Computing chronodisruption: how to avoid potential chronobiological errors in epidemiological studies of shift work and cancer.

We ask if epidemiological studies into shift work and cancer may be prone to chronobiological errors. We illustrate how ignoring internal time (IT), or chronotype, may lead to what we call IT errors. Errors from truncating relevant external time (ET) information (activities start before and do not end with the shift) we call ET errors. We develop how observational research may avoid potential chronobiological biases and how chronodisruption can be computed. We assess how IT and ET errors may have affected studies published so far with a focus on those that considered chronobiological information but were confined to night work.