To test the hypothesis that occupational exposure to chemical agents-p
articularly organic solvents in certain industries-may cause primary l
iver cancer (PLC), a nested case-control study of PLC cases from the D
anish Cancer Registry and an age- and sex-stratified random sample of
controls from the Central Population Register in Denmark were linked w
ith files of a national supplementary pension fund. Employment histori
es since April 1964 were obtained for 973 cases histologically classif
ied as hepatocellular carcinoma, cholangiocarcinoma or combined hepato
cellular and cholangiocarcinoma and 15,348 controls. Men from 35 diffe
rent industrial branches, women from 7 branches, and both men and wome
n from 3 branches had an excess risk of PLC, with an odds ratio of (OR
) >1.0; 29 branches had an OR of liver cancer in excess of 3.0. Women
from book printing and offset printing industries had an OR above 10.
Only male farmers had an OR below unity (0.41). Employees from breweri
es, restaurants, hotels, motels, and distilleries had an increased OR
of both PLC and esophageal cancer. (C) 1997 Wiley-Liss, Inc.