Background: In international overviews Sweden stands out as a country with low incidence of tobacco-related cancers and a unique pattern of tobacco use with record low prevalence of smoking and record high prevalence among men of use of snus, the Swedish kind of moist oral tobacco. This raises many questions about the relationship between changes of the patterns of tobacco use and changes of incidence of tobacco-related cancers.
Aim: Time series of cancer incidence and tobacco use prevalence cannot take into account the effects of differences between different birth cohorts with respect to initiation and cessation of different tobacco products. The current study was aimed at getting a more detailed analysis by comparing different birth cohorts with regard to tobacco use and age-specific cancer incidence rates.
Methods: Birth-cohort-specific incidence data were retrieved from the NORDCAN database1 and matched against tobacco use data for corresponding birth cohorts retrieved from large nationwide representative surveys. Incidence rates included cancer of the bladder, larynx, lip, lung, mouth and pancreas. Tobacco use data included primary initiation of daily smoking and primary initiation of daily snus use.
Results: For cohorts born in 5-year spans around 1942, 1947, 1952 and 1957 we found a decrease of initiation of daily smoking from 56% in the oldest cohort to 39% in the youngest one, and an increase of initiation of snus use from 6% to 21%. For all cancers there was a shift towards lower incidence rates when going from older to younger cohorts. For example, the mouth cancer incidence rate at age 50-54 was 3.760/105 in the oldest cohort and 2.239/105 in the youngest one.
Conclusions: The current study demonstrates that decrease of smoking is associated with decreasing incidence of smoking-related cancers also when there is a concomitant increase of the use of Swedish snus.