Lead modelling — A decision-making tool

The Science of the Total Environment, 16 (1980) 231--237 Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands

LEAD MODELLING

- A DECISION-MAKING

231

TOOL

JOHN E. McEVOY

Technology Policy Unit, The University of Aston in Birmingham, Aston Street, Birmingham, B4 7ET (England) (Received February 19th, 1960; accepted May 18th, 1980)

ABSTRACT

The difficulties of decision-making in a state of ignorance regarding the problem area, true for many questions of environmental pollution, are discussed with reference to lead. Before irrevocably committing a considerable amount of both capital and material resources, the policy-maker would like to make a prediction as to the potential effectiveness of any control measure. To this end a mathematical model is suggested which represents the transport of lead within a standard man. It may be used at two discrete levels, either to analyse a number of control strategies in terms of their relative effectiveness in reducing blood lead concentrations, or, dependent upon the model's accuracy and the availability of data concerning the exposure to, and metabolism of, lead by a specific population, to select between two, or more, directly competing strategies for the control of that population's lead exposure, the selection of the optimum strategy being based on the model's output. Modelling may be seen, therefore, to satisfy, at least in part, the need for decision-making information in an area where time may not permit the acquisition of appropriate, unequivocable scientific evidence.

INTRODUCTION

A recurring problem facing the policy-maker in the field of environmental pollution control is that of predicting the effects of control measures in advance of their implementation in order that the optimal control policy may be developed. In the Technology Policy Unit of the University of Aston in Birmingham a methodology has been developed, in connection with

Editor's note. A formalism developed for the application of the exposure commitment method to the assessment of the transport of chemical pollutants in the environment has recently been discussed by B. J. O'Brien, Monitoring and Assessment Research Centre, Chelsea College, University of London. Report 13 "The exposure commitment method with application to exposure of man to lead pollution", pp 86 + App., 1979. Price: UK £2.00 ; US $4.00.

0048-9697/80/0000--0000/$02.25 0 1 9 8 0 Elsevier Scientific Publishing Company

232 doctoral research into the control of environmental lead, to assist the policy/ decision-maker in the elimination of this difficulty [1]. With respect to pollutants believed to be harmful to man, it is necessary to know the reduction in harm caused to the population brought a b o u t by a particular control strategy, rather than the reduction in the concentration of the pollutant in the environment. This situation is exemplified by the case of environmental lead which, although it is picked up by man through two primary routes, the lungs and the gut, has a multiplicity of sources, particularly for dietary exposure. It is exceptional to the extent that airborne lead has virtually one source, emissions from m o t o r vehicles employing fuel containing lead-based antiknock agents [2]. However, should, for example, a control strategy be suggested to totally eliminate lead from canned food products, a source of dietary lead exposure [ 3 ] , the policy-maker needs to ascertain the impact of this strategy on the overall population's lead exposure and, more precisely, on the harm caused to the population by such exposure. At this point an insurmountable difficulty arises in that it is n o t currently possible to measure the harm caused to populations, or individuals, as a result of exposure to environmental concentrations of lead [4]. If some measure of the suspected harm is required, it is necessary to introduce a surrogate for the above quantity and in this case blood lead concentration presents itself as an available substitute. This is due to the fact that blood lead level reflects the individual's current exposure to lead and thence, it is assumed, the harm currently being caused by lead. However, it should be noted that blood lead does not reflect cumulative lead exposure and so cann o t be used to ascertain the effects of long-term exposure to the metal. Ideally the policy-maker would like to know the change in the population's blood lead distribution, thus reflecting the change in harm caused by environmental lead exposure, brought about by a. given strategy. However, this is impossible for two reasons; firstly the current population blood lead distribution is n o t known, and, secondly, if it were, the alteration to that distribution brought about by a particular control strategy, e.g. eliminating lead from canned food, cannot be calculated [5]. This latter difficulty arises because of the fact that a number of people who exhibit identical blood leads m a y have widely differing sources of exposure. For example an urban, single worker might use a high proportion of canned foods in a convenience food oriented diet whereas a rural resident might use fresh foods exclusively, although both display similar blood lead concentrations. Hence the strategy of eliminating lead from canned food would have a more pronounced effect on the lead exposure, and thus blood lead concentration, of the urban resident as opposed to his rural counterpart. It is thus apparent that, even if the blood lead distribution of the population were to be known, it is impossible t o calculate the effect of any given strategy on this distribution. These t w o difficulties, firstly ignorance regarding the population's blood lead distribution and, secondly, the impossibility of determining the impact of any strategy u p o n the distribution necessitated the adoption of a standard man approach.

233 METHODOLOGY

A mathematical model of lead transport within a standard man was developed and used to predict the blood lead levels employed as a surrogate for the harm caused b y lead to the population as a whole. The model may be used to derive the blood lead concentration of a standard man attributed to any given level of lead exposure. Despite that currently suspected risk groups for lead include pre-school children [6, 7] and expectant mothers [8], the model was based on the transport of lead within a standard man due to the paucity of information in the literature regarding the metabolism of lead in w o m e n and children [5, 9 ] . Detailed information on those factors controlling the movement of lead within the b o d y for a standard man may be obtained from the ICRP Reports of 1959 [10] and 1975 [ 1 1 ] . This is further supported b y the majority of epidemiological studies carried o u t on environmental lead exposure, these being based on male subjects [12--14]. An exception to this rule is the 7-City Study o f Tepper and Levin [9] which is based on female subjects. In this instance no relationship between airborne lead exposure and blood lead level could be found, b u t the authors suggest that this could be due to w o m e n metabolising lead in a different manner from men. Assuming that a constant fraction of inhaled and ingested lead, tl~e latter including lead ejected from the lung and subsequently swallowed, passes directly to blood, the TPU model exhibits a first order dynamic function with respect to blood lead concentration; d(PbB) _ F~ F B ( P b A ) + F~ (PbG) + DB(PbA) -- kK(PbB) dt Given conditions of dynamic equilibrium, i.e. uptake to, and excretion from, blood being equal, d(PbB)/dt becomes zero and the model assumes the form shown in Table 1, which also defines the terms used in the model. Table 2 shows the most recent and hence, in terms of experimental technique TABLE 1 MODEL AND NOTATION

(PbB) =

F1 F B PbA PbG D K PbB

------= = = = ---= =

F 1 F B ( P b A ) + F1 ( P b G ) + D B ( P b A )

~K E l i m i n a t i o n r a t e o f lead f r o m b l o o d , d a y s - I F r a c t i o n o f i n g e s t e d lead t a k e n u p t o b l o o d F r a c t i o n o f i n h a l e d lead e j e c t e d f r o m t h e lung a n d s u b s e q u e n t l y s w a l l o w e d P u l m o n a r y v e n t i l a t i o n , m 3 d a y -1 A i r b o r n e lead c o n c e n t r a t i o n , p g P b m -3 Daily d i e t a r y lead i n t a k e , p g P b d a y -! F r a c t i o n o f i n h a l e d lead r e t a i n e d in t h e lung B l o o d v o l u m e , 1 0 0 ml units B l o o d lead c o n c e n t r a t i o n , / ~ g P b 1 0 0 m 1 - 1

234 TABLE 2 PARAMETER VALUES Parameter

Value

Source

k*

0.033

B D F1 F*

15 m 3 d a y - 1 0.37 0.10 0.044

K

52 × 100 ml

Chamberlain [ 15, 16 ] Griffin [ 16 ] Rabinowitz [17--20] Chamberlain [16] NAS [ 21] K e h o e [ 22 ] Danielson [ 23 ] Lee [ 24 ] ICRP [11]

* Those parameters m a r k e d thus * are derived rather than cited directly. An e x p l a n a t i o n of their derivation is given in the text.

and precision, the most appropriate values for the parameters employed in the model together with their source of reference. From Lee [24], 5% of a typical urban lead aerosol has a mean equivalent diameter (MED) of greater than 0.5 pm. Larger particles such as these tend to become trapped in the upper lung where they are transported to the back of the throat by ciliary action and are subsequently swallowed. Danielson [23] gives the mean retention rate for these particles as 80--95%, and assuming all particles with a MED greater than or equal to 0.5 pm are so trapped and swallowed, a value of 0.044 is obtained for F as shown in Table 2. It has been suggested t h a t these particles, being tightly bound to mucous may be subject to a lower absorption rate. than other ingested lead. However, taking the worst case, it has been assumed that the typical gut absorption factor applies; i.e. 10% of this lead is taken up to blood through the gut. h, the elimination rate of lead from blood was determined from values of the mean and half-life of lead in blood cited in the literature. A value of 0.033 was f o u n d for this parameter. Inserting these, and the other values given in Table 2, into the model initialises it for use on a standard man. However, there remains the question of inputs to the model. For the U.K. situation a value of dietary lead exposure was obtained from the MAFF Report [3], and for airborne lead exposure from the Birmingham study [5] of 1--2 ~ g m -3

RESULTS

These inputs produced a predicted blood lead concentration in the range 1 4 - - 1 7 p g P b 1 0 0 m l -l , values which are within the expected range for the U.K., as shown in Table 3. The model was statistically tested against the epidemiological studies of Tsuchiya [12], Azar [13], Goldsmith [14] and Tepper

235 TABLE 3 RESULTS (Dietary lead exposure set at 1 8 0 p g P b day -l from M A F F Report [3].) Airborne lead conc. (gg m -3 )

Predicted blood lead (/~g 100 m l - l )

0.50 0.75 1.00

12.1 12.9 13.8

Rural

1.50

Sub-urban

2.00

Urban

2.50 3.00

Motorway

Typical U.K. air lead levels

15.4 17.0 18.7 20.3

[9]. In all b u t the last case, where the data was collected from female subjects, the model was found to give a statistically satisfactory output; i.e. no significant difference could be determined, using the t-test for paired observations, between the o u t p u t from the model and the experimentally measured blood leads. In each case a suitable value for dietary lead exposure was obtained from the literature, the remaining standard man parameters being assumed constant.

DISCUSSION

Whilst every effort was made to develop as accurate a model as possible; i.e. one which represents the typical behaviour of lead in a male subject, the on-going development of the model will depend to a large extent on its proposed utilisation. Should it be desired to employ modelling, albeit of a standard man, to derive air quality criteria for lead, particular attention will have to be given to the experimental measurement of those parameters to which the o u t p u t is found to be sensitive. For t h e TPU model these are )~ and B. Furthermore, the contribution of airborne lead to dietary lead exposure through aerosol fallout will have to be accurately determined. In connection with exposure to airborne lead, the 50--55% lung--blood translocation factor suggested b y Chamberlain [16] will also have to be thoroughly examined. However, if, as opposed to producing an absolute value, such as an ambient air quality criterion for lead, the model is to be used to derive a relative value, such as the relative effectiveness of several proposed lead control strategies, its absolute accuracy assumes a secondary role. In fact until the model's lack of accuracy introduces an error to the true ranking order of the above strategies, it is of no real significance; it is to this latter

236

use that the author has applied the model, see McEvoy [4]. In this study the costs and effects of a number of control strategies are examined to determine which of those options may be considered to be dominant in terms of their cost-effectiveness; i.e. if one strategy produces a greater or equal effect for a lower cost than another, then the former may be said to dominate the latter, which may then be discarded by the policy-maker. The model may also provide further information if it is considered to be sufficiently accurate; e.g. if one strategy dominates another, although at many times the former's cost, and provides an only marginally greater effect, of less than 1 p g P b 100m1-1 blood as predicted by the model, it may be decided to reject the technically dominant strategy on the grounds that the difference in effect could n o t be experimentally measured and therefore does not justify the much greater cost. This decision relies to a great extent on h o w much reliability the policy-maker can place on the model's output. In the case of the TPU model, it would appear to be acceptably accurate in the light of its validation against available epidemiological data [12--14]. In conclusion, the lead transport model developed within the TPU may be seen to provide the decision/policy-maker with information at two discrete levels. Firstly, in the production of comparative analyses of control measures and, secondly, with increasing accuracy, a means to select between two directly competing options as outlined above.

REFERENCES 1 D. Collingridge, K. Edwards and J. McEvoy, in D. Black and G. Thomas (Editors), Providing for the Health Service, Croom Helm, 1978, p. 76. 2 R. Engle, Environmental Lead and Public Health, U.S. Environmental Protection Agency, Air Pollution Control Office, Research Triangle Park, NC, 1971. 3 Working Party on the Monitoring of Foodstuffs for Heavy Metals, Survey of Lead in Food, 2nd and 5th Reports, Ministry of Agriculture, Fisheries and Food, H.M.S.O., London, 1972 and 1975. 4 J. McEvoy, An Analysis of Some Control Strategies for Environmental Lead, PhD Thesis, The University of Aston, Birmingham, 1979. 5 Central Unit on Environmental Pollution, Lead Pollution in Birmingham, Department of the Environment, H.M.S.O., London, 1978. 6 P. Landrigan et al., The Lancet, 1 (1975) 708. 7 B. de la Burde et al., J. Pediatr., 81 (1972) 1088. 8 A. Beattie et al., Lancet, 1 (1975) 589. 9 L. Tepper and L. Levin, A Survey of Air and Population Lead Levels in Selected American Communities, in Environmental Quality and Safety Supplement Vol. II, Academic Press, London, 1975, p. 152. 10 International Commission for Radiological Protection, Recommendations Report of Committee II on Permissable Dose for Internal Radiation, Pergamon Press, New York, 1959. 11 International Commission for Radiological Protection, Report of the Task Group on Reference Man, Pergamon Press, New York, 1975. 12 K. Tsuchiya et al., Study of Lead Concentrations in Atmosphere and Population in Japan, in Environmental Quality and Safety Supplement Vol. II, Academic Press, London, 1975, p. 95.

237 13 A. Azar et al., An Epidemiological Approach to Community Air Lead Exposure Using Personal Air Samplers, in Environmental Quality and Safety Supplement Vol. II, Academic Press, London, 1975, p. 254. 14 J. Goldsmith and A. Hexter, Respiratory Exposure to Lead: Epidemiological and Experimental Dose-Response Relationships, Science, 158 (1967) 132--134. 15 A. Chamberlain et al., Uptake of Inhaled Lead from Motor Exhaust, Postgrad. Med. J., 51 (1975) 790--794. 16 A. Chamberlain et al., Investigations into Lead from Motor Vehicles, H.M.S.O., London, 1978. 17 M. Rabinowitz et al., Lead Metabolism in the Normal Human-Stable Isotope Studies, Science, 182 (1973) 725--727. 18 M. Rabinowitz et al., Studies of Human Lead Metabolism by Use of Stable Isotope Tracers, Environ. Health Perspect., 7 (1974) 145--153. 19 M. Rabinowitz et al., Kinetic Analysis of Lead Metabolism in Healthy Humans, J. Clin. Invest., 58 (1976) 260--270. 20 M. Rabinowitz et al., Magnitude of Lead Intake from Respiration by Normal Man, J. Lab. Clin. Med., 90 (1977) 238--248. 21 National Academy of Sciences, Airborne Lead in Perspective, NAS, New York, 1972. 22 R. Kehoe, The Metabolism of Lead in Health and Disease, The Harben Lectures, 1960, J. R. Inst. Pub. Health and Hyg., 24 (1961) 81--97, 101--120, 129--143, 177-203. 23 L. Danielson, Gasoline Containing Lead, Swedish Natural Science Research Council, Stockholm, 1970. 24 R. Lee et al., Concentration and Particle Size Distribution of Particulate Emissions in Automobile Exhaust, Atmos. Environ., 5 (1971) 225.