Beyond point estimates: Probabilistic burden of disease modelling of lead-related IQ loss in Germany

Abstract Background Environmental burden of disease (EBD) assessments typically represent uncertainty by confidence intervals for individual parameters, resulting in scenarios difficult to interpret. This approach fails to propagate uncertainties throughout the assessment process, limiting their utility for public health decision-making. Methods We propose a probabilistic Monte Carlo simulation approach to EBD enabling us to work with joint probability distributions across multiple input parameters. We conducted a case study estimating IQ loss and cases of mild intellectual disability (IQ < 70) as well as years lived with disability (YLDs) attributable to lead exposure in German children using nationally representative human biomonitoring data, incorporating uncertainty distributions of the exposure-response function and disability weights. Results Preliminary results show ≈ 20.4 million (uncertainty interval [UI]: 12-28 Million) IQ points lost and 85,000 (95% UI: 0 - 187,000) cases of mild intellectual disability attributable to lead exposure. For threshold-based outcomes like mild intellectual disability, the Monte Carlo approach reveals substantially broader uncertainty compared to deterministic approaches, more accurately reflecting interactions between parameter uncertainties and threshold-based outcomes. Conclusions Our preliminary findings demonstrate how probabilistic EBD methods offer important advantages for public health policy development. By jointly quantifying uncertainty across all parameters and propagating it throughout the entire assessment, our methods allow us to report full probability distributions in the final burden estimates. We encourage epidemiological studies to adopt probabilistic methods and to report full posterior distributions of parameters, as these enable uncertainty propagation in probabilistic EBD. This methodological advancement promotes transparency and supports evidence-based decision-making in environmental public health. Key messages • Lead exposure in Germany remains a relevant public health concern; our EBD estimates highlight the need for continued efforts to further reduce children’s exposure to lead. • EBD estimates with full uncertainty distributions offer comprehensive insights for public health decision-making; epidemiological studies using probabilistic methods are important to support this.