Epidemiologic similarities leading to etiologic hypotheses.

Analogy is a useful means of generating etiologic hypotheses which can be tested by epidemiologic, clinical, or laboratory studies. We describe how the use of analogy has furthered the understanding of three disease entities--Hodgkin's disease (HD), testicular cancer (TC), and acquired immune deficiency syndrome (AIDS). HD is probably caused by a virus of low pathogenicity with clinical expression dependent on social factors. TC may have a viral etiology, although clinical disease may be triggered by other factors. AIDS is caused by a virus which may act as an initiator. AIDS-related Kaposi's sarcoma (KS) may be the result of exposure to a carcinogenic chemical acting as a promoter. In all three diseases, the clinical manifestation could be regarded as varying expressions of host response to a viral agent, modified by an array of socioeconomic and/or biologic cofactors.