OBJECTIVES: According to epidemiologic surveys, the number of deaths from prostate cancer in Japanese men increased rapidly from 1970 to 2006. However, it is difficult to know the real incidence of, and mortality due to, prostate cancer because the reliability of death certificates and the cancer registry system in Japan are poor. Recently, several studies have demonstrated that baseline prostate-specific antigen (PSA) levels could be one of the most important predictive factors for developing prostate cancer. Therefore, we hypothesized that changes in the baseline PSA distribution in the screening population could reflect trends in the true incidence rate of prostate cancer. METHODS: From 1988 to 2003, 32,274 men, aged 50-79 years, participated in population-based screening for prostate cancer for the first time in Gunma Prefecture, Japan. Changes in the baseline PSA distributions, stratified by a 5-year age range and calendar year, were investigated. The relationships between age and log(10) PSA levels were also investigated and stratified by calendar year. RESULTS: The median baseline PSA level was 0.9-1.2 ng/mL and had not recently increased. No specific trends were found in the percentages of participants with a PSA level greater than 2.0, 4.0, or 10.0 ng/mL within the same age range during the 16-year period. CONCLUSIONS: The increase in the incidence of, and mortality rates for, prostate cancer demonstrated by epidemiologic research might have been misleading in Japan. Investigational changes in the baseline prostate-specific antigen (PSA) distribution of the screened populations revealed that the true incidence rate of prostate cancer in Japan might have been almost the same during the past 16 years.
OBJECTIVES: According to epidemiologic surveys, the number of deaths from prostate cancer in Japanese men increased rapidly from 1970 to 2006. However, it is difficult to know the real incidence of, and mortality due to, prostate cancer because the reliability of death certificates and the cancer registry system in Japan are poor. Recently, several studies have demonstrated that baseline prostate-specific antigen (PSA) levels could be one of the most important predictive factors for developing prostate cancer. Therefore, we hypothesized that changes in the baseline PSA distribution in the screening population could reflect trends in the true incidence rate of prostate cancer. METHODS: From 1988 to 2003, 32,274 men, aged 50-79 years, participated in population-based screening for prostate cancer for the first time in Gunma Prefecture, Japan. Changes in the baseline PSA distributions, stratified by a 5-year age range and calendar year, were investigated. The relationships between age and log(10) PSA levels were also investigated and stratified by calendar year. RESULTS: The median baseline PSA level was 0.9-1.2 ng/mL and had not recently increased. No specific trends were found in the percentages of participants with a PSA level greater than 2.0, 4.0, or 10.0 ng/mL within the same age range during the 16-year period. CONCLUSIONS: The increase in the incidence of, and mortality rates for, prostate cancer demonstrated by epidemiologic research might have been misleading in Japan. Investigational changes in the baseline prostate-specific antigen (PSA) distribution of the screened populations revealed that the true incidence rate of prostate cancer in Japan might have been almost the same during the past 16 years.