Abstract
Simple SummaryThe age at cancer onset varies worldwide and is intuitively linked to the age structure of the population originating the cases. The exact amount by which age explains this pattern has been estimated for breast cancer to give clues about the 10-year difference in age at cancer onset between low/middle and high income countries. The age contribution for 20 selected cancer types was calculated, using cancer incidence data from all worldwide cancer registries, through linear regression models. For all cancers but skin, age explained 56% of the pattern in men and 65% in women. The percentages varied by cancer types. Since age at cancer onset is embedded with the age structure of the underlying population, it is possible to avoid misinterpretations of a high frequency of cancer onset in specific age groups by inspecting the age pyramid of the population to check for the plausibility of the excess.Background The linear association between median age at cancer onset and median age of the underlying population has been described only for breast cancer. We quantified the shape and strength of such association for 20 cancer types using data from all population-based cancer registries (CRs) worldwide. Methods The patients’ median age at cancer onset and of the underlying population were extracted from all CRs listed in volumes VI (1983–1987 years) and XI (2008–2012 years) of Cancer Incidence in Five Continents. The association was assessed at cross-sectional level by linear regression models and longitudinally considering only the long-standing CRs active throughout the study period (i.e., 25-year span). Results During 2008–2012, each one-year increase in median population ageing was associated in men with a nearly half year increase of median age at onset of all cancers, but skin; and a 2/3 year increase in women. Variance explained by linear model was around 60%. In long-standing CRs a decrease in median age at cancer onset was observed for prostate and cervical cancers throughout the 25-year span. Conclusions Population ageing reflected 60% of the median age at cancer onset. Misinterpretation of peaks of cancer incidence in specific age groups may be avoided by examining population pyramids.
Highlights
IntroductionIt has been observed that the median age at onset of female breast cancer (BC) was a decade higher in high income countries than in low/middle income countries (LMIC)
The association between the median age at onset of 20 cancer sites or groups and the median age of the corresponding population are represented by means of scatter plots and linear regressions for males (Figure 1a), and females (Figure 1b) for the period 2008–2012
The observed pattern firstly gave insights on the worldwide diagnostic changes occurred for prostate and cervical cancers during the examined period, and secondly that for most types of cancer, a younger age structure of a population is linked to a high frequency of young age at cancer onset, whereas an older age structure of a population is linked to a higher frequency of older age at cancer onset
Summary
It has been observed that the median age at onset of female breast cancer (BC) was a decade higher in high income countries than in low/middle income countries (LMIC). There is not much evidence to what amount age explains this association in other cancer sites or the shape of this association The knowledge of this quantity may help interpreting high frequencies of cancer onset in specific age groups worldwide. The linear association between median age at cancer onset and median age of the underlying population has been described only for breast cancer. In long-standing CRs a decrease in median age at cancer onset was observed for prostate and cervical cancers throughout the 25-year span
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
