CONTRACT RESEARCH & TOXICOLOGY described as a permitted daily exposure( PDE), acceptable daily intake( ADI), tolerable intake( TI), derived no-effect level( DNEL) or other similar term depending on the specific industry. It is obtained by dividing the PoD by the appropriate uncertainty factors( Figure 2).
Health risk assessment
Comparing a PDE or ADI, etc. with measured or estimated human exposure gives a ratio that we call the margin of safety( MoS). If the PDE or ADI is above the potential human exposure, then the health risks can be considered adequately controlled.
An exposure greater than the PDE or ADI is not necessarily an immediate cause for concern( the use of uncertainty factors is inherently health-precautionary) but some of the built-in conservatism starts to be eroded. An MoS of < 1 can indicate that further investigation is necessary to fully quantify the potential exposures or that additional risk management measures are needed to protect the relevant human population.
Uncertainty factors
Most industries have expert guidance on how to adapt NOAELs and other PoDs to determine a tolerable human level. However, there is a lack of consistency in which specific uncertainty factors should be used.
This can lead to the challenging situation where one key study – and one key PoD for toxicological evaluation – indicates different levels of health risk at the same exposure depending on the industry for which the assessment is conducted.
Let’ s consider the uncertainty factors that are usually applied, and the differences that exist between industries.
The uncertainty factor to account for the variability in the human population, inter-individual variation is probably the most consistent across industries. There is a need to ensure that any toxicological risk assessment applies equally to the most vulnerable individuals, such as the very young or very old, or people with underlying health conditions.
A default factor of 10 is normally considered sufficient, and is used for industrial chemicals1, pharmaceuticals2, 3, medical devices4 and food5.( A factor of 30 might be applied for preterm, neonate and very young infants, for medical devices.) For industrial chemicals, when assessing exposure only to workers, a factor of 5 can apply, as the working population is considered to be generally healthy and less variable( e. g. in terms of age) than the population as a whole.
When it comes to inter-species variation, in contrast, it seems that almost every industry approaches extrapolation in a slightly different way. Guidance in the food industry – some of the longest-standing assessments in toxicology – is to use a standard factor of 10 independent of the species in the study( though this default can be adjusted if speciesspecific data are at hand).
Medical device evaluations take the same approach. However, under guidance for REACH and in the pharmaceutical sector, PoDs are divided by species-specific uncertainty factors( Table 1).
A third factor is study variation. Most human tolerable levels – PDEs, ADIs, etc. – are intended to protect against lifetime, daily exposures to a chemical. However, most chemicals have not been tested in studies covering the whole lifetime of a laboratory animal, so there is a need to apply an uncertainty factor for this extrapolation.
REACH guidance suggests a factor of 2 to extrapolate between a sub-chronic( i. e. 90-day) study and a chronic( lifetime) exposure and a factor of 6 between sub-acute( i. e. 28-day) and lifetime durations. Recent updates in the medical device sector allow for a choice of factors of 2-6( sub-chronic to chronic) and 6-10( sub-acute to chronic), though the upper end is considered the default starting point.
Figure 1- Common uncertainties in laboratory animal studies
JUL / AUG 2025 SPECCHEMONLINE. COM
31