The serum pepsinogen test as a predictor of gastric cancer: The Hisayama study

Yoshinori Oishi, Yutaka Kiyohara, Michiaki Kubo, Keiichi Tanaka, Yumihiro Tanizaki, Toshiharu Ninomiya, Yasufumi Doi, Kentaro Shikata, Koji Yonemoto, Tomoko Shirota, Takayuki Matsumoto, Mitsuo Iida

Research output: Contribution to journalArticlepeer-review

110 Citations (Scopus)


The authors examined whether a serum pepsinogen test (SPT) based on the combination of the serum pepsinogen I level and pepsinogen I/II ratio is a good predictor of gastric cancer occurrence in a general Japanese population. A total of 2,446 subjects aged ≥40 years were classified into negative, positive, and strong-positive SPT groups and were followed prospectively for 14 years (1988-2002). Compared with that for the negative SPT group (26 men, 10 women), gastric cancer incidence increased significantly for both men (n = 17; age-adjusted hazard ratio = 4.56, 95% confidence interval: 2.42, 8.60) and women (n = 6; age-adjusted hazard ratio = 5.84, 95% confidence interval: 2.00, 17.11) in the strong-positive SPT group. It was also significantly higher in the positive SPT group for men (n = 23; age-adjusted hazard ratio = 3.91, 95% confidence interval: 2.23, 6.86). These associations did not attenuate even after adjustment for other comprehensive risk factors. Stratified analysis revealed significant associations between the SPT and development of intestinal-type gastric cancer as well as of cancer in both Helicobacter pylori-negative and -positive subjects. These findings suggest that the SPT can serve as a predictor of intestinal-type gastric cancer, irrespective of H. pylori infection.

Original languageEnglish
Pages (from-to)629-637
Number of pages9
JournalAmerican Journal of Epidemiology
Issue number7
Publication statusPublished - 04-2006
Externally publishedYes

All Science Journal Classification (ASJC) codes

  • Epidemiology


Dive into the research topics of 'The serum pepsinogen test as a predictor of gastric cancer: The Hisayama study'. Together they form a unique fingerprint.

Cite this