The association between fluoride concentrations and spontaneous humeral fracture in first-lactation dairy cows: results from two New Zealand studies.

Study 1

The median fluoride concentration of the ground bone samples in the nine case cows from the Waikato region was 599 (IQR 562.7–763.5) mg/kg and 296.6 (IQR 191.2–391.7) mg/kg, in the seven control cows from the Waikato, Bay of Plenty and Manawat-Whanganui regions (p = 0.007, Mann-Whitney test). A summary of the data is shown in

Study 2

Case samples were from dairy farms located in the Otago (n = 8), Canterbury (n = 4), Southland (n = 9), West Coast (n = 2), Waikato (n = 2) and Manawat-Whanganui regions (n = 1). Although some control animals were from dairy farms located in the Manawat-Whanganui region, the majority were from an animal rendering plant, which while located in this region, collected dead stock from much of the lower North Island. As a result, the farm and region of origin for many of the control samples were not accurately known. The median fluoride concentration of the ground bones from the case cows was 415 (IQR 312.5–515) mg/kg and from the control cows was 290 (IQR 262.5–410) mg/kg, (p = 0.04, exact Mann-Whitney test). A summary of the data is presented in Figure 1 and shows some overlap of values between cases and controls.

Discussion

The results show that for both studies, the median bone fluoride concentration of case cows was significantly higher than that of control cows. However, the bone fluoride concentrations of case cows were well below the 2,500–3,200 mg/kg DM consistent with chronic fluorosis, and instead fell within the range that is considered normal or that is expected to cause no adverse effects (Grace Citation2010).

Fluoride intake via pasture herbage consumption is much lower than through soil ingestion because of the very low fluoride concentration in pasture herbage (generally < 5 mg/kg DM) (Loganathan et al. Citation2001; Kim et al. Citation2016). In addition, the fluoride content of soils varies with the type of rock the soil is derived from, precipitation patterns (due to leaching and/or accumulation of fluoride in soils), soil pH (which affects solubility and mobility of fluoride), presence of organic matter affecting the bioavailability, the region/topography of the soil, and the type of phosphate fertiliser used (e.g. single superphosphate contains 42 times more fluoride than potassium dihydrogen phosphate, while diammonium phosphate contains 57 times more (Li et al. Citation2023)). The majority of fluoride remains in the topsoil, with little moving below a depth of 20–30 cm, and it has been estimated that an application rate of 30 kg P/ha/year will double the fluoride concentrations of a farm’s topsoil every 50 years (Hedley et al. Citation2007; Kim et al. Citation2016).

The estimated annual soil ingestion rate for cattle in New Zealand is around 0.7 kg/day, with an average of 0.35 kg/day in summer and 1.2 kg/day in winter (Healy Citation1968). Unfortunately, when cattle are grazing a winter crop, they often experience high grazing pressure and restricted feed allowances, usually leaving low residuals, meaning soil intake is likely to be higher than Healy (Citation1968) estimated. With dairy cattle regularly wintered on crops there is the potential for the fluoride content of bones to accumulate over one or more seasons.

Bone fluoride concentration increases over time with numerous factors affecting the development of toxic effects, including the amount and duration of ingestion, and animals’ age and nutritional status (Shupe et al. Citation1992; Ranjan and Ranjan Citation2015). Given that none of the affected cows in this study had bone fluoride concentration > 1,200 mg/kg, caution is warranted in interpreting these findings considering that the threshold value for toxic fluoride concentration in bone is 2,500–3,200 mg/kg DM (Grace Citation2010). Indeed, many cattle can have up to 1,000 mg/kg of fluoride in their bones (Suttie et al. Citation1958; Bunce Citation1985) without experiencing a humeral fracture.

Furthermore, in other studies describing the gross and histologic lesions affecting the bones of dairy cows with spontaneous humeral fractures, the observed changes were not consistent with those seen in chronic fluorosis in cattle and were not accompanied by the pathognomonic dental lesions (Shupe et al. Citation1992; Wehrle-Martinez et al. Citation2023a). Nonetheless, bone lesions have been described in animals with mild or no dental lesions (Shupe Citation1980; Ranjan and Ranjan Citation2015).

Studies have shown that bone collagen synthesis and collagen cross-linking are reduced with high fluoride exposure (Susheela and Mukerjee Citation1981; Sharma Citation1982; Ranjan and Ranjan Citation2015), but while there is decreased total collagen content in the bone of cows with spontaneous humeral fractures there are increased collagen cross-links (Wehrle-Martinez et al. Citation2022). Finally, the molecular structure of bone minerals characterised by Raman spectroscopy in mouse femora after in vitro fluoride treatment showed increased bone crystallinity (Freeman et al. Citation2001). In contrast, cases of spontaneous humeral fractures in New Zealand have lower crystallinity (Wehrle-Martinez et al. Citation2023b).

One limitation of this study is that while the control cows for Study 1 mostly came from similar regions as the case cows, for Study 2 the origin of the control cows was either Manawat-Whanganui or unknown but from across the North Island. This is an important limitation for Study 2 as there may be other factors (independent of fluoride concentration) associated with the region that are themselves causatively associated with humeral fracture. This is a good reason to repeat this case-control study using dairy cows from all over New Zealand and perform an analysis of the association of fluoride concentration and humeral fracture stratified on the level of the different potential confounders. For Study 2, the fluoride concentration of the control bones was more similar to the case bones than for Study 1, with obvious overlapping of fluoride concentrations (Figure 1). This shows that the control cows were likely exposed to fluoride-contaminated topsoil as well. This is not surprising given how widespread fluoride contamination of topsoil in New Zealand is, and it is always possible that some of the control cows may have eventually sustained a humeral fracture had they not been culled for another reason. Another limitation of this study is the lack of information regarding the nutrition of the case and control cows and the fertiliser history of the farms of origin for the case and control cows.