Dose-Response Relationships

A dose-response relationship describes how the probability or magnitude of an effect changes as the dose of an agent increases. In environmental health it is the quantitative bridge between an estimated exposure and an expected health effect, and it shapes how thresholds and acceptable levels are reasoned about.

Definition

A dose-response relationship is the association between the amount of an agent received (dose) and the resulting biological response, expressed as a curve relating dose to the probability (quantal) or magnitude (graded) of an effect.

Scope

The topic covers the shape and interpretation of dose-response (and exposure-response) curves, the distinction between threshold and non-threshold (linear no-threshold) models, the points of departure used to summarise them such as the no-observed-adverse-effect level and the benchmark dose, and the debate over low-dose behaviour. It is a reference treatment and does not set or recommend exposure limits.

Core questions

• How does the magnitude or probability of effect change with dose?
• Is there a threshold below which no effect occurs, or does risk extend to the lowest doses?
• How are dose-response data summarised into a point of departure for assessment?
• Why do low-dose extrapolations carry the greatest uncertainty?

Key concepts

• Graded versus quantal dose-response
• Threshold dose
• No-observed-adverse-effect level (NOAEL)
• Lowest-observed-adverse-effect level (LOAEL)
• Benchmark dose (BMD)
• Linear no-threshold model
• Low-dose extrapolation and uncertainty

Mechanisms

As dose rises, more molecular targets are engaged and the response increases, producing a curve that is typically sigmoidal on a log-dose scale. Graded responses describe increasing magnitude in an individual or population mean; quantal responses describe the increasing proportion of a population crossing a defined effect. For many non-carcinogenic effects a threshold is assumed, summarised by a NOAEL/LOAEL or, more robustly, by a modelled benchmark dose (Crump 1984; Klaassen 2018). For some agents, particularly genotoxic carcinogens, a linear no-threshold model is used, implying risk persists at low doses. Empirically, exposure-response gradients such as the inverse relation between waterborne manganese and children's cognitive scores illustrate how observational data inform the curve's shape (Bouchard 2011).

Clinical relevance

The dose-response relationship determines how exposure estimates translate into expected population health effects and frames whether a 'safe' level can be defined. This entry explains how that reasoning works for the purpose of appraising evidence; it does not provide exposure limits, screening thresholds, or individual treatment advice.

Epidemiology

Observed exposure-response gradients are a key strand of evidence for environmental causation, since a monotonic relation strengthens causal interpretation (Bouchard 2011). At the population scale, the large environmental contribution to disease implies that even modest per-unit risks can carry substantial aggregate burden (Rappaport & Smith 2010).

Evidence & guidelines

Dose-response assessment is the second formal step of the 1983 NRC risk-assessment framework, where points of departure such as the NOAEL and the benchmark dose are derived and low-dose extrapolation assumptions are made explicit (NRC 1983; Crump 1984).

History

The dose-response idea descends from Paracelsus's dictum that the dose makes the poison and was formalised in twentieth-century pharmacology and toxicology. The NOAEL/LOAEL approach dominated regulatory practice until Crump's 1984 benchmark-dose method offered a model-based alternative that uses the whole curve, and debate over threshold versus linear no-threshold behaviour at low doses has continued since.

Debates

Threshold versus non-threshold (and hormetic) low-dose behaviour

Whether effects have a true threshold, follow a linear no-threshold relation, or even reverse direction at very low doses is unresolved and agent-specific; the choice strongly affects estimated low-dose risk and the interpretation of 'safe' levels.

Key figures

• Kenny Crump
• Maryse Bouchard
• Curtis Klaassen

Seminal works

• crump-1984
• nrc-1983

Frequently asked questions

What is the difference between a graded and a quantal dose-response?

A graded dose-response measures how the magnitude of an effect increases with dose in an individual or as a population average, whereas a quantal dose-response measures the proportion of a population that crosses a defined all-or-none effect as dose rises.

What is a benchmark dose?

A benchmark dose is a dose estimated from a model of the whole dose-response curve to produce a small, predefined increase in response; it is a more statistically grounded point of departure than a single NOAEL because it uses all of the data.