The absence of correlation between Na in diet duplicates and stomach cancer mortality in Japan.

Abstract

The 24-hr diet duplicates were collected from 488 nonsmoking and nondrinking women at the ages of 30-59 years in 33 regions in Japan in winters, 1977-1981. The daily intake of Na, K, Cl and Na/K ratio for each individual were obtained by the chemical analysis of the duplicates, from which the regional means were calculated and subjected to regression analysis with 1969-1978 regional mortality for stomach cancer (SC), cerebrovascular disease (CVD) and liver cirrhosis for the middle-aged, 1969-1978 regional standardized mortality ratios (SMRs) for cancers of 9 sites and 9 other diseases (including SC and CVD), and 1978-1982 regional SMRs for SC, uterine cancer and CVD. No significant correlation was observed between intake of Na, K, Cl or Na/K and SC mortality or SMR, in contrast to a significant correlation of CVD mortality and SMR with Na intake (and to a lesser extent with K and Cl intake). Correlation of other cancers with Na was generally insignificant except for pancreatic cancer and possibly breast cancer. Thus, it was concluded that the correlation of daily Na intake with stomach cancer mortality is weaker, if present, than that with cerebrovascular disease mortality.