Background In 1950, landmark epidemiology contributed to identifying smoking as a key carcinogen [Wynder and Graham; Doll and Hill]. In 2007, IARC classified shiftwork involving circadian disruption [CD] as probably carcinogenic; however, epidemiological evidence in regards to the carcinogenicity of shift-work that involves nightwork is conflicting.

Objective To compare smoking research from the 1950’s to shiftwork research for exploring the hypothesis that today’s shiftwork epidemiology is lacking chronobiological and methodological rigour and to develop metrics to facilitate improvement.

Methods Comparing smoking and chronobiological insights and deriving CD metrics.

Results If doses had been limited to number of cigarettes smoked at work rather than over 24 hours, smoking insights could have been delayed or disallowed. Similarly, restricting exposures to, let alone doses of, CD from work at night may prove insufficient to elucidate effects of cumulative CD. CD doses may be obtained by comparing how activities overlap with individuals’ biological nights (BNs: predicted by chronotype), yielding CD_BNhours. Total CDhours may be obtained by summing up CD_BNhours due to activities at and off work. As a more easily applicable metric, how much sleep overlaps with the individual biological day (BD) may yield CD_BDhours.

Conclusions Epistemologically, shiftwork epidemiology is lacking chronobiological and methodological rigour. CD - like smoking - must be assessed at and off work to consider cumulative doses in studies of carcinogenicity. Epidemiological research before and after IARC 2007, based on (night)shifts alone, may have delayed or disallowed detection/measurement of the existence/magnitude of possibly carcinogenic effects of cumulative CD.