"The dose makes the poison" is the oldest principle in the field and the most frequently skipped step in litigation. Almost every causation dispute eventually becomes an argument about extrapolation.
Start a conversation with our AI Research Concierge, already scoped to dose-response toxicology. Pick a starting point, or describe your situation directly.
Dose-response describes how the probability or severity of an effect changes with the amount of exposure, and it sits underneath both causation showings. It is what determines whether studies conducted at one exposure level say anything useful about a claimant exposed at a very different one — and since occupational and animal studies are usually conducted at concentrations far above environmental exposures, that extrapolation is where a great deal of the contested reasoning lives. Whether an effect has a threshold below which it does not occur, how to reason from animals to humans, and how to handle exposures below the range any study has examined are all genuinely unresolved in parts of the science, which is precisely why they are argued so hard.
Each carries real scientific uncertainty, which is what makes each of them contestable in a way that is not merely adversarial.
The shape of the relationship — linear, sublinear, or threshold — which governs what low-dose exposure implies.
Whether a level exists below which no effect occurs. Assumed absent for genotoxic carcinogens, contested for much else.
Reasoning from study concentrations down to environmental ones, across a gap that is often several orders of magnitude.
Whether an effect in a test species predicts an effect in humans, and how differences in metabolism are handled.
Whether cumulative dose, peak concentration, or exposure pattern drives the effect — they can imply different conclusions from the same total.
Interaction between substances, which most single-compound literature does not address at all.
How dose-response evidence is evaluated.
It is the hinge connecting the literature to the claimant, and it is where opinions most often over-reach.
Screening levels and reference doses are set conservatively to protect populations, incorporating deliberate safety factors. Exceeding one is not evidence that harm occurred — treating it as though it were is a recurring and highly attackable error.
No, and the conflation is one of the most common errors in these matters. Screening levels, reference doses, and maximum contaminant levels are risk-management values set to be protective across whole populations including sensitive members, typically incorporating substantial uncertainty factors below any observed effect level. An exceedance signals that further evaluation is warranted — it is a trigger, not a finding. Presenting it as evidence that a claimant was harmed misdescribes what the value is, and opposing experts will say so.
With explicit assumptions, stated as assumptions. Where no study has examined the exposure range at issue, any statement about effects there rests on a model of the curve rather than on observation — commonly linear-no-threshold for genotoxic carcinogens, which is a protective default rather than an established fact about low-dose biology. An opinion that adopts such a model should say which model, why it is appropriate for this substance's mode of action, and what conclusion would follow under a reasonable alternative.
Real weight, with the interspecies question always live. Animal studies allow controlled dosing that human studies cannot, and for many substances they are the primary evidence of effect. The contested step is whether the species metabolises the compound comparably, whether the effect has a human analogue, and whether the doses used bear any relation to environmental exposure. Findings supported by a mechanism plausible in humans travel considerably further than findings at extreme doses with no established human pathway.
It is common in these matters and poorly served by the literature, which is overwhelmingly single-compound. Interactions can be additive, more than additive, or antagonistic, and for most real-world mixtures the data simply does not exist. Honest analyses treat co-exposure as a source of uncertainty and address whether other substances present could explain the outcome — which is also a defensive argument, since a co-exposure the plaintiff's expert has not addressed is an alternative explanation left standing.
Describe the substance and the exposure levels at issue. We will scope it and connect you with the right expert.