Introduction. A long-standing belief is that higher testosterone (T) will increase the risk of prostate cancer (PCa), yet recent studies do not support this view.
Aim. To identify the key historical and scientific events leading to the establishment and persistence of the belief in a T-dependent model of PCa growth, despite evidence to the contrary.
Methods. Review of key historical scientific articles regarding T and PCa.
Results. The T-dependent model of PCa growth arose from the work of Huggins and coworkers, who in 1941 demonstrated dramatic responses to castration among men with advanced PCa. These authors and others also reported a rapid clinical progression with T administration. This led to the concept that T was like “food for a hungry tumor” for men with PCa. Fowler and Whitmore recognized in 1981 that the negative effect of T administration did not occur unless men had been previously castrated. However, this critical observation was either forgotten or dismissed amid major changes in PCa diagnosis and management during the 1980s. More recent studies have failed to provide clinical evidence supporting the belief that higher T represents a risk for PCa. Factors contributing to the persistence of the T-dependent model included dramatic effects of castration, continued use of androgen deprivation for treatment of PCa, an influential spokesperson (Huggins), groupthink (failure to acknowledge evidence inconsistent with the prevalent ideology), and an imprecise formulation of the model (“more T, more cancer growth”), making refutation difficult.
Conclusions. The fear that higher T will increase PCa growth stems from a theory of T-dependent PCa growth that originated with observations in a special population (castrated men) that is not particularly relevant to T therapy in hypogonadal men. The negative view of T with regard to PCa should be recognized for what it is—guilt by association. Morgentaler A. Guilt by association: A historical perspective on Huggins, testosterone therapy, and prostate cancer. J Sex Med 2008;5:1834–1840.