of refrigerator use, educational level, fruit and starchy vegetable intakes. Analyses adjusted only for age and gender showed essentially the same results (data not shown). Values in bold are statistically significant (p<0.05). Statistical analysis was not performed for the dysplasia group among cagA negative subjects because of the very small numbers.
A1 includes homozygotes A1/A1 and heterozygotes A1/O
A2 includes heterozygotes A2/A1, heterozygotes A2/O and homozygotes A2/A2
P-value of interaction between Hp cagA status and blood type A