Journal of Public Health Medicine
Vol. 18, No. 4, pp. 432-436
Printed in Great Britain
Assessing the risk from environmental
cadmium exposure
Annette L. Wood
Background Previous soil surveys by Environmental Health
Officers had found high soil cadmium (Cd) concentrations
in gardens next to a battery factory in Worcestershire. This
study was set up to determine whether this had resulted in
high Cd levels in the blood and urine of local residents.
Methods A sample of residents (n = 39) living next to the
factory were matched by age and sex to employees of North
Worcestershire Health Authority. A questionnaire was used
to determine potential Cd exposure. The levels of Cd in
blood, urine and garden soil were measured.
Results None of the members of the study group had a
blood or urine Cd concentration above the levels estimated
to cause harm. Only one member of the comparison group,
but all members of the study group, had soil in their gardens
with a Cd concentration above the recommended level.
Adjusting for smoking status and other confounders by
using logistic regressional analysis showed that being in the
study group did not confer a greater risk of having an
elevated blood or urine Cd concentration. The greatest
influence on Cd concentrations was a current smoking
habit.
Conclusions No evidence was found to show that the high
soil cadmium concentrations had adversely affected the
health of local residents. Specific issues raised during the
implementation of this study were the resource implications
of assessing environmental exposure and the difficulties in
recruiting the study group. Health Authorities and local
government need to be fully aware of similar problems they
might encounter before investigating a potential environmental health hazard.
Keywords: cadmium; chronic environmental exposure;
cross-sectional study with matching
Introduction
There are three main sources of exposure to cadmium
(Cd): inhalation, either from ambient air or from the
tobacco smoke of cigarettes,1 3 and also eating food
grown in Cd-contaminated soil.4 Exposure to excessive
levels of Cd can eventually lead to renal damage, the
first sign of which is a tubular type of proteinuria.5
Once fully established, the renal dysfunction is irreversible, even if exposure subsequently ceases.6
Numerous epidemiological studies of workers have
documented the spectrum of effects that Cd exposure
can produce.6"10 Less information is available on
chronic environmental exposure to doses of Cd smaller
than those experienced in occupational settings. Even
so, recommended levels for soil Cd concentrations"
and daily Cd intake have been issued in the past.12 In
1980, the World Health Organization (WHO) recommended that the Cd concentration of urine should not
exceed 10 /ig/g creatinine and that control measures be
applied as soon as the individual Cd level in urine
exceeded 5 jig/g creatinine.13 However, these values
were based on occupational Cd exposure and the
validity of their extrapolation to environmental
exposure in the general population has been
questioned.14^16
Battery production in Worcestershire
In 1920, a factory producing Cd-containing batteries
was first established in Redditch, Worcestershire. Soil
sampling around the perimeter of the factory in the
mid-1980s by Redditch Environmental Health Department had indicated higher than normal concentrations
of Cd. When re-landscaping of the area was due in the
late 1980s, a sampling exercise involving local gardens
found that only four from 126 samples were under the
recommended level for Cd in vegetable gardens of 3
mg/kg dry soil weight." Produce sampling also found
high Cd levels in vegetables. The soil monitoring
continued and the soil cadmium concentration
remained high. Local residents were advised at this
point not to eat vegetables grown in their gardens.
However, it was not known how much of the cadmium
had been ingested and absorbed, and the effect of the
Cd on the health of the local population was not
North Birmingham Health Authority, Birmingham Communicable
Disease Unit, Bordesley House, 45 Bordesley Green East, Birmingham
B9 5ST.
ANNETTE L. WOOD, Senior Registrar in Public Health Medicine
© Oxford University Press 1996
Downloaded from http://jpubhealth.oxfordjournals.org/ at Pennsylvania State University on March 3, 2014
Abstract
433
ENVIRONMENTAL CADMIUM EXPOSURE
known. This cross-sectional study with matching was
set up to compare the Cd concentrations in the blood
and urine of residents living next to the battery factory,
where the soil Cd concentration was highest, with those
of a group of non-residents, matched for age and sex.
The study was carried out between October 1993 and
April 1994.
Method
data collected were analysed using EPI INFO 5.6 and
SPSS18 software packages.
Results
The study group consisted of 16 women and 23 men.
Their ages ranged from 19 to 64 years. Nine were aged
under 30 years and four were aged over 59 years. The
mean age was 40 years for the men and 39 years for the
women. The length of residence ranged from less than
one year to 62 years. Ten people had lived in the area
for over ten years, six of whom were men.
Fifteen male members and nine female members
of the study group were current smokers. In the
comparison group, five of the men and none of the
women were current smokers. Ten matched pairs from
the male groups and seven from the female groups had
similar smoking habits. Fifteen people in the study
group and ten in the comparison group ate vegetables
grown in their own garden. No-one had a past medical
history that would suggest a compromised renal
system.
The results of the sampling survey found that only
one soil sample from the comparison group but all the
samples from the study group were above the advised
limit for Cd concentrations in vegetable gardens of 3
mg/kg dry soil weight (Table 1). None of the urine or
blood Cd concentrations for members of both study
and comparison groups were above the levels consid113 61925
ered hazardous to
'
Analysis using logistic regression
Logistic regression analysis was used to investigate the
effect of confounders. All subjects with a urine Cd
concentration of greater than 0-5 /xg/g creatinine and a
blood Cd concentration of greater than 1 ng/\ were
considered 'high'. Those with levels below this were
considered 'low', thus creating a binary variable.
Logistic regression models were then constructed
TABLE 1 Summary of cadmium concentrations
Women
Blood (M9/0
Urine (w3/g creatinine)
Soil (mg/kg dry soil)
Men
Blood (/ig/l)
Urine (/ig/g creatinine)
Soil (mg/kg dry soil)
Range 1for groups
Median for groups
Mean for groups
Study
Comparison
Study
Comparison
Study
Comparison
0-3-0
0-1-25
76-78 7
0-1-0
0-0-5
0-5-2
0-8
0
26-3
0
0
0
0-98
032
356
02
005
0-59
0-3-1
0-0-88
7-6-78-7
0-5-1
0-1-8
0-1-7
1
0
31
1
0
0
089
0-21
37-5
0-69
0-14
0-38
Downloaded from http://jpubhealth.oxfordjournals.org/ at Pennsylvania State University on March 3, 2014
Introductory and reminder letters explaining the aim of
the study were sent to those residents on the electoral
register living nearest the factory who were aged
between 18 and 65 years (n = 116). Residents who
had worked in industries where they were likely to be
exposed to cadmium were excluded. Thirty-nine eligible
people (34 per cent) agreed to take part. A comparison
group was obtained from the employees register of
North Worcestershire Health Authority who were
matched to the study group by age and sex and who
did not live near the battery factory. Again, employees
who had worked in industries where they were likely to
be exposed to cadmium were excluded.
Both groups were sent a letter giving more details of
the study and a consent form. They were asked to
complete a standard structured questionnaire to
determine their medical, smoking and occupational
histories and whether they ate vegetables grown in their
gardens. Instructions on how to collect a soil sample
from the garden were given. Arrangements were made
to take a sample of venous blood, and at the same time
an early morning midstream urine specimen and a soil
sample were collected.
All blood and urine samples were analysed at the
West Midlands Regional Toxicology Laboratory. The
limit of detection of Cd for both blood and urine assays
was 0-5/xg/l. All soil samples were analysed at the
County Scientific Laboratory. The limit of detection
for the soil samples was 0-2 mg/kg dry soil weight. All
17
434
JOURNAL OF PUBLIC HEALTH MEDICINE
TABLE 2 Logistic regression analysis
Blood cadmium high
Urine cadmium high
Relative risk
95% Cl
Relative risk
•
vegetables
226
6-46
1-70
1-12
0-46-11-14
0-86-48-53
0-26-11-16
1-04-1-21
062
19-76
1-40
106
0-15-2 62
3 77-103-68
0-30-656
1-00-1-13
vegetables
2-33
2-26
3-65
1-11
0-49-11 03
0-41-12-47
0-69-19-23
1-04-1-19
0-77
8-12
4-55
105
0-20-2-99
1 -82-36-20
1-25-16-60
1-00-1-11
vegetables
2-35
11-44
028
1-30
1-13
044-12 52
1-09-119-76
0-05-1-62
0-17-9-75
1-04-1-23
0-66
2647
0-47
117
106
0-15-6-51
4-27-163-97
0-11-1-97
0-24-5-81
1-00-1-13
vegetables
2-40
3-24
0-38
3-49
112
0-48-1200
0-46-22-57
0-07-1 -96
0-64-19-10
1 03-1 -20
0-79
8-65
0-74
4-42
105
0-20-3-10
1-86-40-23
0-22-2-54
1-21-16-14
1-00-1-11
vegetables
invalid model
0-66
106
0-47
2528
118
107
0-15-2-87
0-08-13-50
0-11-1 98
1 -80-355-24
0-23-5-98
1-00-1-13
Cl, confidence interval.
using the SPSS software package in which variables
were entered in a stepwise fashion (Table 2).
The conclusions from the analysis were that current
smoking and age were of borderline significance
in relation to having a urine Cd level of more than
0-5 jig/g creatinine. After adjusting for other factors,
only current smoking remained a significant risk factor
for having a blood Cd concentration of over 1 /xg/1.
Discussion
An essential element of any investigation of exposure to
potential environmental health hazards will be the
assessment of risk.26 This study was set up to determine
the Cd concentrations in the population potentially
most at risk because they lived in the area with high soil
Cd levels. The sample size was large enough to
determine statistically significant results but would
have had a detrimental effect on the power of the study,
which is dependent on both the sample size and the
magnitude of the effect being measured.
The proportion of residents who were willing to take
part in this study was disappointingly low, despite the
fact that the issue of cadmium concentrations in the soil
was known to local residents, some of whom had
demanded a response from the local council. There was
no evidence found to show that the high soil cadmium
concentrations had resulted in high blood and urine Cd
concentrations in local residents. This was probably as a
result of the short length of time some residents had
lived in the area and also the small proportion who ate
vegetables grown in their gardens. However, if the study
results had revealed blood and urine Cd concentrations
that potentially could induce ill health, any subsequent
study would have needed to overcome the logistical
problems of establishing a larger study group.
Downloaded from http://jpubhealth.oxfordjournals.org/ at Pennsylvania State University on March 3, 2014
Model 1
Ate home-grown
Current smoker
Study group
Age
Model 2
Ate home-grown
Ever smoked
Study group
Age
Model 3
Ate home-grown
Current smoker
Male
Study group
Age
Model 4
Ate home-grown
Ever smoked
Male
Study group
Age
Model 5
Ate home-grown
Ever smoked
Male
Current smoker
Study group
Age
95% Cl
ENVIRONMENTAL CADMIUM EXPOSURE
Acknowledgements
I would like to thank Dr Robin Braithwaite at the West
Midlands Regional Toxicology Laboratory, Mr Bev
Dickens at the Environmental Health Department of
Redditch Borough Council, Dr Sarah Walters at the
University of Birmingham and Dr Alan Tweddell at
North Worcestershire Health Authority for their help
and advice during the development and implementation of this study.
References
1
Friberg L, Kjellstrom T, Nordberg GF. Cadmium. In:
Friberg L, Nordberg GF, Vouk V, eds. Handbook on the
toxicology of metals, 2nd edn. Amsterdam: Elsevier,
1986.
2
Elinder CG, Kjellstrom T, Lind B, el al. Cadmium exposure
from smoking cigarettes: variations with time and
country where purchased. Environ Res 1983; 32: 220227.
3
Lewis GP, Jusko WJ, Coughlin LL, Hartz S. Contribution
of cigarette smoking to cadmium accumulation in man.
Lancet 1972; 1: 291-292.
4
Tjell Jch, Christensen TH, BroRamussen F. Cadmium in
soil and terrestrial biota with emphasis on the Danish
situation. Ecotoxicol Environ Safety 1983; 7: 122-140.
5
Elinder CG. Normal values for cadmium in human tissues,
blood and urine in different countries. In: Friberg L,
Elinder CG, Kjellstrom T, Nordberg GF, eds. Cadmium
and health - a toxicologicaJ and epidemiological appraisal'.
Boca Raton, FL: CRC Press, 1985.
6
Roels H, Djubang J, Buchet JP, Bernard A, Lauwerys R.
Evolution of cadmium induced renal dysfunction in
workers removed from exposure. Scand J Work Environ
Hlth 1982; 8: 191-200.
7
Friberg L. Health hazards in the manufacture of alkaline
accumulators with special reference to chronic cadmium
poisoning. Ada Med Scand 1950; 138(240): 1-124.
8
Katanzis G, Flynn FV, Spourage J, Trott DG. Renal
tubular malfunction and pulmonary emphysema in
cadmium pigment workers. Quart J Med 1963; 32:
165-192.
9
Adams RG, Harrison JF, Scott P. The development of
cadmium-induced proteinuria, impaired renal function
and osteomalacia in alkaline battery workers. Quart J
Med 1969; 152: 425-443.
10
Lauwerys R, Roels H, Regniers M. Significance of
cadmium concentrations in blood and urine in workers
exposed to cadmium. Environ Res 1979; 20: 375-391.
" Interdepartmental Committee on the redevelopment of
contaminated land. Guidance on the assessment and
redevelopment of contaminated land. ICRCL 59(83), 2nd
edn. London: ICRCL, 1987.
12
Joint FAO-WHO Expert Committee on Food Additives
1972. Evaluation of certain food additives and the
contaminants mercury, lead and cadmium. FAO Nutrition
Meeting Report Series Number 51. WHO Technical
Report Series Number 505. Geneva: WHO, 1972.
13
World Health Organization Study Group. Recommended
health based limits in occupational exposure to heavy
metals. Tech Rep Ser Wld Hlth Org 1980; 647: 21-35.
14
Mueller PW, Jay Smith S, Steinberg KK, Thun MJ.
Chronic renal tubular effects in relation to urine
cadmium levels. Nephron 1989; 52: 45-54.
13
Jung K, Pergande M, Graubaum HJ, et al. Urinary proteins
Downloaded from http://jpubhealth.oxfordjournals.org/ at Pennsylvania State University on March 3, 2014
Before the investigation of a possible association
between a source of chemical pollution or contamination and ill health in a population is initiated, both the
human and financial resource implications need to be
taken into account. Joint working between local
government and District Health Authorities can lead
to logistical problems when the boundaries of authority
either overlap or never meet. In this study, the local
authority had a small budget to monitor the soil
cadmium concentrations, whereas the Heath Authority
did not have a specific budget for the investigation of
environmental health hazards. As the laboratory costs
alone for this study were £30 per person, the financial
consequences of having to carry out a number of
similar studies can be overwhelming.
The investigation of environmental health hazards
potentially exposes a human and financial resource
deficit in District Health Authorities, which are
expected to ensure they have adequate arrangements
to deal with the health aspects of chemical contamination of air, soil, water and food.27"29 Health Authorities are also expected to designate an appropriate
individual to be responsible for ensuring there is access
to the necessary advice and expertise on the public
health hazards arising from such chemical contamination.29 At present, responsibility for this often falls to
the Consultant in Communicable Disease Control.30'31
However, the legitimacy of this has been questioned, in
part because there is no recognized training in
environmental health for public health physicians and
microbiologists.
Access to, and the availability of, expert advice can
also be problematic. Examining the literature to
determine whether some sort of 'past precedent' exists
can lead to difficulties. As an example, studies
investigating the effect of environmental Cd exposure
have conflicting conclusions, in part because of the
variety of biological parameters used, differences in the
information collected about the study and comparison
groups, and sample sizes which have ranged from over
1500 to less than 50.32"39 Measures to control the
release of Cd into the environment have also varied
over time and place, resulting in difficulties in estimating environmental Cd exposure.
There are many potential hazards in the environment
that it might be beneficial to know more about.40
However, if after initial assessment, no substantial risk
can be found, it has to be considered whether time and
resources are better spent elsewhere. This was considered to be the outcome of this study.
435
436
16
17
18
19
21
22
23
24
25
26
27
and enzymes as indicators of renal dysfunction in
of the NHS and the role of others. H S G (93) 56. L o n d o n :
chronic exposure to cadmium. Clin Chem 1993; 39(5):
NHSME, 1993.
29
757-763.
N H S M a n a g e m e n t Executive. Arrangements
to deal with
Bernard A, Thielemans N, Roels H, Lauwerys R. Associahealth aspects of chemical contamination incidents. H S G
tion between NAG-B and cadmium in urine with no
(93) 38. L o n d o n : N H S M E , 1993.
30
evidence of a threshold. Occup Environ Med 1995; 52:
Gunnel] DJ. T h e public health physician's role in chemical
177-180.
incidents. J Publ Hlth Med 1993; 15(4): 3 5 2 - 3 5 7 .
31
Dean AD, Dean JA, Burton JH, Dicker DR. EPl INFO
Ayres PJ. Major chemical incidents - a response, the role of
Version 5: a word processing, database and statistics
the C o n s u l t a n t in Communicable Disease C o n t r o l a n d
program for epidemiology on microcomputers. Atlanta,
the case of need for a national surveillance resource
G A : C e n t e r for Disease C o n t r o l , 1990.
centre. J Publ Hlth Med 1995; 17(2): 164-170.
32
N o r u s i s M J . SPSS for Windows advanced statistics release
Staessen J, Y e o m a n W B , Fletcher A E , et al. Blood
6.0. C h i c a g o , IL: SPSS Inc., 1993.
c a d m i u m in L o n d o n civil servants. Int J Epidemiol
Buchet J P , L a u w e r y s R, Roels H , et al. Renal effects o f
1990; 19(2): 3 6 2 - 3 6 6 .
33
c a d m i u m b o d y b u r d e n of the general population. Lancet
I n s k i p H , Beral V, M c D o w a l l M . M o r t a l i t y o f S h i p h a m
1990; 336: 6 9 9 - 7 0 2 .
residents: 40 years follow-up. Lancet 1982; 1: 8 9 6 - 8 9 9 .
34
Roels H , B e r n a r d A , C a r d e n a s A , et al. M a r k e r s o f early
L a u w e r y s R , B e r n a r d A , Buchet J P , et al. D o e s e n v i r o n m e n t a l e x p o s u r e t o c a d m i u m represent a health risk?
renal changes induced b y industrial pollutants. I l l
C o n c l u s i o n s from t h e C a d m i b e l s t u d y . Ada Clin Belg
Application t o workers exposed t o c a d m i u m . Br J Ind
1991; 46(4): 2 1 9 - 2 2 5 .
Med 1993; SO: 3 7 - 4 8 .
35
K a w a n d a T , T o h y a m a C, Suzuki S. Significance o f t h e
K i d o T , N o g a w a K , O h m i c h i M , et al. Significance o f
excretion o f u r i n a r y indicator proteins for a low level of
urinary cadmium concentration in a Japanese population
o c c u p a t i o n a l exposure t o c a d m i u m . Int Arch
Occup
environmentally exposed to cadmium. Arch Environ Hlth
Environ Hlth 1990; 62: 9 5 - 1 0 0 .
1992; 47(3): 196-202.
36
Verschoor M , H e r b e r R, v a n H e m m e n J, W i b o w o A ,
I w a t a K , Saito H , M o r i y a m a M , N a k a n o A . Follow u p
Zielhuis R. R e n a l function o f workers with low-level
study of renal tubular dysfunction and mortality in
c a d m i u m e x p o s u r e . ScandJ Work Environ Hlth 1987; 13:
residents of an area polluted with cadmium. Br J Ind Med
232-238.
1992; 49: 736-737.
37
C h i a K S , O n g C N , O n g H Y , E n d o G . Renal t u b u l a r
Nishijo M , N a k a g a w a H , M o r i k a m a Y , et al. M o r t a l i t y of
function of w o r k e r s exposed to low levels of c a d m i u m . Br
inhabitants in an area polluted by cadmium: 15 year
J Ind Med 1989; 4 6 : 165-170.
follow up. Occup Environ Med 1995; 52: 181-184.
M
Sangster B, de G r o o t G, Loeber JG, et al. Urinary excretion
Mueller P W , Paschal D C , H a m m e l R R , et at. C h r o n i c renal
of cadmium, protein, beta-2-microglobulin and glucose
effects in three studies of m e n a n d w o m e n occupationally
in individuals living in a cadmium-polluted area. Human
exposed t o c a d m i u m . Arch Environ Contam Toxicol 1992;
Toxicol 1984; 3: 7 - 2 1 .
23: 125-136.
39
K a w a n d a T , K o y a m a H, Suzuki S. C a d m i u m , N A G
Vather M . Assessment of exposure to lead and cadmium
activity, a n d /J-2-microglobulin in the urine of c a d m i u m
through biological monitoring - results of a U N E P /
pigment w o r k e r s . Br J Ind Med 1989; 46: 5 2 - 5 5 .
W H O study. Environ Res 1983; 30: 9 5 - 1 2 8 .
40
C o g g o n D . Assessment o f exposure t o environmental
Archibald C P , Kosatsky T . Public health response t o a n
p o l l u t a n t s . Occup Environ Med 1995; 52: 5 6 2 - 5 6 4 .
identified environmental toxin: m a n a g i n g risks t o t h e
James Bay Cree related t o c a d m i u m in caribou a n d
N H S M a n a g e m e n t Executive. Emergency planning in the
m o o s e . Can J Publ Hlth 1991; 82: 2 2 - 2 6 .
NHS: Health Service arrangements for dealing with major
incidents. E L (91) 137. Heywood: N H S M E , 1991.
N H S M a n a g e m e n t Executive. Public health: responsibilities
Accepted on I May 1996
Downloaded from http://jpubhealth.oxfordjournals.org/ at Pennsylvania State University on March 3, 2014
20
JOURNAL OF PUBLIC HEALTH MEDICINE