Environmental standard limit fluoride exposure prioritizes neurotoxicity over osteotoxicity in larval zebrafish: A benchmark dose analysis.

Introduction

Fluorine, acknowledged as the most electronegative halogen, primarily exists in the environment as fluoride compounds. These compounds originate from both natural sources, such as mineral leaching, volcanic emissions, and marine aerosols, and anthropogenic activities, including industrial and agricultural production as well as coal combustion (Kashyap et al., 2021). Due to its high solubility and environmental persistence, fluoride readily contaminates water bodies worldwide, with groundwater pollution reported in over 100 countries (Schlesinger et al., 2020; Shaji et al., 2024). Furthermore, public water fluoridation in many regions represents another significant route of exposure (National Health and Medical Research Council, 2017; U.S. CENTERS FOR DISEASE CONTROL AND PREVENTION, 2024; Yee et al., 2025). Collectively, these factors make water the primary route of human exposure to fluoride.

The impact of fluoride on human health is predominantly dose dependent(Solanki et al., 2022). At low levels, fluoride is important for bone mineralization and dental enamel formation; however, concentrations exceeding safety thresholds (e.g., >1.5 mg/L in drinking water) can lead to pathological conditions such as skeletal and dental fluorosis (Srivastava and Flora, 2020). To safeguard public health, numerous national and international organizations have established environmental standards for fluoride, using its adverse effects on hard tissues as the critical endpoint. Representative limits include 1.5 mg/L, as established by the World Health Organization (WHO) (WORLD HEALTH ORGANIZATION, 2022a) and the European Union (European Union, 2020); as well as 4.0 mg/L (primary) and 2.0 mg/L (secondary), as set by the U.S. Environmental Protection Agency (US EPA) (U.S. ENVIRONMENTAL PROTECTION AGENCY, 2021, 2024). In this study, we collectively term these regulatory limits the ‘Environmental Standard Limit Concentration (ESLC)’.

However, these standards may not adequately address the toxic effects of fluoride on non-skeletal tissues, particularly the nervous system. The developing central nervous system (CNS) is particularly susceptible to environmental pollutants, likely due to the increased permeability of the blood-brain barrier during development and the incomplete maturation of innate defense mechanisms (Bear et al., 2015; Farmus et al., 2021). Fluoride has the ability to cross the blood-brain barrier, potentially altering the structure and function of neural circuits. These alterations may impair learning and memory, and contribute to neuropsychiatric changes (Shalini and Sharma, 2015; Dec et al., 2019; Xin et al., 2023). Epidemiological studies indicate early exposure to fluoride impairs intellectual development in children (Yu et al., 2018; Xia et al., 2024; Taylor et al., 2025). An increasing body of evidence suggests that the safe intake level for fluoride may be lower than current regulatory limits. For example, studies conducted in 2023 in China and the United States reported that water fluoride levels exceeding approximately 1.0 mg/L—the regulatory limit for various water sources in China—were associated with significant reductions in children’s intelligence (Zhao et al., 2021; Veneri et al., 2023). Importantly, the fluoride levels associated with these neurodevelopmental effects are significantly lower than those known to cause skeletal damage (National Research, 2006). Therefore, it is essential to investigate whether the CNS demonstrates greater sensitivity to fluoride than the skeletal system.

The term ‘ESLC’ refers to the fluoride concentration limits established in various water quality standards. Traditionally, these limits have been derived using the No Observed Adverse Effect Level (NOAEL) and the Lowest Observed Adverse Effect Level (LOAEL) (More et al., 2022). However, the NOAEL/LOAEL approach is subject to several constraints, including a high dependence on experimental design and an inability to model continuous dose response relationships (U.S. Environmental Protection Agency, 2012). To overcome these shortcomings, guidelines from the U.S. Environmental Protection Agency (EPA) and the European Food Safety Authority (EFSA) recommend the Benchmark Dose (BMD) or Benchmark Concentration (BMC) method for health risk assessment (HRA) (U.S. Environmental Protection Agency, 2012; More et al., 2022). The BMD/BMC approach applies statistical models to dose response data to estimate the concentration (BMD/BMC) that produces a predetermined benchmark response (BMR). The lower confidence limit of this estimate is referred to as the Benchmark Dose Lower Limit (BMDL) or Benchmark Concentration Lower Limit (BMCL) (U.S. Environmental Protection Agency, 2012; More et al., 2022). Given that different statistical models can yield varying BMD/BMC estimates, the Bayesian Model Averaging (BMA) method was developed to integrate results from multiple plausible models, weighted by their statistical support (Madigan and Raftery, 1994; Kaplan, 2021).

This study focuses on gene expression as a molecular level indicator, as it provides an early response, continuous quantitative data of high precision, and the potential to serve as an early biomarker for adverse outcomes (Bourdon-Lacombe et al., 2015). These features enabled us to apply BMD/BMC modeling to derive toxicity thresholds (BMCL??) for both the CNS and skeletal systems, thereby establishing a quantitative basis for comparing their relative sensitivity to fluoride.

Equally pivotal is the selection of an appropriate model organism. The zebrafish serves as an ideal model for this purpose, providing three principal advantages in neurotoxicity research: high genetic and signaling pathway homology with humans, functional parallels in brain regions and behavior, and embryonic transparency that facilitates direct, real time observation of the CNS. (Martin and Plavicki, 2020; Fontana et al., 2022; Wang et al., 2023). For skeletal research, zebrafish offer the dual advantage of mammalian-like developmental processes and optical accessibility, allowing for real-time visualization of bone formation and mineralization through live staining. (Valenti et al., 2020). Due to these combined strengths, the zebrafish has become the third most prevalent model organism in the life sciences.

To address the critical data gap regarding the comparative susceptibility of the CNS and skeletal system to fluoride at ESLC, this study systematically evaluated and compared the toxicity of fluoride to these two systems using a zebrafish model. We exposed 5 dpf zebrafish to a gradient of fluoride concentrations that encompassed the ESLC. A multilevel assessment, including neurobehavioral and osteogenic phenotypes, transcriptomic profiling, and mRNA expression of key molecular markers, was conducted, complemented by a quantitative risk assessment utilizing BMA BMC modeling. This integrated approach aimed to determine which system, the CNS or the skeletal system, is more sensitive to fluoride, thereby providing a scientific basis for refining water standards of fluoride.