This study found that the water fluoride levels in villages (A-D) after the water source change were much lower than the average pre-intervention level (3.475 mg/l) and the post-intervention levels in all intervention villages declined to 0.11 mg/l or lower. These reduced fluoride levels are lower than the Chinese National Fluoride Standard (1.0 mg/l), and the World Health Organization (WHO) recommended drinking water fluoride levels of 0.5~1.5 mg/l. The fluoride concentration of drinking water is the only objective indicator reflecting drinking water source of endemic fluorosis and environmental fluoride, and for evaluating the policy effect of water change/treatment in the endemic fluorosis areas. As most fluoride intake in the endemic fluorosis areas of China were from drinking water, it is appropriate to monitor the fluoride level in drinking water in endemic areas and over time.
We also found that urinary fluoride levels among the residents living in the villages after 14-15 years of water source change were similar or lower than the level in the control village. In other words, over the years that the high fluoride water was replaced by lower fluoride water (public water), urinary fluoride excretion had gradually decreased. Urinary fluoride (UF), is an objective measure of fluoride intake, because it is related to the water fluoride and reflects the human level of fluoride metabolism . Our study also found that the urinary fluoride concentrations declined the longer the villages had access to the public water supply, which was supported by similar results found in other studies [25, 26]. Gongju Yin et. al. conducted a survey among children aged 8-12 in the fluorosis areas and found that drinking water de-fluoridation resulted in water fluoride concentration reduction and consequent reduction in urinary fluoride . The results from these prior studies are consistent with the findings of our study. The level of urinary fluoride among children in intervention villages having a public water supply for more than 10 years (villages B, C and D) was similar to the level of UF in the control village. However, the children’s UF in village A is still higher than the control village probably due to a shorter time (6 years) of having a public water supply. An interesting finding from our study was that urinary fluoride levels are higher than drinking water fluoride which may be due to the accumulation of fluoride in the body; i.e., fluoride released from bone or other body tissues and excreted through the kidneys. It could also be due to the body accumulation of fluoride from other sources such as tea, local-grown vegetables, fruits or animals. This process could take place over a long period of time.
Long-term intake of excessive fluoride has adverse effects on health including dental fluorosis, change of bone metabolism indicators, and finally bone damage [28–32]. After a long period of low-fluoride public water exposure, all the indicators had reached to the normal level after the water changed for six years or more except for CT. These findings imply the success of the policy of fluoride water change and are consistent with the relationship between urinary fluoride and fluoride-induced bone metabolism indicators including BGP, CT, ALP, and bone mineral density reported by one of the few studies in this area [33, 34].
To select the best bone metabolism indicators related to high fluoride exposure, this study also examined the dose-response relationship between children’s urinary fluoride level and abnormal bone indicators. We found that children’s abnormal levels of BGP and CT increased as the levels of their urinary fluoride contents increased, but this dose-response relationship was not as apparent with respect to serum ALP levels and bone mass density. All of these four indicators can reflect the impact of fluoride on bone metabolism, including bone formation and bone resorption, which is accelerated and is the reason of osteosclerosis and bone softening due to the skeletal fluorosis.
Serum BGP, CT and ALP are commonly used as markers reflecting bone generation. As Liang et.al. described , BGP, a collagen protein produced and secreted by osteoblasts, could regulate the bone metabolism through binding with hydroxyapatite (HA) to form fluoride phosphate stone. This results in the increase of BGP secretion, destruction of normal bone mineralization rate, and induction of osteomalacia and osteoporosis . Therefore, serum BGP is a sensitive, specific, accurate and simple biochemical indicator to measure instantaneous change of bone metabolism [37–40]. ALP has been considered as a marker of osteoblast activity. Fluoride can affect ALP activity in two ways. i.e., changing the structure of the enzyme to directly affect the activity of the enzyme, and accumulating fluorine in the bone to promote bone cells in mitosis, and thus influence vitality of the bone cells. Therefore, the increase of serum ALP, to some extent, reflects an active cell proliferation of the osteoblasts, and suggests that the body that may have fluorosis-related damage. Wan stated that after exclusion of other metabolic bone diseases, elevated serum ALP can be used as a diagnostic indicator of skeletal fluorosis . Liu has shown that when the fluoride concentration in drinking water reaches 0.58mg/l-1.59 mg/l, ALP activity is increased . When the water fluoride concentration is between 1.60mg/l and-3.37 mg/l, ALP activity decreases slowly. It could decrease to as low as 57% of the normal amount, but it can return to normal levels some time later . ALP has been thus recognized as an early and important indicator in the evaluation of bone formation and bone turnover. In addition, the main role of CT is to stimulate osteoblast formation and indirectly affect bone metabolism [44, 45]. The sensitivity and specificity of CT are lower than those of BGP and ALP. Finally, bone mineral density, reflecting per unit volume of bone mineral content, is an important quantitative indicator for evaluating bone mass. Skeletal fluorosis caused by osteoporosis reduces per unit volume of bone mass, i.e., decreases the bone mineral density. The measurement of bone mineral density can be used as a more objective and accurate evaluation indicator of bone changes . In summary, BGP, ALP, and CT reflect pre-bone formation status, but bone mass density indicates final bone mineral change. Thus, use of all these four indicators not only demonstrate early changes of bone metabolism, but also describe the final mineral characteristics in order for us to evaluate the new regulation of supplying public water in the endemic fluorosis areas.
The benchmark dose–response models were first mentioned by Crump . A BMD is defined as the dose that corresponds to a specified change in adverse response compared to the response in untreated situation. The dose is associated with a given incidence (e.g., 1%, 5% or 10% incidence) of effect, the Benchmark Response, based on the best fitting dose-response curve in the region of the dose-response relationship where biologically observable data are available. The resulting BMD in this study is termed BMD10 for a 10% incidence.
Because these four bone metabolism indicators are the physiological continuity indicators of the bone metabolism or damage, we need to identify a cut-off point or threshold to define abnormal value based on the dose-response relationship. This study found that CT has the lowest threshold or BMD among all bone metabolism indicators, but ALP has the highest BMD. Unfortunately, no available literature can be found to compare with our findings. This study also found that most bone damage markers went back to normal level after six years of water change intervention except for CT which significantly declined after 10-14 years. Bone mineral density is the most stable indicator. No other studies are available to compare with our findings.
This study is one of the only or few research efforts to evaluate how this environmental intervention affected public health. We not only examined fluoride change (external exposure marker) in public water after providing low fluoride water, but also assessed UF (internal exposure marker) and bone damage related to fluoride exposure (health markers). Different from some prior studies which relied completely on self-reported information, use of these comprehensive biomarkers from exposure to health endpoints minimized reporting bias and provided reliable data for intervention evaluation. This study may also be the first to describe and compare the changes of several pre-clinical and clinical bone metabolism markers after intervention, which could help in understanding the potential biological mechanism and identify sensitive or appropriate health indicators for future public health surveillance.
One of the major limitations in this study is lack of baseline data for UF and bone metabolism indicators prior to water source change. However, a control village was carefully selected for comparison based on similar status as the intervention villages (A-D) on demographic characteristics (annual income, insurance type, and all rural, farming communities), dietary habits (eating porridge, rice, fish, and drinking a lot of high concentration of tea), similar geographic areas (all near river), and similar health status for participant students by excluding those with infectious diseases and bone metabolism disorders. In other words, all living conditions between the intervention villages and the control village were similar except for water sources. In this study, since the public water supply to all fluorosis endemic areas had similar concentrations of fluoride, and the fluoride levels in all intervention villages were reduced to levels similar to the non-fluorosis endemic areas, it is not possible to assess the dose - response relationship between water fluoride and the indicators. Nevertheless, we did use “the years of having the new water source” as an indicator to assess the health impact of duration of water intervention. In addition, as most previous studies in this area were conducted in China or other countries, limited English literature was available to compare with our findings. Finally, although this study demonstrated the success of the environmental intervention program, i.e., providing low-fluoride water supply in all endemic villages in Guangdong province, similar or even lower values of fluoride in some intervention villages compared to the control village or under the national health standards (1.0 mg/l) caught the attention of local environmental and public health agencies with respect to dental caries. Therefore, the possibility that the drinking water with very low concentration of fluoride could induce dental caries in these area and whether fluoride levels in the water supply of these villages should be modified, is worthy of the further study.