Summary of main themes from participant’s short presentations and discussion:
Any experimental approaches in this area of science need to take into account and control for the known and well described potential confounders of test outcomes such as the test diet, migration of chemicals from cage and water bottles etc.
Any work should be hypothesis-driven meaning that well-constructed hypotheses are required and need to be agreed.
Work should be statistically robust and include power calculations for the endpoints of concern and the degree of change required to be detected.
Work in this area is likely to have the best chance of success in addressing hypotheses around thresholds and the nature of low dose responses if it is focused on chemicals where the MoA is known or can be inferred.
Focus should be on effects known to be adverse, or predicted to be markers of adverse outcomes from a known MoA.
It is intrinsically not possible to investigate threshold events where effects sizes are small using traditional approaches that employ apical adverse outcomes as the endpoint. It is much more likely that focusing on key events in a MoA will be informative
Non-monotonic dose responses for adverse events in vivo are rarely seen and, if seen, the underlying mechanisms have not been described.
Synergistic effects at low dose are rarely if ever seen in vivo for endocrine active chemicals or those that work through other MoAs.
Adverse effects occur when homeostatic mechanisms are overcome.
There is therefore a need for reproducibility of the findings to be established for these low-dose effects in order to provide more convincing evidence.