Jacob Bell
New Member
PHILIP J. LANDRIGAN, MD, MSC, KENNETH E. POWELL, MD, MPH,
LEVY M. JAMES, AND PHILIP R. TAYLOR, MD
Abstract: In March 1978, 13 (21 per cent) of 61 marijuana
samples from the southwestern United States were found to be
contaminated with the herbicide paraquat, a pulmonary toxin, in
concentrations from 3 to 2,264 parts per million. The source of the
contamination was an aerial spraying program in Mexico, supported
indirectly by United States funds. To evaluate US exposure, a
nationwide survey of the paraquat content of confiscated marijuana
was conducted. The survey found 33 (3.6 per cent) of 910 marijuana
specimens to contain detectable paraquat. In states adjacent to
Mexico (Census Region VI), 23 (12.8 per cent) of 180 specimens
were contaminated. Combustion testing indicated that approximately
0.2 per cent of paraquat on marijuana passes into smoke. From
these data, we projected that 100-200 marijuana smokers in Census
Region VI would be exposed by inhalation to 500 micrograms or
more of paraquat per year, a dose judged to represent a health
hazard; nationally, between 150 and 300 smokers were projected to
have such exposure. Another 6,000 persons in Region VI and 9.000
nationally were projected to be at risk of exposure to between 100
and 499 micrograms of paraquat annually. The risk of paraquat
exposure was greatest among those smokers who make one large
purchase of marijuana per year. No clinical cases of paraquat
poisoning were recognized among marijuana smokers during these
studies, but no systematic national search for such cases was
undertaken. (Am J Public Health 1983; 73:784-788.)
Introduction
A perennial problem in public health is the need to
establish disease prevention policy on the basis of incomplete
data. In issues as diverse as the setting of the occupational
lead standard,' the adoption of guidelines for work
with recombinant DNA,2 and the distribution of potassium
iodide to persons living near nuclear reactors,3 policy decisions
were required well before all desired data were available.
Epidemiologic risk assessment offers a quantitative approach
to the development of public health policy.4-6 In this
technique, the number of persons who may be affected by a
policy (or the lack of one) is calculated through the use of
carefully defined decision rules and explicitly stated assumptions.
The effects of policy options can be estimated by
altering the decision rules or assumptions. As new data
become available, the model may be modified.
In 1978, following the discovery that marijuana entering
the United States from Mexico was contaminated with the
herbicide paraquat, Congress directed the Secretary of the
Department of Health, Education, and Welfare (HEW) to
determine whether or not residual paraquat on marijuana
represented a hazard to the health of marijuana users. This
report describes the ensuing epidemiologic risk assessment
undertaken by the Centers for Disease Control (CDC) and
the National Institute of Environmental Health Sciences
(NIEHS).
Toxicity of Paraquat
Paraquat (1,1 '-dimethyl-4,4'-dipyridilium dichloride) is
a widely used, rapidly acting herbicide.7 It appears to react
within plant cells in concert with sunlight to generate superoxide
radicals (02) which in turn destroy plant cellular
membranes8 and desiccate the leaves. If sprayed plants are
harvested before exposure to sunlight, desiccation does not
occur, and unaltered paraquat remains on the plant.
Paraquat is highly toxic to man. The lung is the principal
target organ. Accidental or suicidal ingestion of only a very
small dose (less than one teaspoon) of concentrated paraquat
may produce irreversible progressive lung fibrosis.9"'0 The
case fatality ratio from pulmonary fibrosis following ingestion
of concentrated solutions has been estimated to approach
50 per cent." Other lesions produced by paraquat
include irritative dermatitis, nail damage, eye injury, and
severe epistaxis. 12-15
The pulmonary toxicity of paraquat is enhanced when
paraquat is inhaled or is administered directly into the lungs.
Thus in experimental instillation studies, paraquat doses as
low as 0.5 mg/kg in rats'6and one picogram (10-12 gram) in
rabbits'7 have been shown to produce localized areas of
pulmonary edema and epithelial proliferation which progressed
to fibrosis.'3 The toxicity of inhaled paraquat appears
to be a function of dose, duration of exposure, and
particle size.'8
Paraquat Spray Program
In 1975, the Government of Mexico began an aerial
spraying program to destroy marijuana fields with paraquat,
particularly in the Sierra Madre. From 1975 to 1978, the
United States contributed approximately $30 million per
year to this program and to a companion program for
eradication of poppies through spraying of 2,4-dichlorophenoxyacetic
acid (2,4-D).* United States monies were used
solely for indirect support, such as for purchase and maintenance
of aircraft; herbicides were purchased directly by the
Government of Mexico.'9
Mexican farmers apparently observed that the action of
paraquat depended on sunlight. Accordingly, they began to
harvest marijuana rapidly after spraying and to wrap the
leaves in dark cloths for export to the United States.20 No
accurate estimates are available as to the quantity of paraquat-
contaminated marijuana harvested.'9
Materials and Methodology
Initial Health Surveys
In March 1978, analyses conducted at the University of
Mississippi showed that 13 (21 per cent) of 61 marijuana
samples confiscated in California, Arizona, and Texas from
October 1976 through August 1977 contained paraquat in
concentrations from 3.3 to 2,264 parts per million (ppm).2'
Also, preliminary data from the National Institute of Drug
Abuse (NIDA) indicated that intact paraquat was detectable
in the smoke from a one-gram sample of marijuana which
had been treated with paraquat to a concentration of 450
ppm.22
To evaluate the possible significance of those findings to
public health, the CDC undertook localized surveys to
determine if acute or chronic illness had been caused by
paraquat-contaminated marijuana.23 In an initial case-finding
effort, CDC arranged for three West Coast clinics to interview
marijuana smokers who presented with cough or other
pulmonary symptoms and to obtain a sample of their marijuana
for paraquat analysis. Thirteen persons with suspected
toxicity were interviewed, and 12 samples were submitted;
however, none of the samples was found to contain paraquat.
Additionally, CDC investigators conducted a telephone
survey of 548 persons who had anonymously submitted
marijuana samples to a California laboratory for paraquat
testing.23 This survey was intended to gather data on symptoms
of possible paraquat toxicity and to correlate symptom
prevalence rates with paraquat exposure; it was designed
also to obtain information on patterns of marijuana purchase
and use. The anonymity of participants was preserved.
When the data were analyzed, symptom prevalence could
not be correlated with paraquat exposure, because only one
of the samples was found to contain paraquat. Data on
marijuana consumption patterns (Table 1) were obtained,
however. From those data, it was calculated that 17 per cent
of the marijuana smokers who participated in this survey
made only one purchase per year, whereas 35 per cent made
nine or more purchases annually (Table 2).
The Percy Amendment
In August 1978, Senator Charles Percy (R-Ill.) introduced
an amendment to Section 481 of the Foreign Assistance
Act of 1961, prohibiting US support for programs
involving spraying of herbicide to eradicate marijuana, if the
use of herbicide "is likely to cause serious harm to the health
of persons who may use or consume the sprayed marijuana."'
Under the Percy Amendment, the Secretary of State
was required to inform the Secretary of Health, Education,
and Welfare of the use or intended use of any herbicide in a
US-funded program. The Secretary of HEW was then required
to determine potential health hazards. On October 13,
1978 the Department of State notified HEW that the United
States was providing assistance to the Mexican program.
In response to that notification, then HEW Secretary
Joseph Califano directed CDC to conduct a systematic
sampling program to determine the extent and severity of
paraquat contamination of marijuana. At the same time, he
directed NIEHS to undertake additional combustion testing
of paraquat-contaminated marijuana.24 HEW legal staff reviewed
the legislative history of the Percy Amendment to
define the phrase "serious harm". They determined that no
documentation of persons who had been damaged by exposure
to paraquat was required, and that "actual harm need
not be shown if it is reasonable to predict that it will
occur. "25
Prevalence Survey
In November 1978, CDC initiated a nationwide program
to obtain data on the prevalence of paraquat contamination
of marijuana. Samples of confiscated marijuana were provided
through the seven regional laboratories of the Drug
Enforcement Administration (DEA) and by 14 local and
state law enforcement agencies. Samples were received from
910 seizures covering the period 1975 through January 1979;
of the 832 seizures for which the date of seizure was
provided, 524 (63 per cent) were confiscated in the last
quarter of 1978. DEA seizures accounted for only 15 per cent
of all confiscations, but they were generally larger than the
local interceptions and constituted 95 per cent of the total
weight.
One-gram samples of each lot were analyzed by a
reversed-phase, paired-ion high performance liquid chromatographic
method developed at CDC with a detection
limit of 5 ppm.26 In interpreting the data, it was presumed
that all of the marijuana confiscated in a seizure was from the
same original source, and that the sample received was
representative of the entire seizure.
Nationally, 33 (3.6 per cent) of 910 seizures or 1,059
(0.63 per cent) of 168,424 kilograms were found to contain
detectable paraquat.27 Seventy per cent of the contaminated
seizures and 98.6 per cent of the total weight of contaminated
marijuana was obtained in Census Region VI (New
Mexico, Texas, Oklahoma, Arizona, and Louisiana). No
contamination was found in specimens from the Eastern
Seaboard or the Pacific Northwest. Paraquat concentrations
in the 33 contaminated seizures ranged from 10 to 461 ppm
(mean 111 ppm, median 52 ppm). Twenty-three (69.7 per
cent) of the 33 contaminated samples contained less than 100
ppm of paraquat, and only 2 (6.1 per cent) contained more
than 400 ppm. The origin of 95 of the seized lots was known;
88 (92.6 per cent) had originated in Mexico.
Combustion Testing
The percentage of sprayed paraquat which passes unchanged
into marijuana smoke was estimated by NIDA in
July 1978 to be 0.2 per cent.22 Further combustion testing
data produced under contract to the NIEHS also indicated
that the 0.2 per cent figure was reasonably accurate.28 Most
of the paraquat was found to be pyrolyzed to 4,4'-dipyridyl.
Exposure Assessments
To estimate the potential exposure of smokers to paraquat
in marijuana smoke, CDC undertook a series of computer-
based epidemiologic risk assessments. These projections
were based on the prevalence survey data, the combustion
test results, and the telephone survey data on marijuana
consumption patterns (Tables 1 and 2).
We recognized in making these projections that several
factors could influence exposure to paraquat:
* Amount of marijuana smoked
* Amount of smoke inhaled
* Amount of paraquat in the smoke
* Amount of inhaled paraquat retained in the lungs
* The frequency of marijuana purchases
* The prevalence of paraquat-contaminated marijuana
* The concentration of paraquat in contaminated marijuana
* The geographic distribution of contaminated specimens
To obtain a national risk estimate of exposure to paraquat,
we made the following, obviously somewhat arbitrary
assumptions: **
1) That patterns of marijuana usage reported by persons
who participated in the telephone survey were representative
of those for marijuana users in general and
tha one marijuana cigarette contains one gram of
marijuana;
2) That all of the smoke from a marijuana cigarette is
inhaled;
3) That approximately 0.2 per cent of the paraquat on
marijuana leaf passes unchanged into the smoke;
4) That all inhaled paraquat is retained in the lungs;
5) That regular marijuana smokers purchase 12 batches
of marijuana per year;
6) That the national prevalence of paraquat-contaminated
marijuana was 3.6 per cent;
7) That the distribution of paraquat concentrations in all
contaminated marijuana was identical to that found
in the 33 contaminated specimens examined by CDC;
and
8) That contaminated batches of marijuana were uniformly
distributed throughout the United States.
The number of micrograms of paraquat inhaled by one
person during I year, assuming that one batch of marijuana
is purchased and used during each month, was then calculated
by the following equation:
Inhaled paraquat = .002 (n) (C + C2 + C3 ... C 12),
where .002 is the fraction of paraquat that passes unchanged
into the smoke (0.2 per cent);
n is the average number of grams of marijuana smoked per
month;
and C, through C,2 are the concentrations of paraquat on
12 separately obtained batches of marijuana purchased
in the course of a year.
To replicate the probable distribution of lots which
might confront the population of marijuana users who make
12 purchases per year, we generated randomly by computer
40,000 combinations of 12 batches each. These combinations
were selected from a pool of 910 hypothetical marijuana
batches, representing all of the batches received at CDC; 877
of these batches were considered untainted, and 33 (3.6 per
cent) were considered contaminated with paraquat at the
same concentrations as the specimens analyzed. All specimens
in the pool were considered to have an equal likelihood
of being selected; no allowance was made for variation in
weight.
This model predicted that during one year 68.5 per cent
of marijuana smokers in the United States would be exposed
to no paraquat, that 31.3 per cent would inhale less than 100
micrograms, and that approximately 0.1 per cent would
inhale 100 micrograms or more (Table 3).
To evaluate possible paraquat exposure in Census Region
VI, we undertook a second risk assessment. To reflect
the regional nature of this survey, we modified assumptions
6 and 7 above, as follows:
6) The prevalence of paraquat-contaminated marijuana
was considered to be 12.8 per cent; and
7) The distribution of paraquat concentrations was considered
identical to that found in the specimens from
Region VI.
In this assessment, 100,000 random combinations of 12
batches each were generated, and each combination was
drawn randomly from a pool of 180 hypothetical batches (the
number seized in Region VI), of which 23 were considered
contaminated.
Not surprisingly, this model predicted that the pattern
of exposure in Region VI would be quite different from that
encountered nationally. Only 29.5 per cent of marijuana
smokers in Region VI would be exposed to no paraquat in a
year, 69.5 per cent (695,000) would inhale 1-100 micrograms,
and approximately 1.0 per cent (10,000) would inhale 100
micrograms or more (Table 3).
We undertook further to assess the exposure risk of two
subgroups of marijuana smokers in Region VI: 1) those who
make only one large purchase per year; and 2) those who
make 52 small purchases. In these assessments, we modified
assumption 5 of the previous model such that the number of
batches purchased per year was either I or 52, rather than
12. Each simulation was then based on 100,000 randomly
selected combinations.
In contrast to the preceding assessments, we found
among those marijuana users who make only one large
purchase per year that there exists a small subgroup who
have the potential to be quite heavily exposed to paraquat
(Table 3). The actual size of this group is unknown, because
only limited data are available as to the proportion of
marijuana smokers who make a single purchase per year.
Data from the CDC telephone survey suggest, however, that
as many as 17 per cent of marijuana smokers may purchase
their marijuana in this fashion (Table 2). Assuming that there
are I million regular marijuana smokers in Region VI, and
that purchasing patterns there are similar to those of the
survey participants, then there may be approximately
170,000 once-a-year buyers in Region VI, of whom 100 to
200 would fall into the group whose estimated annual exposure
to paraquat is 500 micrograms or more (Table 3). In
addition, there would be approximately 6,000 persons among
this group in Region VI who would be exposed each year to
between 100 and 499 micrograms of paraquat. Similar calculations
for the nation project that 150-300 marijuana users
might inhale 500 or more micrograms of paraquat per year
and that approximately 9,000 might inhale between 100 and
499 micrograms of paraquat annually.
The projections pertaining to buyers who make 52
purchases per year differ little from the results on those who
make 12 purchases per year, except that there was a slightly
lower prevalence in this group of zero exposures, a finding
consistent with their 4.3-fold greater number of separate
purchases.
Toxic Dose Estimate
To relate the above exposure estimates to possible
health outcomes, we attempted to estimate a minimal toxic
dose for chronic human exposure to inhaled paraquat.
Accurate human data on the quantitative toxicity of chronically
inhaled paraquat do not exist. Previous studies of
agricultural workers using paraquat have not examined the
airborne concentrations to which the workers may have
been exposed.'2,29 Further, although one of those studies'2
performed serial chest x-rays on exposed workers and found
no abnormalities over a 12-week period, it is important to
note that the workers studied were using low-pressure spray
equipment ("back-pack" sprays) which would have produced
airborne droplets too large to have reached the
workers' pulmonary alveoli'2.'8; almost all inhaled paraquat
would have precipitated in the workers' upper airways.
In the absence of relevant human data, we were required
to extrapolate an estimate of minimal toxic dose from
the available animal data. We were guided in that effort by
the observation that man appears to be more sensitive to the
effects of paraquat than are most animal species.9 Further,
we knew that as little as one picogram of paraquat instilled
directly into the lungs of rabbits had been shown capable of
producing discrete pulmonary fibrosis.'7
Finally, we recognized that our estimates of single-year
inhalation exposure would tend to underestimate lifetime
paraquat exposure, because many users smoke marijuana
for more than one year. Based on these factors, we judged
arbitrarily that an annual dose of 500 micrograms of inhaled
paraquat was capable of producing pulmonary damage in
exposed persons. We considered an annual dose of 100 to
499 microgram to convey a less serious, but still heightened
risk of pulmonary injury. We did not consider the possible
additional pulmonary injury which might result from the
inhalation of 4,4'-dipyridyl, the principal combustion product
of paraquat and a compound which has recently been
shown to be itself a pulmonary toxin.30
The Califano Decision
On April 2, 1979, HEW Secretary Califano concluded
"that the spraying of marijuana with paraquat is likely to
cause serious harm to the health of persons who may
consume the sprayed marijuana."3' That decision was conveyed
to the Department of State, and United States support
for the Mexican program ceased.
Discussion
The major finding of the risk assessments described in
this report is that each year from 1975 to 1979 100 to 200
marijuana smokers in the southwestern United States and
150 to 300 nationally were projected to have been exposed
by inhalation to 500 micrograms or more of unaltered
paraquat in marijuana smoke.
We found additionally that persons who make a single
yearly purchase of marijuana are at greatest risk of paraquat
exposure. This observation may have implications even for
persons who make more frequent purchases, for if lines of
supply are relatively stable, then users who make several
purchases per year may obtain all of their marijuana from the
same seller and same ultimate source.
During the present studies, we detected no acute or
chronic cases of paraquat poisoning among marijuana smokers.
This failure cannot, however, be taken as reassurance
that exposure to paraquat-contaminated marijuana did not or
might not in the future cause damage to the lungs of certain
persons. There are several possible explanations for our
failure to detect clinical cases. First, no active nationwide
surveillance for such illness and no systematic screening of
lung function in marijuana smokers was undertaken. Persons
made ill by exposure to an illicit drug are not likely to come
voluntarily to the attention of public health authorities or to
admit drug exposure even if they seek medical attention.
Secondly, chronic pulmonary fibrosis is the most likely
adverse effect of exposure to paraquat,32 and neither the
time interval from onset of exposure to appearance of
fibrosis, nor the cumulative dose which must be inhaled to
produce pulmonary symptoms is known; the time from the
beginning of the Mexican spray program to the time of this
study may have been insufficient for chronic disease to have
appeared. Finally, it appears that in the late 1970s many
marijuana smokers in the United States may have changed
from Mexican to domestic marijuana because of fear of
paraquat exposure33; thus population exposure to paraquat
may have been reduced.
Marijuana smoke has itself been shown to be a pulmonary
toxin. A recent report by a Committee of the National
Academy of Sciences concluded that "chronic heavy smoking
of marijuana causes inflammation and pre-neoplastic
changes in the airways, similar to those produced by smoking
of tobacco. . . This suggests the strong possibility that
prolonged heavy smoking of marijuana, like tobacco, will
lead to cancer of the respiratory tract and to serious impairment
of lung function."32 (The NAS Committee also concurred
in our assessment of the potential health hazard to
smokers of residual paraquat on marijuana.32) The CDC, the
NIEHS, and the other agencies of the US Public Health
Service are seriously concerned about the hazards of marijuana
and subscribe to the position that marijuana smoking is
a hazard to health and ought to be discouraged. It should be
noted, however, in the context of the present study, that
HEW was not asked by the Congress to weigh the hazards of
marijuana smoke against those of paraquat or otherwise to
consider the benefits which might result from a herbicide
spray program. Under the Percy Amendment, HEW was
directed by the Congress solely to evaluate the possible
health hazards of exposure to paraquat.
Epilog
On December 15, 1981, Congress repealed the Percy
Amendment and appropriated $37.7 million each year for
fiscal years 1982 and 1983 for the renewed use of herbicides
in international narcotics control. These funds may be
expended in various nations and apparently will not be
limited to Mexico.
The US Department of Health and Human Services
(successor to HEW) has been given responsibility for monitoring
the impact on health of the renewed spraying program.
The techniques of epidemiologic risk assessment
described in this report can be applied to that monitoring.
However, the model may need to be altered by forthcoming
survey data to reflect: 1) possible increases in the prevalence
of paraquat contamination which may result from multinational
spraying; and 2) possible resultant changes in the
regional distribution of paraquat-contaminated marijuana
within the United States.
REFERENCES
1. Occupational Safety and Health Administration, US Department of
Labor: Occupational Exposure to Lead, Final Standard. Federal Register
November 14. 1978: 43:52952-53014.
2. National Institutes of Health: Guidelines for Research on Recombinant
DNA Molecules. Bethesda. MD: NIH, 1976.
3. Kolata G: Potassium iodide and nuclear accidents. Science 1982:
215:1485.
4. Kates RW: Risk Assessment of Environmental Hazards. New York: John
Wiley and Sons, 1977.
5. McLean AEM: Assessment and Evaluation of Risks to Health from
Chemicals. In: Warner F, Slater DH (eds): The Assessment and Perception
of Risk-A Royal Society Discussion. London: The Royal Society,
1981.
6. Fox AJ: Mortality Statistics and the Assessment of Risk. In: Warner F,
Slater DH (eds): The Assessment and Perception of Risk-A Royal
Society Discussion. London: The Royal Society, 1981.
7. Smith P, Heath D: Paraquat. CRC Crit Rev Toxicol 1976; 4:411-445.
8. Harley JB, Grinspan S, Root RK: Paraquat suicide in a young woman:
results of therapy directed against the superoxide radical. Yale J Biol Med
1977; 50:481-488.
9. Kimbrough RD: Toxic effects of the herbicide paraquat. Chest 1974;
65:65S-67S (Supplement).
10. Paraquat poisoning. (Editorial) Lancet 1971; 1:1018-1019.
11. Fairshter RD, Wilson AF: Paraquat and marijuana: assessing the hazard.
(Editorial) Chest 1978; 74:357-358.
12. Swan AAB: Exposure of spray operators to paraquat. Brit J Industr Med
1969; 26:322-329.
13. Baran RL: Nail damage caused by weed killers and insecticides. Arch
Dermatol 1974; 110:467.
14. Hearn CED, Keir W: Nail damage in spray operators exposed to
paraquat. Brit J Industr Med 1971; 28:399-403.
15. Samann PD, Johnston ENM: Nail damage associated with handling of
paraquat and diquat. Brit Med J 1969; 1:818-819.
16. Kimbrough RD, Gaines TB: Toxicity of paraquat to rats and its effect on
rat lungs. Toxicol Appl Pharmacol 1970; 17:679-690.
17. Zavala DC, Rhodes ML: An effect of paraquat on the lungs of rabbits: its
implications for smoking contaminated marihuana. Chest 1978; 74:418-
420.
18. Gage JC: Toxicity of paraquat and diquat aerosols generated by a size
selective cyclone: effect of particle size distribution. Brit J Indust Med
1968; 25:304-314.
19. US House of Representatives, Select Committee on Narcotic Abuse and
Control, 96th Congress, Second Session: The Use of Paraquat to Eradicate
Illicit Marihuana Crops and the Health Implications of Paraquat-
Contaminated Marihuana on the US Market. Washington, DC: Govt
Printing Office, 1980.
20. Smith RJ: Spraying of herbicides on Mexican marijuana backfires on US.
Science 1978; 199:861-864.
21. Turner CE, Elsohly MA, Cheng FP, Torres LM: Marijuana and paraquat.
(Letter to the Editor) JAMA 1978; 240:1857.
22. Hawks RL: Chemistry and Toxicology of Paraquat-Contaminated Marijuana.
Rockville, MD: National Institute of Drug Abuse, 1978.
23. Centers for Disease Control: Epidemiology Report 78-57-2, Paraquat-
Contaminated Marijuana. Atlanta: CDC, October 27, 1980.
24. Califano JA Jr: (Memorandum) Contamination of Marijuana with Herbicides.
Washington, DC: Department of Health, Education, and Welfare,
Office of the Secretary, November 3, 1978.
25. Dietz RL, Solet I: (Memorandum) Legal Opinion on Paraquat. Washington,
DC: Department of Health, Education, and Welfare, Office of the
Secretary, May 14, 1979. (Reprinted as Appendix A to Reference 19).
26. Needham LL, Paschal DC, Rollen ZJ, Liddle JA, Bayse DD: Determination
of paraquat in marijuana by reversed-phase paired-ion high performance
liquid chromatography. J Chromatographic Sci 1979; 17:87-90.
27. Liddle JA, Needham LL, Rollen ZJ, Roark BR, Bayse DD: Characterization
of the contamination of marijuana with paraquat. Bull Environ
Contam Toxicol 1980; 24:49-53.
28. Brine DR, David KH, Wall ME: Combustion of Paraquat Contaminated
Marihuana-Analyses of Paraquat and its Degradation Products: Final
Report, Contract No. 273-79-C-0003, RTI/1755/OF. Research Triangle
Park, NC: Research Triangle Institute, 1981.
29. Howard JK: A clinical survey of paraquat formulation workers. Brit J
Industr Med 1979; 36:220-223.
30. Groce DF, Kimbrough RD: Acute and subacute toxicity in Sherman
strain rats exposed to 4,4'- and 2,2'-dipyridyl. J Toxicol Environ Health
1982; 10:363-372.
31. Califano JA Jr: Letter to C. R. Vance, Secretary of State, Washington,
DC: August 2, 1979.
32. Marijuana and Health: Report of a Study by a Committee of the Institute
of Medicine. Washington, DC: National Academy Press, 1982.
33. Chapple S: The sweet smoke of success: guess what's become California's
No. I cash crop? Mother Jones April 1982; 7:35-44.
ACKNOWLEDGMENTS
We would like to thank Dr. Renate D. Kimbrough of the Centers for
Disease Control and Dr. Raymond E. Shapiro of the National Institute of
Environmental Health Sciences for their advice and encouragement throughout
this study.
Source: Paraquat And Marijuana: Epidemiologic Risk Assessment
LEVY M. JAMES, AND PHILIP R. TAYLOR, MD
Abstract: In March 1978, 13 (21 per cent) of 61 marijuana
samples from the southwestern United States were found to be
contaminated with the herbicide paraquat, a pulmonary toxin, in
concentrations from 3 to 2,264 parts per million. The source of the
contamination was an aerial spraying program in Mexico, supported
indirectly by United States funds. To evaluate US exposure, a
nationwide survey of the paraquat content of confiscated marijuana
was conducted. The survey found 33 (3.6 per cent) of 910 marijuana
specimens to contain detectable paraquat. In states adjacent to
Mexico (Census Region VI), 23 (12.8 per cent) of 180 specimens
were contaminated. Combustion testing indicated that approximately
0.2 per cent of paraquat on marijuana passes into smoke. From
these data, we projected that 100-200 marijuana smokers in Census
Region VI would be exposed by inhalation to 500 micrograms or
more of paraquat per year, a dose judged to represent a health
hazard; nationally, between 150 and 300 smokers were projected to
have such exposure. Another 6,000 persons in Region VI and 9.000
nationally were projected to be at risk of exposure to between 100
and 499 micrograms of paraquat annually. The risk of paraquat
exposure was greatest among those smokers who make one large
purchase of marijuana per year. No clinical cases of paraquat
poisoning were recognized among marijuana smokers during these
studies, but no systematic national search for such cases was
undertaken. (Am J Public Health 1983; 73:784-788.)
Introduction
A perennial problem in public health is the need to
establish disease prevention policy on the basis of incomplete
data. In issues as diverse as the setting of the occupational
lead standard,' the adoption of guidelines for work
with recombinant DNA,2 and the distribution of potassium
iodide to persons living near nuclear reactors,3 policy decisions
were required well before all desired data were available.
Epidemiologic risk assessment offers a quantitative approach
to the development of public health policy.4-6 In this
technique, the number of persons who may be affected by a
policy (or the lack of one) is calculated through the use of
carefully defined decision rules and explicitly stated assumptions.
The effects of policy options can be estimated by
altering the decision rules or assumptions. As new data
become available, the model may be modified.
In 1978, following the discovery that marijuana entering
the United States from Mexico was contaminated with the
herbicide paraquat, Congress directed the Secretary of the
Department of Health, Education, and Welfare (HEW) to
determine whether or not residual paraquat on marijuana
represented a hazard to the health of marijuana users. This
report describes the ensuing epidemiologic risk assessment
undertaken by the Centers for Disease Control (CDC) and
the National Institute of Environmental Health Sciences
(NIEHS).
Toxicity of Paraquat
Paraquat (1,1 '-dimethyl-4,4'-dipyridilium dichloride) is
a widely used, rapidly acting herbicide.7 It appears to react
within plant cells in concert with sunlight to generate superoxide
radicals (02) which in turn destroy plant cellular
membranes8 and desiccate the leaves. If sprayed plants are
harvested before exposure to sunlight, desiccation does not
occur, and unaltered paraquat remains on the plant.
Paraquat is highly toxic to man. The lung is the principal
target organ. Accidental or suicidal ingestion of only a very
small dose (less than one teaspoon) of concentrated paraquat
may produce irreversible progressive lung fibrosis.9"'0 The
case fatality ratio from pulmonary fibrosis following ingestion
of concentrated solutions has been estimated to approach
50 per cent." Other lesions produced by paraquat
include irritative dermatitis, nail damage, eye injury, and
severe epistaxis. 12-15
The pulmonary toxicity of paraquat is enhanced when
paraquat is inhaled or is administered directly into the lungs.
Thus in experimental instillation studies, paraquat doses as
low as 0.5 mg/kg in rats'6and one picogram (10-12 gram) in
rabbits'7 have been shown to produce localized areas of
pulmonary edema and epithelial proliferation which progressed
to fibrosis.'3 The toxicity of inhaled paraquat appears
to be a function of dose, duration of exposure, and
particle size.'8
Paraquat Spray Program
In 1975, the Government of Mexico began an aerial
spraying program to destroy marijuana fields with paraquat,
particularly in the Sierra Madre. From 1975 to 1978, the
United States contributed approximately $30 million per
year to this program and to a companion program for
eradication of poppies through spraying of 2,4-dichlorophenoxyacetic
acid (2,4-D).* United States monies were used
solely for indirect support, such as for purchase and maintenance
of aircraft; herbicides were purchased directly by the
Government of Mexico.'9
Mexican farmers apparently observed that the action of
paraquat depended on sunlight. Accordingly, they began to
harvest marijuana rapidly after spraying and to wrap the
leaves in dark cloths for export to the United States.20 No
accurate estimates are available as to the quantity of paraquat-
contaminated marijuana harvested.'9
Materials and Methodology
Initial Health Surveys
In March 1978, analyses conducted at the University of
Mississippi showed that 13 (21 per cent) of 61 marijuana
samples confiscated in California, Arizona, and Texas from
October 1976 through August 1977 contained paraquat in
concentrations from 3.3 to 2,264 parts per million (ppm).2'
Also, preliminary data from the National Institute of Drug
Abuse (NIDA) indicated that intact paraquat was detectable
in the smoke from a one-gram sample of marijuana which
had been treated with paraquat to a concentration of 450
ppm.22
To evaluate the possible significance of those findings to
public health, the CDC undertook localized surveys to
determine if acute or chronic illness had been caused by
paraquat-contaminated marijuana.23 In an initial case-finding
effort, CDC arranged for three West Coast clinics to interview
marijuana smokers who presented with cough or other
pulmonary symptoms and to obtain a sample of their marijuana
for paraquat analysis. Thirteen persons with suspected
toxicity were interviewed, and 12 samples were submitted;
however, none of the samples was found to contain paraquat.
Additionally, CDC investigators conducted a telephone
survey of 548 persons who had anonymously submitted
marijuana samples to a California laboratory for paraquat
testing.23 This survey was intended to gather data on symptoms
of possible paraquat toxicity and to correlate symptom
prevalence rates with paraquat exposure; it was designed
also to obtain information on patterns of marijuana purchase
and use. The anonymity of participants was preserved.
When the data were analyzed, symptom prevalence could
not be correlated with paraquat exposure, because only one
of the samples was found to contain paraquat. Data on
marijuana consumption patterns (Table 1) were obtained,
however. From those data, it was calculated that 17 per cent
of the marijuana smokers who participated in this survey
made only one purchase per year, whereas 35 per cent made
nine or more purchases annually (Table 2).
The Percy Amendment
In August 1978, Senator Charles Percy (R-Ill.) introduced
an amendment to Section 481 of the Foreign Assistance
Act of 1961, prohibiting US support for programs
involving spraying of herbicide to eradicate marijuana, if the
use of herbicide "is likely to cause serious harm to the health
of persons who may use or consume the sprayed marijuana."'
Under the Percy Amendment, the Secretary of State
was required to inform the Secretary of Health, Education,
and Welfare of the use or intended use of any herbicide in a
US-funded program. The Secretary of HEW was then required
to determine potential health hazards. On October 13,
1978 the Department of State notified HEW that the United
States was providing assistance to the Mexican program.
In response to that notification, then HEW Secretary
Joseph Califano directed CDC to conduct a systematic
sampling program to determine the extent and severity of
paraquat contamination of marijuana. At the same time, he
directed NIEHS to undertake additional combustion testing
of paraquat-contaminated marijuana.24 HEW legal staff reviewed
the legislative history of the Percy Amendment to
define the phrase "serious harm". They determined that no
documentation of persons who had been damaged by exposure
to paraquat was required, and that "actual harm need
not be shown if it is reasonable to predict that it will
occur. "25
Prevalence Survey
In November 1978, CDC initiated a nationwide program
to obtain data on the prevalence of paraquat contamination
of marijuana. Samples of confiscated marijuana were provided
through the seven regional laboratories of the Drug
Enforcement Administration (DEA) and by 14 local and
state law enforcement agencies. Samples were received from
910 seizures covering the period 1975 through January 1979;
of the 832 seizures for which the date of seizure was
provided, 524 (63 per cent) were confiscated in the last
quarter of 1978. DEA seizures accounted for only 15 per cent
of all confiscations, but they were generally larger than the
local interceptions and constituted 95 per cent of the total
weight.
One-gram samples of each lot were analyzed by a
reversed-phase, paired-ion high performance liquid chromatographic
method developed at CDC with a detection
limit of 5 ppm.26 In interpreting the data, it was presumed
that all of the marijuana confiscated in a seizure was from the
same original source, and that the sample received was
representative of the entire seizure.
Nationally, 33 (3.6 per cent) of 910 seizures or 1,059
(0.63 per cent) of 168,424 kilograms were found to contain
detectable paraquat.27 Seventy per cent of the contaminated
seizures and 98.6 per cent of the total weight of contaminated
marijuana was obtained in Census Region VI (New
Mexico, Texas, Oklahoma, Arizona, and Louisiana). No
contamination was found in specimens from the Eastern
Seaboard or the Pacific Northwest. Paraquat concentrations
in the 33 contaminated seizures ranged from 10 to 461 ppm
(mean 111 ppm, median 52 ppm). Twenty-three (69.7 per
cent) of the 33 contaminated samples contained less than 100
ppm of paraquat, and only 2 (6.1 per cent) contained more
than 400 ppm. The origin of 95 of the seized lots was known;
88 (92.6 per cent) had originated in Mexico.
Combustion Testing
The percentage of sprayed paraquat which passes unchanged
into marijuana smoke was estimated by NIDA in
July 1978 to be 0.2 per cent.22 Further combustion testing
data produced under contract to the NIEHS also indicated
that the 0.2 per cent figure was reasonably accurate.28 Most
of the paraquat was found to be pyrolyzed to 4,4'-dipyridyl.
Exposure Assessments
To estimate the potential exposure of smokers to paraquat
in marijuana smoke, CDC undertook a series of computer-
based epidemiologic risk assessments. These projections
were based on the prevalence survey data, the combustion
test results, and the telephone survey data on marijuana
consumption patterns (Tables 1 and 2).
We recognized in making these projections that several
factors could influence exposure to paraquat:
* Amount of marijuana smoked
* Amount of smoke inhaled
* Amount of paraquat in the smoke
* Amount of inhaled paraquat retained in the lungs
* The frequency of marijuana purchases
* The prevalence of paraquat-contaminated marijuana
* The concentration of paraquat in contaminated marijuana
* The geographic distribution of contaminated specimens
To obtain a national risk estimate of exposure to paraquat,
we made the following, obviously somewhat arbitrary
assumptions: **
1) That patterns of marijuana usage reported by persons
who participated in the telephone survey were representative
of those for marijuana users in general and
tha one marijuana cigarette contains one gram of
marijuana;
2) That all of the smoke from a marijuana cigarette is
inhaled;
3) That approximately 0.2 per cent of the paraquat on
marijuana leaf passes unchanged into the smoke;
4) That all inhaled paraquat is retained in the lungs;
5) That regular marijuana smokers purchase 12 batches
of marijuana per year;
6) That the national prevalence of paraquat-contaminated
marijuana was 3.6 per cent;
7) That the distribution of paraquat concentrations in all
contaminated marijuana was identical to that found
in the 33 contaminated specimens examined by CDC;
and
8) That contaminated batches of marijuana were uniformly
distributed throughout the United States.
The number of micrograms of paraquat inhaled by one
person during I year, assuming that one batch of marijuana
is purchased and used during each month, was then calculated
by the following equation:
Inhaled paraquat = .002 (n) (C + C2 + C3 ... C 12),
where .002 is the fraction of paraquat that passes unchanged
into the smoke (0.2 per cent);
n is the average number of grams of marijuana smoked per
month;
and C, through C,2 are the concentrations of paraquat on
12 separately obtained batches of marijuana purchased
in the course of a year.
To replicate the probable distribution of lots which
might confront the population of marijuana users who make
12 purchases per year, we generated randomly by computer
40,000 combinations of 12 batches each. These combinations
were selected from a pool of 910 hypothetical marijuana
batches, representing all of the batches received at CDC; 877
of these batches were considered untainted, and 33 (3.6 per
cent) were considered contaminated with paraquat at the
same concentrations as the specimens analyzed. All specimens
in the pool were considered to have an equal likelihood
of being selected; no allowance was made for variation in
weight.
This model predicted that during one year 68.5 per cent
of marijuana smokers in the United States would be exposed
to no paraquat, that 31.3 per cent would inhale less than 100
micrograms, and that approximately 0.1 per cent would
inhale 100 micrograms or more (Table 3).
To evaluate possible paraquat exposure in Census Region
VI, we undertook a second risk assessment. To reflect
the regional nature of this survey, we modified assumptions
6 and 7 above, as follows:
6) The prevalence of paraquat-contaminated marijuana
was considered to be 12.8 per cent; and
7) The distribution of paraquat concentrations was considered
identical to that found in the specimens from
Region VI.
In this assessment, 100,000 random combinations of 12
batches each were generated, and each combination was
drawn randomly from a pool of 180 hypothetical batches (the
number seized in Region VI), of which 23 were considered
contaminated.
Not surprisingly, this model predicted that the pattern
of exposure in Region VI would be quite different from that
encountered nationally. Only 29.5 per cent of marijuana
smokers in Region VI would be exposed to no paraquat in a
year, 69.5 per cent (695,000) would inhale 1-100 micrograms,
and approximately 1.0 per cent (10,000) would inhale 100
micrograms or more (Table 3).
We undertook further to assess the exposure risk of two
subgroups of marijuana smokers in Region VI: 1) those who
make only one large purchase per year; and 2) those who
make 52 small purchases. In these assessments, we modified
assumption 5 of the previous model such that the number of
batches purchased per year was either I or 52, rather than
12. Each simulation was then based on 100,000 randomly
selected combinations.
In contrast to the preceding assessments, we found
among those marijuana users who make only one large
purchase per year that there exists a small subgroup who
have the potential to be quite heavily exposed to paraquat
(Table 3). The actual size of this group is unknown, because
only limited data are available as to the proportion of
marijuana smokers who make a single purchase per year.
Data from the CDC telephone survey suggest, however, that
as many as 17 per cent of marijuana smokers may purchase
their marijuana in this fashion (Table 2). Assuming that there
are I million regular marijuana smokers in Region VI, and
that purchasing patterns there are similar to those of the
survey participants, then there may be approximately
170,000 once-a-year buyers in Region VI, of whom 100 to
200 would fall into the group whose estimated annual exposure
to paraquat is 500 micrograms or more (Table 3). In
addition, there would be approximately 6,000 persons among
this group in Region VI who would be exposed each year to
between 100 and 499 micrograms of paraquat. Similar calculations
for the nation project that 150-300 marijuana users
might inhale 500 or more micrograms of paraquat per year
and that approximately 9,000 might inhale between 100 and
499 micrograms of paraquat annually.
The projections pertaining to buyers who make 52
purchases per year differ little from the results on those who
make 12 purchases per year, except that there was a slightly
lower prevalence in this group of zero exposures, a finding
consistent with their 4.3-fold greater number of separate
purchases.
Toxic Dose Estimate
To relate the above exposure estimates to possible
health outcomes, we attempted to estimate a minimal toxic
dose for chronic human exposure to inhaled paraquat.
Accurate human data on the quantitative toxicity of chronically
inhaled paraquat do not exist. Previous studies of
agricultural workers using paraquat have not examined the
airborne concentrations to which the workers may have
been exposed.'2,29 Further, although one of those studies'2
performed serial chest x-rays on exposed workers and found
no abnormalities over a 12-week period, it is important to
note that the workers studied were using low-pressure spray
equipment ("back-pack" sprays) which would have produced
airborne droplets too large to have reached the
workers' pulmonary alveoli'2.'8; almost all inhaled paraquat
would have precipitated in the workers' upper airways.
In the absence of relevant human data, we were required
to extrapolate an estimate of minimal toxic dose from
the available animal data. We were guided in that effort by
the observation that man appears to be more sensitive to the
effects of paraquat than are most animal species.9 Further,
we knew that as little as one picogram of paraquat instilled
directly into the lungs of rabbits had been shown capable of
producing discrete pulmonary fibrosis.'7
Finally, we recognized that our estimates of single-year
inhalation exposure would tend to underestimate lifetime
paraquat exposure, because many users smoke marijuana
for more than one year. Based on these factors, we judged
arbitrarily that an annual dose of 500 micrograms of inhaled
paraquat was capable of producing pulmonary damage in
exposed persons. We considered an annual dose of 100 to
499 microgram to convey a less serious, but still heightened
risk of pulmonary injury. We did not consider the possible
additional pulmonary injury which might result from the
inhalation of 4,4'-dipyridyl, the principal combustion product
of paraquat and a compound which has recently been
shown to be itself a pulmonary toxin.30
The Califano Decision
On April 2, 1979, HEW Secretary Califano concluded
"that the spraying of marijuana with paraquat is likely to
cause serious harm to the health of persons who may
consume the sprayed marijuana."3' That decision was conveyed
to the Department of State, and United States support
for the Mexican program ceased.
Discussion
The major finding of the risk assessments described in
this report is that each year from 1975 to 1979 100 to 200
marijuana smokers in the southwestern United States and
150 to 300 nationally were projected to have been exposed
by inhalation to 500 micrograms or more of unaltered
paraquat in marijuana smoke.
We found additionally that persons who make a single
yearly purchase of marijuana are at greatest risk of paraquat
exposure. This observation may have implications even for
persons who make more frequent purchases, for if lines of
supply are relatively stable, then users who make several
purchases per year may obtain all of their marijuana from the
same seller and same ultimate source.
During the present studies, we detected no acute or
chronic cases of paraquat poisoning among marijuana smokers.
This failure cannot, however, be taken as reassurance
that exposure to paraquat-contaminated marijuana did not or
might not in the future cause damage to the lungs of certain
persons. There are several possible explanations for our
failure to detect clinical cases. First, no active nationwide
surveillance for such illness and no systematic screening of
lung function in marijuana smokers was undertaken. Persons
made ill by exposure to an illicit drug are not likely to come
voluntarily to the attention of public health authorities or to
admit drug exposure even if they seek medical attention.
Secondly, chronic pulmonary fibrosis is the most likely
adverse effect of exposure to paraquat,32 and neither the
time interval from onset of exposure to appearance of
fibrosis, nor the cumulative dose which must be inhaled to
produce pulmonary symptoms is known; the time from the
beginning of the Mexican spray program to the time of this
study may have been insufficient for chronic disease to have
appeared. Finally, it appears that in the late 1970s many
marijuana smokers in the United States may have changed
from Mexican to domestic marijuana because of fear of
paraquat exposure33; thus population exposure to paraquat
may have been reduced.
Marijuana smoke has itself been shown to be a pulmonary
toxin. A recent report by a Committee of the National
Academy of Sciences concluded that "chronic heavy smoking
of marijuana causes inflammation and pre-neoplastic
changes in the airways, similar to those produced by smoking
of tobacco. . . This suggests the strong possibility that
prolonged heavy smoking of marijuana, like tobacco, will
lead to cancer of the respiratory tract and to serious impairment
of lung function."32 (The NAS Committee also concurred
in our assessment of the potential health hazard to
smokers of residual paraquat on marijuana.32) The CDC, the
NIEHS, and the other agencies of the US Public Health
Service are seriously concerned about the hazards of marijuana
and subscribe to the position that marijuana smoking is
a hazard to health and ought to be discouraged. It should be
noted, however, in the context of the present study, that
HEW was not asked by the Congress to weigh the hazards of
marijuana smoke against those of paraquat or otherwise to
consider the benefits which might result from a herbicide
spray program. Under the Percy Amendment, HEW was
directed by the Congress solely to evaluate the possible
health hazards of exposure to paraquat.
Epilog
On December 15, 1981, Congress repealed the Percy
Amendment and appropriated $37.7 million each year for
fiscal years 1982 and 1983 for the renewed use of herbicides
in international narcotics control. These funds may be
expended in various nations and apparently will not be
limited to Mexico.
The US Department of Health and Human Services
(successor to HEW) has been given responsibility for monitoring
the impact on health of the renewed spraying program.
The techniques of epidemiologic risk assessment
described in this report can be applied to that monitoring.
However, the model may need to be altered by forthcoming
survey data to reflect: 1) possible increases in the prevalence
of paraquat contamination which may result from multinational
spraying; and 2) possible resultant changes in the
regional distribution of paraquat-contaminated marijuana
within the United States.
REFERENCES
1. Occupational Safety and Health Administration, US Department of
Labor: Occupational Exposure to Lead, Final Standard. Federal Register
November 14. 1978: 43:52952-53014.
2. National Institutes of Health: Guidelines for Research on Recombinant
DNA Molecules. Bethesda. MD: NIH, 1976.
3. Kolata G: Potassium iodide and nuclear accidents. Science 1982:
215:1485.
4. Kates RW: Risk Assessment of Environmental Hazards. New York: John
Wiley and Sons, 1977.
5. McLean AEM: Assessment and Evaluation of Risks to Health from
Chemicals. In: Warner F, Slater DH (eds): The Assessment and Perception
of Risk-A Royal Society Discussion. London: The Royal Society,
1981.
6. Fox AJ: Mortality Statistics and the Assessment of Risk. In: Warner F,
Slater DH (eds): The Assessment and Perception of Risk-A Royal
Society Discussion. London: The Royal Society, 1981.
7. Smith P, Heath D: Paraquat. CRC Crit Rev Toxicol 1976; 4:411-445.
8. Harley JB, Grinspan S, Root RK: Paraquat suicide in a young woman:
results of therapy directed against the superoxide radical. Yale J Biol Med
1977; 50:481-488.
9. Kimbrough RD: Toxic effects of the herbicide paraquat. Chest 1974;
65:65S-67S (Supplement).
10. Paraquat poisoning. (Editorial) Lancet 1971; 1:1018-1019.
11. Fairshter RD, Wilson AF: Paraquat and marijuana: assessing the hazard.
(Editorial) Chest 1978; 74:357-358.
12. Swan AAB: Exposure of spray operators to paraquat. Brit J Industr Med
1969; 26:322-329.
13. Baran RL: Nail damage caused by weed killers and insecticides. Arch
Dermatol 1974; 110:467.
14. Hearn CED, Keir W: Nail damage in spray operators exposed to
paraquat. Brit J Industr Med 1971; 28:399-403.
15. Samann PD, Johnston ENM: Nail damage associated with handling of
paraquat and diquat. Brit Med J 1969; 1:818-819.
16. Kimbrough RD, Gaines TB: Toxicity of paraquat to rats and its effect on
rat lungs. Toxicol Appl Pharmacol 1970; 17:679-690.
17. Zavala DC, Rhodes ML: An effect of paraquat on the lungs of rabbits: its
implications for smoking contaminated marihuana. Chest 1978; 74:418-
420.
18. Gage JC: Toxicity of paraquat and diquat aerosols generated by a size
selective cyclone: effect of particle size distribution. Brit J Indust Med
1968; 25:304-314.
19. US House of Representatives, Select Committee on Narcotic Abuse and
Control, 96th Congress, Second Session: The Use of Paraquat to Eradicate
Illicit Marihuana Crops and the Health Implications of Paraquat-
Contaminated Marihuana on the US Market. Washington, DC: Govt
Printing Office, 1980.
20. Smith RJ: Spraying of herbicides on Mexican marijuana backfires on US.
Science 1978; 199:861-864.
21. Turner CE, Elsohly MA, Cheng FP, Torres LM: Marijuana and paraquat.
(Letter to the Editor) JAMA 1978; 240:1857.
22. Hawks RL: Chemistry and Toxicology of Paraquat-Contaminated Marijuana.
Rockville, MD: National Institute of Drug Abuse, 1978.
23. Centers for Disease Control: Epidemiology Report 78-57-2, Paraquat-
Contaminated Marijuana. Atlanta: CDC, October 27, 1980.
24. Califano JA Jr: (Memorandum) Contamination of Marijuana with Herbicides.
Washington, DC: Department of Health, Education, and Welfare,
Office of the Secretary, November 3, 1978.
25. Dietz RL, Solet I: (Memorandum) Legal Opinion on Paraquat. Washington,
DC: Department of Health, Education, and Welfare, Office of the
Secretary, May 14, 1979. (Reprinted as Appendix A to Reference 19).
26. Needham LL, Paschal DC, Rollen ZJ, Liddle JA, Bayse DD: Determination
of paraquat in marijuana by reversed-phase paired-ion high performance
liquid chromatography. J Chromatographic Sci 1979; 17:87-90.
27. Liddle JA, Needham LL, Rollen ZJ, Roark BR, Bayse DD: Characterization
of the contamination of marijuana with paraquat. Bull Environ
Contam Toxicol 1980; 24:49-53.
28. Brine DR, David KH, Wall ME: Combustion of Paraquat Contaminated
Marihuana-Analyses of Paraquat and its Degradation Products: Final
Report, Contract No. 273-79-C-0003, RTI/1755/OF. Research Triangle
Park, NC: Research Triangle Institute, 1981.
29. Howard JK: A clinical survey of paraquat formulation workers. Brit J
Industr Med 1979; 36:220-223.
30. Groce DF, Kimbrough RD: Acute and subacute toxicity in Sherman
strain rats exposed to 4,4'- and 2,2'-dipyridyl. J Toxicol Environ Health
1982; 10:363-372.
31. Califano JA Jr: Letter to C. R. Vance, Secretary of State, Washington,
DC: August 2, 1979.
32. Marijuana and Health: Report of a Study by a Committee of the Institute
of Medicine. Washington, DC: National Academy Press, 1982.
33. Chapple S: The sweet smoke of success: guess what's become California's
No. I cash crop? Mother Jones April 1982; 7:35-44.
ACKNOWLEDGMENTS
We would like to thank Dr. Renate D. Kimbrough of the Centers for
Disease Control and Dr. Raymond E. Shapiro of the National Institute of
Environmental Health Sciences for their advice and encouragement throughout
this study.
Source: Paraquat And Marijuana: Epidemiologic Risk Assessment
