Cancer risk analysis and assessment of trihalomethanes in drinking water

(1)

O R I G I N A L P A P E R

Han-Keng Lee Æ Yir-Yarn Yeh Æ Wei-Ming Chen

Cancer risk analysis and assessment of trihalomethanes in drinking water

Published online: 22 February 2006 Springer-Verlag 2006

Abstract This study conducts risk assessment for an ar-ray of health effects that may result from exposure to disinfection by-products (DBPs). An analysis of the relationship between exposure and health-related out-comes is conducted. The trihalomethanes (THMs) spe-cies have been verified as the principal DBPs in the drinking water disinfection process. The data used in this study was collected from the Taiwan Water Cor-poration (TWC) from 1998 to 2002. Statistical analysis, multistage of Benchmark model, Monte Carlo simula-tion (MCS) and sensitive analysis were used to estimate the cancer risk analysis and assessment. This study in-cluded the statistical data analysis, epidemiology inves-tigation and cancer risk assessment of THMs species in drinking water in Taiwan. It is more significant to establish an assessment procedure for the decision making in policy of drinking water safety predomi-nantly.

Keywords DBPs Æ Monte Carlo simulation Æ Multistage of benchmark model Æ Sensitive analysis

1 Introduction

Chlorination has been the major, economical and eﬀective drinking water disinfection strategy from microorganisms. This disinfection process may induce serious waterborne infectious diseases dangerous to public health. Research consequence of Rook (1974) and

Bellar et al. (1974) exhibited that disinfection by-prod-ucts (DBPs) were produced in the disinfection process. Nowadays, such disinfection process is most adopted in drinking water treatment commonly (Houston 1913; Yang et al.1998; Hsu et al.2001).

Disinfection by-products are defined as hazardous materials with carcinogenic risk by Taiwan USEPA. Animal and epidemiology study evaluations have shown that developmental toxicity and adverse effects are the main potential risks to humans. The result from animal studies demonstrated evidence of liver, kidney, intestinal tumor genesis, urinary bladder, rectum and colon cancer (Morris et al.1992; Doyle et al.1997; Cantor et al.1998) and some associated effects of intrauterine growth and retardation (Kramer et al. 1992). Low birth weight, small for gestational age, central nervous system defects, oral cleft defects and cardiac defects (Bove et al.1995), retarded fetal growth (Gallagher et al. 1998) and spon-taneous abortion (Waller et al.1998) that are caused by disinfected water. Epidemiologic studies were conducted that examined the possible associations between con-sumption of chlorinated drinking water and cancer mortality, risk or incidence (Page et al. 1976; Cantor et al. 1978, 1987, 1998; Yang et al. 1998, 2000). The results suggest a positive association between consumption of chlorinating drinking water and cancer of the rectum, lung, bladder and kidney (Yang et al.

1996).

This study has been carried out from 1998–2002, in order to develop risk assessment and management for THMs species in drinking water for the purpose of preserving a safe environment and protecting human health in Taiwan. Risk assessment is a systematic, ana-lytical method used to determine the probability of ad-verse eﬀects. The purpose of this study conferred the risk assessment to process the derived THMs species in the drinking water of Taiwan. By following the estimation procedure of risk assessment, the outcomes will interpret the condition of the level of impact by THM species. The consequence may be a good decision-making process for risk management in the drinking water.

H.-K. Lee Æ W.-M. Chen

Department of Water Source Engineering,

100, Wnehwa Road, Seatwen, Taichung, 407, Taiwan, ROC Y.-Y. Yeh (&)

Graduate Institute of Civil and Hydraulic Engineering, 100, Wnehwa Road, Seatwen, Taichung,

407, Taiwan, ROC E-mail: [email protected] Tel.: +886-4-24517250 Fax: +886-4-24513797 DOI 10.1007/s00477-006-0039-4

(2)

2 Methods

The risk assessment paradigm was developed by the US NRC (National Research Council 1983) to evaluate the procedures of framework. It contained hazard identiﬁ-cation, dose-response assessment, exposure assessment and risk characterization, mainly.

2.1 THMs species data in Taiwan since 1998–2002 This study assumed and divided the geographical dis-tribution of Taiwan into ﬁve parts (northern, midland, southern, eastern and external islands). Oﬃcial Data obtained from Taiwan Water Corporation (TWC) since 1998 to 2002. There are 35, 45 and 13 water treatment plants and 25, 52 and 54 supply systems in Northern, Midland and Eastern regions. 30 water treatment plants and 45 supply systems in Southern and External islands regions, respectively. The moni-toring stations examined the temperature, pH per month and THMs species for three months. Four thousand nine hundred and forty water quality moni-toring data are obtained from those monimoni-toring sta-tions that was published in the annual TWC subscriber drinking water reports.

2.2 Cancer risk analysis and assessment 2.2.1 Hazard identiﬁcation of THMs species

Hazard identification involves a qualitative assessment of the presence of, and the degree of hazard that an agent could have on potential receptors. USEPA has developed a scheme that contains two broad categories of sufficient and insufficient evidence in Table1. Hossein (1995) defined THM species as TCM, BDCM, DBCM and TBM, respectively. Animal and epidemiology studies exhibited THMs species by considered weight of evidence in EPA reports on cancer guideline descriptions about Group B2 as TCM, BDCM, DBCM and Group C is TBM (USEPA1999), respectively.

2.2.2 Dose-response assessment

Dose-response relationships are then used to quantita-tively evaluate the toxicity information, and to charac-terize the relationship between dose of the contaminant administered or received and the incidence of adverse eﬀects on the exposed population. Specially, the purpose of the assessment developed for the risk management ensures the safety and oﬀers procedures to control the quality of drinking water.

Table 1 Basic processes involved in USEPA carcinogenesis

Group A Human carcinogen Suﬃcient human evidence for causal association between exposure and cancer Group B1 Probable human Limited evidence in humans

Group B2 Probable human Inadequate evidence in humans and suﬃcient evidence in animals Group C Possible human carcinogen Limited evidence in animals Group D Not classiﬁable as to

human carcinogenicity

Inadequate evidence in animals Group E No evidence of carcinogenicity

in humans

At least two adequate animal tests or both negative epidemiology and animal studies

Table 2 Animal experimental carcinogenic data derived from THMs species

IRIS (2003)

Chemicals Data set Data values Reference

Dose (mg/kg/day) N Incidence TCM Moderate or marked fatty

cysts in males plus females

0 27 1 Heywood et al. (1979) 15 15 9 30 15 13 25 50 1 50 50 8 BDCM B6C3F1 mice, male 0 46 1 NTP (1987) 25 49 2 50 50 9 DBCM Mouse/B6C3F1, female 0 50 6 NTP (1985) 50 49 10 100 50 19 TBM F344/N rat, female 0 50 0 NTP (1988) 25 50 1 50 50 8

(3)

In the dose-response assessment step, the goal deter-mined the relationships between the route, dose, fre-quency and duration of exposure conditions and the health that eﬀect chemical hazards. Additionally, apply the uncertainty or safety factors and mathematical model may an approach by USEPA.

Benchmark model (USEPA http://www.epa.gov/ ncea, 2003b) supported the assessment tool focused on the low-dose in the animal experiment that may cause an observed adverse inﬂuence. Generally, the adverse ef-fects included reproductive, developmental toxicity or mortality phenomenon etc.

In this study, we adopted the USEPA risk assessment guidance (1986) and the reference data (Table2) of animal studies from Integrated Risk Information System (IRIS, 2003) to process the BMD/BMDL value of THM species mathematically. BMD/DMDL value can inter-pret the toxicity information of THM species in the low-dose level.

2.2.3 Exposure assessment

Exposure is deﬁned as human contact with THM species through diﬀerent pathways. Referring to the exposure factor data handbook (USEPA 1997), USEPA risk

assessment guidance for superfund Volume-Human Health Evaluation Manual (USEPA 1989), and Risk Assessment Information System (RAIS2003a) assumed the pathways to reasonable maximum exposure (RME) to THM species as ingestion, inhalation and dermal intake, evaluated based on chronic daily intake (CDI).

Table 3 References data and formula for exposure assessment

Parameters Value Reference

Weight of population, C¼CiPi

Ptotal

Ci: concentration of i region

Pi: population of water supply in the i region

Ptotal: total population of water supply in the i region

Exposure pathway of ingestion, CDI¼ðCW0:8IREFEDÞ_ðATBWÞ Chronic daily intake (CDI) [mg (kg day1)]

THMs concentration of drinking water (CW)

Intake quantity (IR) 2.5 (L day1) Wu (1999)

Average exposure time (AT) 70 (year)· 365 (day/year) USEPA (1989)

Exposure during (ED) 70 (year) USEPA (1989)

Exposure frequency (EF) 365 (day year1) USEPA (1989)

Body weight (BW) Male: 64.8±10 (kg) Taiwan DOH

Female: 56.3±9.09 (kg) http://www.doh.gov.tw/statistic/index.htm

Absorptivity of body 100% Assumption

Exposure pathway of inhalation, CDI¼ðCairVREFETEDÞ

ðATBWÞ

THMs vapor concentration in the bathroom (Cair)

Mean vapor quantity of daily inhalation (adult) (VR) 12.3 (m3day1) Wu et al. (2003) Flow velocity (QL) 0.032 (L min1) Wu et al. (2003)

Air ﬂow velocity (QGS) 50 (L min1) Little (1992) Volume of bathroom (Vs) 6.6 (m3) Wu et al. (2003)

Henry constant (H) TCM: 0.150 RAIS (2003a)

BDCM: 0.087 DBCM: 0.032 TBM: 0.022 Transferred coeﬃcient of liquid mass· valid

air/surface area, KOLA

0.019 Little (1992)

Exposure pathway of dermal intake, CDI¼ðCWPCSAEFETEDÞ_ðATBWÞ

Dermal intake permeable coeﬃcient (PC) TCM: 8.9·103_{(cm h}1₎ _{USEPA (}₁₉₉₇₎

BDCM: 5.8·103_{(cm h}1₎

DBCM: 3.9·103_{(cm h}1₎

TBM: 2.6·103(cm h1)

Surface area dermal intake contact (SA) (4BW+7) (BW+90)1 USEPA (1997)

Exposure time (ET) 20 (min day1) MCKone (1989)

Northern Middle Southern Eastern External Island Colon _Rectum Bladder 0.00 0.05 0.10 0.15 Death/Population ( ‰ ) Locations Cancers

Fig. 1 Announced consequences in death of colon, rectum and bladder cancers by DOH of Taiwan

(4)

Table 4 The ranges of variables, means, maximum, minimum and standard deviations of subscriber Location Northern Midland Southern Eastern External island Yi-Lan aJi-Long Tai-Pei Tao-Yuan Shin-Chu Miao-Li Tai-Chung Nan-Tou Chang Hua Yun-Lin Chia-Yi Tai-Nan Kao-Chiong Ping-Dong Hua-Lian aTai-Dong Peng-Hu Temperature ( C) Mean 21.1 20.4 22.1 21.7 21.5 22.6 23.9 23.2 26.2 26.9 21.6 27.1 25.5 23.1 23.6 23.8 25.0 Max 24.0 28.5 30.5 27.5 28.0 27.0 31.5 30.5 29.0 31.8 30.0 31.0 31.0 29.0 31.0 29.0 29.5 Min 15.5 15.0 15.5 16.0 14.5 19.0 16.0 16.0 23.5 22.3 6.0 20.0 15.0 14.0 17.0 19.0 17.0 Stdev 2.2 3.5 3.6 3.5 3.5 2.2 2.2 2.7 1.2 2.2 6.5 3.3 2.6 3.5 3.2 2.6 4.0 PH Mean 7.5 7.1 7.2 7.5 7.6 6.7 6.8 8.0 7.4 7.6 7.9 8.3 7.6 7.2 7.8 7.8 8.2 Max 8.1 7.7 8.9 8.2 8.4 8.3 8.2 6.7 7.8 8.2 8.7 8.8 8.2 8.1 8.4 8.4 8.9 Min 6.5 6.4 6.3 6.3 7.0 5.2 5.6 6.0 6.3 6.4 6.7 8.0 6.9 5.3 7.1 7.1 7.4 Stdev 0.5 0.3 0.5 0.5 0.3 1.0 0.6 5.6 0.4 0.4 0.5 0.3 0.3 0.7 0.4 0.4 0.4 Number 39 45 88 30 50 54 123 170 52 116 59 11 169 78 56 62 45 Trichloromethane [(TCM) l g/L] Mean – 9.3 4.4 7.6 5.2 7.1 4.8 3.5 3.8 5.7 7.9 18.5 11.4 3.6 – 2.5 1.3 Max – 38.1 18.8 67.5 24.9 24.6 24.5 32.5 14.4 26.4 35.0 47.4 57.3 32.8 – 34.8 21.0 Min – 0.2 0.1 0.2 0.3 0.3 0.3 0.3 0.3 0.2 0.2 1.8 0.2 0.2 – 0.1 0.2 Stdev – 10.0 5.2 8.1 5.7 5.8 6.9 5.1 3.5 6.9 10.5 12.5 14.1 6.4 – 19.4 2.9 Dichlorobromomethane [(DCBM) l g/L] Mean – 6.9 2.6 2.1 0.3 0.5 0.5 0.2 0.2 2.2 1.5 6.2 4.2 1.3 – 1.9 3.4 Max – 27.1 13.6 17.4 1.0 4.2 9.1 1.1 1.1 12.7 5.8 17.5 26.5 9.6 – 29.3 21.0 Min – 0.8 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.1 1.0 0.2 0.2 – 0.1 0.2 Stdev – 6.6 2.6 2.3 0.3 1.0 1.3 0.2 0.3 2.7 1.7 3.9 5.2 2.4 – 16.4 5.1 Number 0 2 0 7 3 189 37 28 57 54 19 89 59 31 144 66 0 6 1 6 1 Dibromochloromethane [(DBCM) l g/L] Mean – 4.3 1.6 0.6 0.2 0.2 0.4 0.1 0.3 1.3 0.7 5.6 1.6 0.9 – 2.2 11.2 Max – 13.1 10.2 6.4 1.1 0.4 6.4 0.7 2.5 13.1 5.6 35.3 11.0 6.3 – 20.0 37.3 Min – 0.8 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.2 – 0.1 0.2 Stdev – 3.0 1.7 0.9 0.3 0.1 0.9 0.1 0.6 2.5 1.2 9.8 2.2 1.6 – 10.9 12.4 Tribromomethane [(TBM) l g/L] Mean – 0.7 0.3 0.4 0.2 0.1 0.1 0.1 0.4 2.6 0.4 3.4 0.4 0.7 – 1.0 24.8 Max – 2.6 1.6 8.2 2.1 0.4 0.7 0.4 5.5 37.4 2.5 26.8 1.5 24.5 – 41.3 65.1 Min – 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.2 – 0.0 0.4 Stdev – 0.8 0.3 1.1 0.4 0.1 0.1 0.1 1.2 6.5 0.5 7.0 0.3 3.0 – 23.6 17.7 Total Trichloromethane [(TTHM) l g/L] Mean – 21.2 8.9 10.8 5.8 7.9 5.8 3.9 4.7 11.9 10.5 33.7 17.5 6.5 – 7.6 40.6 Max – 80.5 32.1 86.0 25.5 28.7 28.2 33.1 17.7 71.9 41.3 84.0 96.2 49.0 – 66.7 107.2 Min – 2.0 0.2 0.8 0.4 0.6 0.4 0.4 0.8 0.7 0.7 6.8 0.7 0.7 – 0.2 0.9 Stdev – 19.1 7.9 10.3 5.7 6.7 7.9 5.1 4.0 14.5 11.7 17.9 20.4 10.5 – 36.4 29.9 Number 0 2 0 7 3 189 37 28 57 54 19 89 59 31 144 66 0 6 1 6 1 Mean average value, Max maximum value, Min minimum value, Stdev standard deviation, Number sam pling number aYi-Lan and Hua -Lian lacked TTHMs data

(5)

For many drinking water DBPs, the potential for exposure and uptake occurs by ingestion but also through dermal absorption or inhalation (Lee et al.

2004). Since 1990, scientists proposed that inhalation and dermal absorption were considered in the risk assessment of drinking water (Jo et al. 1990; Maxwell et al.1991; Weisel and Jo1996; Weisel et al. 1999; Lin and Hoang2000).

Moreover, the study assumed the behaviors of ‘‘drinking water’’, ‘‘take a shower’’, and ‘‘skin contact in the shower’’ represent the exposure pathways of inges-tion, inhalation and dermal intake simplistically (Dan

2003; Chen 2003). The mathematic model by Little (1992) exhibited THM species concentration within the air was influenced by many parameters adopted to evaluate the concentration in the bathroom. All expo-sure pathway formulas and parameters are displayed in Table3. We adopted the weight of population pattern to calculate the concentration of THM species in the five regions of Taiwan. It is more reasonable to interpret the weight of population in the exposure concentration le-vel. Utilized the concentration, it can obtain the CDI values from different exposure routes, respectively. 2.2.4 Risk characterization

In this step, the hazardous identiﬁcation, dose-response and exposure assessment procedures are summarized and integrated into quantitative and qualitative expres-sions of the risk level. For carcinogenic eﬀects, the risk is expressed as the probability that an individual will ex-hibit dose-response characteristics. Under the assump-tion that the slope factor is a constant, the risk related to the intake pathways in this study directly.

Linear low-dose cancer risk equation Risk ¼ CDI SF;

where

Risk a unitless of an individual developing cancer, CDI chronic daily intake averaged over 70 years

[mg (kg day1)],

SF slope factor, expressed in milligram (kg day1).

Estimating the risk or hazard potential requires a combination of simultaneous exposures to more than one pathway and carcinogenic eﬀect. In this paper we assumed THM species dose are additivity. And there are no synergistic or antagonistic interactions. Equally, the total cancer risk assumes that all carcinogens are equal, and the slope factors derived from the animal data are given the same weights as factors derived from the hu-man data. It can express into below:

Total exposure cancer risk¼ Riskðexposure pathway 1Þ

þ Risk ðexposure pathway 2Þ þ . . . þ Risk ðexposure pathway iÞ

Table 5 Physical–chemical properties of total THMs species Chemicals CAS No. a Physical–chemical property a Ha rmful-ness Slope factor [mg (kg day 1 )] Benchmark Model d MW SG D BP CP VP VPD Diss. R CI b Rf D c Ingestion Inhalation Dermal BMD BMDL Stage TCM 067-6 6-3 119.4 1.485 1.484 61.2 63.5 159.6 4.12 0.80 1.4422 B2 0.01 6.10 · 10 3 (IRIS) 3.05 · 10 2 (IRIS) 8.05 · 10 2 0.41–3.58 (IRIS) 0.49 3.58 0.28 2.44 0.28 2.44 First second BDCM 075-2 7-4 163.8 – 1.971 90.1 57.1 – – 0.67 1.4953 B2 0.02 6.20 · 10 2 (IRIS) 6.33 · 10 2 (RAIS) 6.20 · 10 2 (Dan) 1.17 3.48 0.55 4.84 0.38 2.67 0.32 2.86 First second DBCM 124-4 8-1 208.3 – 2.451 120 22 – – 0.40 1.5465 C 0.02 8.40 · 10 2 (IRIS) 1.40 · 10 1 (RAIS) 8.40 · 10 2 (Dan) 1.23 4.43 0.45 3.99 0.30 2.65 0.29 2.54 First second TBM 075-2 5-2 252.8 2.089 – 151.2 9 – – – 1.6005 B2 0.02 7.90 · 10 2 (RAIS) 1.32 · 10 2 (IRIS) 6.10 · 10 3 (IRIS) 8.36 24.8 3.21 36.4 3.21 19.3 2.50 22.0 First second aMSDS, bUSEPA, cUSEPA 2003a , dBMR (Benchmark response, set BMR=2.5, 5, 10 and 20% in the confidence 95% at one or two stage degree) MW Molecular weight, SG Specific gravity (25/4 C), D Density (20 C), BP Boiling point (1 atm, C), CP Congeal point (1 atm, C), VP Vapor pressure (20 C, mmHg), VPD Vapor pressure density, Diss Dissolution (20 C, g/100 mL H2 O), R Refraction (25 C), CI Carcinogenetic identification, Rf D Reference dose (mg/kg/day), BMD Benchmark dose (mg/kg/day), BMDL Lower-bound confi dence limit on BMD (mg/kg/day)

(6)

2.2.5 Uncertainty and sensitivity analysis

There are several types of uncertainty parameters. An important task in risk analysis is to determine what kinds of uncertainty are likely to aﬀect the MCS ﬁnding sug-gested by USEPA (1997) in processing the uncertainty and sensitivity analysis. Essentially, MCS involves conducting and comparing repeated inputs that sample the system parameter distributions. This study utilized @ Risk view (version 4.5) software to execute the data probability distribution and simulate the sensitivity using MCS.

3 Results and discussion

3.1 Investigated result of epidemiologic studies

The consequence of epidemiologic studies that inhibited several cancers caused by THM species are colon,

rectum and bladder cancers, respectively. Annual re-ports from the Department of Health, Taiwan (DOH http://www.doh.gov.tw/statistic/index.htm) announced the mean numbers for these cancers, which were dis-criminated by location from 1996 to 2000. Fig.1

exhibited the ratio of death count versus water supply population (TWC http://www.water.gov.tw/sample1/ about/data1.asp#3). The investigation results exhibited Northern region has a higher death count (colon (1,535), rectum (1,188) and bladder (566)), but Southern region displays higher ratio (colon (0.15&), rectum (0.14&) and bladder (0.09&)) in these cancers evidently.

3.2 Statistic analysis of THMs species data

Statistical results in Table4 exhibit the means, maxi-mum, minimum and standard deviations values in

sub-Fig. 2 BMD/BMDL values from a TCM, b BDCM, c DBCM and d TBM data by first stage multistage model fit with 95% confidence limits

(7)

scriber levels. The TCM level is the predominant deri-vation of THMs in Taiwan. The Northern and Southern regions presented higher mean concentrations, shown by the epidemiology study investigation of Taiwan DOH.

In the BDCM, DBCM and TBM levels are pre-dominant in external island. Previous research (Garcia-Villanova et al.1997; Golﬁnopoulos et al.1996) veriﬁed groundwater and seawater contain bromide compounds if the water sources are near the seacoast. It conformed to the situation of external island of Taiwan.

3.3 Hazard identiﬁcation

A number of epidemiological studies were performed to investigate adverse eﬀects in human exposed to TCM, BDCM, DBCM and TBM, respectively.

Table5 collates the physical–chemical properties, harmfulness, slope factors and quantity of Benchmark dose (BMD) of THMs species completely. Obviously, evidences of animal study revealed THMs species may carcinogenic hazardous materials.

3.4 Dose-response assessment

This study adopted a multistage type of benchmark model approved by USEPA (2003b) to process the dose-response assessment. The chronic toxicity and carcino-genic potential of total THMs species at low dose situa-tion were interpreted. Figure 2shows the BMD/BMDL value for the total THMs species calculated from the Benchmark model from animal data (95% confidence limits and first stage model fit). Furthermore, the range

(8)

of BMD/BMDL value in first/second multistage model had shown in Table 5. Generally, the BMD values of first are higher than second stage, but the BMDL values are similar between first and second stage model. USEPA proposed the standard BMD and BMDL values of TCM is 1.69, 1.15; BDCM is 2.28, 1.35; DBCM is 1.88, 1.20; and BMD is 17.6, 10.3 mg/kg day1, respectively.

3.5 Exposure assessment

THM species data processed the weighted average method via the regional population and speciﬁc year to acquire the statistical analysis. The MCS method was used to evaluate the CDI dose and obtain the proba-bility distribution results by performing 1,000 frequency calculations.

Figure3 shows the MCS consequence exhibited in the ingestion pathway, southern region exist the higher CDI values, in opposite to female examined higher CDI endured than male (CDI ranges are 1.07·104 to 1.63·103 and 9.22·105 to 1.42·103mg (kg day1), respectively). The variance in body weight between females and males is the main reason for the CDI ingestion level. Moreover, in the respired estimated, consequence exhibited in ingestion pathway, external island exist higher CDI values. Similarly, females exhibited higher levels than males (CDI range is 7.28·106to 1.42·103 mg (kg day1) and 4.29·105to

4.96·103mg (kg day1), respectively) because of their shower behavior. This included the diﬀerence between municipal and rural areas and gender. In the estimated skin contact, the dermal intake pathway was similar to the ingestion pathway. The Southern region exhibited higher CDI values and females were higher than males (CDI ranges are 2.55·107to 2.95·105 and 2.42·107 to 2.81·105 mg (kg day1), respectively). In the con-tact time during showers, females took longer showers than males. The Taiwan DOH announced and suggested that showers not exceed 12 minutes, otherwise, the health risk will increase.

3.6 Risk characterization

The THM species slope factor from the dose-response curve showed a low dose situation from the linear model (Table5). Assumption the cancer risk is CDI multipli-cation slope factor and total cancer risks include dif-ferent exposure pathways. The total cancer risk assessment order was southern, northern, central over the external island segment. The average value for fe-males was 4.04·106 _{to 4.67·10}4 _{and 9.25·10}6 _to 4.07·104_{for men, respectively. Figures}₄_and ₅ exhib-ited the THM species distribution via diﬀerent exposure pathways to estimate the CDI value, risk assessment and contribution percentage simultaneously. The cancer risk CDI x 10-6 (mg/kg/day) P ro b a b ility x 1 0 2 Risk x 10-6 Probability x 10 4

Fig. 3 MCS results of CDI values in the ingestion, inhalation, dermal intake exposure pathways and total risk probability distribution

(9)

quantitative analysis evaluation for Taiwan (Arc View plot) is shown in Fig.6. It is shown that the Southern Taiwan region presents a higher risk.

In term of the inhalation pathway distinct revealed the magniﬁcent in the risk assessment of THMs species, continuous is ingestion and dermal intake pathways, respectively. TCM is the main contribution to the risk assessment in Taiwan (50% approximately), and TBM is predominance in external island (50% approximately).

3.7 Sensitivity analysis

The sensitivity analysis processed the ±20% extra risk to interpret the eﬀective THM species parameters, including body weight, intake quantity and exposure duration in formula of exposure assessment. Analysis was performed using the radar plots exhibited in Fig.7. The research regions displayed a negative correlation consistent with the exposure duration and positive cor-relation in body weight dramatically. Furthermore, the TCM concentration is the predominant inﬂuence parameter in Taiwan, whereas the external islands are

inﬂuenced by the DBCM concentration shown in Table6.

4 Conclusions

In the mean concentration distribution of total THMs in Taiwan external islands (48.39 lg L1), southern (17.28 lg L1), northern (12.11 lg L1) and middle segments (9.59 lg L1). By investigation consequences, the TCM concentration is the major DBP species in the local regions of Taiwan, and the external islands is characterized by TBMs, respectively.

A multistage Benchmark model (USEPA 2004) was used to evaluate the dose-response assessment. Conse-quence exhibited at the 95% conﬁdence level, the BMD and the quantity of lower-bound conﬁdence limit for the BMD (BMDL) of TCM were 1.69 and 1.15 mg (kg day1), dibromochloromethane (DBCM) are 1.88 and 1.20 mg (kg day1), dichloromethane (DCBM) are 2.28 and 1.35 mg (kg day1) and TBM are 17.6 and 10.3 mg (kg day1), respectively. The exposure was compared with the reaction dose concentration of 0 1000 2000 3000 4000 5000 6000 JL TP TY SC ML TC NT CH YL CY TN KC PD PH Area CD I 1 0 -6(m g/kg-day ) Ingestion Inhalation Dermal (a) 0 50 100 150 200 250 300 350 400 450 JL TP TY SC ML TC NT CH YL CY TN KC PD PH Area Risk 10 -6 Dermal Ingestion Inhalation (b)

Fig. 4 Special distribution via diﬀerent exposure pathways in CDI (a) and risk (b) assessment

(10)

THMs species in drinking water. In terms of lower quantity of BMD/BMDL showed doses opposite to the cancer risk obviously.

In the exposure assessment calculated by MCS, inhalation was found as the principal pathway. The next pathway was ingestion followed by dermal intake. The quantity of average risk in male and female is 3.14·105 to 1.04·104 and 3.64·105 to 1.16·104 in northern, 9.25·106 to 7.25·105 and 4.04·106to 8.40·105in middle, 1.14·104 to 4.07·104 and 1.33·104 to 4.67·104in southern, and 6.07·105and 7.09·105in external island, respectively. The Southern region pre-sented a high cancer risk and corresponded with the

result of epidemiology. Furthermore, females presented higher CDI values (intention risk level) than males in Taiwan.

Consequence of sensitivity analysis exhibited body weight and exposure duration are provided the inﬂuence in cancer risk analysis and assessment predominantly. Exposure time and body weight are the eﬀective parameters used in the sensitivity analysis. The greater the exposure time, the greater the cancer risk endured. A negative correlation exists between body weight and the unit dose sustained risk probability.

Quantifying the risk factors is important for popu-lation and decision-making policy for drinking water

Fig. 5 Diﬀerent contribution percentage of THMs species and exposure pathways in cancer risk assessment [(a) is male and (b) is female]

(11)

Fig. 6 Quantity of cancer risk assessment of Taiwan (Arc View plot)

Table 6 Cancer risk assessment sensitivity analysis

Numbers exhibited are the extent of sensitivity, and 1 is the most sensitivity

BWBody weight, IR intake quantity, ET exposure time

Regions Parameters TCM BDCM DBCM TBM BW IR ET Northern Ji-Long 3 4 5 7 1 6 2 Tai-Pei 3 4 5 6 1 5 2 Tao-Yuan 3 4 6 6 1 5 2 Shin-Chu 3 4 5 6 1 5 2 Middle Miao-Li 3 4 6 7 1 5 2 Tai-Chung 3 4 5 6 1 5 2 Nan-Tou 3 4 6 7 1 5 2 Chang-Hua 3 6 4 7 1 5 2 Yun-Lin 3 4 5 7 1 6 2 Southern Chia-Yi 3 4 6 7 1 5 2 Tai-Nan 3 4 5 7 1 6 2 Kao-Chiong 3 5 4 7 1 6 2 Ping-Dong 3 4 5 7 1 6 2

External island Peng-Hu 7 6 3 5 1 4 2

Table 7 Legislation limit values for diﬀerent counties in DBPs level

Chemicals Taiwan USA WHO Japan Sweden Australia

TCM – – 0.20 0.06 – –

DBCM – – 0.10 0.03 – –

DCBM – – 0.06 0.10 – –

TBM – – 0.10 0.09 – –

Total THMs species 0.10 Stage one: reduced to 0.08; Stage two: reduced to 0.04 1.0 (mg/L)a 0.10 0.05 0.25

Trichloroacetic acid – – 0.10 0.3 – 0.10

Dichloroacetic acid – – 0.05 0.04 – 0.10

HAAs – Stage one: reduced to 0.06; Stage two: reduced to 0.03 – – – 0.15

a

(12)

safety. Fortunately, the Benchmark model and MCS @Risk supply the methodology were used for risk cal-culation. The standard for the total THMs species in Taiwan was 100 ppb presently. Table7 displays the legislation limit values for diﬀerent countries for DBPs levels. We suggest that the standard be separated using separate TCM, BDCM, DBCM and TBM standards. This may establish a control management for individual material to reduce the harmful risk.

Acknowledgements The authors gratefully acknowledge the fund support by National Science Council (NSC-93-2621-Z-035-002)and the assistance of the Taiwan Water Corporation in this study.

References

Bellar TA, Lichtenberg JJ, Kroner RC (1974) The occurrence of organohalides in chlorinated drinking waters. J Am Water Works Assoc 66:703–706

Bove FJ, Fulcomer MC, Klotz JB, Esmart J, Duﬃcy EM, Savrin JE (1995) Public drinking water contamination birth outcome. Am J Epidemiol 141:850–862

Cantor KP, Hoover R, Mason TJ, McCabe LJ (1978) Association of cancer mortality with halomethanes in drinking water. J Natl Cancer Inst 61:979–985

Cantor KP, Hoover R, Hartge P (1987) Bladder cancer, drinking water sources, and tap water consumption: a case control study. J Natl Cancer Inst 79:1269–1279

Cantor KP, Lynch CF, Hildesheim ME, Dosemeci M, Lubin J, Alavanja M, Craun G (1998) Drinking water source and chlorination byproducts I Risk of bladder cancer. Epidemiol-ogy 9:21–28

Chen WM (2003) Health risk analysis and assessment of trihalo-methanes in drinking water of Taiwan, Master Thesis, Feng Chia University

Dan YJ (2003) Formation and risk assessment of trihalomethanes in drinking water, Master Thesis, National Taiwan University Doyle TJ, Zheng W, Cerhan JR, Hong CP, Sellers TA, Kushi LH,

Polsom AR (1997) The association of drinking water source and chlorination by-products with cancer incidence among postmenopausal women in Iowa: a prospective cohort study. Am J Public Health 87:1168–1176

Garcia-Villanova RJ, Garcia C, Gomez JA, Garcia MP, Ardanny R (1997) Formation, evolution and modeling of trihalome-thanes in the drinking water of a town: at the muncipal treat-ment utilities. Water Res 31:1299–1308

Gallagher MD, Nuckols JR, Stallones L, Savitz DA (1998) Exposure to trihalomethanes and adverse pregnancy outcomes. Epidemiology 9:484–489

Golﬁnopoulos SK, Kostopoulou MN, Lekkas TD (1996) THMs formation in the high bromide water supply of Athens. J Environ Sci Health A 31:67–81

Heywood R, Sortwell RJ, Noel PRB (1979) Safety evaluation of toothpaste containing chloroform: III. Long-term study in beagle dogs. J Environ Pathol Toxicol 2:835–851

Hossein P, Stevens AA (1995) Relationship between trihalome-thanes and haloacetic acids with total organic halogen during chlorination 29:2059–2062

Houston AC (1913) Studies in water supply. Macmillan & Co, London

Hsu CH, Jeng WL, Chang RM, Chien LC, Han BC (2001) Esti-mation of potential lifetime cancer risk for trihalomethanes from consuming chlorinated drink water in Taiwan. Environ Res 85:77–82

Jo WK, Weisel CP, Lioy PJ (1990) Routes of chloroform exposures and body burden from showering with chlorinated tap water. Risk Anal 10:575–580

Kramer MD, Lynch CF, Isacson P, Hanson JW (1992) The asso-ciation of waterborne chloroform with intrauterine growth retardation. Epidemiology 3:407–413

Lee SC, Guo H, Lam SMJ, Lau SLA (2004) Multipathway risk assessment on disinfection by-products of drinking water in Hong Kong. 94:47–56

Lin TF, Hoang SW (2000) Inhalation exposure to THMs from drinking water in south Taiwan. Sci Total Environ 246:41–49 Little, John C (1992) Applying the two-resistance theory to

con-taminant volatilization in showers. Environ Sci Technol 26:1341–1349

Maxwell NI, Burmaster DE, Ozonoﬀ D (1991) Trihalomethanes andmaximum contaminant levels: the signiﬁcance of inhalation and dermal exposures to chloroform in household water. Regul Toxicol Pharmacol 14:297–312

MCKone TE (1989) Household exposure models. Toxicol Lett 49:321–339

Morris RD, Audet AM, Angelillo IF, Chalmers TC, Mosteller F (1992) Chlorination, chlorination by-products, and cancer: a meta-analysis. Am J Public Health 82:955–963

National Research Council (1983) Risk assessment in the federal government: managing the process. NAS Press, Washington

-5 -4 -3 -2 -1 0 1 2 3 4 5 TCM BDCM DBCM TBM BW IR ET (a) -5 -4 -3 -2 -1 0 1 2 3 4 5 TCM BDCM DBCM TBM BW IR ET (b)

Fig. 7 Radar plots for sensitivity analysis in Taiwan area (a) and external island (b)

(13)

National Toxicology Program (NTP) (1985) Toxicology and car-cinogenesis studies of chlorodibromomethane (CAS. No. 124– 48–1) in F344/N rats and B6C3F1 mice (gavage studies). Na-tional Toxicology Program Technical Report Series No. 282. DHHS Publications No. (NIH), 85–2538

National Toxicology Program (NTP) (1987) Toxicology and car-cinogenesis studies of bromodichloromethane (CAS. No. 75– 27–4) in F344/N rats and B6C3F1 mice (gavage studies). Na-tional Toxicology Program Technical Report Series No. 321. DHHS Publications No. (NIH), 85–2537

National Toxicology Program (NTP) (1988) Toxicology and carci-nogenesis studies of bromoform (CAS. No. 75–25–2) in F344/N Page T, Harris RH, Epstein SS (1976) Drinking water and cancer

mortality in Louisiana. Science 193:55–57

Rook JJ (1974) Formation of haloforms during chlorination of natural water. Water Treat Exam 23:234–243

U.S. Environmental Protection Agency (USEPA) (2004) The Benchmark Dose Software 1.3.2. Available on the Internet at: http://www.epa.gov/ncea

U.S. Environmental Protection Agency (USEPA) (1989) Risk assessment guidance for superfund. USEPA, Washington DC U.S. Environmental Protection Agency (USEPA) (1997) Exposure

factors handbook. USEPA, Washington DC

USEPA (1986) Guidelines for carcinogen risk assessment. U.S. Environmental Protection Agency, Washington DC. EPA/600/ 8–87/045

USEPA (1999) Guidelines for carcinogen risk assessment. Risk Assessment Forum, U.S. Environmental Protection Agency, Washington DC. NCEA-F-0644

USEPA (2002) Integrated risk information system (electronic data base). U.S. Environmental Protection Agency, Washington DC. Available on Internet at: http://www.epa.gov/iris USEPA (2003a) Risk assessment information system (electronic

data base). U.S. Environmental Protection Agency, Washing-ton DC. Available on Internet at: http://www.cfpub.epa.gov/ iris Cited Aug 2003.

USEPA (2003b) The Benchmark Dose Software 1.3.2. EPA NCEA. Available on Internet at:http://www.epa.gov/ncea Waller K, Swan SH, DeLorenze G, Hopkins B (1998)

Trihalome-thanes in drinking water and spontaneous abortion. Epidemi-ology 9:134–140

Wang KS, Dan YJ (2003) Health risk assessment and distribution of trihalomethanes in drinking water of Taiwan. In: Twenty-ﬁrst conference, Taiwan Water Cooperation

Weisel CP, Jo WK (1996) Ingestion, inhalation and dermal expo-sure to chloroform andtrichlor oethene from tap water. Environ Health Perspect 104:48–51

Weisel CP, Kim H, Haltmeier P, Klotz JB (1999) Exposure esti-mates to disinfection by-products of chlorinated drinking water. Environ Health Perspect 107:103–110

Wu KY (1999) Applicated toxicological mechanism in the risk assessment (in Chinese). Taiwan National Science Council re-port (NSC 89-2621-Z-039-001)

Wu KY, Chen MJ, Chang L (2003) A new approach to estimate the volatilization rates of volatile organic compounds during showering. Atmos Environ 37:4325–4333

Yang CY, Chiu JF, Chiu HF, Wang TN, Lee CH, Ko YC (1996) Relationship between water hardness and coronary mortality in Taiwan. J Toxicol Environ Health 49:1–9

Yang CY, Chiu HF, Cheng MF, Tsai SS (1998) Chlorination of drinking water and cancer mortality in Taiwan. Environ Res 78:1–6

Yang CY, Cheng BH, Tsai SS, Wu TN, Lin MC, Lin KC (2000) Association between chlorination of drinking water and adverse pregnancy outcome in Taiwan. Environ Health Perspect 108:765–768