Spatial distribution of fluoride in drinking water and health risk assessment of children in typical fluorosis areas in north China

Chemosphere 239 (2020) 124811

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Spatial distribution of fluoride in drinking water and health risk assessment of children in typical fluorosis areas in north China Lei Zhang a, Liang Zhao a, Qiang Zeng b, c, Gang Fu a, Baojia Feng a, Xiaohui Lin d, Zhonghui Liu a, Yang Wang a, Changchun Hou a, * a

Institute of Environment and Health, Tianjin Centers for Disease Control and Prevention, Tianjin, 300011, China Institute of Occupational and Health, Tianjin Centers for Disease Control and Prevention, Tianjin, 300011, China Department of Occupational and Environmental Health, School of Public Health, Tianjin Medical University, Tianjin, 300070, China d Department of Sanitation Detecting Laboratory, Tianjin Centers for Disease Control and Prevention, Tianjin, 300011, China b c

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


Fluoride content in drinking water for fluorosis areas in Tianjin was investigated.
MWS system has a better effect on reducing water fluoride concentration.
Younger children (1e4 years old) are more vulnerable to high fluoride exposure.
Special attention should be paid to health education strategies against fluorosis.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 26 July 2019 Received in revised form 25 August 2019 Accepted 6 September 2019 Available online 7 September 2019

China has been suffering from endemic fluorosis for the past 30 years. This study investigated fluoride concentrations in 10 districts of Tianjin, China, to illustrate their spatial distribution characteristics and potential human health risks. The results showed fluoride concentration of 0.01e6.30 mg L1 with a mean value of 0.99 mg L1, and 78.82% of water fluoride reaches the standard for drinking water (1.5 mg L1). Higher fluoride levels were recorded in deep well pumps supply zones, and more potential changes in fluoride occurred was positively correlated with pH in groundwater. Mean value of fluoride in drinking water in 10 districts followed the order of WQ > BC > JZ > NH > BD > BH > JN > JH > DL > XQ. Estimations of non-carcinogenic risk for drinking water indicated that mean hazard quotient values of fluoride for combined pathways (i.e., oral ingestion and dermal absorption) were >1.0 for all age groups of WQ and BC. The results also showed that the estimated risk primarily came from the ingestion pathway. Risk levels for children varied obviously, generally in the order of 1-4y > 4-7y > 7-9y (years old). In the central tendency center and reasonable maximum exposure conditions, estimated risks were 1.25, 1.12, 0.771 and 3.66, 3.29, 2.27, respectively. The results supply material information for health authorities in fluorosis areas to put forward more efficient policies to control the endemic diseases. Attention should be paid to the formulation of health promotion strategies and measures to reduce fluoride intake in order to protect the health of residents. © 2019 Elsevier Ltd. All rights reserved.

Handling Editor: A. Gies Keywords: Exposure Risk Drinking water Water pollutants

* Corresponding author. Room 208, building 5, 76 hualong road, hedong district, Tianjin, China. E-mail address: magicfi[email protected] (C. Hou). https://doi.org/10.1016/j.chemosphere.2019.124811 0045-6535/© 2019 Elsevier Ltd. All rights reserved.

2

L. Zhang et al. / Chemosphere 239 (2020) 124811

1. Introduction Fluorine is one of the most important elements closely related to human health (Fawell et al., 2006). Fluoride is absorbed mainly through the gastrointestinal tract in the form of drinking water, and there is still no global consensus on recommended maximum intake of fluoride in drinking water (Khan et al., 2004; MOH, 2006b; USEPA, 2009). The effects of fluoride intake on health are significantly different (Erdal and Buchanan, 2004). Appropriate fluoride intake is beneficial to the growth of teeth and can prevent dental caries, while chronic excessive fluoride intakes lead to dental and skeletal fluorosis (Pendrys and Stamm, 1990; Mohammadi et al., 2017). Dental fluorosis usually shows as staining of teeth and bone damage is more severe in the form of skeletal fluorosis (Pendrys and Katz, 1989; Mohammadi et al., 2017; Death et al., 2018). Not only can it cause fluorosis, various tissues and organs such as the thyroid
nez-Co
rdova et al., 2018) (Barberio et al., 2017), kidney (Jime affected by chronic high fluoride exposure, which has become a hot topic in recent years. Studies in typical fluorosis areas have shown that the developing children are more vulnerable to fluoride neurotoxicity, high fluoride exposure in drinking water can cut children's intelligence quotient(IQ) by up to 16 percent, and dental fluorosis may even be an early indicator of impaired IQ in children (Choi et al., 2015; Yu et al., 2018). In view of the above-mentioned fact, restriction of high fluoride intake has very important practical significance on the health of the population. The recommended concentration of fluoride in drinking water is 0.5e1.5 mg L1 on the basis of the guideline from World Health Organization (WHO) (WHO, 2011). Countries around the world that suffer from high levels of fluoride intake from drinking water, such as India (Ahada and Suthar, 2017), Tunisia (Guissouma et al., 2017), Mexico (Martinez-Acuna et al., 2016) and South Africa (Elumalai et al., 2019), have worked hard to bring fluoride concentration in their drinking water down to this level. In China, drinking water defluoridation projects have been launched since the 1960s. By 2010, about 82% of endemic villages have implemented water improvement projects to reduce fluoride content. However, fluoride concentrations in drinking water are rarely systematically investigated, and the corresponding health risks are rarely documented after these projects have been in operation for a period of time. Tianjin is one of the most water - deficient cities in northern China (Hu et al., 2002). In the past few decades, groundwater has been an important source of drinking water and makes it be a traditional fluorosis area (Rongshu et al., 1995; Lu et al., 2000; Li et al., 2017). The implementation of various strategies to reduce water fluoride in this area provides an ideal environment for studying the health effects of high fluoride intake (Lu et al., 2000; Yu et al., 2018). Although previous researchers have identified concentrations of water fluoride and population health risks in alternating regions, an citywide distribution and risk assessments of fluoride exposure are lacking. Therefore, this study intends to achieve the following research purposes through the investigation of fluoride concentration within the city: a) To describe the spatial distribution of fluoride in drinking water in Tianjin city based on townships; B) Using the assessment model to evaluate children health risks through drinking water fluoride exposure; C) Evaluates the effect of different types of projects on drinking water fluoride reduction. 2. Material and methods 2.1. Study area Located in the northeastern part of the North China Plain,

Tianjin is a national central city and mega city with the location between 116 430 to 118 040 east longitude and 38 340 to 40150 north latitude. The annual average temperature is about 14  C, and the annual average precipitation is 360e970 mm. Being an important part of China's national strategy-Beijing-Tianjin-Hebei integration, and has a total area of 1.195  104 km2, Tianjin comprises sixteen districts and has a population of over 15 million by the end of 2017. The ten regions involved in agriculture have been identified as endemic drinking water fluorosis areas in earlier studies (Yu et al., 2018), which has a population of about 6 million (Table S1).

2.2. Sample collection 2.2.1. Monitoring points This study covers all 171 towns/streets in 10 agriculture-related areas in Tianjin to make sure the stability and representatives of the results. An average of three monitoring points was selected per town/street. There were 11, 9, 13, 8, 29, 24, 14, 18, 27 and 18 townships in DL, XQ, BC, JN, WQ, BD, NH, BH, JZ respectively, and the number of corresponding monitoring points was 27, 42, 48, 26, 93, 66, 42, 45, 79 and 56, respectively. A total of 524 water samples were collected. The water-supply engineering was divided into the following types: ⅰ Municipal Water Supply (MWS) system: With sanitary permission, the water supply is more than 1000 tons/day. The main sources of water are rivers and lakes; groundwater is used in a few engineering; usually with perfect technology process and equipment. ⅱ Deep Well Pumps (DWP): Daily water supply less than 1000 tons without sanitation permit; deep groundwater is the dominant source of water, with a convenient technological process. The geographical location of each sampling point was shown in Fig. 1. The terminal water samples were collected from March to June (dry season) and July to September (Wet season) separately in 2016.

2.2.2. Sample treatment and test Water samples were stored in clean high-quality plastic bottles (polyethylene), which was protected from light. The bottles had undergone a series of special treatments. In the preparation phase, they were washed with clean water and then soaked in 10% nitric acid for 8 h before being washed with distilled water. At each sampling site, the bottles were rinsed with terminal water three times before sampling. It would take at least 15 min to draw off some possibly polluted water, which would make sure that the collected water could accurately reflect the status of local water before starting to take water samples. The pH of the samples was measured in situ using pH meters. Once collected, samples were stored in the dark at 4  C and taken to the laboratory and tested as soon as possible. Fluoride is detected by ion chromatography according to the “Standard examination methods for drinking Water” (MOH, 2006a). The samples were filtered by 0.20 mm filter membrane to prevent blockage of the protection and separation column system. Quality assurance was achieved by implementing standard laboratory procedures and quality control techniques. Parallel samples were collected simultaneously to ensure the accuracy of the test results, and the relative error was within 5%. For the samples with fluoride concentration exceeding 1.2 mg L1, the samples were reexamined (MOH, 2006b).

L. Zhang et al. / Chemosphere 239 (2020) 124811

3

Fig. 1. The geographical location of each sampling point.

2.3. The health risk assessment 2.3.1. Assessment model Health risk assessment (HRA) is a method of assessing the potentially harmful effects of exposure to certain chemicals over a period of time. HRA usually consists of five components, including: problem formulation, hazard and dose response assessment, exposure assessment, risk characterization and uncertainty analysis. Humans are potentially exposed to pollutants from drinking water though oral ingestion, dermal contact and inhalation absorption. Based on Environmental Protection Agency (USEPA) guidelines, fluorides are rarely absorbed via the respiratory tract. Therefore, in this study, human non-carcinogenic risks for fluoride occurrences were carried out via two different exposure routes: oral ingestion and dermal absorption (USEPA, 1989). The population exposure to fluoride was determined by estimated daily intakes (EDI), which were calculated for two exposure pathway as follows:

EDIing ¼

Cw  IRw  EF  ED BW  AT

EDIderm ¼

Cw  SA  Kp  F  ETs  EF  ED  CF BW  AT

1

2

Hazard quotient (HQ) was calculated by dividing the EDI by the reference dose (RfD). In this study, HQ represents the risk of fluoride intake from drinking water:

HQ ¼

EDI RfD

3

RfD represents the reference dose for a specific exposure pathway (mg kg1 d1). The value of oral RfD for fluoride was adopted to be 0.06 mg kg1 d1, which was specified by the integrated risk information system. The HQ limit is placed at 1, less than 1 means the risk of non-carcinogenic damage is negligible, while greater than the limit representative of potential health issue. The reference dose associated with the absorption of fluoride through dermal is

4

L. Zhang et al. / Chemosphere 239 (2020) 124811

samples were monitored, of which 78.82% had a water fluoride less than 1.5 mg L1. The integral estimation of the situation at the regional level shows that the mean and median values are both lower than the limits according to the standard of small centralized water supply sanitation in rural areas of China (MOH, 2006b). The difference of fluoride concentration in terminal water of MWS and DWP was analyzed by non-parametric test, and the results indicate that water quality of MWS is better than that of DWP. The maximum, mean and median of water fluoride concentration are all show the same trend in the order of MWS < DWP. The compliance rate was calculated at a limit of 1.5 mg L1, and the results indicate that the qualified rate of MWS is higher than that of DWP. As can be seen from Table 1, the pH values of drinking water from the MWS and DWP projects are 7.23e8.50 and 6.74e8.95 with a median value of 8.00 and 7.98, respectively. The relationship between pH value and the concentration of fluoride in drinking water from the two types of projects has also been assessed. The results showed that there was a correlation between pH value and water fluoride concentration in the DWP project (P < 0.001), while waters from the MWS with surface water as the main source of water did not show similar characteristics, which suggested that the high fluoride in local groundwater might be related to the alkaline environment (Fig. S1).

given in Eq. (4).

RfDderm ¼ RfDo  ABSgi

4

where RfDderm and RfDo represent the dermal RfD and oral RfD respectively. ABSgi denotes the gastrointestinal absorption factor (unit less). Central tendency exposure (CTE) and reasonable maximum exposure (RME) were used in characterizing potential exposures (Erdal and Buchanan, 2004). The CTE is estimated by mean value, while RME is estimated by 90th percentile value of the data. 2.3.2. Grouping and parameter The parameters applicable to the survey region play a comparatively more important role in the health risk assessment. Therefore, values directly obtained from Tianjin area were granted the highest priority. We collected data of water ingestion rate, dermal surface area and body weight of Tianjin population from Highlight of Chinese Children's Exposure Factors Handbook (Table S2) (Duan, 2012). Research has shown that water fluoride levels between 0.5 and 1.5 mg L1 is beneficial to the growth of teeth and the prevention of dental caries, between 1.5 and 5 mg L1 can cause dental fluorosis, and 5e40 mg d1 fluoride intake can lead to skeletal fluorosis especially in adolescents exposed to water fluoride concentration up to 10 mg L1 from birth, the risk will be significantly increased (Fordyce et al., 2007; Goodarzi et al., 2016; Yousefi et al., 2018). Therefore, we divide the assessment age into three stages: 1e4 years old, 4e7 years old and 7e9 years old. Health risks of different ages were calculated under CET and RME scenarios.

3.2. Spatial distribution of drinking water fluoride In all 171 monitoring townships, the proportion of fluoride concentration in the safe range is about 77.78%. The fluoride in drinking water shows concentration fluctuation in the different administrative district. In BC and DL districts, the average concentration of fluoride in drinking water exceeds the permissible concentration (1.5 mg L1). The overall compliance rate of monitored water samples in each region is shown in Table 2 (see Table 3). Except for DL and XQ District, the maximum water fluoride in the remaining areas exceeds 1.5 mg L1. The highest value appears in the BH District that adjacent to the Bohai Bay, where regional geological condition is complicated, and the groundwater contains a high concentration of fluoride, iodide, chloride, sulfate, etc (Dong et al., 2013). The mean and median of water fluoride in WQ and BC District are significantly higher than the contact limit 1.5 mg L1 for fluoride in drinking water, which has been set up by WHO. The mean value in JZ district is between 1.0 mg L1 and 1.2 mg L1; these two values are the limits of fluoride in water supply for urban and rural areas of China(MOH, 2006b), respectively. The mean and median values of water fluoride in other districts are both lower than the limit of 1.0 mg L1, which is generally believed no harmful for human health. Using Arcgis software, we draw the distribution map of fluoride in tap water based on township strategy (Fig. 2). The results shows that DL, XQ and JN are in accordance with the standard of 1.5 mg L1, indicating that the lower risk in this area is mainly due to the high quality of MWS water supply. Water fluoride

2.4. Statistical analysis For each water-quality parameter, SPSS (IBM SPSS statistics version 22.0) was utilized to calculate the mean, standard error of mean, median, minimum, and maximum values. It puts up quantitative analysis on the otherness of projects types by means of rank test of non-parametric test in order to find out the extent of the otherness change in water quality improvement strategy. The geospatial map of drinking water fluoride is drawn based on townships using ArcGIS Desktop (ESRI ArcGIS version 10.5). The level of p < 0.05 was selected as statistically significant. 3. Result 3.1. Properties of the drinking water The data on water quality like the minimum, maximum, mean, standard error of mean, and median obtained from the results of terminal water samples are summarized in Table 1. The fluoride concentration in drinking water in the survey area is in the range from 0.01 to 6.30 mg L1, with a mean value and median value of 0.99 mg L1 and 0.47 mg L1 respectively. A total of 524 water

Table 1 Fluoride concentration and related indicators of MWS and DWP. Type

N

Min (mg L1)

Max (mg L1)

P90 (mg L1)

Median (mg L1)

Meana (mg L1)

SE (mg L1)

pHb

Standard (mg L1)

Attainment ratec (%)

MWS DWP Total

129 395 524

0.01 0.01 0.01

1.79 6.30 6.30

0.90 3.49 2.91

0.21 0.70 0.47

0.38 1.19 0.99

0.03 0.07 0.05

7.23e8.50 6.74e8.95 6.74e8.95

1.50 1.50 1.50

98.45 72.41 78.82

Abbreviation: N: Number of samples; Min: Minimum; Max: Maximum; SE: Standard Error; P90: 90th percentile. a Compared MWS with DWP: Mann-Whitney U test, Z ¼ 4.949, p < 0.001. b With median value of: MWS: 8.00; DWP: 7.98. c Compared MWS with DWP: Chi-square test, x2 ¼ 39.51, p < 0.001, Attainment rate: DWP < MWS, Standard: 1.50 mg L1.

L. Zhang et al. / Chemosphere 239 (2020) 124811

5

Table 2 Fluoride-related indicators in drinking water in ten regions involved in agriculture collected from Tianjin, China. N

Max (mg L1) P90 (mg L1) Mean (mg L1) SE (mg L1) Median (mg L1) pH

Standard (mg L1) Samples achieved rate Compliance rate of townships

DL

27

0.79

0.30

0.22

0.03

0.21

1.5

100%

100%(11/11)

XQ

42

0.25

0.23

0.18

0.004

0.19

1.5

100%

100%(9/9)

BC

48

4.62

4.10

1.75

0.24

1.58

1.5

50.00%

46.15%(6/13)

JN

26

1.70

0.80

0.33

0.07

0.10

1.5

96.15%

100%(8/8)

WQ 93

4.98

3.82

1.96

0.13

1.67

1.5

49.46%

48.28%(14/29)

BD

66

3.50

2.03

0.80

0.09

0.53

1.5

81.82%

75.00%(18/24)

NH

42

3.17

1.57

0.89

0.09

0.79

1.5

90.48%

92.86%(13/14)

BH

45

6.30

0.77

0.74

0.21

0.39

1.5

93.34%

88.89%(16/18)

JZ

79

5.95

3.16

1.15

0.16

0.61

1.5

79.75%

77.78%(21/27)

JH

56

2.25

1.14

0.26

0.07

0.10

1.5

92.85%

94.44%(17/18)

Total 524 6.30

2.91

0.99

0.05

0.47

1.5

78.82%

77.78%(133/171)

7.47 e8.05 7.48 e8.50 7.83 e8.27 6.74 e8.13 6.83 e8.76 7.23 e8.01 8.05 e8.86 7.44 e8.95 7.34 e8.49 7.52 e7.88 6.74 e8.95

Abbreviation: N: Number of samples; Min: Minimum; Max: Maximum; SE: Standard Error; P90: 90th percentile.

Table 3 Comparison of fluoride in drinking water with other endemic fluorosis areas.

Tianjin, China Tunisia Nigeria Mexico Sweden India Iran Shanxi, China Tianjin, China Heilongjiang, China Jilin, China

N

Min (mg L1)

Max (mg L1)

Mean (mg L1)

Attainment rate (%)

Sampling year

Source

524 100 63 47 4800 34 112 126 30 585 2373

0.01 0.05 0.48 0.4 0.1 0.06 0.23 0.10 0.64 0.03 0.5

6.30 2.40 1.84 3.0 15 4.33 10.3 5.70 5.90 7.83 10

0.99 e 1.23 1.4 1.0 1.13 1.70 0.70 2.23 1.95 2.33

78.82 e 67 57.45 76.00 80 43 88.30 e e e

2016 2014 e 2012 1978e2007 2011 2014e2016 2008e2016 2009 e 1998

This study Guissouma et al. (2017) Emenike et al. (2018) Martinez-Acuna et al. (2016) Erdal and Buchanan (2004) Narsimha and Sudarshan (2018) Yousefi et al. (2018) Li et al. (2019) Zhang et al. (2017) Liu et al. (2014) Zhang et al. (2003)

Abbreviation: N: Number of samples; Min: Minimum; Max: Maximum.

concentrations in BC and WQ regions are at higher risk, with about half of the townships exceeding the limit. The compliance rates of towns and villages in other regions are shown in Table 2.

1e4 y and 4e7 y were higher than 1, and BC, WQ, BD, NH and JZ in the age group of 7e9 y met the limit.

3.3. Health risks to children

3.3.2. Risk assessment The health risks of fluoride absorbed through oral and skin were in the order of 7e9 y < 4e7 y < 1e4 y. Table S3 shows the health risks of different types of the water supply project through drinking water consumption and skin absorption. Overall, the health risk of children in the study area was greater than 1 in both models except for children aged 7e9 y. In The CTE model, the health risk values for 1e4 y, 4e7 y and 7e9 y were 1.25, 1.12 and 0.771. In The RME model, the health risks were 3.66, 3.29 and 2.27 respectively. Different from drinking water intake, the non-carcinogenic risk of fluoride absorbed through the skin after water quality improvement can be neglected, and the health risk value of fluoride absorbed through the skin is significantly lower than the limit, which is 3e4 orders of magnitude lower than the risk of oral intake. The results are consistent with those of Zhang et al. (2017). The risk assessment values of different regions are shown in Table S4. In The CTE model, the health risk assessment values in WQ and BC district are all higher than 1, suggesting that further efforts should be made to reduce fluoride concentration in this region. The health risk assessment values of different populations in each region are shown in Table S4.

3.3.1. Estimated daily intake The estimated values of fluoride intake of children at different ages in water supply projects and diverse parts in Tianjin are shown in Fig. 3. CTE and RME were selected for estimation. The former was estimated by the mean of the concentration of fluoride, while the latter was estimated by the 90th percentile value of the data (Erdal and Buchanan, 2004). Overall, except for the age of 7e9 y, the fluoride intake of children at different ages in the study area was greater than the limit of 0.06 mg kg1 d1 in both models, which was consistent with the high incidence of dental fluorosis in the study area. The average daily intake of DWP and MWS showed a decreasing trend with the increase of age. It is noteworthy that the intake of children aged 1e4 y in MWS exceeded the limit slightly in the RME mode (Fig. 3c). We also estimated the daily intake in different regions. In CTE scenario, the intake of BC, WQ children at all stages are higher than 1. JZ exceeded the limit in the age group of 1e4 y and 4e7 y. NH exceeded the limit in the age group of 4e7 y. In the worst-case scenario (RME), the intake of DL, XQ, JN, BH in the age group of

6

L. Zhang et al. / Chemosphere 239 (2020) 124811

Fig. 2. Spatial distribution of fluoride concentration in drinking water in Tianjin, China.

4. Discussion

relatively good level in China. It shows that all efforts to improve water quality in this area are effective and efficient.

4.1. The status and spatial distribution of fluoride The results of this study are significantly better than previous studies (Zhang et al., 2017), indicating that the continuous construction of various water supply projects in this region has effectively improved the quality of drinking water, although some regions have higher fluoride concentrations. The overall mean of the region has fallen to the Chinese drinking water standard of 1.0 mg L1 (MOH, 2006b) and recommendation value of other countries and regions (Craig et al., 2015). Compared with many countries and regions with fluorosis in the world, the concentration of fluoride in water in the study area is lower, and it is also at a

4.2. The alkaline environment of the regional groundwater may be the cause of high fluoride in drinking water The results from our study suggest that groundwater samples with elevated water fluoride concentration are related to higher pH values, which is similar to the cases in various locations around the world. Geological data from Li et al. (2017) shows that the quaternary aquifer is the most dominating source of groundwater in Tianjin and consists of four groups: Holocene aquifer, the upper, middle and lower Pleistocene aquifers. High concentrations of fluoride were detected in water samples collected from various

L. Zhang et al. / Chemosphere 239 (2020) 124811

7

Fig. 3. Estimates of fluoride intake for children of all age groups in two water supply projects and different regions. CTE: central tendency exposure. RME: reasonable maximum exposure. Dotted lines represent the recommended daily intake value of USEPA.

aquifers. The high fluoride in water may be as a consequence of dissolution from quartzite and paleosols near the underlying groundwater table (Edmunds and Smedley, 2004). The geological attributes of the region may be responsible for the positive relationship existing between fluoride and pH. 4.3. MWS is superior to DWP in improving water quality Compared with the DWP project, MWS has the better effect in reducing water fluoride concentration. We encourage the construction of MWS projects in long-run planning. In addition, Large Water Plants have a better management and operation mechanism and water quality can be guaranteed even in the long-term operation (Jiang, 2015); Nevertheless, our results show that the mean value of DWP project is also around the rural water supply standard of 1.2 mg L1 (MOH, 2006b). If limited by economic conditions and equipment facilities, the construction of DWP projects will still be a good choice. 4.4. Uncertainty Numerous existing studies have used uncertain simulation techniques such as Monte Carlo to assess the uncertainty of the health risks (Fallahzadeh et al., 2017; Saha et al., 2017). The results have confirmed that pollutant concentration is important factors affecting the results. In this risk assessment, the specific survey

parameters from northern China were selected, but other uncertainties still exist during the risk assessment given the residents' drinking habits. In addition, we believe that the health risks remain underestimated by substituting total drinking water for other forms of exposure, regardless of beverage consumption, eating fluoride-containing foods and exposure to fluoride-containing toothpaste (Erdal and Buchanan, 2004). 4.5. Health education strategies are as important as reducing fluoride concentrations in drinking water In order to explore the relationship between fluoride concentration in water and fluorosis, we collected studies on related diseases and government work reports (TMHB, 2017). In 2016, 2017, the prevalence of dental fluorosis among children was showing an overall downward trend, and the rate was 45.50% and 39.27%, respectively. The prevalence of dental fluorosis among children in various areas of Tianjin is shown in Fig. 4. The fact that the concentration of fluoride in water is not completely consistent with the prevalence of dental fluorosis suggests that even with clean water supply, the health risks of fluoride still cannot be ignored. The reason for the discrepancy may be: 1. The construction period of all kinds of projects may last 1e5 years with a relatively long cycle. From the beginning of construction to the final water supply, the surrounding residents

8

L. Zhang et al. / Chemosphere 239 (2020) 124811

Fig. 4. Health risk of fluoride intake and overall incidence of dental fluorosis in children in Tianjin, China.

still drink high-fluoride groundwater during this period (Liu et al., 2016). 2. For small drinking water safety projects, due to inconvenient use of water, inhabitant's habits and economic conditions, residents prefer not to safe water but to other sources of water, such as self-prepared wells and machine wells (Zhang et al., 2017). This situation still leads to health risks. In areas with endemic fluorosis, it is equally important to reduce the health risks, not only by controlling the concentration of water fluoride, but also for residents to bettering drinking habits and implement effective health education and health promotion

strategies (Xu et al., 2015; Yu et al., 2015). 4.6. Recommendations for the prevention and control of fluorosis It is suggested that appropriate drinking water projects should be constructed according to the economic conditions and geological conditions of the regions. Reducing the concentration of fluoride in water and the duration of continuous exposure are necessary to control population health risks. In addition, in endemic fluorosis areas, even if the concentration of fluoride in water is well controlled, health education should be carried out for the residents.

L. Zhang et al. / Chemosphere 239 (2020) 124811

5. Conclusion 1) Although some drinking water with high fluoride still exists, the concentration of fluorine in drinking water in Tianjin has been controlled. Through the implementation of drinking water safety measures in this area, fluoride in drinking water has been impressively reduced. 2) Both DWP and MWS can effectively reduce the concentration of fluoride in water. MWS is the key to control fluorosis, and should be the first choice for long-term stable construction. 3) In endemic fluorosis areas, even if the water fluoride level is well controlled, health education and health promotion strategies are still necessary, and their importance must be highly valued. Acknowledgments This research was supported by the Tianjin Science and Technology Committee (18ZXJMTG00250), the Tianjin Municipal Bureau of Public Health (2014KY22), the National Natural Science Foundation of China (81573107) and Medical Talent Program of Tianjin (2018). Appendiix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.chemosphere.2019.124811. References Ahada, C.P.S., Suthar, S., 2017. Assessment of human health risk associated with high groundwater fluoride intake in southern districts of Punjab, India. Expo. Heal. 1, 1e9. ~ onez, C., McLaren, L., 2017. Fluoride exposure and Barberio, A.M., Hosein, F.S., Quin indicators of thyroid functioning in the Canadian population: implications for community water fluoridation. J. Epidemiol. Community Health 71, 1019e1025. Choi, A.L., Zhang, Y., Sun, G., Bellinger, D.C., Wang, K., Yang, X.J., Li, J.S., Zheng, Q., Fu, Y., Grandjean, P., 2015. Association of lifetime exposure to fluoride and cognitive functions in Chinese children: a pilot study. Neurotoxicol. Teratol. 47, 96e101. Craig, L., Lutz, A., Berry, K.A., Yang, W., 2015. Recommendations for fluoride limits in drinking water based on estimated daily fluoride intake in the Upper East Region, Ghana. Sci. Total Environ. 532, 127e137. Duan, X.L., 2012. Highlight of Chinese Children's Exposure Factors Handbook. China Environmental Science Press, Peking (in Chinese). Dong, D., Sun, W., Zhu, Z., Xi, S., Lin, G., 2013. Groundwater risk assessment of the third aquifer in Tianjin City, China. Water. Resour. Manag. 27, 3179e3190. Death, C., Coulson, G., Kierdorf, U., Kierdorf, H., Ploeg, R., Firestone, S., Dohoo, I., Hufschmid, J., 2018. Chronic excess fluoride uptake contributes to degenerative joint disease (DJD): evidence from six marsupial species. Ecotoxicol. Environ. Saf. 162, 383e390. Edmunds, M., Smedley, P., 2004. Fluoride in natural waters. In: Selinus, O., Alloway, B., Centeno, J., Finkelman, R., Fuge, R., Lindh, U., Smedley, P. (Eds.), Esssential of Medical Geology. Impacts of the Natural Environment on Public Health. Elsevier, Amsterdam, pp. 301e329. Erdal, S., Buchanan, S.N., 2004. A quantitative look at fluorosis, fluoride exposure, and intake in children using a health risk assessment approach. Environ. Health Perspect. 113, 111e117. Emenike, C.P., Tenebe, I.T., Jarvis, P., 2018. Fluoride contamination in groundwater sources in Southwestern Nigeria: assessment using multivariate statistical approach and human health risk. Ecotoxicol. Environ. Saf. 156, 391e402. Elumalai, V., Nwabisa, D.P., Rajmohan, N., 2019. Evaluation of high fluoride contaminated fractured rock aquifer in South AfricaeGeochemical and chemometric approaches. Chemosphere 235, 1e11. Fawell, J., Bailey, K., Chilton, J., Dahi, E., Magara, Y., 2006. Fluoride in Drinking-Water. IWA Publishing. Fordyce, F.M., Vrana, K., Zhovinsky, E., Povoroznuk, V., Toth, G., Hope, B.C., Iljinsky, U., Baker, J., 2007. A health risk assessment for fluoride in Central Europe. Environ. Geochem. Health 29, 83e102. Fallahzadeh, R.A., Ghaneian, M.T., Miri, M., Dashti, M.M., 2017. Spatial analysis and health risk assessment of heavy metals concentration in drinking water resources. Environ. Sci. Pollut. Res. 24, 24790e24802. Goodarzi, F., Mahvi, A.H., Hosseini, M., Nedjat, S., Nabizadeh Nodehi, R., Kharazifard, M.J., Parvizishad, M., Cheraghi, Z., 2016. The prevalence of dental fluorosis and exposure to fluoride in drinking water: a systematic review. J. Dent. Res. Dent. Clin. Dent. Prospects 10, 127e135.

9

Guissouma, W., Hakami, O., Al-Rajab, A.J., Tarhouni, J., 2017. Risk assessment of fluoride exposure in drinking water of Tunisia. Chemosphere 177, 102e108. Hu, R., Wang, S., Lee, C., Li, M., 2002. Characteristics and trends of land subsidence in Tanggu, Tianjin, China. Bull. Eng. Geol. Environ. 61, 213e225. Jiang, Y., 2015. China's water security: current status, emerging challenges and future prospects. Environ. Sci. Policy 54, 106e125.
nez-Co
rdova, M.I., Ca
rdenas-Gonza
lez, M., Aguilar-Madrid, G., SanchezJime
Domínguez-Guerrero, I.A., Gonz
~ a, L.C., Barrera-Hern
Pen andez, A., alez-Horta, C., Barbier, O.C., Del Razo, L.M., 2018. Evaluation of kidney injury biomarkers in an adult Mexican population environmentally exposed to fluoride and low arsenic levels. Toxicol. Appl. Pharmacol. 352, 97e106. Khan, A.A., Whelton, H., O'Mullane, D., 2004. Is the fluoride level in drinking water a gold standard for the control of dental caries? Int. Dent. J. 54, 256e260. Lu, Y., Sun, Z., Wu, L., Wang, X., Lu, W., Liu, S., 2000. Effect of high-fluoride water on intelligence in children. Fluoride 33, 74e78. Liu, H., Gao, Y., Sun, L., Li, M., Li, B., Sun, D., 2014. Assessment of relationship on excess fluoride intake from drinking water and carotid atherosclerosis development in adults in fluoride endemic areas, China. Int. J. Environ. Health Res. 217, 413e420. Liu, Y., Yao, T., Bai, Y., Liu, Y., 2016. The sustainability of drinking water supply in rural China: does the provision of drinking water investment mismatch the demand of residents? Phys. Chem. Earth, Parts A/B/C 96, 34e40. Li, J.X., Zhou, H.L., Qian, K., Xie, X.J., Xue, X.B., Yang, Y.J., Wang, Y.X., 2017. Fluoride and iodine enrichment in groundwater of North China Plain: evidences from speciation analysis and geochemical modeling. Sci. Total Environ. 598, 239e248. Li, Y., Wang, F., Feng, J., Lv, J.P., Liu, Q., Nan, F.R., Zhang, W., Qu, W.Y., Xie, S.L., 2019. Long term spatial-temporal dynamics of fluoride in sources of drinking water and associated health risks in a semiarid region of Northern China. Ecotoxicol. Environ. Saf. 171, 274e280. MOH, 2006a. Standards Examination Method for Drinking Water(GB 5750 - 2006). National Health Commission of the People’s Republic of China, Peking (in Chinese). MOH, 2006b. Standards for Drinking Water Quality (GB 5749 - 2006). National Health Commission of the People’s Republic of China, Peking (in Chinese). Martinez-Acuna, M.I., Mercado-Reyes, M., Alegria-Torres, J.A., Mejia-Saavedra, J.J., 2016. Preliminary human health risk assessment of arsenic and fluoride in tap water from Zacatecas, Mexico. Environ. Monit. Assess. 188, 476. Mohammadi, A.A., Yousefi, M., Yaseri, M., Jalilzadeh, M., Mahvi, A.H., 2017. Skeletal fluorosis in relation to drinking water in rural areas of West Azerbaijan, Iran. Sci. Rep. 7, 17300. Narsimha, A., Sudarshan, V., 2018. Drinking water pollution with respective of fluoride in the semi-arid region of Basara, Nirmal district, Telangana State, India. Data Brief 16, 752e757. Pendrys, D.G., Katz, R.V., 1989. Risk of enamel fluorosis associated with fluoride supplementation, infant formula, and fluoride dentifrice use. Am. J. Epidemiol. 130, 1199e1208. Pendrys, D., Stamm, J., 1990. Relationship of total fluoride intake to beneficial effects and enamel fluorosis. J. Dent. Res. 69, 529e538. Rongshu, W., Haiming, L., Ping, N., Ying, W., 1995. Study of a new adsorbent for fluoride removal from waters. Water Qual. Res. J. Can. 30, 81e88. Saha, N., Rahman, M.S., Ahmed, M.B., Zhou, J.L., Ngo, H.H., Guo, W., 2017. Industrial metal pollution in water and probabilistic assessment of human health risk. J. Environ. Manag. 185, 70e78. TMHB, 2017. Annual Report on the Health Status of Tianjin Residents. Tianjin Municipal Health Bureau, Tianjin. USEPA, 1989. Risk Assessment Guidance for Superfund Volume I: Human Health Evaluation Manual (Part A). EPA/540/1e89/002. U.S. Environmental Protection Agency, Office of Emergency and Remedial Response, Washington, D.C. USEPA, 2009. National Primary Drinking Water Regulations. U.S. Environmental Protection Agency. U.S. Environmental Protection Agency, Washington, D.C. WHO, 2011. Guidelines for Drinking-Water Quality, fourth ed. World Health Organization, Geneva. Xu, J., Feng, P., Yang, P., 2015. Research of development strategy on China's rural drinking water supply based on SWOTeTOPSIS method combined with AHPEntropy: a case in Hebei Province. Environ. Earth Sci. 75, 58. Yu, X., Geng, Y., Heck, P., Xue, B., 2015. A review of China's rural water management. Sustainability 7, 5773e5792. Yousefi, M., Ghoochani, M., Hossein Mahvi, A., 2018. Health risk assessment to fluoride in drinking water of rural residents living in the Poldasht city, Northwest of Iran. Ecotoxicol. Environ. Saf. 148, 426e430. Yu, X., Chen, J., Li, Y., Liu, H., Hou, C., Zeng, Q., Cui, Y., Zhao, L., Li, P., Zhou, Z., Pang, S., Tang, S., Tian, K., Zhao, Q., Dong, L., Xu, C., Zhang, X., Zhang, S., Liu, L., Wang, A., 2018. Threshold effects of moderately excessive fluoride exposure on children's health: a potential association between dental fluorosis and loss of excellent intelligence. Environ. Int. 118, 116e124. Zhang, B., Hong, M., Zhao, Y., Lin, X., Zhang, X., Dong, J., 2003. Distribution and risk assessment of fluoride in drinking water in the west plain region of Jilin province, China. Environ. Geochem. Health 25, 421e431. Zhang, L.E., Huang, D., Yang, J., Wei, X., Qin, J., Ou, S., Zhang, Z., Zou, Y., 2017. Probabilistic risk assessment of Chinese residents' exposure to fluoride in improved drinking water in endemic fluorosis areas. Environ. Pollut. 222, 118e125.