Objectives When an exposure is protracted or repeated over time, questions arise regarding variation in the effect for different temporal patterns of exposure. We review approaches to describe variation in an exposure’s effect as a function of age-at-exposure, time-since-exposure, and exposure rate. These models implicitly assume that the effect of an exposure increment on subsequent disease risk depends upon its intensity and time-since-exposure, but not upon the intensities of prior exposures. Then, we consider the possibility that the effect of an exposure is dependent upon earlier exposures. We use the term ‘desensitisation’ to refer to the scenario in which a person’s response to an exposure diminishes if they have been previously exposed to it. We use the term ‘sensitisation’ to refer to the scenario in which a person becomes more susceptible to the effect of an exposure if they have been previously exposed to it.
Method We propose a general model for analysis of disease rates in a setting of protracted or repeated exposure that encompasses (de)sensitisation. We illustrate the model using empirical data from a cohort mortality study.
Results The presentation provide examples of how such models can offer a insights into a notion of interaction between an exposure at one point in time and later exposure to the same agent.
Conclusions The possibility that an exposure’s effect may depend upon prior exposure to it is often considered in narrative descriptions of etiological processes, yet not readily accommodated by most standard approaches for analysis of protracted occupational and environmental data.