Fluoride Risk Assessment of Different Brands of Coffee Commercialized in the Canary Islands

Abstract

Coffee is part of our daily lives since it is a drink with a symbolic value, both in our social relationships and as a stimulating drink. The main objective of this paper is to conduct a risk characterization of fluoride overexposure from the consumption of these beverages; for this purpose, in the present study, the fluoride concentration was determined in 80 samples of different brands of coffee commonly consumed in the Canary Islands, and the risk derived from their intake was then evaluated. Significant differences (p < 0.05) were found between the coffee brands soluble Nescafé Classic^®, Saimaza^®, Auchan^®, Tirma^® soft blend, and Hacendado^® and the other analyzed brands. The brand that recorded the highest mean fluoride levels is “Hacendado descafeinado^®” (0.308 mg/L). The problem with coffee is not the fluoride it contains but the fact that, according to the survey results, most coffee drinkers brew their coffee with tap water (over 51% of respondents use tap water). The results of the fluoride content in coffee prepared with tap water from the different areas of Tenerife in the Canary Islands (metropolitan, northern, and southern areas) have reported higher concentrations of fluoride: north (max. concentration found of 16.68 mg/L), south (max. concentration found of 8.47 mg/L) and metropolitan (max. concentration found of 4.73 mg/L). Coffee has not been identified as a risk of dietary overexposure to fluoride; however, the use of tap water for its brewing has been identified as a factor that considerably increases its concentration. Therefore, the use of bottled water is recommended, especially in regions such as the Canary Islands, which have high concentrations of fluoride in their water.

1. Introduction

Fluorine is the most electronegative and reactive element in the periodic table. Therefore, it is not found in its elemental state in nature. In water, fluorine is mainly found as fluoride ion (F⁻) and fluorinated organic compounds. The main natural source of fluorides in water bodies is the emission of gases rich in hydrofluoric acid (HF) from volcanic activity. Similarly, the erosion of fluorinated minerals is noteworthy [1,2] fundamentally in fluorapatite (Ca₅(PO₄)₃F), fluorite (CaF₂), and cryolite (Na₃AlF₆). Furthermore, fluoride levels can increase because of anthropogenic activity such as industrial activity (coal burning, electroplating, glass, steel, ceramics) as well as the use of fertilizers [3].

The presence of environmental fluoride can cause terrestrial plants close to pollution sources to accumulate inorganic fluorides [4], with the highest accumulation rate being recorded as a result of the deposition of fluorides from the air and subsequent absorption by the stomata. Fluorapatite is the main source of fluoride in soils that have been fertilized with phosphate fertilizers containing this mineral.

Although fluorine is not considered an essential element, it does have a hormetic effect, providing certain benefits to humans [5]. An adequate intake of fluorides offers protection against dental caries because, among other mechanisms, it acts by inhibiting the metabolism of cariogenic bacteria, reducing the formation of acids (acetic and butyric) that would end up decomposing dental hydroxyapatite [6,7]. Furthermore, hydroxyapatite reacts in the presence of fluoride, yielding fluorhydroxyapatite, which is more resistant to erosion [8]. At certain doses, it has also been related to an improvement in osteoporosis [9,10].

However, when fluoride exposure is high and sustained over time, the risk of suffering adverse effects such as dental fluorosis (mottled teeth, with weakness and malformation of teeth) increases, as well as bone fluorosis in extreme cases [11]. Various studies conducted in areas of India and China have shown that when the concentration of fluoride in water exceeds 1 mg/L, the risk of suffering from bone fluorosis increases, and the intelligence quotient (IQ) in children is reduced [12,13].

In the Spanish Royal Decree (RD) 3/2023, which regulates parametric values in supply water in Spain, a value of 1.5 mg/L of fluoride is established for water for human consumption. When this value is exceeded, the water is not considered suitable for human consumption, and the health authority, at the proposal of the operator, is obliged to take the appropriate measures [14]. This parametric value is coincident with that established in RD 1798/2010 of December 30, which regulates the exploitation and marketing of natural mineral waters and spring waters bottled for human consumption [15] and the Directive (EU) 2020/2184 on the quality of water intended for human consumption [16].

Fluoride, coming from both natural sources and human activities, reaches humans mainly through water and food. The main known dietary sources of fluoride are drinking water, tea, and soft drinks [17,18].

Coffee is a drink that is part of daily life for a large part of the Spanish and world population, as it is one of the most marketed products in the world. Currently, it is, along with water and tea, among the three most consumed beverages in the world [18,19]. Therefore, a risk assessment of the presence of fluoride in this foodstuff is of great interest.

Coffee is obtained from the roasted and ground grains of the fruits of the coffee plant (Coffea spp.), which is a genus of plants belonging to the Rubiaceae family [20], C. arabica and C. canephora being the two most tare noteworthy varieties. These plants are shrubs or small trees with alternate, persistent, leathery, and shiny leaves, native to tropical and southern Africa and tropical Asia. The white and fragrant flowers are grouped in whorls in the axils of the leaves. The fruit, which requires high humidity and high-altitude terrain, is a green drupe that develops in about fifteen weeks from flowering and turns red at maturity, normally containing two plano-convex seeds joined by their flat face [21].

The most widespread coffee preparation method in the study population was the Italian coffee maker, through which lower amounts of fluoride are also obtained in the final preparation compared to the Turkish method. Likewise, there are studies that report that roasting coffee could result in the formation of less soluble fluoride compounds, which are less able to be infused [22].

Soluble or instant coffee is a dry coffee powder that is capable of dissolving directly. Two different processes are used to produce soluble coffee: spray-drying and freeze-drying. In both cases, the coffee is roasted at a lower temperature (between 190 °C and 210 °C) and is then ground and solubilized in hot water. The liquid obtained is centrifuged and then dried. In spray drying, hot air is used, while in freeze-drying, a sudden freezing of the extract at low temperatures and subsequent sublimation of the water is performed. The coffee obtained is equivalent to approximately 40% of the weight of the green coffee fruit [23].

This paper’s objectives were to determine the fluoride content of different coffee brands, assess the dietary exposure of this anion from its consumption brewed both with distilled and tap water and study the risk characterization in a given intake scenario.

2. Material and Methods

2.1. Samples

Eighty samples of caffeinated or decaffeinated coffee bought between October 2022 and March 2023 were analyzed in different presentations (coffee “from a pot”, soluble and in capsules), from different brands marketed on the island of Tenerife (Canary Islands, Spain) (

Table 1. Characteristics of the analyzed coffee samples.

Brand	No. of Samples	Type (Caffeinated/Decaffeinated)	Presentation
Bio Lidl® (San Borja, Peru)	3	Caffeinated	“from a pot”
Orix® (Ciutadella de Menorca, Spain)	3	Caffeinated	“from a pot”
Hacendado Natural Ground Coffee® (Logroño, Spain)	8	Caffeinated	“from a pot”
Kynia® (Llanera, Spain)	3	Caffeinated	“from a pot”
Marcilla® (Amsterdam, The Netherlands)	3	Caffeinated	“from a pot”
Caracol® (San Cristobal de La Laguna, Spain)	3	Caffeinated	“from a pot”
Tirma Mild Blend® (Las Palmas de Gran Canaria, Spain)	3	Caffeinated	“from a pot”
Sabanda® (San Cristobal de La Laguna, Spain)	3	Caffeinated	“from a pot”
Carioca® (Amsterdam, The Netherlands)	3	Caffeinated	“from a pot”
Hacendado Caracolillo® (San Cristobal de La Laguna, Spain)	3	Caffeinated	“from a pot”
JSP® (Las Palmas de Gran Canaria, Spain)	3	Caffeinated	“from a pot”
Hacendado Natural Espresso® (Logroño, Spain)	3	Caffeinated	“from a pot”
Saimaza® (Amsterdam, The Netherlands)	3	Caffeinated	“from a pot”
Auchan Coffee® (Madrid, Spain)	3	Caffeinated	“from a pot”
Bellarom Natural® (Barcelona, Spain)	8	Caffeinated	“from a pot”
Bellarom blend® (Barcelona, Spain)	3	Caffeinated	“from a pot”
Bellarom Decaffeinated® (Barcelona, Spain)	8	Decaffeinated	“from a pot”
Hacendado Decaffeinated® (Palencia, Spain)	5	Decaffeinated	“from a pot”
Hacendado Soluble Coffee® (Palencia, Spain)	3	Caffeinated	Soluble
Nescafé Classic Soluble® (Barcelona, Spain)	3	Caffeinated	Soluble
Auchan Intense Espresso® (Madrid, Spain)	3	Caffeinated	Capsule
N Total	80

).

2.2. Sample Preparation

A first study was carried out by brewing this drink with distilled water to determine the fluoride content present exclusively in the coffee. The following study was conducted by preparing the coffee samples with tap water from each of the three areas into which the island of Tenerife is divided: metropolitan area (Santa Cruz–La Laguna), northern area (El Ravelo), and southern area (San Isidro).

Eight grams of coffee from each sample were weighed and prepared following the manufacturer’s instructions (indicated on the packaging of each sample) using 100 mL of distilled water. In the case of ground coffee, the sample was prepared using an Italian coffee maker or pot. The soluble coffee samples were prepared using 100 mL of distilled water in a glass beaker that was heated to a temperature of 70–75 °C. In the case of coffee in capsules, it was prepared using a single-dose capsule coffee maker (NESCAFÉ^® Dolce Gusto^®, Krups KP160, Girona, Spain). This process was performed in triplicate for each sample.

2.3. Analytical Method

The determination of the fluoride content was performed using a potentiometer HACH SensION-MM340 (HACH, Düsseldorf, Germany) with a fluoride ion selective electrode (ISE) HACH ISE F-9655C (HACH, Düsseldorf, Germany) and a magnetic stirrer [24]. The instrumental parameters are: measuring range (0.001–19,000 mg/L), pH range (4–8), linear range (0.1–19,000 mg/L), slope (59 mV/pF), working temperature (5–50 °C) and possible interferences, eliminated with the conditioning solution (Fe³⁺ and Al³⁺).

Firstly, a calibration curve of fluoride concentrations from 10⁻⁵ to 10⁻² M was prepared from a 10⁻¹ M sodium fluoride stock solution for each measurement day. The stock solution was prepared by dissolving 0.4199 g of sodium fluoride (NaF) of analytic purity (Merck, Germany) and previously dried in an oven at 120 °C for 24 h in 100 mL of orthophosphoric acid buffer solution (H₃PO₄) at 0.75 M prepared from a concentrated solution of 85% purity (Honeywell-Fluka, Germany).

Twenty-five mL of preparation was then taken for each sample, and 5 mL of the indicated buffer solution was added. Subsequently, the determination was performed by potentiometry. The concentration of fluoride in the samples was obtained by extrapolation of the equation of the previously standard line obtained and is expressed as the milligrams of fluoride contained in each kilogram of coffee used for the preparation of the beverage to be consumed (mg/kg).

2.4. Statistical Analysis

The GraphPad Prism 8.0.1 (GraphPad Software, Boston, MA, USA) program was used to study possible significant differences (p < 0.05) in the fluoride content between the brands and types of coffee analyzed. When the data did not follow a normal distribution, non-parametric independent variable tests (Kruskal-Wallis, Mann-Whitney) were applied.

2.5. Dietary Fluoride Intake Calculations

Risk assessment requires calculating the estimated daily intake (EDI) of fluoride from coffee consumption. The exposure expressed in mg of fluoride ingested per day is obtained as follows:

$$\mathrm{EDI} = \mathrm{Consumption} \mathrm{of} \mathrm{coffee} \left({\displaystyle \frac{\mathrm{L}}{\mathrm{day}}}\right)\times \mathrm{Concentration} \mathrm{of} \mathrm{fluoride} \left({\displaystyle \frac{\mathrm{mg}}{\mathrm{L}}}\right)$$

2.6. Consumption Survey

A single-choice, closed-ended question format survey was carried out in order to get a better understanding of the coffee consumption of Tenerife’s population. The survey was distributed online through the university’s internal survey channels, posted in different virtual classrooms of the university, and promoted through different social networks. A total of 195 responses were obtained.

The questions asked aimed to obtain the following information: area of residence on the island, age range, gender, method of preparation, and amount drunk in cups of coffee per day. No reasons were found to exclude any survey, finding that all participants answered the survey correctly. The questions were asked in the local language (Spanish).

3. Results and Discussion

3.1. Survey Results

Since all questions offered discrete variables as answers, the results are presented as percentages. The high percentage of participation in the metropolitan area (61.5%), women (71.5%), and young people 21–30 (64.3%) can be attributed to the population density distribution and the way the survey was distributed (

Table 2. Survey results.

Question/Possible Answers	Results	Question/Possible Answers	Results
Where do you live in Tenerife?		What is your gender?
Metropolitan area	61.5%	Female	71.5%
Northern area	22.1%	Male	28%
Southern area	11.8%	I would rather do not say	0.5%
I do not live in Tenerife	4.6%
Are you a regular coffee consumer?		How old are you?
Yes	73.3%	Under 20	3.4%
No	17.4%	21–30	64.6%
Occasional consumer	9.2%	31–40	13.3%
		41–50	9%
		Older than 50	9.7%
How many cups of coffee do you drink a day?		Which water do you use to prepare the coffee?
0	5.9%	Tap water	51.8%
1–2	64.1%	Bottled water	46.2%
2–3	21%	I do not drink coffee	2%
More than 3	9%

).

Most of the participants reported a consumption of 1 to 2 cups of coffee per day (64.1%), followed by 2 to 3 cups (21%). A similar proportion of coffee is brewed with tap water (51.8%) and bottled water (46.2%), reflecting the traditionally high consumption of bottled water in the islands (Table 2).

These results are similar to those found in studies that did collect this information as part of their objectives, unlike the present work, which seeks to characterize the risk of overexposure to fluoride due to coffee consumption. Sotos-Prieto et al. recorded a consumption of 0.93 ± 0.74 cups per day in a population aged 55–80 years with diabetes and/or cardiovascular risk in Valencia [25]. Papandreou et al. also recorded an average consumption of one cup per day [26]. Stepaniak et al., in their study among women from three central European regions, recorded consumption of 0 cups in 35.7%, 1–2 in 38.4%, and 3 or more in 19.4% of participants [27]. However, studies such as the one by Miranda et al. in a Brazilian population reported a consumption of three cups of coffee per day [28]. This shows the relevance of cultural factors in the greater or lesser consumption of coffee. Therefore, the present paper considers the consumption of 1, 2, and 3 cups of coffee as possible scenarios.

3.2. Coffee Fluoride Content

Table 3. Fluoride content (mg/L), standard deviations, and maximum and minimum values of the coffees prepared with distilled water.

Brand	Mean (mg/L)	DS (mg/L)	Max (mg/L)	Min (mg/L)
Bio Lidl®	0.011	0.001	0.012	0.010
Orix®	0.007	0.001	0.008	0.006
Hacendado Natural Ground Coffee®	0.291	0.457	1.650	0.025
Kynia®	0.015	0.004	0.021	0.011
Marcilla®	0.006	0.001	0.008	0.005
Caracol®	0.008	0.001	0.009	0.006
Tirma Mild Blend®	0.142	0.143	0.331	0.031
Sabanda®	0.010	0.002	0.013	0.009
Carioca®	0.018	0.009	0.030	0.011
Hacendado Caracolillo®	0.010	0.002	0.011	0.008
JSP®	0.057	0.033	0.093	0.018
Hacendado Natural Espresso®	0.077	0.041	0.126	0.030
Saimaza®	0.253	0.008	0.264	0.246
Auchan Coffee®	0.205	0.022	0.230	0.179
Bellarom Natural®	0.032	0.087	0.280	0.003
Bellarom blend®	0.004	0.000	0.004	0.003
Bellarom Decaffeinated®	0.009	0.002	0.012	0.007
Hacendado Decaffeinated®	0.308	0.020	0.340	0.290
Hacendado Soluble Coffee®	0.065	0.025	0.098	0.046
Nescafé Classic Soluble®	0.274	0.237	0.771	0.087
Auchan Intense Espresso®	0.005	0.001	0.006	0.004

shows the descriptive data of the fluoride concentrations recorded in each of the coffee beverages brewed with distilled water in which a large variability was detected depending on the type of coffee.

Fluoride content found in the Hacendado decaffeinated^® brand stands out, with a mean concentration of 0.308 mg/L. Other brands, such as “Hacendado café molido^®” are worth mentioning with a mean average concentration of 0.291 mg/L, followed by the Nescafé Classic Soluble^®, with a mean average concentration of 0.274 mg/L.

A study by Wolska et al. (2017) recorded concentrations of 0.013–0.502 mg/L of fluoride in different types of coffee, highlighting the content in green bean Turkish coffee [29]. Jáudenes Marrero (2019) reported concentrations of 0.02 mg/L of fluoride in samples of infused roasted coffee. Both studies obtained results consistent with those of the present study [17].

Similarly, Satou et al., in their study on the fluoride content in different beverages, recorded 0.03–0.15 mg/L in coffee, differing greatly from the concentrations found in different types of tea 0.26–6.68 mg/L [30]. Galvis-Sánchez et al. also presented similar results in a methodological study, where the fluoride concentrations found were 0.09 mg/L and 0.11 mg/L in normal coffee and 0.29 mg/L in instant coffee [31]. These results follow the same trend as found in the present paper, where the highest concentrations are found in instant/soluble coffee.

The statistical analysis revealed the existence of significant differences (p < 0.05) between the brands Hacendado^® (natural ground coffee, natural espresso, caracolillo, decaffeinated, soluble coffee), Nescafé Classic^®, Tirma soft blend^®, Auchan^® and Saimaza^® and the rest of the analyzed brands (

Table 4. Obtained p-values from the statistical study.

Brand	B1	B2	B3	B4	B5	B6	B7	B8	B9	B10	B11	B12	B13	B14	B15	B16	B17	B18	B19	B20	B21
B1		0.087	0.011	0.067	0.231	0.393	0.0068	0.2799	0.0892	0.0011	0.063	0.0002	<0.0001	0.003	0.7959	0.0753	0.0892	0.0288	0.0232	0.0003	0.0011
B2	0.067		0.015	0.054	0.32	0.0753	0.0012	>0.9999	0.0693	<0.0001	0.2973	0.0028	0.0106	0.0003	0.9314	0.7021	0.0693	0.0185	0.0028	0.0106	0.0078
B3	0.003	0.012		0.001	0.013	0.001	0.02	0.015	0.022	0.3	0.042	0.153	0.0524	0.315	0.024	0.014	0.03	0.153	0.234	0.5288	0.7959
B4	0.07	0.082	<0.0001		0.326	0.056	0.0068	0.078	0.32	0.0024	0.153	<0.0001	0.0024	0.0002	0.1236	0.0892	0.7959	0.0012	0.003	<0.0001	0.0288
B5	0.067	0.7959	0.0015	0.315		0.326	<0.0001	0.2799	0.0753	0.0002	0.087	0.011	0.0003	0.0028	0.064	0.0693	0.2799	<0.0001	0.0012	0.004	0.024
B6	0.331	0.2799	0.0068	0.2176	0.393		0.0106	0.063	0.2176	0.0063	0.07	0.015	0.001	<0.0001	0.058	0.321	0.153	0.0011	0.0028	0.0012	0.0068
B7	0.0147	<0.0001	0.2799	0.0185	0.02	0.022		0.042	0.002	>0.9999	0.0028	0.063	>0.9999	0.2799	0.0392	0.0022	0.001	0.321	0.2799	0.064	0.315
B8	0.315	0.0892	0.0185	0.0955	0.0892	0.067	0.0003		0.321	0.0067	0.054	0.002	0.003	0.0067	0.2799	0.1236	0.067	0.0028	0.0024	0.0028	0.0001
B9	0.795	0.5538	0.0024	0.1236	0.2799	0.054	0.001	0.0693		<0.0001	0.0753	<0.0001	0.0003	0.013	0.0068	0.063	0.1236	0.014	0.0003	0.014	0.013
B10	0.0001	0.0078	0.1236	0.0024	0.0068	0.001	0.2799	0.001	0.0067		0.0024	>0.9999	0.153	0.087	0.0001	0.0106	0.0003	0.087	0.0523	0.0524	0.082
B11	0.0892	0.0693	0.0433	0.2799	0.0753	>0.9999	>0.9999	0.2799	0.0892	<0.0001		0.0067	0.024	0.014	0.393	0.0892	0.2973	0.004	0.001	0.0002	0.0002
B12	0.0185	0.0024	0.315	<0.0001	0.0068	0.0185	0.153	0.0002	0.0003	0.087	<0.0001		0.315	0.9314	0.0003	0.0011	0.0028	0.054	0.153	0.1236	0.0753
B13	0.0067	0.0185	0.001	0.0078	0.0012	0.0147	0.082	0.0028	0.001	0.067	0.0078	0.153		0.7021	0.001	0.003	0.003	0.315	0.063	0.2799	0.392
B14	0.0147	0.0024	0.003	0.001	0.02	0.0024	0.067	0.003	0.002	0.07	0.001	0.321	0.7021		0.002	0.001	0.0024	0.132	0.321	0.087	0.2973
B15	0.331	0.2799	0.315	0.054	0.0753	0.2973	0.003	0.315	0.2973	0.001	0.0892	0.001	0.0002	0.03		0.321	0.063	0.0025	0.0028	0.0002	0.014
B16	0.315	0.1236	0.2176	0.062	0.7959	0.153	0.0003	0.227	0.153	0.0028	0.0693	0.0032	0.0028	0.042	0.0002		0.0892	0.003	<0.0001	0.0068	0.0022
B17	0.795	0.0955	0.2799	0.067	0.9314	0.331	0.015	0.2799	0.331	<0.0001	0.321	<0.0001	<0.0001	0.0022	0.0028	0.0753		0.0012	0.03	0.02	0.042
B18	0.0001	0.0185	0.011	0.0024	0.024	<0.0001	0.0753	0.0011	0.003	0.7959	0.0024	0.2799	0.087	0.082	0.0011	<0.0001	0.015		0.067	>0.9999	0.085
B19	0.012	0.0024	0.015	0.02	<0.0001	0.0106	0.1236	0.0067	0.0003	0.9314	0.02	>0.9999	0.067	0.054	<0.0001	0.0106	0.0003	0.0524		0.7021	0.321
B20	<0.0001	0.001	0.2799	0.0068	0.0011	0.0024	0.078	0.003	0.015	0.087	0.0068	0.231	0.07	0.067	0.0002	0.0024	0.001	0.087	0.315		0.7959
B21	0.001	0.001	0.1236	0.015	0.013	0.0003	0.3	0.03	0.0067	0.067	0.015	0.32	0.2799	0.7959	0.0028	0.0003	0.0012	0.2799	0.153	0.054

). The brand codes are: Bio Lidl^® (B1), Orix^® (B2), Hacendado natural ground coffee^® (B3), Kynia^® (B4), Marcilla^® (B5), Caracol^® (B6), Tirma soft blend^® (B7), Sabanda^® (B8), Carioca^® (B9), Hacendado caracolillo^® (B10), JSP (B11), Hacendado natural espresso (B12), Saimaza (B13), Auchan coffee (B14), Bellarom Natural^® (B15), Bellarom Blend^® (B16), Bellarom Decaffeinated^® (B17), Hacendado Decaffeinated^® (B18), Hacendado Soluble Coffee^® (B19), Nescafé Classic^® (B20), Intense espresso Auchan^® (B21).

This could be related to the soil, water or fertilizers used in the agricultural process, furthermore, the water used to wash the grains may have contained fluorides. However, the origin of the grain is unknown, so it is not possible to identify the possible factors responsible for these differences.

3.3. Fluoride in Coffee Samples Prepared with Tap Water

As indicated by surveys, more than half of the population may prepare coffee with tap water. Considering the historical background of the Canary Islands, especially those of the island of Tenerife, which has a high concentration of fluoride in this kind of water [3,32,33], it would be interesting to know the fluoride concentrations under these conditions in order to estimate the fluoride exposure.

The use of tap water for the preparation of coffee markedly increases the concentration of fluoride in these beverages regardless of the area of origin. The highest mean concentration was recorded in coffee made with water from the southern part of the island with 2.73 mg/L, followed by the northern part with 1.35 mg/L. Finally, in the metropolitan area, where part of the supply comes from desalination water, an average of 0.39 mg/L was recorded (

Table 5. Mean fluoride content is in the tap water of the different areas of Tenerife and in the coffee brewed with this coffee.

Fluoride by Area	Mean (mg/L)	SD (mg/L)	Min (mg/L)	Max (mg/L)
Metropolitan	0.25	0.691	0.02	3
South	1.04	0.599	0.1	7
North	1.12	2.677	0.12	15.1
Coffee Prepared with Tap Water	Mean (mg/L)	SD (mg/L)	Min (mg/L)	Max (mg/L)
Metropolitan	0.39	0.48	0.01	4.73
South	2.73	2.83	0.19	8.47
North	1.35	3.37	0.21	16.68

).

These results show that although coffee has been identified in some studies as a possible dietary source of fluoride [17,29], unlike tea [30,34,35], the main source of fluoride in these beverages is the water from which they are made. The variations in concentrations by area are based on the different water supplies on the island of Tenerife; in some cases, the water comes directly from underground galleries (providing high concentrations of fluoride, especially in times of drought) and in other cases. The water comes from the mixing of desalinated water with groundwater. This is why, as can be seen in the table above, there are large differences even between neighborhoods in the same municipality.

These results are similar to those found by Viswanathan et al. in India, where they recorded concentrations from 1.45 mg/L in coffee brewed in areas without endemic fluorosis to 2.70 mg/L and 3.85 mg/L in regions with higher fluorosis [35].

3.4. Exposure Assessment and Risk Characterization

The exposure assessment was calculated by setting the coffee serving volume at 60 mL, based on which the fluoride estimated daily intake (EDI) from these beverages was established for three consumption scenarios following the survey results: 1, 2, and 3 coffees per day. This was carried out for both the data obtained from coffee brewed with distilled water and the ones from coffee brewed with tap water (

Table 6. Fluoride, estimated daily intakes from coffee, prepared with distilled water according to brands.

Brand	EDI 1 Cup (mg/Day)	EDI 2 Cup (mg/Day)	EDI 3 Cup (mg/Day)
Bio Lidl®	0.001	0.001	0.002
Orix®	0.000	0.001	0.001
Hacendado Natural Ground Coffee®	0.017	0.035	0.052
Kynia®	0.001	0.002	0.003
Marcilla®	0.000	0.001	0.001
Caracol®	0.000	0.001	0.001
Tirma Mild Blend®	0.009	0.017	0.026
Sabanda®	0.001	0.001	0.002
Carioca®	0.001	0.002	0.003
Hacendado Caracolillo®	0.001	0.001	0.002
JSP®	0.003	0.007	0.010
Hacendado Natural Espresso®	0.005	0.009	0.014
Saimaza®	0.015	0.030	0.046
Auchan Coffee®	0.012	0.025	0.037
Bellarom Natural®	0.002	0.004	0.006
Bellarom blend®	0.000	0.000	0.001
Bellarom Decaffeinated®	0.001	0.001	0.002
Hacendado Decaffeinated®	0.018	0.037	0.055
Hacendado Soluble Coffee®	0.004	0.008	0.012
Nescafé Classic Soluble®	0.016	0.033	0.049
Auchan Intense Espresso®	0.000	0.001	0.001

and

Table 7. Fluoride, estimated daily intakes from coffee, prepared with tap water according to Tenerife areas.

Areas	EDI (1 Cup) mg/Day	EDI (2 Cup) mg/Day	EDI (3 Cup) mg/Day
Metropolitan	0.02	0.05	0.07
South	0.16	0.33	0.49
North	0.08	0.16	0.24

).

The tolerable upper intake level established by EFSA for the adult population (7 mg/day) was used to perform the risk characterization for the different scenarios [36] (

Table 8. Fluoride contribution to the EFSA UL from coffee prepared with distilled water according to brands.

Brand	UL 1 Cup (%)	UL 2 Cup (%)	UL 3 Cup (%)
Bio Lidl®	0.01	0.02	0.03
Orix®	0.01	0.01	0.02
Hacendado Natural Ground Coffee®	0.25	0.50	0.75
Kynia®	0.01	0.03	0.04
Marcilla®	0.01	0.01	0.02
Caracol®	0.01	0.01	0.02
Tirma Mild Blend®	0.12	0.24	0.37
Sabanda®	0.01	0.02	0.03
Carioca®	0.02	0.03	0.05
Hacendado Caracolillo®	0.01	0.02	0.03
JSP®	0.05	0.10	0.15
Hacendado Natural Espresso®	0.07	0.13	0.20
Saimaza®	0.22	0.43	0.65
Auchan Coffee®	0.18	0.35	0.53
Bellarom Natural®	0.03	0.05	0.08
Bellarom blend®	0.00	0.01	0.01
Bellarom Decaffeinated®	0.01	0.02	0.02
Hacendado Decaffeinated®	0.26	0.53	0.79
Hacendado Soluble Coffee®	0.06	0.11	0.17
Nescafé Classic Soluble®	0.23	0.47	0.70
Auchan Intense Espresso®	0.00	0.01	0.01

and

Table 9. Fluoride contribution to the EFSA UL from coffee prepared with tap water, according to Tenerife areas.

Areas	UL (1 Cup) %	UL (2 Cup) %	UL (3 Cup) %
Metropolitan	0.33	0.66	0.99
South	2.34	4.68	7.02
North	1.16	2.32	3.48

).

The highest percentage contribution to the reference value by brand for the intake of a daily cup of coffee was found for Hacendado decaffeinated coffee (0.25%), followed by Hacendado natural ground (0.25%). As for coffees prepared with tap water for the same scenario, the highest percentage was determined for coffee prepared with water from the southern part of the island (2.34%).

It is necessary to consider that these percentages refer to the average fluoride content recorded in each area of Tenerife since, if these percentages were calculated with the maximum concentrations recorded in each area, there would be a risk (specific areas of the island of Tenerife). However, this has been calculated with the averages to generalize the results.

These percentages of contribution to the UL are small, so it could be hypothesized that there would not be a risk of overexposure if the overall dietary contribution is considered.

3.5. Limitations of the Study

Limitations to this study could be the following: not all brands of coffee marketed in Europe are sold in the Canary Islands; the way coffee is prepared may vary in the population due to the high presence of foreigners in the Canary Islands; the study carried out using public water supply, being an area with endemic fluorosis, has limitations due to the differences in fluoride compared to other European regions.

4. Conclusions

Low concentrations of fluoride have been found in the analyzed coffees prepared with distilled water. Therefore, based on the established consumption, coffee is not considered to pose a risk of exposure to fluoride.

Furthermore, the samples do not exceed the parametric value set for tap water (1.5 mg/L). Significant differences (p < 0.05) were found between the coffee brands soluble Nescafé classic^®, Saimaza^®, Auchan coffee^®, Tirma soft blend^® and Hacendado^® (natural ground and soluble coffee) and the rest of the analyzed brands.

The coffee prepared with tap water from Tenerife (metropolitan, northern, and southern areas) greatly increases the concentrations of fluoride in this stimulating beverage. However, its intake does not pose a toxicological risk of exposure to fluoride, considering the mean concentration recorded. However, there are areas in Tenerife where the concentrations recorded in coffee prepared with tap water pose a severe health risk.

Regular coffee consumers should consider the water with which they brew their coffee with, as it is the main source of fluoride in these beverages. Based on the results obtained, it is recommended to use bottled water to prepare coffee, especially if it is a region with high fluoride concentrations in its water. Also, legal requirements for better traceability should be introduced to include the origin of the coffee bean on the label.