Current practice and recommendations for advancing how human variability and susceptibility are considered in chemical risk assessment

Abstract

A key element of risk assessment is accounting for the full range of variability in response to environmental exposures. Default dose-response methods typically assume a 10-fold difference in response to chemical exposures between average (healthy) and susceptible humans, despite evidence of wider variability. Experts and authoritative bodies support using advanced techniques to better account for human variability due to factors such as in utero or early life exposure and exposure to multiple environmental, social, and economic stressors.This review describes: 1) sources of human variability and susceptibility in dose-response assessment, 2) existing US frameworks for addressing response variability in risk assessment; 3) key scientific inadequacies necessitating updated methods; 4) improved approaches and opportunities for better use of science; and 5) specific and quantitative recommendations to address evidence and policy needs.Current default adjustment factors do not sufficiently capture human variability in dose-response and thus are inadequate to protect the entire population. Susceptible groups are not appropriately protected under current regulatory guidelines. Emerging tools and data sources that better account for human variability and susceptibility include probabilistic methods, genetically diverse in vivo and in vitro models, and the use of human data to capture underlying risk and/or assess combined effects from chemical and non-chemical stressors.We recommend using updated methods and data to improve consideration of human variability and susceptibility in risk assessment, including the use of increased default human variability factors and separate adjustment factors for capturing age/life stage of development and exposure to multiple chemical and non-chemical stressors. Updated methods would result in greater transparency and protection for susceptible groups, including children, infants, people who are pregnant or nursing, people with disabilities, and those burdened by additional environmental exposures and/or social factors such as poverty and racism.

Keywords: Adjustment factors; Chemicals; Cumulative risk; EPA; Environmental justice; NAMs; Risk assessment; Susceptibility; Variability; Vulnerability.

Fig. 1

Sources of human variability and susceptibility to disease risk from exposure to environmental chemicals and pollutants

Fig. 2

How intrinsic and extrinsic factors can influence risk of an adverse outcome In this illustration, a physiological parameter value greater than the clinically-defined threshold results in an adverse outcome or diagnosed disease in the general population (baseline risk). With intrinsic factors alone, some additional proportion of the population may have a parameter value above the clinically-defined threshold and experience an adverse outcome (baseline risk + intrinsic factors). With the addition of extrinsic factors, such as exposure to hazardous chemicals and/or non-chemical stressors, values of the physiological parameter in the population increase, resulting in an increased proportion of the population above the clinically-defined threshold and thus experiencing the adverse outcome

Fig. 3

Depiction of US construct for deriving oral reference dose (RfD) and inhalation reference concentration (RfC) RfD and RfC defined as exposure estimates that are “likely to be without an appreciable risk of deleterious effects during a lifetime” for non-cancer endpoints in the human population (including susceptible subgroups), when using experimental animal data. The RfD or RfC is derived from an experimental animal point of departure (POD), such as the statistical lower limit on a benchmark dose (BMDL) that is associated with a pre-determined change in response. Adjustment factors (AFs) to account for inter-species (experimental animal-to-human) differences and intra-species (healthy humans to susceptible subgroups) variability are then applied to the experimental animal POD. The POD is divided by the animal-to-human AF (AFA) to extrapolate from animals to humans and by the human variability AF (AFH) to account for within-human population variability [73]

Fig. 4

Additive risk of chronic kidney disease (CKD) due to cadmium exposure Cadmium risk assessed in relation to background risk (blue curve) of CKD for a single age group (47.8-year-old women). CKD is diagnosed with a glomerular filtration rate (GFR) of less than 60 ml/min/1.73m2. A reduction in the GFR distribution (red curve) from a chronic cadmium exposure of 1 µg/kg/d increases the population risk of CKD (red box represents increased portion of the population with GFR below the diagnostic threshold) by 3.7%. Reproduced with permission from Ginsberg 2012