Correction of misclassification bias induced by the residential mobility in studies examining the link between socioeconomic environment and cancer incidence

Abstract

Background: Many international ecological studies that examine the link between social environment and cancer incidence use a deprivation index based on the subjects' address at the time of diagnosis to evaluate socioeconomic status. Thus, social past details are ignored, which leads to misclassification bias in the estimations. The objectives of this study were to include the latency delay in such estimations and to observe the effects.

Methods: We adapted a previous methodology to correct estimates of the influence of socioeconomic environment on cancer incidence considering the latency delay in measuring socioeconomic status. We implemented this method using French data. We evaluated the misclassification due to social mobility with census data and corrected the relative risks.

Results: Inclusion of misclassification affected the values of relative risks, and the corrected values showed a greater departure from the value 1 than the uncorrected ones. For cancer of lung, colon-rectum, lips-mouth-pharynx, kidney and esophagus in men, the over incidence in the deprived categories was augmented by the correction.

Conclusions: By not taking into account the latency period in measuring socioeconomic status, the burden of cancer associated with social inequality may be underestimated.

Keywords: Bias (epidemiology); Cancer incidence; Misclassification bias; Poisson regression; Residential mobility; Socioeconomic status.