Less overdiagnosis of kidney cancer? an age-period-cohort analysis of incidence trends in 16 populations worldwide

Abstract

The increasing rates of kidney cancer incidence, reported in many populations globally, have been attributed both to increasing exposures to environmental risk factors, as well as increasing levels of incidental diagnosis due to widespread use of imaging. To better understand these trends, we examine long-term cancer registry data worldwide, focusing on the roles of birth cohort and calendar period, proxies for changes in risk factor prevalence and detection practice respectively. We used an augmented version of the Cancer Incidence in Five Continents series to analyze kidney cancer incidence rates 1978-2007 in 16 geographically representative populations worldwide by sex for ages 30-74, using age-period-cohort (APC) analysis. The full APC model provided the best fit to the data in most studied populations. While kidney cancer incidence rates have been increasing in successive generations born from the early twentieth century in most countries, equivalent period-specific rises were observed from the late-1970s, although these have subsequently stabilized in certain European countries (the Czech Republic, Lithuania, Finland, Spain) as well as Japan from the mid-1990s, and from the mid-2000s, in Colombia, Costa Rica and Australia. Our results indicate that the effects of both birth cohort and calendar period contribute to the international kidney cancer incidence trends. While cohort-specific increases may partly reflect the rising trends in obesity prevalence and the need for more effective primary prevention policies, the attenuations in period-specific increases (observed in 8 of the 16 populations) highlight a possible change in imaging practices that could lead to mitigation of overdiagnosis and overtreatment.

Keywords: age-period-cohort analysis; incidence trends; kidney cancer.