Comparison of drinking water, raw rice and cooking of rice as arsenic exposure routes in three contrasting areas of West Bengal, India

Page 1

ORIGINAL PAPER

Comparison of drinking water, raw rice and cooking of riceas arsenic exposure routes in three contrasting areas of WestBengal, India

Debapriya Mondal • Mayukh Banerjee • Manjari Kundu •

Nilanjana Banerjee • Udayan Bhattacharya • Ashok K. Giri •

Bhaswati Ganguli • Sugata Sen Roy • David A. Polya

Received: 31 August 2009 / Accepted: 26 February 2010 / Published online: 27 May 2010

� Springer Science+Business Media B.V. 2010

Abstract Remediation aimed at reducing human

exposure to groundwater arsenic in West Bengal, one

of the regions most impacted by this environmental

hazard, are currently largely focussed on reducing

arsenic in drinking water. Rice and cooking of rice,

however, have also been identified as important or

potentially important exposure routes. Quantifying the

relative importance of these exposure routes is criti-

cally required to inform the prioritisation and selection

of remediation strategies. The aim of our study,

therefore, was to determine the relative contributions

of drinking water, rice and cooking of rice to human

exposure in three contrasting areas of West Bengal

with different overall levels of exposure to arsenic, viz.

high (Bhawangola-I Block, Murshidibad District),

moderate (Chakdha Block, Nadia District) and low

(Khejuri-I Block, Midnapur District). Arsenic expo-

sure from water was highly variable, median exposures

being 0.02 lg/kg/d (Midnapur), 0.77 lg/kg/d (Nadia)

and 2.03 lg/kg/d (Murshidabad). In contrast arsenic

exposure from cooked rice was relatively uniform,

with median exposures being 0.30 lg/kg/d (Midna-

pur), 0.50 lg/kg/d (Nadia) and 0.84 lg/kg/d (Mursh-

idabad). Cooking rice typically resulted in arsenic

exposures of lower magnitude, indeed in Midnapur,

median exposure from cooking was slightly negative.

Water was the dominant route of exposure in Mursh-

idabad, both water and rice were major exposure routes

in Nadia, whereas rice was the dominant exposure

route in Midnapur. Notwithstanding the differences in

balance of exposure routes, median excess lifetime

cancer risk for all the blocks were found to exceed the

USEPA regulatory threshold target cancer risk level of

10-4–10-6. The difference in balance of exposure

routes indicate a difference in balance of remediation

approaches in the three districts.

Keywords Arsenic � Rice � Exposure routes �Probabilistic risk assessment � West Bengal

Introduction

It is well known that groundwater arsenic hazard

represents a serious threat to human populations in

many countries. Contaminated drinking water supplies

D. Mondal � M. Banerjee � M. Kundu �N. Banerjee � U. Bhattacharya � D. A. Polya (&)

School of Earth Atmospheric and Environmental

Sciences, University of Manchester, Manchester M13

9PL, UK

e-mail: [email protected]

D. Mondal � M. Banerjee � M. Kundu �N. Banerjee � U. Bhattacharya � A. K. Giri

Molecular and Human Genetics Division, Indian Institute

of Chemical Biology, 4 Raja S.C. Mullick Road, Kolkata,

West Bengal 700 032, India

B. Ganguli � S. Sen Roy

Department of Statistics, University of Calcutta,

35 Bullygunge Circular Road, Kolkata, West Bengal

700 019, India

123

Environ Geochem Health (2010) 32:463–477

DOI 10.1007/s10653-010-9319-5

Page 2

are reported from over 70 countries, posing a serious

health hazard to an estimated 150 million people

worldwide (Polya and Charlet 2009; Ravenscroft et al.

2009).

Though drinking water is widely considered to be

the main exposure route, recent studies have reported

significant exposure from rice intake for the exposed

population, especially in southern and south-eastern

Asia, which in some cases exceeds that from drinking

water (Ohno et al. 2007; Roychowdhury et al. 2002;

Signes et al. 2008a; Williams et al. 2006). Cooking of

rice with arsenic contaminated water may lead to an

increase in the arsenic content of rice (Ackerman

et al. 2005; Bae et al. 2002; Laparra et al. 2005;

Signes et al. 2008b), whereas cooking with uncon-

taminated water and a large excess water discarding

the gruel may lead to a decrease in arsenic content

(Pal et al. 2009; Raab et al. 2009; Rahman et al. 2006;

Sengupta et al. 2006).

However, the importance of exposure routes other

than drinking water for arsenic is still being dis-

cussed. For example, van Geen and Duxbury (2009)

and van Geen et al. (2006) note that for Bangladesh,

where many people still continue to drink ground-

water with arsenic [100 lg/L, exposure from drink-

ing water is more important than from eating rice,

even when large quantities of rice are consumed, and

emphasized the importance of focusing remediation

efforts on improving drinking water quality. How-

ever, Mondal and Polya (2008), note that not only is

rice the most important exposure route for some areas

of West Bengal, but also that its relative importance

will increase as effective drinking water remediation

technologies are put in place and as the arsenic

content of arsenic-bearing groundwater irrigated

paddy fields increases over time (Dittmar et al.

2007; Roberts et al. 2007). Quantifying the relative

importance of drinking water, rice and cooking of

rice as exposure routes is clearly important to

selecting the focus of arsenic remediation measures

in impacted countries.

In West Bengal, millions of people are currently at

risk from arsenic contaminated water, with estimates

as high as 3.9 million (Nickson et al. 2007), 26 million

(Chakraborti et al. 2009) and 28.6 million (PHED

2008). Differences in these estimates may reflect, in

part, changing exposure due to mitigation efforts and

also due to difference in the areas covered by various

surveys. The aim of our study was to estimate the

contribution towards total exposure from different

exposure routes—drinking water, rice and cooking of

rice—using a probabilistic model for three contrasting

study areas of West Bengal, viz.

1. The Bhawangola-I block of Murshidabad district,

having high arsenic in groundwater (Rahman

et al. 2005) with 61.2% of the analysed samples

by SOES exceeding the WHO provisional value

of 10 lg/L and 30.6% exceeding the Indian

permissible limit of 50 lg/L (SOES 2007)

2. The Chakdha block of Nadia district having

reduced arsenic in drinking water with 44% of

the analysed samples by Mondal and Polya

(2008) exceeding 10 lg/L and 11% exceeding

50 lg/L

3. The Khejuri-I block of Midnapur district, an area

of lower exposure with groundwater arsenic

mostly less than 10 lg/L (Pal et al. 2009).

Finally, the excess lifetime cancer risk attributable

to arsenic considering the total exposure from

drinking water, rice and cooking of rice was also

calculated. The detrimental health outcomes of

arsenic are numerous and the dose-response relation-

ships for many sequela are poorly known (Adamson

and Polya 2007; Mondal et al. 2008). A comprehen-

sive review of the epidemiology of arsenic-attribut-

able diseases is beyond the scope of this paper.

Hence, we have chosen instead to report cancer-

related risks calculated by a simple model—the

probabilistic risk model based on the USEPA one-

hit model (USEPA 1989)—as a guide figure to

indicate the relative importance of various arsenic

exposures routes.

Materials and methods

Sampling strategy

In each of the three field areas, we selected approx-

imately 50–100 adult volunteers, selected largely on

the basis of practicalities of sample collection as

detailed below. No attempt was made to implement a

completely objective randomised sampling strategy

as the aim of this study was not to determine overall

groundwater arsenic attributable risks (cf. Mondal

and Polya 2008), but rather to determine the relative

importance of various exposure routes.

464 Environ Geochem Health (2010) 32:463–477

123

Page 3

For both the Bhawangola-I and Khejuri-I blocks,

medical camps were organized by one of the authors

(AKG, IICB) as part of another project on a pre-

selected date and time, and villagers were informed in

advance of the camp. A total of 53 volunteers from

Bhawangola-I block and 68 volunteers from Khejuri-I

block were interviewed and respective samples col-

lected. In Chakdha block, seven villages were selected

from a register of all rural villages for which ground-

water arsenic data already existed (PHED/UNICEF;

R. Nickson, personal communication), on the basis of

obtaining a wide range of drinking water arsenic

concentrations. Approximately five to ten households,

mostly from which both male and female volunteers

were available, were then surveyed in each village.

From each household drinking water, cooking water,

raw rice and the corresponding cooked rice samples

were collected. Use of different cooking water rather

than using the same source as the drinking water was

hardly observed (observed only in 10% of the surveyed

households from Chakdha block).

Data and sample collection

Age, gender, and rice and water consumption data

were obtained by questionnaire from volunteers in

Bhawangola-I block (n = 53) from March to July

2008, in Chakdha Block (n = 111) from August to

September 2008 and in Khejuri-I block (n = 68)

during September 2008. Questionnaires for Bhawan-

gola-I and Khejuri-I were administered by IICB field

staff, those for Chakdha by DM.

To better quantify the daily water intake by direct

drinking, a method similar to the water diary method

(Ohno et al. 2007; Watanabe et al. 2004) was used, in

which direct water intake was estimated by asking the

volunteer how many units of water are consumed in a

day from the container they normally use to drink

water. At the first visit, the container, which

happened to be a glass or bottle, used for drinking

water was identified and the capacity of the container

was measured. Each volunteer was provided with a

recording sheet to self-record the number of units

drunk in 24 h by marking the sheet every time they

drank water from their own container. About 24 h

later, the sheets were collected, and the number of

units marked was multiplied by the capacity of that

volunteer’s container to estimate the water consump-

tion rate. This data was obtained for 51 of 53

volunteers (96%) from Bhawangola-I block, 104 of

the 111 volunteers (93%) from Chakdha block and 68

volunteers (100%) from Khejuri-I block.

Raw rice, cooked rice, cooking water and drinking

water samples were collected from each household

surveyed at the same time. After collection, the rice

samples were placed individually in polythene bags.

Raw rice samples were stored at room temperature

while cooked rice samples were oven dried at 65�C for

48 h and finally stored in polythene bags before

shipping to the University of Manchester for analysis.

The unfiltered drinking and cooking water samples

were transferred to acid cleaned translucent HDPE

bottles followed by acidification to 1% v/v HNO3

(Aristar grade) and then heated to 55�C for 1 h to

deactivate any polio wild virus, which is still endemic

in parts of India, and finally stored at room temperature

until transported to the University of Manchester.

Sample analysis

Drinking water, cooking water (if different from

drinking water), raw and cooked rice samples col-

lected from all three areas were analysed for total

arsenic. Ground sub samples of rice (0.1–0.2 g)

weighed into Pyrex glass digestion tubes were added

with 2 ml of concentrated sub-distilled HNO3 (Aristar

grade) and pre-digested overnight at room tempera-

ture. Samples were then heated to 120�C on a heating

block, until clear, and then evaporated to dryness and

the residue was dissolved in deionised water (18 MX)

to a weight of 10 g before filtering it through 0.45-lm

nylon membrane syringe filters (VWR).

The total arsenic in drinking water samples were

analysed by inductively coupled plasma optical

emission spectrometry (ICP-OES) (Perkin Elmer

Optima 5300, DV-Dual view) and the total arsenic

in rice samples were analysed by inductively coupled

plasma mass spectrometry (ICP-MS) (Agilent 7500

Series ICP-MS). The operating conditions of the

instruments are summarized in Table 1. Method of

detection and correction for drift, blank and sensitiv-

ity were the same as described by Gault et al. (2005).

Calculated exposure and arsenic attributable

cancer risks

Lifetime average daily doses (LADD) ([lg/kg]/day)

(USEPA 1992) were calculated from the equation

Environ Geochem Health (2010) 32:463–477 465

123

Page 4

LADD ¼Xi¼N

i¼1

Ci � Cingi �IRi

BW

� �� BCFi

� EDi

LT

� �ð1Þ

where i = refers to the different potential exposure

routes, viz. drinking water (dw), raw rice (rr) and

cooking of rice (ckr), N is the total number of

exposure routes, viz. 3, Ci is the total arsenic

concentration in the subscripted medium (in the case

of cooking rice as an exposure route this was taken to

be the difference between the concentration in cooked

and raw rice), Cingi is the percentage of inorganic

arsenic in the subscripted medium, IRi is the ingestion

rate for the subscripted medium, BW is the body

weight of the exposed person, BCFi is the bioconcen-

tration factor for the subscripted medium, EDi is the

exposure duration for the ingestion pathway, and LT is

the life expectancy of the exposed person.

Excess lifetime cancer risks, TR (per person), were

then calculated from the LADD values, based on the

USEPA one-hit model (USEPA 1989):

TR ¼ CPSo� 10�3LADD ð2Þ

CPSo is the oral cancer potency slope factor for

arsenic, taken here to be 1.5 ([mg/kg]/day)-1 (USEPA,

IRIS 1998).

This equation is only valid for low risk levels

(\10-2) and is based on the assumption that the dose-

response relationship is linear in the low dose portion

of the multistage model (USEPA 1988) and that the

risk is directly related to intake.

The parameters Cdw, Crr, Ccr (cr = cooked rice)

and IRdw were determined in this study, whilst

previously published values were used for the

parameters Cingdw, Cingrr, Cingcr (Mondal and Polya

2008), BCFdw (Gault et al. 2005) and BCFrr (Juhasz

et al. 2006). Parameterising probability distributions

for these variables was done by fitting the distribu-

tions of these input variables to standard probability

distributions by statistical (k2 test) methods. The

software @Risk (Version 5, Students Edition, Pali-

sade Corp., USA) was used in Microsoft Excel for

this analysis. For this model the values of each of the

parameters IR, Cing and BCF were taken to be the

same for raw rice and cooked rice. The sources of

these and other parameter values are summarized in

Table 6.

Finally, the Monte Carlo simulations were run

(100,000 iterations) and the output exposures and

cancer risks rendered as distributions. In order to

obtain data more representative of the entire com-

munities studied, these iterations involved sampling a

model population with an age, gender, age- and

gender-dependent bodyweight and bodyweight

dependent rice-ingestion rate distribution as detailed

by Mondal and Polya (2008) after various India-wide

surveys (Table 6). We note that the absolute LADD

and cancer risk values calculated are sensitive to the

nature of age and gender distributions in this model,

whereas the relative exposures and relative cancers

risks ascribable to each of the three exposure routes

considered are independent of these distributions in

this model.

Table 1 ICP-MS and ICP-

OES operating conditions

ICP-MS inductively

coupled plasma mass

spectrometry, ICP-OESinductively coupled plasma

optical emission

spectrometry

Parameters ICP-MS ICP-OES

Instrument Agilent 7500 Series ICP-MS

with Octopole reaction system

Perkin Elmer Optima

5300 (DV-Dual view)

Instrument power/W 1500 1350

Reflected power/W B1

Coolant gas flow rate/l min-1 15 15

Auxiliary gas flow rate/l min-1 1.0 0.2

Nebulizer gas flow rate/l min-1 0.95 0.75

Nebulizer Concentric glass Concentric plastic

Spray chamber Peltier cool (2�C) Scott double

pass spray chamber

Air cooled cyclonic

Detection mode Pulse counting Charge counting

466 Environ Geochem Health (2010) 32:463–477

123

Page 5

Results and discussion

Quality control

The quality of ICP-OES analyses for arsenic was

checked by the analysis of a certified reference

material. In addition, ICP-OES analysis of arsenic

for groundwater samples ARS29, ARS30, ARS31,

ARS32 (indicator arsenic concentrations ranging

from 66 to 330 lg/L) as part of an Inter Laboratory

Quality Evaluation (Berg and Stengel 2009) gave

values within 12% of the indicative values. Replicate

analysis of 46 water samples at the SOES, Jadavpur

University agreed with our results with a mean

deviation of 14%. Accuracy of the total rice digestion

and analysis procedure utilized was confirmed by a

mean of arsenic recovery of 99 ± 9% (n = 5) from a

NIST1568a rice flour reference material.

Arsenic concentrations in drinking water

The mean arsenic concentration in drinking water for

Bhawangola-I block of Murshidabad district was

found to be more than three times higher than that of

Chakdha block of Nadia district (Tables 2 and 4)

whilst concentrations for Khejuria-I block were less

than detection levels (1 lg/L). The range of values

observed in all three areas are broadly comparable

with those reported by previous studies (Pal et al.

2009; Roychowdhry 2008; Signes et al. 2008a; SOES

2007) (Table 2). The difference in arsenic concen-

trations among studies likely reflects, in part, differ-

ences in actual samples taken and temporal

variations, not least of all due to arsenic mitigation

measures implemented over the last decade.

In Chakdha block 2% of the surveyed population

was found to be using drinking water from arsenic

removal treatment plants (median As = 74 lg/L),

28% from deep tube wells (median As = 34 lg/L)

and 15% from supplied tap water (median As =

20 lg/L) while the rest were from private tube wells. In

Bhawangola-I block 10% of the surveyed population

was found to be using drinking water from deep tube

wells (median As = 61 lg/L), 46% from supplied tap

water (median As = 53 lg/L) and the rest from

private tube wells. A significant number of the

arsenic-mitigated water supplies—arsenic removal

treatment plants, deep tube wells and supplied tap

water—from both Bhawangola-I and Chakdha blocks

showed arsenic concentrations substantially higher

than the WHO provisional guide value of 10 lg/L.

Water consumption

Table 3 summarises the estimated daily water intake

for the surveyed areas, grouped by gender. The mean

water intake for the surveyed population was

3.1 ± 1.0 L/day for males and 2.6 ± 0.9 L/day for

females. Our results are in closer agreement with the

studies of Watanabe et al. (2004), Ohno et al. (2007),

Kile et al. (2007) and Khan et al. (2009) for the

population of Bangladesh (Table 3) which gives an

indication of the accuracy of the protocol employed

in this study. But our values are somewhat lower than

the average values for West Bengal males and

females of 4 and 3 L/day, respectively, reported by

Chowdhury et al. (2001), who also speculated that

people working in the field may have drunk as much

as 6 L/day on an average, and in the summertime as

much as 10 L/day; although these estimates may be

reasonable, the methods to derive these values were

not provided.

Arsenic concentrations in rice

The mean arsenic concentration in raw rice and

cooked rice along with cooking water for the samples

collected from the three areas is summarised in

Table 4. The values are broadly comparable with

other studies in West Bengal and Bangladesh (Bae

et al. 2003; Ohno et al. 2009; Roychowdhry 2008;

Signes et al. 2008a) (Fig. 1). The concentrations of

arsenic in drinking water and rice are not correlated

in any significant way (Figs. 2 and 3).

A very good correlation between raw and cooked

rice is only observed for Khejuri-I block when there

is no arsenic in cooking water (Table 5). Poor

correlation between raw and cooked rice for Bha-

wangola block-I is also because of an observed

bimodal distribution for arsenic in cooked rice with

respect to raw rice. It is observed that for a particular

subset of samples with very high arsenic in cooking

water ([150 lg/L) and in cooked rice there is a good

correlation (R2 = 0.63), and the increase in arsenic in

cooked rice is four times that of the raw rice (Fig. 3),

while for the rest of the samples there still exists a

correlation (R2 = 0.27). Poor correlation between

increased arsenic in cooked rice with respect to

Environ Geochem Health (2010) 32:463–477 467

123

Page 6

Ta

ble

2S

um

mar

yta

ble

of

arse

nic

con

cen

trat

ion

ind

rin

kin

gw

ater

sin

Bh

awan

go

la-I

,C

hak

dh

aan

dK

hej

uri

-Ib

lock

Lo

cati

on

Sam

ple

typ

eN

Tim

eo

fsa

mp

lin

gM

edia

nA

s

con

c(l

g/l

)

Sam

ple

s

exce

edin

g

10

lg/L

(%)

Sam

ple

s

exce

edin

g

50

lg/L

(%)

Ref

eren

ce

Bh

awan

go

la-I

blo

ckD

rin

kin

gw

ater

fro

m

ho

use

ho

ldsu

rvey

50

20

08

57

.25

(29

.2,

10

7.1

)a

94

56

Th

isst

ud

y

Bh

awan

go

la-I

blo

ckT

ub

ew

ell

wat

er1

,77

5O

ver

last

20

yea

rs6

13

1S

OE

S(2

00

7)

Ch

akd

ha

blo

ckD

rin

kin

gw

ater

fro

m

ho

use

ho

ldsu

rvey

58

20

08

9.0

(3.0

,4

1.0

)a

43

20

Th

isst

ud

y

Ch

akd

ha

blo

ckD

rin

kin

gw

ater

fro

m

ho

use

ho

ldsu

rvey

41

20

06

and

20

07

3.1

(1.8

,3

0.0

)a

44

11

Mo

nd

alan

dP

oly

a(2

00

8)

Ch

akd

ha

blo

ckT

ub

ew

ell

wat

er2

,62

6O

ver

last

20

yea

rs5

62

2S

OE

S(2

00

7)

Ch

akd

ha

blo

ckP

ub

lic

wel

lsin

stal

led

by

go

ver

nm

ent

2,1

71

20

03

–2

00

52

7.9

(9.0

,6

9.5

)a

72

34

PH

ED

/UN

ICE

Fd

atab

ase.

Mo

nd

alet

al.

(20

08

)

Kh

eju

ri-I

blo

ckD

rin

kin

gw

ater

fro

m

ho

use

ho

ldsu

rvey

62

00

81 (1

,1

)b

00

Th

isst

ud

y

a2

5th

and

75

thp

erce

nti

les

bT

he

min

imu

md

etec

tio

nli

mit

of

the

ind

uct

ivel

yco

up

led

pla

sma

op

tica

lem

issi

on

spec

tro

met

ry(I

CP

-OE

S)

468 Environ Geochem Health (2010) 32:463–477

123

Page 7

Table 3 Mean daily water consumption

Study population Gender N Age range

(years)

Water intake

(L/day)

Survey dates Reference

Bhawangola-I block Male 28 16–68 3.0 ± 0.9 Mar–July This study

Female 23 14–58 2.7 ± 0.7

Chakdha block Male 50 20–74 3.0 ± 1.2 Aug–Sept This study

Female 52 20–60 2.4 ± 1.0

Khejuria-I block Male 35 20–73 3.3 ± 0.5 Sept This study

Female 33 18–65 2.8 ± 0.8

Bangladesh Male 28 16–80 3.1 ± 1.3 June Ohno et al. (2007)

Female 23 20–70 2.9 ± 1.0

Bangladesh Female 47 20–65 2.7 Jan–Mar, Jun–Aug Kile et al. (2007)

Bangladesh Male 9 [20 3.0 ± 1.2 September Watanabe et al.

(2004)Female 9 [20 3.0 ± 0.8

Bangladesh Male 134 [ 14 2.6 2005–2006 Khan et al. (2009)

125 3.3

3.9127

Female 135 [14 2.4

139 2.7

3.0123

All published values rounded to 0.1 L/day

Table 4 Mean arsenic concentration in different exposure routes

Area Area

Code

Raw rice

(mg/kg)

Cooked rice

(mg/kg)

Cooking water

(lg/L)

Drinking water

(lg/L)

Reference

Exposed areas

Bhawangola-I block, Murshidabad

district, West Bengal

1 0.12 ± 0.09a 0.31 ± 0.21 130 ± 128 130 ± 128 This study

Murshidabad district, West Bengal 2 0.23 0.56 110 110 Roychowdhry

(2008)

Chakdha block, Nadia district, West

Bengal

3 0.16 ± 0.05 0.22 ± 0.32 37 ± 95 40 ± 99 This study

Chakdha block, Nadia district, West

Bengal

4 0.13 ± 0.06 0.17 ± 0.11 20 ± 29 17 ± 21 Mondal and

Polya (2008)

Nadia district, West Bengal 5 0.19 0.29 52 52 Roychowdhry

(2008)

North 24 Parganas, West Bengal 6 0.27 ± 0.02 0.22 ± 0.02 50 50 Signes et al.

(2008a)

Nawabganj district, Bangladesh 7 0.22 ± 0.11 0.26 ± 0.15 32 ± 67 32 ± 67 Ohno et al.

(2009)

Bangladesh 8 0.17 0.27 ± 0.05 312 ± 81 312 Bae et al. (2002)

Unexposed area

Khejuri-I block, Midnapur district,

West Bengal

9 0.12 ± 0.02 0.10 ± 0.03 1 1 This study

a Mean ± standard deviation

Environ Geochem Health (2010) 32:463–477 469

123

Page 8

cooking water for Bhawangola-I block (Table 5)

might suggest that increase in arsenic concentration

in cooked rice with increase in arsenic concentration

in cooking water holds true for a certain concentra-

tion range in cooking water beyond which may be the

chelating effect of rice grain (Bae et al. 2002; Signes

et al. 2008b) getting saturated (Fig. 3). There cer-

tainly seems to be little increase in arsenic in rice

through cooking where the concentration of cooking

water is greater than 50 lg/L (Fig. 3), but this needs

further investigation. Alternatively, the poor correla-

tions may reflect differences in rice variety and

cooking methods.

Rice consumption

Although rice consumption data were collected

during questionnaire survey, many answers received

were clearly unrealistic (e.g. estimates of eating over

1 kg dry weight rice per day!). Our preferred

estimates of rice intake for the purposes of calculat-

ing arsenic exposure are therefore based on the

national database of NNMB (2002). This detailed

survey, which was based on a 24-h recall method of

diet surveys and aimed to assess diet and nutritional

status in the rural populations, included data on

nearly 800 households from West Bengal (Table 6).

Arsenic exposure

The distribution of absolute arsenic exposures from

rice, water and cooking for the three studied areas,

calculated from the fitted parameters in Table 6, are

shown in Fig. 4, whilst the relative importance of

these exposure routes are shown in Table 7. The

median contributions to total arsenic exposure from

drinking water were found to be 70% for Bhawan-

gola-I block, 58% for Chakdha block and just 2% for

the unexposed area-Khejuri-I block. Contributions

from rice and cooking of rice were found to be 12%

and 18%, respectively, for Bhawangola-I block, 34

and 8% for Chakdha block, and 113 and -15% for

Fig. 1 Correlation between arsenic in cooked rice and arsenic

in raw rice. In most cases, cooking has resulted in an increase

in arsenic in rice. (WB) indicates samples from West Bengal

and (B) indicates data from Bangladesh. The line joins loci of

equal arsenic concentrations in cooked and raw rice

As in raw rice vrs drinking water

0.00

0.10

0.20

0.30

0.40

0.50

0.60

0 50 100 150 200 250 300 350 400

As in drinking water(ug/l)

As

in r

aw r

ice(

mg

/kg

)

Bhawangola-I block(WB)Chakdha block(WB)Khejuri-I block(WB)Mondal and Polya 2008(WB)Roychowdhury T 2008(WB)Signes et al., 2008a(WB)Ohno et al., 2009(B)Bae et al., 2002(B)

Fig. 2 Correlation between arsenic in raw rice and arsenic in

drinking water. The correlation is very poor indicating that the

concentrations of arsenic in these media are largely indepen-

dent. (WB) indicates samples from West Bengal and (B)indicates samples from Bangladesh

Ascr/rr vrs Ascw

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

0 50 100 150 200 250 300 350 400

Ascw(ug/l)

As c

r/rr

Bhawangola-I block(WB)Chakdha block(WB)Khejuri-I block(WB)Mondal and Polya 2008(WB)Roychowdhury T 2008(WB)Signes et al., 2008a(WB)Ohno et al., 2009(B)Bae et al., 2002(B)

Fig. 3 Correlation between Ascooked rice/Asraw rice and As in

cooking water. (WB) indicates samples from West Bengal and

(B) indicates samples from Bangladesh

470 Environ Geochem Health (2010) 32:463–477

123

Page 9

Table 5 Correlation between arsenic concentration in raw and cooked rice and between increased arsenic in cooked rice with

cooking water

Study area Sample

size (n)

Correlation between arsenic in raw

and cooked rice

Correlation between increase in arsenic on

cooking rice and arsenic in cooking water

(r) Significance

level

95% Confidence

interval for r(r) Significance

level

95% confidence

interval for r

Bhawangola-I block 49 0.13 P = 0.38 -0.15 to 0.39 0.08 P = 0.60 -0.20 to 0.34

Chakdha block 54 0.10 P = 0.44 -0.16 to 0.35 0.89 P \ 0.01 0.81 to 0.93

Chakdha block

(Mondal & Polya 2008)

39 0.52 P \ 0.01 0.24–0.71 0.74 P \ 0.01 0.56 to 0.85

Khejuri-I block 27 0.47 P = 0.01 0.11–0.72 NR NR NR

NR No correlation calculated because all arsenic in cooking water concentrations were below detection limits

Table 6 The input parameters used in calculation of distributions of LADD and arsenic attributable cancer risks

Input variable Study area Fitted distribution Parameter Data source

Aswater (lg/l) 1 Inversegauss Mean 91.18; shape parameter

86.85

This study

2 Point estimate Mean 110 Roychowdhry (2008)

3 Log normal Mean 42.34, SD 126.67 This study

4 BetaGeneral Shape parameters 0.39, 1.89,

Min 0.4, Max 100.96

Mondal and Polya (2008)

5 Point estimate Mean 52 Roychowdhry (2008)

6 Point estimate Mean 50 Signes et al. (2008a)

7 Point estimate Mean 32 Ohno et al. (2009)

8 Point estimate Mean 312 Bae et al. (2002)

9 Point estimate Mean 1 This study

Asraw rice

(mg/kg)

1 Log logistic Location parameter 0.011; shape

parameters 0.085, 2.24

This study

2 Point estimate 0.23 Roychowdhry (2008)

3 Extreme value Location parameter 0.13; scale

parameter 0.05

This study

4 Log logistic Location parameter 0.03; shape

parameters 0.08, 3.21

Mondal and Polya (2008)

5 Point estimate 0.19 Roychowdhry (2008)

6 Point estimate 0.27 Signes et al. (2008a)

7 Point estimate 0.22 Ohno et al. (2009)

8 Point estimate 0.17 Bae et al. (2002)

9 Triangular Min 0.08, continuous mode 0.09,

Max 0.20

This study

Ascooked rice

(mg/kg)

1 Extreme Value Location parameter 0.22, scale

parameter 0.14

This study

2 Point estimate 0.56 Roychowdhry (2008)

3 Log normal Mean 0.16, SD 0.17 This study

4 Log normal Mean 0.17, SD 0.15 Mondal and Polya (2008)

5 Point estimate 0.29 Roychowdhry (2008)

6 Point estimate 0.22 Signes et al. (2008a)

Environ Geochem Health (2010) 32:463–477 471

123

Page 10

Khejuri-I block (Table 7). Thus, in the unexposed

Khejuri-I area, raw rice accounts for almost all the

total exposure for arsenic, and cooking results in a

decreased arsenic exposure from rice by 15%. Our

results are broadly comparable with those of studies

in similar areas of West Bengal (Table 7), with the

exception of one study by Signes et al. (2008a) who

found cooking to reduce the arsenic exposure for an

arsenic exposed area of West Bengal. The relative

exposures estimated for Bangladesh from the data of

Ohno et al. (2009) and Bae et al. (2002) vary

considerably, mostly because the arsenic concentra-

tion in drinking water reported by Bae et al. (2002)

was ten times that of Ohno et al. (2009). These large

variations provide further justification for using a

probabilistic approach, rather than using point values

of parameters, to estimating exposure and health

risks. In both their studies and this study, variability

of arsenic concentrations in water far exceeded that

of arsenic concentrations in rice (Fig. 2), resulting in

the relative importance of water and rice as exposure

routes being most sensitive to variations in drinking

water compositions.

For exposed areas, the contribution towards total

arsenic exposure from cooking was found to be in the

range of -7 to 22% (Table 7); hence the influence of

the cooking towards total arsenic exposure was less

significant than the contribution from arsenic con-

centrations in drinking water (range 48–90%) and

rice (range 6–44%) itself. This confirms that mitigat-

ing arsenic concentration in drinking water (cf. van

Geen and Duxbury 2009) and in irrigation water

(Williams et al. 2006) are quantitatively more

important arsenic mitigation measures than focusing

on cooking methods; however, we note that education

regarding the most appropriate cooking methods

Table 6 continued

Input variable Study area Fitted distribution Parameter Data source

7 Point estimate 0.26 Ohno et al. (2009)

8 Point estimate 0.27 Bae et al. (2002)

9 Uniform Min 0.04, Max 0.16 This study

Inorganic

arsenic

content of

rice (%)

For All Extreme Value Location parameter 67.72, scale

parameter 10.63

Mondal and Polya (2008)

Water intake

rate (L/day)

For All Log normal Male mean 3.60, SD 0.97

Female mean 2.61, SD 0.85

This study

Rice intake rate

normalised to

body weight

(g/kg/day)

For All Constant Male 11.604

Female 11.273

NNMB (2002),

Mondal and Polya (2008)

Body weight

(kg)

For All Log normal Same as Table 2 of Mondal and

Polya (2008)

Burmaster and Crouch

(1997), NNMB (2002)

Bio-

concentration

factor

For All Constant Water 100%

Rice 90%

Gault et al. (2005),

Juhasz et al. (2006)

Exposure

duration

(years)

For All Constant Equal to the age in years for

age \ 40 and equal to 40 for

age [ 40

Assumed after Bagla and

Kaiser (1996)

Life

expectancy

(years)

For 1–6 & 9 Constant Male 61

Female 63

WHO (2006) data for India

Life

expectancy

(years)

For 7 & 8 Constant Male 63

Female 63

WHO (2006) data for

Bangladesh

LADD lifetime average daily doses, SD standard deviation

Study area codes refer to those stated in Table 4

472 Environ Geochem Health (2010) 32:463–477

123

Page 11

nevertheless represents a relatively inexpensive way

of making small but widespread reductions in arsenic

exposure (cf. Carbonell-Barrachina et al. 2009).

Arsenic attributable cancer risks

Age and gender adjusted median excess lifetime

cancer risk for Bhawangola-I block (4.35 9 10-3),

Chakdha block (2.04 9 10-3), and Khejuri-I block

(4.56 9 10-4) were all found to be higher than the

USEPA regulatory threshold target cancer risk level

of 10-4–10-6. The median excess lifetime cancer risk

estimated in this study for Chakdha block is 1.4 times

higher than that estimated by Mondal and Polya

(2008) for the same area; these values are broadly

compared despite being based on different groups of

volunteers.

Study limitations

For future work, detailed recording of cooking

methods and the rice varieties utilized may help

resolve the reasons for the relatively poor correlations

observed here between arsenic concentrations in raw

and cooked rice even when cooking arsenic concen-

trations are considered (Fig. 3).

Worldwide, there are three common methods of

cooking rice: (1) raw rice is washed till the washing

water become clear (5–6 times), water is discarded

and then the rice is boiled in excess water till cooked,

finally discarding the remaining water by tilting the

pan against the lid before serving the rice; (2) the rice

is washed till the washing water become clear (5–6

times) and boiled with low volume of water until no

water is left to discard; (3) unwashed rice is boiled

with low volume of water until no water is left to

discard (Sengupta et al. 2006). Method (1), also

known as the traditional method, is used by more than

90% of the villagers in the Bengal delta (Bae et al.

2002; Rahman et al. 2006; Sengupta et al. 2006).

Raab et al. (2009) observed that cooking rice by

method (1) did effectively remove both total and

inorganic arsenic by 35–45% depending on rice

variety for total and inorganic arsenic content,

respectively, compared to raw rice. Sengupta et al.

(2006) found that around 57% of total arsenic was

removed when cooking by method (1). Both these

studies were using arsenic-free water for cooking the

rice and both these studies noted that when rice was

cooked by methods (2) and (3), the arsenic content of

cooked rice did not modify with respect to raw rice.

The significance of rice variety on the arsenic content

of cooked rice is also considered by other studies.

Ackerman et al. (2005) found that the absorption of

arsenic by rice from the total volume of water was

between 89 and 105% for six different rice varieties.

In an experiment by Signes et al. (2008b), with

cooking water having arsenic concentration of

Bhawangola-I block, Murshidabad district

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

Exposure as LADD(ug/kg/d)

Rel

ativ

e fr

equ

ency

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

Chakdha block, Nadia district

00.050.1

0.150.2

0.250.3

0.350.4

0.45

Exposure as LADD(ug/kg/d)

Rel

ativ

e fr

equ

ency

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

Khejuri-I block, Midnapur district

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

-1 0 1 2 3 4 5 6

-1 0 1 2 3 4 5 6

-1 0 1 2 3 4 5 6

Exposure as LADD (ug/kg/d)

Rel

ativ

e fr

equ

ency

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

@RISK Student VersionFor Academic Use Only

Cooking Rice Water

Fig. 4 Frequency distributions of arsenic exposure from

drinking water, raw rice and cooking of rice in three

contrasting areas of West Bengal. Negative exposures refer

to removal of arsenic in rice by cooking

Environ Geochem Health (2010) 32:463–477 473

123

Page 12

40 lg/l and using two different rice varieties from

West Bengal, it was observed that all three cooking

methods significantly increased the total arsenic

content in cooked rice (27–42%) compared to raw

rice for one rice variety (atab rice); however, cooking

rice with excess water after washing (method 1)

reduced the arsenic content of cooked rice by 12%

with respect to raw rice for the other rice variety

(boiled rice).

Despite the broad comparability of this work with

other studies, there are nevertheless a number of

biases that may have been introduced into this work,

these include: (a) the sample collection strategies for

Bhawangola-I block and Khejuri-I block were based

on selecting volunteers from medical camp attend-

ees—this may have biased the sample towards

individuals seeking free medical advice, which in

turn might be related to increased chances of having

been exposed to arsenic-bearing groundwaters; (b)

although the household-based survey in Chakdha was

likely a more objectively random sample of the

population, the time-of-day of the survey may well

have resulted in under-sampling of those individuals

working in the fields or otherwise not in the

household and concomitant over-sampling of poorer

or unemployed people; (c) only point data for

arsenic-bearing media has been obtained—differ-

ences in the seasons in which data were collected

might therefore be a potential bias, particularly with

regard to water ingestion rates and rice arsenic

concentrations; however, we found no systematic

dependence of these variables on time-of-year.

Fletcher et al. (2006) also points out the lack of a

long-term exposure history reducing confidence in

any long-term predictions of arsenic-attributable

health risks. This is particularly relevant to West

Bengal because massive investment in arsenic miti-

gation is likely to significantly reduce exposure

Table 7 Median percentage contribution towards total arsenic exposure by different exposure routes calculated using Eq. (1) and

parameters from Table 6 (with 5th and 95th percentiles)

Area of study Study type Drinking water Rice Cooking of rice Reference

Bhawangola-I block,

Murshidabad, West

Bengal

Camp based survey,

collecting samples

from households

70%

(58.7, 80.6)

12%

(7.2, 21.5)

18%

(-41.8, 51.2)

This study

Murshidabad district,

West Bengal

Household based survey 62%

(57.8, 66.8)

16%

(13.7, 18)

22%

(6.3, 34.4)

Calculated in this study

using data from

Roychowdhry (2008)

Chakdha block, Nadia,

West bengal

Household based survey 58%

(35.5, 80)

34%

(16.5, 56.2)

8%

(-122, 59)

This study

Chakdha block, Nadia,

West Bengal

Household based survey 48%

(32.1, 61.7)

44%

(26.1, 56.4)

8%

(-30.4, 44.4)

Mondal and Polya

(2008)

Nadia district, West

Bengal

Household based survey 60%

(55.9, 65.1)

26%

(22.5, 29.5)

14%

(-2.4, 26.6)

Calculated in this study

using data from

Roychowdhry (2008)

North 24 Parganas,

West Bengal

Cooking was simulated

in lab, collecting

samples from farm

65%

(60.9, 69.6)

42%

(35.8, 47.5)

-7%

(-25.9, 9.07)

Calculated in this study

using data from Signes

et al. (2008a)

Nawabganj district,

Bangladesh

Household based survey 51%

(46.7, 56.1)

41%

(36, 46.5)

7%

(-6.9, 19.8)

Calculated in this study

using data from Ohno

et al. (2009)

Bangladesh On site study collecting

rice from local market

90%

(89, 92.3)

6%

(4.9, 6.9)

4%

(-21.7, 23)

Calculated in this study,

using data from Bae

et al. (2002)

Unexposed area-

Khejuri-I block,

Midnapur district,

West Bengal

Camp based survey,

collecting samples

from households

2%

(0, 5.5)

113%

(92.6,137.4)

-15%

(-45.9, 0.5)

This study

474 Environ Geochem Health (2010) 32:463–477

123

Page 13

through drinking water, although this might ulti-

mately be counterbalanced by potential secular

increases in rice arsenic as a result of long-term

irrigation of paddy fields with arsenic-bearing

groundwaters (cf. Roberts et al. 2007; Dittmar et al.

2007).

Like any other exposure assessment model there

are various other limitations to the calculations

(Paustenbach 2000), and for the consequent risk

assessment these are compounds by uncertainties

both in the range of arsenic-attributable sequella and

the dose-relationships for impacted communities

(Smith et al. 1992; NRC 1999, 2001; Adamson and

Polya 2007; Smith and Steinmaus 2009; Polya et al.

2010). Nevertheless, calculations of the relative

importance of the various exposure routes, we hope,

will provide useful indicators, in combination with

other detailed studies, to policy makers when assess-

ing where and what remediation strategies may be the

most effective.

Conclusions

Direct and indirect exposure from groundwater

arsenic and consequent arsenic-attributable health

risks have been determined in three contrasting areas

of West Bengal, in order of decreasing impact of

groundwater arsenic: Bhawangola-I (Murshidabad),

Chakdha block (Nadia), and Khejuri-I block (Mid-

napur). Age and gender adjusted median (arsenic

attributable) excess lifetime cancer risks in all three

areas were found to exceed the 10-4–10-6 range

typically used by the USEPA as a threshold to guide

determination of regulatory values, with the highest

risks found for Bhawangola-I. Total exposure is

dominated by drinking water in Bhawangola-I block

(70%), by both drinking water and rice for Chakdha

block, whilst rice is the dominant exposure route in

Khejuri-I block. Cooking was found to be a minor

exposure route, though the difference between its

detrimental impact of cooked rice in Bhawangola-I

and its positive impact on arsenic in cooked rice in

Khejuri-I was substantial. The difference in balance

of exposure routes indicate a difference in balance of

remediation approaches in the three districts.

Acknowledgments PRAMA is a UKIERI (UK India

Education and Research Initiative) project funded by the

British Council, the UK Department for Innovation,

Universities and Skills (DIUS), Office of Science and

Innovation, the FCO, and the Department of Science and

Technology, Government of India. We acknowledge the

Scottish government, Northern Ireland, Wales, GSK, BP,

Shell and BAE for the benefit of the India Higher Education

Sector and the UK Higher Education Sector. DM acknowledges

the receipt of a Dorothy Hodgkins Postgraduate Award. The

views expressed are not necessarily those of the funding bodies.

Ethical approval for this study was obtained from the University

of Manchester Committee on the Ethics of Research on Human

Beings (Ref. 05031/12 May 2005) and from IICB. We

acknowledge and thank Professor Dipankar Chakraborti and

Ross Nickson and their groups for their kind help during the

surveys and discussion. We thank Louise Ander and an

anonymous reviewer for thoughtful comments that led to

improvements in the clarity of the manuscript. Elements of this

work were presented at the ‘‘Practical Applications of Medical

Geology’’ meeting at the British Geology Survey 19th–20th

March 2009. We thank the organisers, and in particular Mark

Cave and Michael Watts, for the opportunity to present our

work.

References

Ackerman, A. H., Creed, P. A., Parks, A. N., Fricke, M. W.,

Schwegel, C. A., Creed, J. T., et al. (2005). Comparison of

a chemical and enzymatic extraction of arsenic from rice

and an assessment of the arsenic absorption from con-

taminated water by cooked rice. Environmental Scienceand Technology, 39(14), 5241–5246.

Adamson, G. C. D., & Polya, D. A. (2007). Critical pathway

analysis to determine key uncertainties in net impacts on

disease burden in Bangladesh of arsenic mitigation

involving the substitution of arsenic bearing for ground-

water drinking water supplies. Journal of EnvironmentalScience and Health, 42, 1909–1917.

Bae, M., Watanabe, C., & Inaoka, T. (2003). Arsenic in cooked

rice in Bangladesh (vol 360, pg 1839, 2002). Lancet,361(9362), 1060 (Erratum).

Bae, M., Watanabe, C., Inaoka, T., Sekiyama, M., Sudo, N.,

Bokul, M. H., et al. (2002). Arsenic in cooked rice in

Bangladesh. Lancet, 360(9348), 1839–1840.

Bagla, P., & Kaiser, J. (1996). Epidemiology—India’s

spreading health crisis draws global arsenic experts. Sci-ence, 274(5285), 174–175.

Berg, M., & Stengel, C. (2009). ARS 29-32, Arsenic reference

samples interlaboratory quality evaluation (IQE). Report

to Participants.EAWAG, October 2004.

Burmaster, D. E., & Crouch, E. A. C. (1997). Lognormal

distribution for body weight as a function of age for males

and females in the United States, 1976–1980. Risk Anal-ysis, 17(4), 499–505.

Carbonell-Barrachina, A. A., Signes-Pastor, A. J., Vazquez-

Araujo, L., Burio, F., & Sengupta, B. (2009). Presence of

arsenic in agricultural products from arsenic-endemic

areas and strategies to reduce arsenic intake in rural vil-

lages. Molecular Nutrition & Food Research, 53(5), 531–

541.

Environ Geochem Health (2010) 32:463–477 475

123

Page 14

Chakraborti, D., Das, B., Rahman, M. M., Chowdhury, U. K.,

Biswas, B., Goswami, A. B., et al. (2009). Status of

groundwater arsenic contamination in the state of West

Bengal, India: A 20-year study report. Molecular Nutri-tion & Food Research, 53, 542–551.

Chowdhury, U. K., Rahman, M. M., Mondal, B. K., Paul, K.,

Lodh, D., Basu, G. K., et al. (2001). Groundwater arsenic

contamination and human suffering in West Bengal, India

and Bangladesh. Environmental Science, 8, 393–415.

Dittmar, J., Voegelin, A., Roberts, L. C., Hug, S. J., Saha, G.

C., Ali, M. A., et al. (2007). Spatial distribution and

temporal variability of arsenic in irrigated rice fields in

Bangladesh. 2. Paddy Soil. Environmental Science &Technology, 41(17), 5967–5972.

Fletcher, T., Leonardi, G., Hough, R., Goessler, W., Gurzau,

E., Koppova, K., et al. (2006). Long-term arsenic expo-

sure and cancer risk-sensitivity to choice of indicators

based on recent and lifetime arsenic intake. Epidemiology,17(6), S307.

Gault, A. G., Jana, J., Chakraborty, S., Mukherjee, P., Sarkar,

M., Nath, B., et al. (2005). Preservation strategies for

inorganic arsenic species in high iron, low-Eh ground-

water from West Bengal, India. Analytical and Bioana-lytical Chemistry, 381(2), 347–353.

Juhasz, A. L., Smith, E., Weber, J., Rees, M., Rofe, A., Kuchel,

T., et al. (2006). In vivo assessment of arsenic bioavail-

ability in rice and its significance for human health risk

assessment. Environmental Health Perspectives, 114(12),

1826–1831.

Khan, N. I., Bruce, D., Naidu, R., & Owens, G. (2009).

Implementation of food frequency questionnaire for the

assessment of total dietary arsenic intake in Bangladesh:

Part B, preliminary findings. Environmental Geochemistryand Health, 31, 221–238.

Kile, M. L., Houseman, E. A., Breton, C. V., Smith, T., Qua-

mruzzaman, O., Rahman, M., et al. (2007). Dietary

arsenic exposure in Bangladesh. Environmental HealthPerspectives, 115(6), 889–893.

Laparra, J. M., Velez, D., Barbera, R., Farre, R., & Montoro, R.

(2005). Bioavailability of inorganic arsenic in cooked

rice: Practical aspects for human health risk assessments.

Journal of Agricultural and Food Chemistry, 53(22),

8829–8833.

Mondal, D., Adamson, G. C. D., Nickson, R., & Polya, D. A.

(2008). A comparison of two techniques for calculating

groundwater arsenic-related lung, bladder and liver cancer

disease burden using data from Chakdha block, West

Bengal. Applied Geochemistry, 23(11), 2999–3009.

Mondal, D., & Polya, D. A. (2008). Rice is a major exposure

route for arsenic in Chakdaha block, Nadia district, West

Bengal, India: A probabilistic risk assessment. AppliedGeochemistry, 23(11), 2987–2998.

Nickson, R., Sengupta, C., Mitra, P., Dave, S. N., Banerjee, A.

K., Bhattacharya, A., et al. (2007). Current knowledge on

the distribution of arsenic in groundwater in five states of

India. Journal of Environmental Science and Health Parta-Toxic/Hazardous Substances & Environmental Engi-neering, 42(12), 1707–1718.

NNMB. (2002). Diet and nutritional status of rural population.

Report by National Institute of Nutrition. Indian Council

of Medical Research, Hyderabad.

NRC. (1999). Arsenic in drinking water. Washington, DC:

National Academy Press.

NRC. (2001). Arsenic in drinking water : 2001 Update.

Washington, DC: National Academy Press.

Ohno, K., Matsuo, Y., Kimura, T., Yanase, T., Rahman, M. H.,

Magara, Y., et al. (2009). Effect of rice-cooking water to

the daily arsenic intake in Bangladesh: results of field

surveys and rice-cooking experiments. Water Science andTechnology, 59(2), 195–201.

Ohno, K., Yanase, T., Matsuo, Y., Kimura, T., Rahman, M. H.,

Magara, Y., et al. (2007). Arsenic intake via water and

food by a population living in an arsenic-affected area of

Bangladesh. Science of the Total Environment, 381(1–3),

68–76.

Pal, A., Chowdhury, U. K., Mondal, D., Das, B., Nayak, B.,

Ghosh, A., et al. (2009). Arsenic burden from cooked rice

in the populations of Arsenic affected and nonaffected

areas and Kolkata City in West-Bengal, India. Environ-mental Science and Technology, 43(9), 3349–3355.

Paustenbach, D. J. (2000). The practice of exposure assess-

ment: A state-of-the-art review (reprinted from Principles

and Methods of Toxicology, 4th edition, 2001). Journal ofToxicology and Environmental Health-Part B-CriticalReviews, 3(3), 179–291.

PHED. (2008). Website of the Public Health Engineering

Department, Government of West Bengal, India.

http://www.wbphed.gov.in. Accessed 11 May 2010.

Polya, D., & Charlet, L. (2009). Rising arsenic risk? NatureGeoscience, 2(6), 383–384.

Polya, D. A., Mondal, D., & Giri, A. K. (2010). Quantification

of deaths and DALYs arising from chronic exposure to

arsenic in groundwaters utilized for drinking, cooking and

irrigation of food crops. In VR Preedy, RR Watson (Eds.),

Handbook of disease burdens and quality of life measures,

Springer.

Raab, A., Baskaran, C., Feldmann, J., & Meharg, A. A. (2009).

Cooking rice in a high water to rice ratio reduces inor-

ganic arsenic content. Journal of Environmental Moni-toring, 11(1), 41–44.

Rahman, M. A., Hasegawa, H., Rahman, M. A., Rahman, M.

M., & Miah, M. A. M. (2006). Influence of cooking

method on arsenic retention in cooked rice related to

dietary exposure. Science of the Total Environment,370(1), 51–60.

Rahman, M. M., Sengupta, M. K., Ahamed, S., Lodh, D., Das,

B., Hossain, M. A., et al. (2005). Murshidabad—One of

the nine groundwater arsenic-affected districts of West

Bengal, India. Part I: Magnitude of contamination and

population at risk. Clinical Toxicology, 43(7), 823–834.

Ravenscroft, P., Brammer, H., & Richards, K. S. (2009). Arsenicpollution: A global synthesis. UK: Wiley-Blackwell.

Roberts, L. C., Hug, S. J., Dittmar, J., Voegelin, A., Saha, G.

C., Ali, M. A., et al. (2007). Spatial distribution and

temporal variability of arsenic in irrigated rice fields in

Bangladesh. 1. Irrigation water. Environmental Scienceand Technology, 41(17), 5960–5966.

Roychowdhry, T. (2008). Impact of sedimentary arsenic

through irrigated groundwater on soil, plant, crops and

human continuum from Bengal delta: Special reference to

raw and cooked rice. Food and Chemical Toxicology,46(8), 2856–2864.

476 Environ Geochem Health (2010) 32:463–477

123

Page 15

Roychowdhury, T., Uchino, T., Tokunaga, H., & Ando, M.

(2002). Survey of arsenic in food composites from an

arsenic-affected area of West Bengal, India. Food andChemical Toxicology, 40(11), 1611–1621.

Sengupta, M. K., Hossain, M. A., Mukherjee, A., Ahamed, S.,

Das, B., Nayak, B., et al. (2006). Arsenic burden of

cooked rice: Traditional and modern methods. Food andChemical Toxicology, 44(11), 1823–1829.

Signes, A., Mitra, K., Burlo, F., & Carbonell-Barrachina, A. A.

(2008a). Contribution of water and cooked rice to an

estimation of the dietary intake of inorganic arsenic in a

rural village of West Bengal, India. Food Additives andContaminants, 25(1), 41–50.

Signes, A., Mitra, K., Burlo, F., & Carbonell-Barrachina, A. A.

(2008b). Effect of cooking method and rice type on

arsenic concentration in cooked rice and the estimation of

arsenic dietary intake in a rural village in West Bengal,

India. Food Additives and Contaminants Part a-ChemistryAnalysis Control Exposure & Risk Assessment, 25(11),

1345–1352.

Smith, A. H., Hopenhaynrich, C., Bates, M. N., Goeden, H. M.,

Hertzpicciotto, I., Duggan, H. M., et al. (1992). Cancer

risks from arsenic in drinking-water. EnvironmentalHealth Perspectives, 97, 259–267.

Smith, A. H., & Steinmaus, C. M. (2009). Health effects of

arsenic and chromium in drinking water: Recent human

findings. Annual Review of Public Health, 30, 107–122.

SOES. (2007). Groundwater arsenic contamination in West

Bengal-India (19 years study). www.soesju.org/arsenic/

arsenicContents. Accessed 11 May 2010.

USEPA. (1988). Special report on ingested inorganic arsenic.

Risk Assessment Forum, Washington DC; 20460 EPA/

625/3-87/013.

USEPA. (1989). Risk assessment guidance for Superfund.

Volume I. Human health evaluation manual (part A).

Interim final. Washington, DC: U.S. Environmental Pro-

tection Agency, Office of Emergency and Remedial

Response EPA/540/1-89/002.

USEPA. (1992). Guidelines for exposure assessment. NoticeFederal Register, 57(104), 22888–22938.

USEPA, IRIS. (1998). Arsenic, inorganic. Integrated risk

information system. CASRN 7440-38-2(04/10/1998) 991.

van Geen, A., & Duxbury, J. M. (2009). Comment on ‘‘Growing

rice aerobically markedly decreases arsenic accumulation’’.

Environmental Science and Technology, 43(10), 3971.

van Geen, A., Zheng, Y., Cheng, Z., He, Y., Dhar, R. K., Garnier,

J. M., et al. (2006). Impact of irrigating rice paddies with

groundwater containing arsenic in Bangladesh. Science ofthe Total Environment, 367(2–3), 769–777.

Watanabe, C., Kawata, A., Sudo, N., Sekiyama, M., Inaoka, T.,

Bae, M., et al. (2004). Water intake in an Asian population

living in arsenic-contaminated area. Toxicology andApplied Pharmacology, 198(3), 272–282.

WHO. (2006). World Health Organisation countries information.

www.who.int/countries/ind/en/. Accessed 11 May 2010.

Williams, P. N., Islam, M. R., Adomako, E. E., Raab, A.,

Hossain, S. A., Zhu, Y. G., et al. (2006). Increase in rice

grain arsenic for regions of Bangladesh irrigating paddies

with elevated arsenic in groundwaters. EnvironmentalScience and Technology, 40(16), 4903–4908.

Environ Geochem Health (2010) 32:463–477 477

123