Associations between Dietary Patterns, Fluoride Intake and Excretion in Women Exposed to Fluoridated Salt: A Preliminary Study.

In our study, we found an association between a diet high in convenience foods and dairy (termed the “Urban Convenience Diet”) and the intake and excretion of dietary F in a sample of Mexican women. Our novel approach considers fluoride intake as part of dietary patterns, rather than in isolation, and questions the accuracy of crude dietary intake estimates in reflecting biologically relevant exposures. Interestingly, while the Urban Convenience Diet was positively associated with F intake, it was negatively associated with F excretion. This contradicts the assumption of a linear relationship between intake and excretion [6].

There are two potential explanations for our findings: (1) The Urban Convenience Diet may include foods that when eaten together with fluoride, inhibit its intestinal absorption, resulting in lower urinary excretion [12], or (2) The Urban Convenience Diet may promote the reabsorption of fluoride in the renal tubules, leading to a longer re-circulation time in the plasma before its excretion [14]. The first explanation is supported by our finding of a moderate, significant correlation between the Urban Convenience Pattern, and estimated dietary calcium intake (r = 0.37, p = 0.02). It is well-known that simultaneous intake of fluoride and calcium-rich foods reduces fluoride’s gastric and intestinal absorption; this is because fluoride and calcium react to form insoluble salts that are excreted in the feces [14]. The second explanation ?higher fluoride reabsorption in the renal tubules, could be due to the overall acidity of the diet. Alkaline diets are associated with higher urinary excretion of fluoride, whereas acidic diets have been linked to lower urinary excretion due to increased recirculation of fluoride in the plasma [14,25]. The Urban Convenience pattern is characterized by the consumption of foods such as dairy, sodas and salsas, which are high in both salt (and therefore, F), but also in acid precursors. It is important to note that in this study, the dietary intake of fluoride was estimated from a dietary questionnaire, which like any dietary instrument, is subject to measurement error [16]. Although, we cannot definitively state the reasons behind the observed associations, we can say that the urinary excretion of fluoride may not necessarily follow a monotonic relationship with F intake, and that this relationship will depend on the combination of foods with which fluoride is ingested, and the overall eating pattern. For instance, there are individual foods with high fluoride levels (such as fast foods and sea food) [20], however it is the context of their ingestion the one determining the bioavailability of the fluoride ion. When using dietary questionnaires to estimate F intake in exposure assessment, it is crucial to understand that a higher estimated intake does not necessarily mean a higher biologically relevant (bioavailable) exposure to F and interpret the data within these limitations.

Dietary recalls and questionnaires are well-established tools for estimating dietary exposures in nutritional epidemiology [16] and are becoming a cost-effective alternative for assessing F exposure in environmental epidemiology studies [5,26]. However, our findings highlight that crude estimates do not account for factors affecting nutrient bioavailability, potentially increasing the risk of exposure misclassification. Using a Bland-Altman plot (Figure 2), we observed better agreement between fluoride intake estimates and urinary fluoride measurements at exposure levels below 1.5 mg/day. Beyond these levels, the likelihood of error increased. For fluoride exposure assessment, it is desirable to use instruments with higher precision also at high exposure levels, as these data points could provide valuable information on potential side effects. Our data suggests that the 24-h recall may not be suitable to substitute 24-h urinary fluoride measurements; however, this was a preliminary study, and a larger sample would be needed to confirm this finding.

This study had both strengths and limitations. Among its strengths, we collected 24-h urine samples and had a professional nutritionist administer the dietary recall. Our fluoride database was specifically developed for the population under study, and our study personnel visited the participants’ homes to make individual estimates of salt intake. On the other hand, limitations include that the current analyses were nested within another study that included a sodium intake reduction intervention [18], and that participants with metabolic disease were not excluded from participation. Although it is known that acid/base disorders can influence the excretion of fluoride [14], families served by Centro Meneses were under medical treatment, which decreases the likelihood of confounding. Lastly, the age range of the participating women was wide (while it is well known that the excretion of fluoride increases with age) [14]. To address these limitations, our quantile regression models controlled for each participant’s age and the study visit (1st or 2nd visit) at which the sample was collected. In an abundance of precaution, we also measured urinary creatinine and included it as a variable in our urinary excretion models to account for the effects of urine dilution.

We found that in this sample of Mexican women exposed to fluoridated salt, a dietary pattern characterized by dairy and convenience foods influences both the intake and excretion of fluoride. Under these conditions, the dietary intake of fluoride as estimated with a 24-h dietary recall and its 24-h urinary excretion were not linearly related and had better agreement at levels below 1.5 mg F/day.