The core controversy: what is a 'safe' dose?
The biggest source of controversy is that scientists cannot agree on the exposure level below which a chemical causes no harm. For the heavy metal cadmium, a traditional 'tolerable' intake was set at 58 µg/day for a 70 kg person. But a 2025 review found that even much lower intakes—between 10 and 15 µg/day—were linked to increased risks of chronic kidney disease, fragile bones, diabetes, and cancer [1]. This directly challenges the idea that a single threshold can protect everyone, because effects appear at levels once thought harmless.
A similar problem appears with the pesticide chlorpyrifos. A 2022 review notes a 'discrepancy between the actual observations and the final conclusions drawn for the substance's approval' in regulatory reports [4]. In other words, the data showing harm at low doses sometimes conflicts with the official conclusions that declare the chemical safe enough for use. This gap between evidence and policy fuels public distrust and scientific debate.
Why different study methods produce conflicting results
Another layer of controversy comes from the fact that different study designs can yield different 'safe' levels for the same chemical. For the anti-inflammatory drug diclofenac in water, a 2023 study on rainbow trout found that subtle kidney damage occurred at concentrations as low as 0.5 µg/L—a level found in many rivers [2]. Yet earlier studies using less sensitive methods had not detected these effects, leading to disagreement about whether the chemical poses a real risk at environmental levels. The authors argue that using more precise, quantitative methods (like stereology) can reveal harm that cruder techniques miss, which shifts the goalposts for what counts as 'safe.'
Similarly, a 2024 study on metal pollution in river sediments used multiple different risk indices (like the modified hazard quotient and ecological risk index) and found that some indices indicated 'moderate' contamination while others suggested 'low' risk [3]. This means the answer to 'is this site toxic?' depends heavily on which mathematical model you choose, creating room for legitimate scientific disagreement.
A path forward: moving from 'safe thresholds' to probabilistic risk
Many experts argue the controversy stems from an outdated approach that pretends there is a single 'safe' dose. A 2022 article proposes that risk assessment should instead embrace uncertainty by using probabilistic methods—essentially, calculating the probability of harm at different exposure levels rather than declaring a hard cutoff [5]. This would replace the current system of 'uncertainty factors' and worst-case assumptions with more realistic, data-driven estimates. The author argues this shift is 'key for constructing a new toxicology paradigm' and could reduce conflict by making the assumptions and remaining risks explicit [5].
The cadmium review [1] and the diclofenac study [2] both implicitly support this view: they show that effects occur across a range of doses, not at a single threshold. By adopting probabilistic approaches, regulators could acknowledge that low-level exposure carries some risk, rather than arguing over an arbitrary 'safe' line. This would not eliminate controversy, but it would make the debate more transparent and grounded in actual data.
About These Sources
This answer is built on 5 peer-reviewed studies — published from 2022 to 2025, 2 from 2024 or later, 3 in Q1 journals, collectively cited 377 times — selected as the most relevant from 5 studies that passed quality screening, drawn from 47 papers retrieved from a database of over 500 million.
Sources used in this answer
Challenges in Toxicological Risk Assessment of Environmental Cadmium Exposure
A 2025 review found that cadmium intakes as low as 10-15 µg/day—far below the traditional 'tolerable' level of 58 µg/day—are linked to increased risks of kidney disease, diabetes, and cancer, challenging threshold-based risk assessment [1].
Detection of Diclofenac-Induced Alterations in Rainbow Trout (<i>Oncorhynchus mykiss</i>) Using Quantitative Stereological Methods
A 2023 study on rainbow trout found that the drug diclofenac caused subtle kidney damage at concentrations as low as 0.5 µg/L, a level commonly found in rivers, using sensitive quantitative methods that earlier studies missed [2].
Metal pollution assessment in the surface sediments of a river system in Türkiye: Integrating toxicological risk assessment and source identification
A 2024 sediment study in a Turkish river used multiple risk indices and found that results ranged from 'low' to 'moderate' contamination depending on the index chosen, showing how methodology affects risk conclusions [3].
Chlorpyrifos Occurrence and Toxicological Risk Assessment: A Review
A 2022 review of the pesticide chlorpyrifos highlights a 'discrepancy between actual observations and final conclusions' in regulatory approval reports, indicating that observed harm at low doses is sometimes downplayed in official risk assessments [4].
Probabilistic risk assessment – the keystone for the future of toxicology
A 2022 article argues that probabilistic risk assessment—which calculates the probability of harm rather than a single safe threshold—is the key to resolving controversies and building a more transparent, data-driven toxicology paradigm [5].