Regulatory Systems Must Rely on a Sturdy Foundation of Sound Science

Monday, February 26, 2024

A recent commentary published in the Society of Toxicology’s Toxicological Sciences journal examines the European Food Safety Authority’s (EFSA) reduction of the tolerable daily intake (TDI) level of bisphenol A (BPA) in food contact materials (compared with the temporary TDI established in 2015), which the authors of the commentary say is based on an insufficient foundation for regulatory action.

The commentary notes: “The EFSA hazard assessment review process is not scientifically sound and has led to a conclusion that there are low-dose effects based on very few, lower quality experimental animal studies. These effects and the conclusions drawn are not sufficiently supported by the totality of the available evidence, which includes multiple high quality studies not considered in EFSA’s recent review.”

In their conclusion, the authors make recommendations for future assessments that seek to develop safe dose estimates for BPA (or any substance):

  • Include all available evidence;
  • Provide clear guidance for evaluating study quality;
  • Consider the reliability of study results; and
  • Choose endpoints that are either adverse, apical effects (or their precursors) and that are both consistent and coherent across studies in experimental animals and humans.

The European Medicines Agency (EMA) and the German Federal Institute for Risk Assessment (BfR) have criticized the analysis used by EFSA, and both groups stated they opposed EFSA’s approach. Similarly in the United States, the Food & Drug Administration, which has evaluated scientific evidence around BPA, stated that it “continues to conclude that BPA is safe for the currently authorized uses in food containers and packaging.”

Regulations must have a robust and sturdy scientific foundation to provide for public health and safety and keep the confidence of the public. The commentary was published on January 23, 2024, in Toxicological Sciences and can be accessed at the following link: