NIH Press Advisory
National Cancer Institute

For Response to Inquiries

Friday, November 21, 1997

NCI Press Office
(301) 496-6641

Recent Trends in Prostate Cancer Incidence and Mortality

Increases in U.S. prostate cancer incidence and mortality rates since the 1970s have been followed more recently by declines. Scientists at the National Cancer Institute (NCI) are working to understand these trends, including conducting investigations into the effect of changing patterns in prostate cancer screening on these rates.

Prostate cancer incidence

Incidence rates rose sharply after 1989, but then began falling sharply after 1992.

There is broad agreement that increased use of the prostate specific antigen (PSA) screening test after 1989 created an artificial rise in incidence rates.

Researchers at NCI are studying the fall in incidence, which may also be a result of changing PSA screening patterns. A first round of screening picks up long latent as well as more recent cancers and creates a "spike" in incidence rates. But if the rate of first-time screening drops, the incidence rate can also be expected to drop, since first-time screening yields more cancers than subsequent screening. The researchers are looking at the use of PSA screening among Medicare patients, and hope that by separating out the early screens from later ones, they will better understand the relationship between PSA testing and prostate cancer incidence.

Prostate cancer mortality

The recent decline in prostate cancer mortality rates occurs in the context of a long-term increase.

NCI researchers caution against attributing the recent decrease in reported prostate cancer death rates to the increase in prostate cancer screening. Two epidemiological concepts are important in understanding why:

Researchers would not usually expect to see any significant drop in mortality before the initial five-year average lead time has passed. The fact that the decline in prostate cancer mortality occurred so soon after the popularization of the PSA test casts serious doubt on the plausibility of a direct cause-and-effect link.

Rapid adoption of PSA testing led to a sharp increase in new diagnoses of preclinical prostate cancer after 1989. As the initial "spike" in incidence began to fall off in 1992, there were fewer men with recent diagnoses of prostate cancer, and consequently fewer deaths might have been incorrectly attributed to prostate cancer.

Using computer modeling techniques and other methods, NCI researchers are investigating the relationship between the decline in prostate cancer mortality rates and the rapidly changing patterns in PSA screening.

Investigating the benefits and risks of screening

Screening is considered useful when there is evidence that treatment at an earlier stage of disease will result in fewer overall deaths or reduce the need for aggressive treatment. In prostate cancer, uncertainty about the natural progression of the disease and the efficacy of specific treatments makes it unclear whether early treatment will result in lower mortality.

It is often very difficult to distinguish between prostate cancers that may become life-threatening and those that may not. Prostate cancer is often slow-growing, and autopsy studies reveal that a large percentage of older men who died of other causes also had undiagnosed prostate cancer. But increased screening is certain to lead to more treatment, and thus more treatment-related morbidity, including impotence and incontinence.

NCI is currently

Prostate cancer is the most frequently diagnosed non-skin cancer in U.S. men, but is a distant second to lung cancer as a cause of death. In 1997, the estimated number of new cases of prostate cancer is 209,900, and the estimated number of deaths from this disease is 41,800. For comparison, lung cancer will strike 98,300 men and kill 94,400 men in 1997.

Potosky AL, Miller BA, Albertson PC, Kramer BS, "The role of increasing detection in the rising incidence of prostate cancer," Journal of the American Medical Association, 1995, Feb. 15; 273 (7): 548-552

For more information about cancer visit NCI's Website for patients, public and the mass media at or NCI's main website at

*Rates are age-adjusted to take into account the aging of the U.S. population. All rates cited in this statement are age-adjusted to the 1970 U.S. standard population.