Quantification of the impact of known risk factors on time trends in non-Hodgkin's lymphoma incidence.

The incidence of non-Hodgkin's lymphoma among white men in the United States was measured as 6.9/100,000 person-years in 1947-1950 and as 17.4 in 1984-1988. We have estimated how much the known and suspected diagnostic and risk factors might have contributed to this apparent increase of 152%. Firm conclusions cannot be drawn without more data on risk and changes in prevalence, but a reasonable range of impacts can be constructed. After accounting for the likely effects of misdiagnosis of Hodgkin's disease as non-Hodgkin's lymphoma, of the acceptance of new entities of non-Hodgkin's lymphoma, of familial factors, of human immunodeficiency virus and other immunosuppressive conditions or drugs, and of occupation, we estimate that the percentage increase in incidence was still 80% among all males and 42% among those aged 0-64. An agent carrying a relative risk of 2.0 rising in prevalence from 0 to 42% would account for the latter rise. Diet, hair dyes, and general environmental exposures to pesticides may be contributing, but currently estimated risks and changes in exposure levels do not appear large enough to account for the residual rise. Among men aged 75-84, some of the residual rise of 109% probably is diagnostic, but only further research will clarify the issue.