Helicobacter pylori is thought to be involved in the pathogenesis of gastri
c cancer, but the time point at which it produces its effects (critical tim
e) is unknown. We measured the serum level of H. pylori antibody in 787 gas
tric cancer patients and 1007 controls aged 20 to 69, Odds ratios for diffe
rent gastric cancer types and stages were determined for each 10-year age c
lass, The overall odds ratio for gastric cancer decreased with age, being 7
.0 for those aged 20-29, 14.5 for those aged 30-39, 9.1 for those aged 40-4
9, 3.5 for those aged 50-59, and 1.5 for those aged 60-69 (trend in odds ra
tios: P<0.01). However, there was no such age-dependent trend for early dif
fuse-type cancer; the odds ratios were 12.6, 4.0, 7.2, 6.5, and 18.5 respec
tively (P=0.29), Early cancer tended to show higher seroprevalence than adv
anced cancer, especially in older subjects. No significant difference in se
roprevalence was observed between diffuse and intestinal cancers within eac
h age-class, Seroreversion must have occurred in the time interval between
the critical time and the diagnosis of the cancer, especially in older pati
ents. The age-dependent relationship between H pylori and gastric cancer ma
y be due to seroreversion, which itself may be independent of age. This age
-independence indicates that prolonged exposure to H. pylori does not incre
ase the magnitude of its influence on gastric carcinogenesis. Possible mech
anisms through which H. pylori exerts pathogenic effects are continuous inf
lammation in adulthood and/or irreversible damage to gastric mucosa in chil
dhood or the teenage years.