Background. There have been significant efforts in the United States t
o lower high cholesterol levels. Studies of men, however, have found a
higher total cancer mortality rate at lower levels of plasma choleste
rol. Many of these studies have found that lung cancer is more closely
associated than other cancers with low cholesterol. Of the studies th
at include women, none has demonstrated a statistically significant in
verse association between low cholesterol and lung cancer. Methods. We
examined the relation between very low plasma cholesterol levels (<16
0 mg/dl) and lung cancer death in an 18-year prospective study of 2,01
1 men and 2,327 women. Results. After adjusting for age, body mass ind
ex, smoking, and education, the relative hazard of lung cancer mortali
ty for those with low cholesterol (<160 mg/dl) compared with all other
cholesterol levels (greater than or equal to 160 mg/dl) was 1.75 amon
g men (P = 0.28) and 3.29 among women (P = 0.02). Excluding those who
died within 5 years of baseline did not change the results. Conclusion
s. Both men and women with baseline plasma cholesterol levels <160 mg/
dl were more likely to die of lung cancer. This difference was statist
ically significant in women. The association could not be explained by
occult malignancy, smoking, or socioeconomic status. (C) 1995 Academi
c Press, Inc.