YRTPH 3268
No. of Pages 10, Model 5G
8 April 2015 Regulatory Toxicology and Pharmacology xxx (2015) xxx–xxx 1
Contents lists available at ScienceDirect
Regulatory Toxicology and Pharmacology journal homepage: www.elsevier.com/locate/yrtph 4 5 3
Biomonitoring Equivalents for interpretation of urinary fluoride
6
L.L. Aylward a, S.M. Hays b,⇑, A. Vezina c, M. Deveau c, A. St Amand c, A. Nong c
7 8 9 10 11 1 2 3 2 14 15 16 17 18 19 20 21
a
Summit Toxicology, LLP, Falls Church, VA, USA Summit Toxicology, LLP, Lyons, CO, USA c Health Canada, Ottawa, ON, Canada b
a r t i c l e
i n f o
Article history: Received 30 January 2015 Available online xxxx Keywords: Biomonitoring Biomonitoring Equivalents Fluoride
a b s t r a c t Exposure to fluoride is widespread due to its natural occurrence in the environment and addition to drinking water and dental products for the prevention of dental caries. The potential health risks of excess fluoride exposure include aesthetically unacceptable dental fluorosis (tooth mottling) and increased skeletal fragility. Numerous organizations have conducted risk assessments and set guidance values to represent maximum recommended exposure levels as well as recommended adequate intake levels based on potential public health benefits of fluoride exposure. Biomonitoring Equivalents (BEs) are estimates of the average biomarker concentrations corresponding to such exposure guidance values. The literature on daily urinary fluoride excretion rates as a function of daily fluoride exposure was reviewed and BE values corresponding to the available US and Canadian exposure guidance values were derived for fluoride in urine. The derived BE values range from 1.1 to 2.1 mg/L (1.2–2.5 lg/g creatinine). Concentrations of fluoride in single urinary spot samples from individuals, even under exposure conditions consistent with the exposure guidance values, may vary from the predicted average concentrations by several-fold due to within- and across-individual variation in urinary flow and creatinine excretion rates and due to the rapid elimination kinetics of fluoride. Thus, the BE values are most appropriately applied to screen population central tendency estimates for biomarker concentrations rather than interpretation of individual spot sample concentrations. Ó 2015 Published by Elsevier Inc.
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
42 43
1. Introduction
44
Fluoride is naturally occurring in food, water, soil and air, and can also be added to drinking water, dental products such as toothpaste and mouthwash, and in supplements for the prevention of dental caries. Thus, exposure is widespread. The public health benefits associated with the use of fluorides in dental products and in drinking water have been established, but undesirable outcomes associated with elevated exposures to fluoride are also recognized (Health Canada, 2010). The undesirable outcomes include cosmetic effects (mottling of teeth) associated with excess fluoride, and, at higher levels, adverse effects on skeletal integrity. Tolerable exposure levels are established to protect against potential adverse effects while optimal levels focus on optimizing dental health benefits while remaining well below tolerable exposure levels. Total intakes from drinking water, food, dentifrice, air and soil are considered to estimate the proportion of fluoride allocated to each source of exposure. Drinking water concentration standards and guidelines are targeted so that total fluoride intake from
45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
⇑ Corresponding author.
all sources does not exceed exposure guidance values such as tolerable daily intakes (TDIs) or reference doses (RfDs) (Health Canada, 2010; USEPA, 2010). Fluoride in urine has also been widely used as a biomarker of exposure (Rugg-Gunn et al., 2011; Villa et al., 2010). Measured concentrations of fluoride in urine cannot be directly interpreted in terms of the available exposure guidance values. The purpose of this evaluation is to derive Biomonitoring Equivalent (BE) values for interpretation of population urinary fluoride concentrations. BE values are estimates of the concentration of a chemical or its metabolite in blood or urine that are consistent with risk assessment-derived exposure guidance values such as RfDs or TDIs (Hays et al., 2007, 2008; Angerer et al., 2011). BE values can be used as screening values for the assessment of biomonitoring data in order to provide an initial evaluation of whether the detected concentrations are below, near, or above the concentrations consistent with current exposure guidance values (for both toxicity and nutritional requirements). This evaluation is directed at fluoride anion in urine, which arises from exposure to fluoride from all sources including drinking water, dentifrices, supplements, food, air and soil.
E-mail address: [email protected] (S.M. Hays). http://dx.doi.org/10.1016/j.yrtph.2015.04.005 0273-2300/Ó 2015 Published by Elsevier Inc.
Please cite this article in press as: Aylward, L.L., et al. Biomonitoring Equivalents for interpretation of urinary fluoride. Regul. Toxicol. Pharmacol. (2015), http://dx.doi.org/10.1016/j.yrtph.2015.04.005
61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
YRTPH 3268
No. of Pages 10, Model 5G
8 April 2015 2 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104
L.L. Aylward et al. / Regulatory Toxicology and Pharmacology xxx (2015) xxx–xxx
Concerns regarding potential adverse effects of exposure to fluoride focus on two main outcomes: moderate dental fluorosis with widespread mottling of tooth enamel, which is considered to be an adverse effect due to cosmetic impacts when it reaches a moderate level (Health Canada, 2010; EPA, 2010), and skeletal fluorosis at elevated fluoride exposure levels over a long period of time (Health Canada, 2010). These effects have been observed in several studies of human populations with well-characterized levels of fluoride intake, allowing for robust derivation of no-observed-adverse-effect-levels and lowest-observed-adverse-effect levels (NOAELs and LOAELs). Fluoride has also been considered to be beneficial to public health because of its ability to protect against the development of caries (IOM, 1997). We derive here BE values that can be used as benchmarks to evaluate measured urinary fluoride concentrations. The fluoride BE values are derived using average age-specific values for urinary parameters such as urine flow rate and creatinine excretion rates, as well as an assumption of steady-state exposure. We also include an evaluation of the variation in urinary fluoride concentrations expected to be observed, due to within- and across-individual variation in urinary spot sample flow and creatinine excretion rates, as well as to temporal variations associated with the rapid excretion of fluoride in urine.
105
2. Materials and methods
106
The approach used for the derivation of the urinary BE value for fluoride follows the same general approach used for many other urinary analytes, the urinary mass balance approach. Specifically, the BE value associated with the following exposure guidance values: Health Canada tolerable daily intake (TDI), the ATSDR minimal risk level (MRL), the US EPA reference dose (RfD), and the tolerable upper intake levels (ULs) and adequate intakes (AIs) by the US Institute of Medicine (IOM) can be calculated using data from studies of human populations that relate urinary excretion of fluoride under steady-state exposure conditions to daily intake of fluoride. The estimated daily urinary fluoride excretion rate (DUFE, mg/kgd) corresponding to an exposure guidance value such as the TDI, is then divided by the estimated body weight-adjusted daily urinary volume (V24, ml/kg-d) or creatinine excretion (CE24, mg/kg-
107 108 109 110 111 112 113 114 115 116 117 118 119
d) to calculate the corresponding BE value (mg/L or mg/g creatinine):
DUFETDI BE ¼ V 24 or CE24
ð1Þ
Study description
Exposure guidance values based on adverse health effects Health Canada, chronic oral Studies of the occurrence of dental TDI (2010) fluorosis EPA, chronic oral RfD (2010) ATSDR, chronic oral MRL (2003) IOM, UL, infant to age 8 (1997)
IOM, UL, ages P 8 (1997)
Studies of tooth mottling in children at varying fluoride concentrations in water Study of risk of bone fractures in elderly persons in China exposed to fluoride in drinking water Multiple studies on occurrence of moderate dental fluorosis Multiple studies of the occurrence of skeletal fluorosis
Exposure guidance values based on health benefits IOM, AI, age 6 6 months Studies indicating no increased risk of (1997) dental caries in infants fed human milk 6 6 months IOM, AI, age > 6 months (1997) Studies of intake in communities with water concentrations previously identified to be optimally fluoridated for protective effect against dental caries
121
122 124
These calculations are conducted for a variety of age ranges using age-specific parameterizations for DUFE, V24, and CE24. To the extent that values differ among age groups, age-specific BE values are reported. Data and parameters required for this approach include healthbased exposure guidance values such as a tolerable daily intake, data allowing for estimation of urinary excretion of fluoride as a function of intake under steady-state conditions, and age-specific and bodyweight-adjusted values for daily urinary flow or creatinine excretion. We reviewed the available literature and government documents to identify exposure guidance values for fluoride from the United States and Canada. We also obtained and reviewed literature on the pharmacokinetics of fluoride in humans, with a focus on studies that examine the urinary mass balance of fluoride in the range of exposure levels likely to be encountered in the general population in the US and Canada. We used recent data from the US National Health and Nutrition Examination Survey (NHANES), as well as data obtained from the literature, to characterize body weight-adjusted urinary flow rates and daily creatinine excretion rates as a function of age. Data on urinary flow rates and creatinine excretion rates from the NHANES 2009–2010 survey cycle was downloaded and descriptive statistics were generated using STATA IC 9 (Stata Corp, College Station, TX).
125
3. Results
150
3.1. Exposure guidance values
151
The health effects of excessive exposure to fluoride have been studied extensively, and exposure guidance values have been established by many organizations, including Health Canada (Health Canada, 2010), the US Environmental Protection Agency (USEPA, 2010), the US Agency for Toxic Substances and Disease Registry (ATSDR, 2003), and the US IOM (1997) (Table 1). The risk
152
Table 1 Available exposure guidance values for fluoride. Organization, criteria (year of evaluation)
120
Critical endpoint and dose
Uncertainty factors
Value
NOAEL for moderate dental fluorosis, accounting for multi-route exposures, 0.105 mg/kg-d NOAEL for objectionable dental fluorosis, a cosmetic effect, 0.08 mg/kg-d NOAEL for increased risk of bone fractures, 0.15 mg/kg-d
1
0.105 mg/kg-d
1
0.08 mg/kg-d
3
0.05 mg/kg-d
LOAEL for moderate, cosmetically objectionable dental fluorosis (see Table 2), 0.1 mg/kg-d NOAEL for development of early signs of skeletal fluorosis, 10 mg/d
1
0.1 mg/kg-d (ages infant to 8)
1
10 mg/d (ages P 8)
Average dietary fluoride intake in human milk-fed infants, 0.01 mg/d
1
0.01 mg/d (0.001 to 0.003 mg/kg-d)
Average dietary fluoride intake in infants and children from fluoridated communities, to provide protection against dental caries without unwanted health effects 0.05 mg/kg-d
1
0.05 mg/kg-d
Please cite this article in press as: Aylward, L.L., et al. Biomonitoring Equivalents for interpretation of urinary fluoride. Regul. Toxicol. Pharmacol. (2015), http://dx.doi.org/10.1016/j.yrtph.2015.04.005
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
153 154 155 156 157
YRTPH 3268
No. of Pages 10, Model 5G
8 April 2015 L.L. Aylward et al. / Regulatory Toxicology and Pharmacology xxx (2015) xxx–xxx Table 2 Characterization of degrees of dental fluorosis (adapted from Health Canada, (2010), based on categories set by the American Dental Association). Classification
Criteria: description of enamel
Normal Questionable Very mild
Smooth, glossy, pale creamy-white translucent surface A few white flecks or white spots Small opaque, paper-white areas covering less than 25% of the tooth surface Opaque white areas covering less than 50% of the tooth surface All tooth surfaces affected; marked wear on biting surfaces; brown stain may be present All tooth surfaces affected; discrete or confluent pitting; brown stain present
Mild Moderate Severe
196
assessments by Health Canada (2010) and USEPA (2010) establish exposure guidance values based on moderate dental fluorosis, which is considered to be an unacceptable cosmetic outcome, (see Table 2). This effect occurs only under conditions of elevated exposure for extended time periods in childhood, and the sensitivity for this outcome essentially ends after age 8 when adult teeth are fully formed (Health Canada, 2010). The ATSDR (2003) evaluation is based on avoidance of the increased risk of skeletal fractures associated with elevated fluoride exposure. The IOM (1997) identified tolerable upper intake levels (ULs) designed to ensure that neither unacceptable levels of cosmetic dental fluorosis (for childhood exposures up to age 8) nor skeletal fluorosis (for ages > 8 years) would occur. Both endpoints of concern with respect to elevated fluoride exposure, namely moderate dental fluorosis and skeletal fluorosis, occur only after long-term elevated exposures and are not consequences of acute exposure events. In addition to an exposure guideline for adverse effects from excessive exposure, the IOM (1997) established adequate intakes (AIs) for sufficient fluoride exposure. The AIs were established as levels of fluoride that would provide maximal protection of a population against dental caries without unwanted health effects. AIs were established for infants 66 months of age and for individuals >6 months of age. The typical approach for derivation of BE values is to identify the point of departure (POD) for the risk assessment (in the case there studies in humans form the basis for the exposure guidance value), estimate urinary concentrations corresponding to the POD (BEPOD), and then apply the composite uncertainty factors from the risk assessment to derive the BE value corresponding to the exposure guidance value (Fig. 1). The risk assessments from Health Canada, the USEPA, and the IOM identified human exposures equal to a NOAEL and applied no additional uncertainty factors, so the POD and final exposure guidance value were the same within each of these risk assessments (see Table 1). The use of no uncertainty factors for deriving the fluoride guidance values was deemed appropriate by the regulatory agencies because the NOAEL was obtained from studies that included sensitive human populations. The ATSDR MRL applied a composite uncertainty factor of 3 to the identified NOAEL.
197
3.2. Mass balance for urinary fluoride excretion
198
The pharmacokinetics of fluoride have been studied extensively in human populations. Fluoride is well absorbed from drinking water and dietary sources, with absorption rates of 80–90% (Whitford, 1994). Absorbed fluoride is partially retained in the skeletal system, with a fraction excreted in urine (Villa et al., 2010). Plasma and urinary half-lives of 5.1 and 5.8 h, respectively, were reported in a controlled administration study by Ekstrand and Ehrnebo (1983). Whitford et al. (2008) tested different fluoride compounds used for water fluoridation (natural sources, sodium
158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195
199 200 201 202 203 204 205 206
3
fluoride and fluorosilicic acid) and in dentifrices and found that the pharmacokinetics were similar for different fluoride compounds. Numerous studies have assessed the quantitative relationships between fluoride intake and urinary excretion, both under conditions of relatively steady intake rates (reviewed in Villa et al., 2010) and under conditions of changing intake rates (see, for example, Ekstrand et al., 1984). For the purposes of derivation of BE values for fluoride associated with chronic exposure guidance values, we focus on the evaluations of fluoride kinetics under conditions of steady intake rates. Villa et al. (2010) collected individual data from multiple studies of children and adults. Data on measured fluoride intake and urinary excretion rates (mg/d) for 212 children ages 0.15–7 years from nine studies in six countries and for 269 adults aged 18–75 years from eight studies in two countries were collected. Based on these data, Villa et al. (2010) constructed linear functions for prediction of daily urinary fluoride excretion (DUFE, mg/d) as a function of total daily fluoride intake (TDFI, mg/d) for children and adults of the form:
DUFE ¼ I þ b TDFI
ð2Þ
207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
226 228
where I is an age-specific intercept representing daily fluoride excretion from body stores in the absence of any fluoride intake, and b is the slope of the linear relationship between fluoride excretion and intake. The fitted age-specific regression parameters are presented in Table 3. The slope relating daily fluoride excretion to intake was lower in children than in adults (slopes of 0.35 and 0.54, respectively). However, the intercept, representing excretion from body stores in the absence of any exposure, was higher in adults than in children; this is consistent with greater stored fluoride in adults than in children. Villa et al. (2010) reported that they did not find data sufficient to derive relationships for adolescents. For the purposes of this evaluation, we assumed that the urinary excretion fraction for adolescents is an average of that predicted for children and adults using the regression parameters in Table 3.
229
3.3. Urinary flow rates
244
3.3.1. Infants and children We identified 13 articles that provided data on 24-h urinary flow rate in infants and children along with data on bodyweight (Table 4), allowing estimation of urinary flow rate on a ml/kg-d basis. In addition, the NHANES 2009–2010 dataset (CDC, 2012) collected data allowing estimation of urinary flow rate for the spot samples collected in that survey. In these datasets, the time since last urinary void, the void volume, and the bodyweight were recorded, allowing calculation of urinary flow on a ml/kg-hr basis. We extrapolated the urine flow rate from the spot samples to 24 h. Table 4 presents a summary of these data, and Fig. 2A presents a scatter plot of the data. For children under age 15, no significant differences between males and females were observed in the studies that reported sex-specific data; therefore, data for both sexes are considered together here. Reported urinary flow rates for newborn and very young infants varied widely. Data for somewhat older infants and children were more consistent from study-to-study. Based on the extreme heterogeneity in the reported flow rates for children at the younger age range, we did not attempt to fit a function to these data. However, for the more stable data beginning at age 3, we fit the logarithm of the flow rate with a linear function of age, extended to age 14, with each point estimate weighted by the size of the dataset. The fit was highly significant, with an R2 value of 0.48, p < 0.001. This equation can be used to predict an average age-specific urinary flow rate (FR, ml/kg-d) for ages 3–14:
245
Please cite this article in press as: Aylward, L.L., et al. Biomonitoring Equivalents for interpretation of urinary fluoride. Regul. Toxicol. Pharmacol. (2015), http://dx.doi.org/10.1016/j.yrtph.2015.04.005
230 231 232 233 234 235 236 237 238 239 240 241 242 243
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
271
YRTPH 3268
No. of Pages 10, Model 5G
8 April 2015 4
L.L. Aylward et al. / Regulatory Toxicology and Pharmacology xxx (2015) xxx–xxx
Fig. 1. Schematic for derivation of Biomonitoring Equivalents for fluoride in urine.
Table 3 Regression parameters for the estimation of daily urinary excretion of fluoride (DUFE, mg/d) as a function of total daily fluoride intake (TDFI, mg/d) from Villa et al. (2010).a
a
Age group
Slope
p value for slope
Intercept
p value for intercept
R2
Children (0.15 to 7 years) Adults (18+ years)
0.35 ± 0.01
No. of Pages 10, Model 5G
8 April 2015 Regulatory Toxicology and Pharmacology xxx (2015) xxx–xxx 1
Contents lists available at ScienceDirect
Regulatory Toxicology and Pharmacology journal homepage: www.elsevier.com/locate/yrtph 4 5 3
Biomonitoring Equivalents for interpretation of urinary fluoride
6
L.L. Aylward a, S.M. Hays b,⇑, A. Vezina c, M. Deveau c, A. St Amand c, A. Nong c
7 8 9 10 11 1 2 3 2 14 15 16 17 18 19 20 21
a
Summit Toxicology, LLP, Falls Church, VA, USA Summit Toxicology, LLP, Lyons, CO, USA c Health Canada, Ottawa, ON, Canada b
a r t i c l e
i n f o
Article history: Received 30 January 2015 Available online xxxx Keywords: Biomonitoring Biomonitoring Equivalents Fluoride
a b s t r a c t Exposure to fluoride is widespread due to its natural occurrence in the environment and addition to drinking water and dental products for the prevention of dental caries. The potential health risks of excess fluoride exposure include aesthetically unacceptable dental fluorosis (tooth mottling) and increased skeletal fragility. Numerous organizations have conducted risk assessments and set guidance values to represent maximum recommended exposure levels as well as recommended adequate intake levels based on potential public health benefits of fluoride exposure. Biomonitoring Equivalents (BEs) are estimates of the average biomarker concentrations corresponding to such exposure guidance values. The literature on daily urinary fluoride excretion rates as a function of daily fluoride exposure was reviewed and BE values corresponding to the available US and Canadian exposure guidance values were derived for fluoride in urine. The derived BE values range from 1.1 to 2.1 mg/L (1.2–2.5 lg/g creatinine). Concentrations of fluoride in single urinary spot samples from individuals, even under exposure conditions consistent with the exposure guidance values, may vary from the predicted average concentrations by several-fold due to within- and across-individual variation in urinary flow and creatinine excretion rates and due to the rapid elimination kinetics of fluoride. Thus, the BE values are most appropriately applied to screen population central tendency estimates for biomarker concentrations rather than interpretation of individual spot sample concentrations. Ó 2015 Published by Elsevier Inc.
23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
42 43
1. Introduction
44
Fluoride is naturally occurring in food, water, soil and air, and can also be added to drinking water, dental products such as toothpaste and mouthwash, and in supplements for the prevention of dental caries. Thus, exposure is widespread. The public health benefits associated with the use of fluorides in dental products and in drinking water have been established, but undesirable outcomes associated with elevated exposures to fluoride are also recognized (Health Canada, 2010). The undesirable outcomes include cosmetic effects (mottling of teeth) associated with excess fluoride, and, at higher levels, adverse effects on skeletal integrity. Tolerable exposure levels are established to protect against potential adverse effects while optimal levels focus on optimizing dental health benefits while remaining well below tolerable exposure levels. Total intakes from drinking water, food, dentifrice, air and soil are considered to estimate the proportion of fluoride allocated to each source of exposure. Drinking water concentration standards and guidelines are targeted so that total fluoride intake from
45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
⇑ Corresponding author.
all sources does not exceed exposure guidance values such as tolerable daily intakes (TDIs) or reference doses (RfDs) (Health Canada, 2010; USEPA, 2010). Fluoride in urine has also been widely used as a biomarker of exposure (Rugg-Gunn et al., 2011; Villa et al., 2010). Measured concentrations of fluoride in urine cannot be directly interpreted in terms of the available exposure guidance values. The purpose of this evaluation is to derive Biomonitoring Equivalent (BE) values for interpretation of population urinary fluoride concentrations. BE values are estimates of the concentration of a chemical or its metabolite in blood or urine that are consistent with risk assessment-derived exposure guidance values such as RfDs or TDIs (Hays et al., 2007, 2008; Angerer et al., 2011). BE values can be used as screening values for the assessment of biomonitoring data in order to provide an initial evaluation of whether the detected concentrations are below, near, or above the concentrations consistent with current exposure guidance values (for both toxicity and nutritional requirements). This evaluation is directed at fluoride anion in urine, which arises from exposure to fluoride from all sources including drinking water, dentifrices, supplements, food, air and soil.
E-mail address: [email protected] (S.M. Hays). http://dx.doi.org/10.1016/j.yrtph.2015.04.005 0273-2300/Ó 2015 Published by Elsevier Inc.
Please cite this article in press as: Aylward, L.L., et al. Biomonitoring Equivalents for interpretation of urinary fluoride. Regul. Toxicol. Pharmacol. (2015), http://dx.doi.org/10.1016/j.yrtph.2015.04.005
61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
YRTPH 3268
No. of Pages 10, Model 5G
8 April 2015 2 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104
L.L. Aylward et al. / Regulatory Toxicology and Pharmacology xxx (2015) xxx–xxx
Concerns regarding potential adverse effects of exposure to fluoride focus on two main outcomes: moderate dental fluorosis with widespread mottling of tooth enamel, which is considered to be an adverse effect due to cosmetic impacts when it reaches a moderate level (Health Canada, 2010; EPA, 2010), and skeletal fluorosis at elevated fluoride exposure levels over a long period of time (Health Canada, 2010). These effects have been observed in several studies of human populations with well-characterized levels of fluoride intake, allowing for robust derivation of no-observed-adverse-effect-levels and lowest-observed-adverse-effect levels (NOAELs and LOAELs). Fluoride has also been considered to be beneficial to public health because of its ability to protect against the development of caries (IOM, 1997). We derive here BE values that can be used as benchmarks to evaluate measured urinary fluoride concentrations. The fluoride BE values are derived using average age-specific values for urinary parameters such as urine flow rate and creatinine excretion rates, as well as an assumption of steady-state exposure. We also include an evaluation of the variation in urinary fluoride concentrations expected to be observed, due to within- and across-individual variation in urinary spot sample flow and creatinine excretion rates, as well as to temporal variations associated with the rapid excretion of fluoride in urine.
105
2. Materials and methods
106
The approach used for the derivation of the urinary BE value for fluoride follows the same general approach used for many other urinary analytes, the urinary mass balance approach. Specifically, the BE value associated with the following exposure guidance values: Health Canada tolerable daily intake (TDI), the ATSDR minimal risk level (MRL), the US EPA reference dose (RfD), and the tolerable upper intake levels (ULs) and adequate intakes (AIs) by the US Institute of Medicine (IOM) can be calculated using data from studies of human populations that relate urinary excretion of fluoride under steady-state exposure conditions to daily intake of fluoride. The estimated daily urinary fluoride excretion rate (DUFE, mg/kgd) corresponding to an exposure guidance value such as the TDI, is then divided by the estimated body weight-adjusted daily urinary volume (V24, ml/kg-d) or creatinine excretion (CE24, mg/kg-
107 108 109 110 111 112 113 114 115 116 117 118 119
d) to calculate the corresponding BE value (mg/L or mg/g creatinine):
DUFETDI BE ¼ V 24 or CE24
ð1Þ
Study description
Exposure guidance values based on adverse health effects Health Canada, chronic oral Studies of the occurrence of dental TDI (2010) fluorosis EPA, chronic oral RfD (2010) ATSDR, chronic oral MRL (2003) IOM, UL, infant to age 8 (1997)
IOM, UL, ages P 8 (1997)
Studies of tooth mottling in children at varying fluoride concentrations in water Study of risk of bone fractures in elderly persons in China exposed to fluoride in drinking water Multiple studies on occurrence of moderate dental fluorosis Multiple studies of the occurrence of skeletal fluorosis
Exposure guidance values based on health benefits IOM, AI, age 6 6 months Studies indicating no increased risk of (1997) dental caries in infants fed human milk 6 6 months IOM, AI, age > 6 months (1997) Studies of intake in communities with water concentrations previously identified to be optimally fluoridated for protective effect against dental caries
121
122 124
These calculations are conducted for a variety of age ranges using age-specific parameterizations for DUFE, V24, and CE24. To the extent that values differ among age groups, age-specific BE values are reported. Data and parameters required for this approach include healthbased exposure guidance values such as a tolerable daily intake, data allowing for estimation of urinary excretion of fluoride as a function of intake under steady-state conditions, and age-specific and bodyweight-adjusted values for daily urinary flow or creatinine excretion. We reviewed the available literature and government documents to identify exposure guidance values for fluoride from the United States and Canada. We also obtained and reviewed literature on the pharmacokinetics of fluoride in humans, with a focus on studies that examine the urinary mass balance of fluoride in the range of exposure levels likely to be encountered in the general population in the US and Canada. We used recent data from the US National Health and Nutrition Examination Survey (NHANES), as well as data obtained from the literature, to characterize body weight-adjusted urinary flow rates and daily creatinine excretion rates as a function of age. Data on urinary flow rates and creatinine excretion rates from the NHANES 2009–2010 survey cycle was downloaded and descriptive statistics were generated using STATA IC 9 (Stata Corp, College Station, TX).
125
3. Results
150
3.1. Exposure guidance values
151
The health effects of excessive exposure to fluoride have been studied extensively, and exposure guidance values have been established by many organizations, including Health Canada (Health Canada, 2010), the US Environmental Protection Agency (USEPA, 2010), the US Agency for Toxic Substances and Disease Registry (ATSDR, 2003), and the US IOM (1997) (Table 1). The risk
152
Table 1 Available exposure guidance values for fluoride. Organization, criteria (year of evaluation)
120
Critical endpoint and dose
Uncertainty factors
Value
NOAEL for moderate dental fluorosis, accounting for multi-route exposures, 0.105 mg/kg-d NOAEL for objectionable dental fluorosis, a cosmetic effect, 0.08 mg/kg-d NOAEL for increased risk of bone fractures, 0.15 mg/kg-d
1
0.105 mg/kg-d
1
0.08 mg/kg-d
3
0.05 mg/kg-d
LOAEL for moderate, cosmetically objectionable dental fluorosis (see Table 2), 0.1 mg/kg-d NOAEL for development of early signs of skeletal fluorosis, 10 mg/d
1
0.1 mg/kg-d (ages infant to 8)
1
10 mg/d (ages P 8)
Average dietary fluoride intake in human milk-fed infants, 0.01 mg/d
1
0.01 mg/d (0.001 to 0.003 mg/kg-d)
Average dietary fluoride intake in infants and children from fluoridated communities, to provide protection against dental caries without unwanted health effects 0.05 mg/kg-d
1
0.05 mg/kg-d
Please cite this article in press as: Aylward, L.L., et al. Biomonitoring Equivalents for interpretation of urinary fluoride. Regul. Toxicol. Pharmacol. (2015), http://dx.doi.org/10.1016/j.yrtph.2015.04.005
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149
153 154 155 156 157
YRTPH 3268
No. of Pages 10, Model 5G
8 April 2015 L.L. Aylward et al. / Regulatory Toxicology and Pharmacology xxx (2015) xxx–xxx Table 2 Characterization of degrees of dental fluorosis (adapted from Health Canada, (2010), based on categories set by the American Dental Association). Classification
Criteria: description of enamel
Normal Questionable Very mild
Smooth, glossy, pale creamy-white translucent surface A few white flecks or white spots Small opaque, paper-white areas covering less than 25% of the tooth surface Opaque white areas covering less than 50% of the tooth surface All tooth surfaces affected; marked wear on biting surfaces; brown stain may be present All tooth surfaces affected; discrete or confluent pitting; brown stain present
Mild Moderate Severe
196
assessments by Health Canada (2010) and USEPA (2010) establish exposure guidance values based on moderate dental fluorosis, which is considered to be an unacceptable cosmetic outcome, (see Table 2). This effect occurs only under conditions of elevated exposure for extended time periods in childhood, and the sensitivity for this outcome essentially ends after age 8 when adult teeth are fully formed (Health Canada, 2010). The ATSDR (2003) evaluation is based on avoidance of the increased risk of skeletal fractures associated with elevated fluoride exposure. The IOM (1997) identified tolerable upper intake levels (ULs) designed to ensure that neither unacceptable levels of cosmetic dental fluorosis (for childhood exposures up to age 8) nor skeletal fluorosis (for ages > 8 years) would occur. Both endpoints of concern with respect to elevated fluoride exposure, namely moderate dental fluorosis and skeletal fluorosis, occur only after long-term elevated exposures and are not consequences of acute exposure events. In addition to an exposure guideline for adverse effects from excessive exposure, the IOM (1997) established adequate intakes (AIs) for sufficient fluoride exposure. The AIs were established as levels of fluoride that would provide maximal protection of a population against dental caries without unwanted health effects. AIs were established for infants 66 months of age and for individuals >6 months of age. The typical approach for derivation of BE values is to identify the point of departure (POD) for the risk assessment (in the case there studies in humans form the basis for the exposure guidance value), estimate urinary concentrations corresponding to the POD (BEPOD), and then apply the composite uncertainty factors from the risk assessment to derive the BE value corresponding to the exposure guidance value (Fig. 1). The risk assessments from Health Canada, the USEPA, and the IOM identified human exposures equal to a NOAEL and applied no additional uncertainty factors, so the POD and final exposure guidance value were the same within each of these risk assessments (see Table 1). The use of no uncertainty factors for deriving the fluoride guidance values was deemed appropriate by the regulatory agencies because the NOAEL was obtained from studies that included sensitive human populations. The ATSDR MRL applied a composite uncertainty factor of 3 to the identified NOAEL.
197
3.2. Mass balance for urinary fluoride excretion
198
The pharmacokinetics of fluoride have been studied extensively in human populations. Fluoride is well absorbed from drinking water and dietary sources, with absorption rates of 80–90% (Whitford, 1994). Absorbed fluoride is partially retained in the skeletal system, with a fraction excreted in urine (Villa et al., 2010). Plasma and urinary half-lives of 5.1 and 5.8 h, respectively, were reported in a controlled administration study by Ekstrand and Ehrnebo (1983). Whitford et al. (2008) tested different fluoride compounds used for water fluoridation (natural sources, sodium
158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195
199 200 201 202 203 204 205 206
3
fluoride and fluorosilicic acid) and in dentifrices and found that the pharmacokinetics were similar for different fluoride compounds. Numerous studies have assessed the quantitative relationships between fluoride intake and urinary excretion, both under conditions of relatively steady intake rates (reviewed in Villa et al., 2010) and under conditions of changing intake rates (see, for example, Ekstrand et al., 1984). For the purposes of derivation of BE values for fluoride associated with chronic exposure guidance values, we focus on the evaluations of fluoride kinetics under conditions of steady intake rates. Villa et al. (2010) collected individual data from multiple studies of children and adults. Data on measured fluoride intake and urinary excretion rates (mg/d) for 212 children ages 0.15–7 years from nine studies in six countries and for 269 adults aged 18–75 years from eight studies in two countries were collected. Based on these data, Villa et al. (2010) constructed linear functions for prediction of daily urinary fluoride excretion (DUFE, mg/d) as a function of total daily fluoride intake (TDFI, mg/d) for children and adults of the form:
DUFE ¼ I þ b TDFI
ð2Þ
207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225
226 228
where I is an age-specific intercept representing daily fluoride excretion from body stores in the absence of any fluoride intake, and b is the slope of the linear relationship between fluoride excretion and intake. The fitted age-specific regression parameters are presented in Table 3. The slope relating daily fluoride excretion to intake was lower in children than in adults (slopes of 0.35 and 0.54, respectively). However, the intercept, representing excretion from body stores in the absence of any exposure, was higher in adults than in children; this is consistent with greater stored fluoride in adults than in children. Villa et al. (2010) reported that they did not find data sufficient to derive relationships for adolescents. For the purposes of this evaluation, we assumed that the urinary excretion fraction for adolescents is an average of that predicted for children and adults using the regression parameters in Table 3.
229
3.3. Urinary flow rates
244
3.3.1. Infants and children We identified 13 articles that provided data on 24-h urinary flow rate in infants and children along with data on bodyweight (Table 4), allowing estimation of urinary flow rate on a ml/kg-d basis. In addition, the NHANES 2009–2010 dataset (CDC, 2012) collected data allowing estimation of urinary flow rate for the spot samples collected in that survey. In these datasets, the time since last urinary void, the void volume, and the bodyweight were recorded, allowing calculation of urinary flow on a ml/kg-hr basis. We extrapolated the urine flow rate from the spot samples to 24 h. Table 4 presents a summary of these data, and Fig. 2A presents a scatter plot of the data. For children under age 15, no significant differences between males and females were observed in the studies that reported sex-specific data; therefore, data for both sexes are considered together here. Reported urinary flow rates for newborn and very young infants varied widely. Data for somewhat older infants and children were more consistent from study-to-study. Based on the extreme heterogeneity in the reported flow rates for children at the younger age range, we did not attempt to fit a function to these data. However, for the more stable data beginning at age 3, we fit the logarithm of the flow rate with a linear function of age, extended to age 14, with each point estimate weighted by the size of the dataset. The fit was highly significant, with an R2 value of 0.48, p < 0.001. This equation can be used to predict an average age-specific urinary flow rate (FR, ml/kg-d) for ages 3–14:
245
Please cite this article in press as: Aylward, L.L., et al. Biomonitoring Equivalents for interpretation of urinary fluoride. Regul. Toxicol. Pharmacol. (2015), http://dx.doi.org/10.1016/j.yrtph.2015.04.005
230 231 232 233 234 235 236 237 238 239 240 241 242 243
246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270
271
YRTPH 3268
No. of Pages 10, Model 5G
8 April 2015 4
L.L. Aylward et al. / Regulatory Toxicology and Pharmacology xxx (2015) xxx–xxx
Fig. 1. Schematic for derivation of Biomonitoring Equivalents for fluoride in urine.
Table 3 Regression parameters for the estimation of daily urinary excretion of fluoride (DUFE, mg/d) as a function of total daily fluoride intake (TDFI, mg/d) from Villa et al. (2010).a
a
Age group
Slope
p value for slope
Intercept
p value for intercept
R2
Children (0.15 to 7 years) Adults (18+ years)
0.35 ± 0.01
