A Bayesian approach to investigate life course hypotheses involving continuous exposures

Abstract

BackgroundDifferent hypotheses have been proposed in life course epidemiology on how a time-varying exposure can affect health or disease later in life. Researchers are often interested in investigating the probability of these hypotheses based on observed life course data. However, current techniques based on model/variable selection do not provide a direct estimate of this probability. We propose an alternative technique for a continuous exposure, using a Bayesian approach that has specific advantages, to investigate which life course hypotheses are supported by the observed data.MethodsWe demonstrate the technique, the relevant life course exposure model (RLM), using simulations. We also analyse data from a case-control study on risk factors of oral cancer, with repeated measurements of betel quid chewing across life. We investigate the relative importance of chewing one quid of betel per day, at three life periods: ≤20 years, 21–40 years and above 40 years of age, on the risk of developing oral cancer.ResultsRLM was able to correctly identify the life course hypothesis under which the data were simulated. Results from the case-control study showed that there was 74.3% probability that betel quid exposure earlier in life, compared with later, results in higher odds of developing oral cancer later in life.ConclusionsRLM is a useful option to identify the life course hypothesis supported by the observed data prior to the estimation of a causal effect.

Latency of tobacco smoking for head and neck cancer among HPV‐positive and HPV‐negative individuals

“…We used a novel Bayesian relevant exposure model that can be viewed as a modification of models previously proposed for latency analysis . The Bayesian approach allowed us to incorporate prior knowledge into the analysis, and we constrained the parameters of the B‐spline basis to be non‐negative (≥0) based on the assumption that smoking cannot reduce the risk of HNC during any period in life.…”

“…The latency of a protracted exposure (e.g., tobacco smoking) is expressed as the effect of exposure (in number of standard packs of cigarettes per day) at different intervals of time before disease development (latency function) . We used a Bayesian relevant exposure model to estimate the time‐dependent relative effects of exposure as a function of time‐at‐exposure . This model can be applied to latency analysis by using retrospective time from disease diagnosis as the time scale and flexible functions (B‐splines) for time‐dependent effects …”

“…We used a Bayesian relevant exposure model to estimate the time‐dependent relative effects of exposure as a function of time‐at‐exposure . This model can be applied to latency analysis by using retrospective time from disease diagnosis as the time scale and flexible functions (B‐splines) for time‐dependent effects …”

“…We used the discrete version of the Bayesian hierarchical model described before to estimate the relative importance of exposure during two periods on the development of HNC, stratified by HPV status. Four Markov Chain Monte Carlo (MCMC) chains were run for each model until convergence and acceptable effective sample sizes were achieved.…”

Disadvantageous Socioeconomic Position at Specific Life Periods May Contribute to Prostate Cancer Risk and Aggressiveness

“…We recently proposed a Bayesian relevant life course exposure model (BRLM) to identify periods of life in which exposure have the highest impact on the outcome ( 39 ). Although we used continuous exposures to demonstrate the technique, it can be used for binary exposures.…”