br Methods Poisson regression age period cohort APC models w
Methods: Poisson regression age-period-cohort (APC) models were conducted to estimate the age, period and cohorts eﬀects on CRC incidence across the West (i.e., the UK, the US and Australia) and Asia (i.e. Japan, Hong Kong, Shanghai, Singapore and India). We maximized the length of the study period according to each country’s data availability.
Results: Western populations show upward inflections for their 1950s–1960s cohorts, while Asian populations (except India) show downward inflections for their 1950s cohorts. Japanese population also shows upward inflections for its 1960s cohorts, similar to the Western populations. There are apparent upward inflections towards the more recent cohorts for Hong Kong, Shanghai and Singapore; nevertheless, the confidence intervals are wider towards the more recent cohorts.
Conclusion: Our findings imply an increasing risk of CRC in both Western and Asian populations as their younger cohorts reach older ages. These findings are consistent with the life course argument that macro-environmental changes associated with socio-economic development have specific eﬀects that extend over the life course. Actions that pertain to altering lifestyle-related exposures over the life course are of great importance in com-bating young CRC risks in the future.
Colorectal cancer (CRC) is the third most common cancer world-wide after lung and breast cancers, with over 1 million new cases being detected yearly . It is also the third leading cause of cancer death with 880,792 deaths recorded in 2018 . CRC has traditionally been more prevalent among developed countries, contributing to around 60% of global CRC cases . However, incidence varied widely across the globe, with low rate traditionally observed in India .
While genetic factors are known to play an important role, CRC is highly sensitive to non-genetic, macro-environmental factors, in parti-cular modifiable risk factors associated with lifestyle, including low Methoxy-X04 and vegetable intake, high-fat diet, overweight and obesity, se-dentary lifestyle, tobacco use and alcohol consumption . Another
well-known example that demonstrated the impact of macro-environ-mental changes on CRC is the rapid increase of the CRC incidence rates of immigrants to reach those of the host country [5,6]. Other important risk factors include increasing age, male gender, positive family history of inflammatory bowel diseases and adenoma or CRC [4,7,8].
To date, there is scarcity of studies comparing diﬀerent countries to give a coherent story of the temporally related macro-environmental risk factors of CRC. Previous studies have demonstrated the importance of considering the eﬀects of biological age, calendar period and birth cohort to identify temporally related macro-environmental risk factors of CRC [9–15]; however, the conclusions were inconsistent. r> In this study, we aimed to give clues on possible broad, macro-en-vironmental risk factors associated with the changes in CRC incidence across diﬀerent countries. We used age-period-cohort (APC) modeling
Corresponding author at: Room 404, School of Public Health and Primary Care, Prince of Wales Hospital, Shatin, Hong Kong. E-mail address: [email protected] (R.Y.-N. Chung).
which is a powerful method to identify risk factors that are temporally coincided with, and potentially contribute to, the changes in patterns of disease by decomposing incidence rates into the contributions of age, calendar period and birth cohort eﬀects . Age is the surrogate of the natural aging process . Period is the surrogate of contemporary population-wide influences that aﬀect all population under study – e.g. implementation of health interventions on the population at risk , or more general population-wide exposures that see no age limit (e.g. regional wars, financial setbacks or population-wide policies). Cohort is the surrogate of long-term life course influences that aﬀect diﬀerent generations diﬀerentially, because the same generations may have very similar macro-environmental exposures at critical periods of their lives such as during gestation, early childhood, adolescence or early adult-hood [17,18]. As such, this study adopts a cross-country comparative study design, which will enable us to observe both the commonalities and dissimilarities in the patterns of age, period and cohort eﬀects across diﬀerent countries. With this intention, for each country in-cluded in our analysis, we estimated the relative contributions of age, period and cohort to the CRC incidence trends for their corresponding observed periods.