A review entitled "The
human toxicity of marijuana" was published in
It is not a pleasant task
to make an open criticism of a paper by colleagues, but we feel compelled to do
so in the case of a paper published in the Journal of Australia in 1992*. The
paper, entitled "The human toxicity of marijuana" which purported to be a review
of the most recent scientific literature on the health effects of cannabis,
claimed that previous knowledge on the subject should now be revised in the
light of "new evidence". However we believe that the authors of this paper were
less than meticulous in their description of the work of others which they
cited. We believe that the readers of this review could be confused as to the
interpretation of the present state of knowledge of the human toxicity of
marijuana.
Under the circumstances we
feel it is of importance to publish a detailed examination, paragraph by
paragraph, of the statements made by Nahas & Latour and the references cited
by them in support of such statements. The review contains 18 paragraphs and six
main headings. The reference numbers cited and the section headings listed here
are those in the paper by Nahas & Latour.
Paragraphs 1 to 6 comprise
the introduction to the review and prepare the reader for the "new" information
which will demonstrate the "new" evidence for the mutagenic effects of cannabis
which "have now been reported in man" (end of para 6). In paragraph 1, the
authors cite a recent (widely cited) major review of the health effects of
cannabis by Hollister (ref 1) as an example of how the adverse
effects of cannabis have been "trivialized".
Nahas & Latour cite
Hollister to claim "Because of its lack of life threatening effects, cannabis
has been called a 'soft drug', no more damaging than coffee or tobacco".
Hollister actually concluded: "Compared with other licit social drugs such as
alcohol, tobacco, and caffeine, marijuana does not pose greater risks. One would
wonder, however, if society were given a choice based on current knowledge,
whether these drugs would have been granted their present status of acceptance".
Contrary to the inference of Nahas & Latour, Hollister's review included the
effects of long-term use and nowhere does he refer to marijuana as a "soft"
drug. This citation misquoted what remains a most authoritative review of the
health effects of cannabis.
To counter Hollister's
review Nahas & Latour claimed that information has emerged on the damaging
effects of cannabis "which have only recently been reported" (ref 2, Nahas & Latour 1991) and
"which confirm experimental observations" (ref 3, Nahas 1986). In fact, ref. 2 is a symposium proceedings edited
by Nahas in 1990. Many of the symposium papers are rewritten versions of other
papers cited below, i.e. the findings were not new. ref 3 was a review by Nahas
published in the Medical Journal of Australia in 1986, and therefore was neither
experimental nor new. Fourteen of the citations in that review (Nahas,
ref
3) were traced
to publications by, or in symposia proceedings organized by, Nahas. Citation of
reviews and symposium proceedings might be considered reasonable when reviewing
a large body of literature, but that process could not be considered in itself
to provide substantial new experimental evidence. Furthermore, 12 of the
references in the Nahas & Latour 1992 paper were dated between 1976 to 1985,
and were all available to Hollister for his 1986 review.
In paragraphs 2 to 6 Nahas
and Latour compare the properties of marijuana and tobacco smoke, and summarize
some of the pharmacological properties of tetrahydrocannabinol (THC). There is
no new information in these paragraphs which include only one citation, that of
Turner, 1980 (ref 4).
Paragraph 6 cites two
references to indicate that cannabis extracts are mutagenic (ref 5; Busch et al 1979) and that "THC
and other cannabinoids impair DNA and RNA synthesis in cell cultures" (ref 6; Leuchtenberger et al 1973). It is
not surprising that marijuana smoke condensates are mutagenic in
"Pulmonary toxicity
and carcinogenicity of the aerodigestive tract"
Paragraph 6 concludes with
the statement that "such properties account for the toxicity which has now been
reported in man". In paragraphs 7 and 8 the claim is made that aerodigestive
cancers are common in young cannabis smokers and is supported by two references,
8 and 9, both of which are dated 1980 and therefore are not "new" evidence.
ref 8 (Tashkin et al 1980) refers
predominantly to large airways obstruction. The link between this finding and
cancer implied by Nahas & Latour is speculative. ref 9 (Tennant et al 1980) was cited to
claim the identification of squamous cell hyperplasia in bronchial biopsies from
heavy hashish users. In fact, the original reference concluded "chronic hashish
smoking combined with cigarette smoking appear to have more deleterious
pulmonary effects than either hashish or cigarettes smoked alone". Paragraph 9,
it seems, provides five references as "documentation" of a link between
aerodigestive tract cancer and cannabis. Of these only ref 14 (Caplan & Briham, 1990) is
correctly cited; ref 10 (Donald, 1991 and is actually from
ref 2) is cited by Nahas & Latour as
reporting 12 cases of advanced head and neck cancer in young patients: "All had
been daily marijuana or hashish users since high school, but did not smoke
tobacco or use much alcohol". Perusal of the 11 cases reported in this paper
revealed that six users also smoked tobacco and three reported "heavy alcohol
intake". To fail to acknowledge the reported tobacco use and heavy alcohol
intake gives the inaccurate implication that cannabis use was the sole risk
factor in these cases.
ref 11 (Endicott & Skipper, 1991), is
from the proceedings of a symposium in
In paragraph 10 Nahas &
Latour cite ref 15 (Robison et al 1989) with these
words: "The study showed there was a tenfold risk of leukemia in the offspring
of mothers who had smoked just before or during pregnancy ... Children exposed
to marijuana in utero developed the disease at a younger age compared with the
controls (at a mean age of 19 months compared with 93 months) and showed clonal
abnormalities". Nahas & Latour did not mention the qualifications expressed
by the authors of the original paper in the interpretation of these data.
They pointed out that their
data may have been influenced by confounding variables and, although the
association was significant, the risk confidence intervals were large (95%
confidence interval of risk was 1. 42 to 85.20) for all mind-altering drugs. Of
greater concern, the authors pointed out that marijuana use (obtained by
telephone interview) was greatly under-reported in the entire sample,
questioning the validity of the data. They point out that rates of use of 5% and
0.5% were reported in the affected and control groups, respectively. Other
studies of use during pregnancy report frequencies greater than 10%. The authors
stressed the need to confirm this observation in a larger sample, presumably
with more accurate reporting of maternal drug use. If these results were
confirmed by a large, properly controlled study there would be cause for
concern. However, Nahas & Latour presented this preliminary and, perhaps,
erroneous, observation as an established fact. They omitted any mention of the
cited authors' cautions.
"Damaging effects of
marijuana on human fetal development"
Paragraph 10
begins with the assertion that "marijuana products were toxic to fetal
development in all species studied: fish, birds, rodents, hamsters, rabbits,
dogs and monkeys", citing ref 16 (Rosenkrantz, 1979) as the source.
In fact, the sum paper, conducted exclusively in that although so me feticidal
effects "No drug-related teratogenic effects were found Nahas & Latour then
state "Offspring also displayed retarded development and behavioral anomalies",
citing in this case ref 17 (Dalterio et al 1981). Ref 17 was
actually concerned with the effects of THC on plasma testosterone concentrations
in adult male mice. The experiments did not consider fetal development in any
way and no reference was made, even in discussion, to 'retarded development' .
No behavioral effects were examined and discussion of' behavioural anomalies'
was confined to anecdotal reports of aphrodisiac effects of marijuana in humans
and copulatory behavior in male mice. This section continues with the assertion
that human birth defects such as "lower weight and head circumference" occur in
marijuana smokers. The support for this statement was in two references, 18 and
19. ref
18 (Qazi et al
1985) was a summary of five case reports of offspring of young women who lived
in the central
In support of their
contentions Nahas & Latour also cited two subsequent studies, ref 20 (Hatch & Bracken, 1986) and
ref
22 (Zuckerman
et al 1989). Both studies found reduced birth-weight or small for gestational
age, but Hatch Bracken (ref 20) made no mention of head
circumference. ref 22 was from the same group of
ref
19 and reported
that low birth-weight and length were associated with validated cannabis use,
but reduced head circumference was not, when potentially confounding variables
were controlled for. This study, published in a highly respected journal, was
the most thorough of those cited. As such, Nahas & Latour's claim of "lower
head circumference" (with what we believe to be an implicit message of brain
damage) could be seen as a misrepresentation of the cited refs 20 and 22.
The study cited to support
"possible impairment of fetal brain development", ref 21 (Lester & Dreher, 1989),
examined the cries of offspring of very heavy chronic cannabis-using mothers in
Regardless of how the
findings of Zuckerman et al (ref 22) and others were viewed, there is
concern that the major observations, of reduced birth-weight and length, may be
valid. The prudent approach suggested by Hollister (ref 1) is worth noting: suggested 'While
no definite clinical association has yet been made between cannabis use during
pregnancy and fetal abnormalities, such events are likely to be rare at best and
could easily be missed. The belated recognition of the harmful effects on the
fetus of smoking tobacco and drinking alcoholic beverages indicates that the
same caution with cannabis is wise".
"Marijuana and the
brain"
Nahas &
Latour claim that the well recognized acute impairment of mental performance
continues well beyond the period of cannabis intoxication, citing ref 23 (Soueif 1976; unrefereed conference
proceedings). This information is neither new nor are the results consistent
with larger studies of cannabis using populations (cf ref 1). Nahas & Latour next cite a
study conducted in 1973 (ref 24; Satz et al 1973) which reports
negative findings for this effect. However, they then cite a follow-up study by
the same research group (ref 25; Page et al 1988), the results of
which they describe in this way: "performed on the same cohort of Latin American
marijuana users and on non-using controls, showed selective impairment of
short-term memory skills in cannabis users, contradicting the results obtained
10 years earlier". Examination of this report (Page et al ref 25) reveals that the study actually
found significant but "not particularly robust" deficits among users for three
performance measures from at least 14 memory and performance tests. None of the
tests showing deficits had been included in the original study, and only 57 of
the original 82 subjects completed these tests. Although this second study found
small effects, the authors described these as "subclinical". The authors also
pointed out that their results may have been complicated by the consequences of
the increased social disapproval of marijuana use in
Nahas & Latour then
cite a study by Varma et al (ref 26; Varma et al 1988) which they
summarize as reporting "short term memory impairment in heavy marijuana smokers
studied in
The next citation was that
of Schwartz et al (ref 27, 1989) which Nahas & Latour
describe as "an exceptionally well controlled study which showed persistent
short term memory impairment in a group of middle class American adolescents who
used marijuana". Schwartz et al actually described their work as a "pilot study"
and commented that "Although it was conducted in a methodologically more
rigorous manner than many past efforts, some problems admittedly remained in the
study design. Larger research efforts with at least 25 subjects in each group
better matched for gender would have improved the results of our study".
They compared performance
of six short-term memory tests between 10 patients enrolled in an out-patient
drug abuse treatment program for adolescents and 17 (presumably) matched
controls. Significant deficits were found for two of the six memory tests.
Although Nahas & Latour reported a persistent deficit in memory tests during
a 6-week drug-free period to support the notion of long-term memory impairment,
this finding was a source of concern for the authors because it could have
indicated initial differences (unrelated to drug use) between groups. It is
nevertheless an important study which should be replicated.
In paragraph 12, Nahas
& Latour seem to attempt to establish by implication that cannabis produces
subtle brain damage. They first claimed (without any experimental evidence) that
disruption of short-term memory produced by cannabis (which is not actually
established) is similar to that found in brain-damaged patients presenting with
memory disorders. We do not agree with their claim that "These memory deficits
.. have been traced to impairment of the hippocampus ... " and "Impaired mental
performance ... has been related to impaired turnover of acetylcholine", as
these were not substantiated by any citations. They cited Herkenham et al
(ref 28,
1990) who
reported that THC binds to the hippocampus (as well as many other brain
structures), which is thought to be involved in memory formation. However,
contrary to the implications of Nahas & Latour, these observations do not in
any way implicate a damaging effect of THC on this brain structure, or memory.
Nahas & Latour then
cited a positron emission tomography (PET) study of glucose utilization
(ref
29; Volkow et
al. 1991, which is contained in ref. 2,
a Nahas &
Latour edited colloquium) to indicate brain regions associated with impairment.
The technique examines metabolic activity in different brain regions to discover
which brain regions are activated during intoxication. It does not measure
impairment. The study was described by the authors as "preliminary" (six
subjects) and no significant differences were actually reported for glucose
utilization in any brain region. Nevertheless, Nahas & Latour concluded that
"THC significantly alters glucose metabolism in the frontal and parietal lobes
and cerebellum for several hours. Such metabolic changes in areas of the brain
which control memory and information processing are associated with impairment
in the performance of complex tasks".
Paragraph 13 cites studies
of the "lingering effects of marijuana on memory and coordination" when
volunteers were tested on a flight simulator. This study, (ref 30; Leirer & Yesavage, 1991)
reported marginal deficits in performance eight and 24 hours after smoking. It
is debatable whether they were statistically significant, since they used a
single tail t-test to test an unexpected outcome; i.e. a previous flight
simulator study had proven negative for the hypothesized long-term effect (cf
ref 1).
"Marijuana and road
accidents"
Nahas &
Latour introduced this section (in paragraph 14) with the implication that
impairment produced by cannabis is more serious than for alcohol. This was
implied by citing a report to suggest "that THC was 4000 times more potent than
alcohol in producing decrements in performance ... " (ref 31; Chesher et al 1985, unrefereed
symposium proceedings). The comparison in the original study was made to
indicate the differences in absolute potency of these drugs tested with a
specified test battery. Doses of alcohol are measured in grams and those of THC
in milligrams, so one would expect a potency ratio of this order.
Nahas & Latour then
describe railway and heavy vehicle accidents and make the claim "several major
railroad accidents have illustrated the impairing effects of marijuana on the
performance of complex tasks". They first mention a freight train crash which
occurred in January 1987 and then another crash which occurred one year later.
In each case railway-men who may have been culpable in these crashes were found
to have cannabinoids in their body fluids. These accident reports are not cited,
so it has not been possible to check whether the investigation actually claimed
a causative role for cannabis. Nahas & Latour have already noted in this
review the retention of cannabinoids in the body for up to a month (paragraph
5), so the detection of cannabinoids in the body fluids of a railwayman does not
provide proof of impairment at the time of the accident.
The authors maintain their
implication of cause- effect in paragraph
Nahas & Latour cite a
journal editorial (ref 33; Negrete, 1989) to support their
claim that "It is well established that marijuana smoking can trigger an acute
psychotic episode in schizophrenics". In fact, the editorial's main focus was to
draw attention to a study of Swedish military conscripts (see ref 34; Andreasson et al 1987). The
weaknesses of this study were criticized, as they have been elsewhere. (eg
Johnson et al Lancet 1988;i:592 593). mentions of triggering psychotic episodes
in this editorial were: "However, Andreasson et al's suggestion to the effect
that cannabis use is a 'stressor' capable of triggering the psychotic breakdown,
is an aspect that requires considerably more exploration and "Some clinical
studies indeed found that cannabis appears to enhance or magnify the of
schizophrenia-particularly those of the positive type-and it is therefore
conceivable that the drug might precipitate the clinical expression of a
disorder that would otherwise remain latent". Although there is general concern
that cannabis might exacerbate symptoms of schizophrenia, Nahas & Latour's
conclusions are not consistent with the overall emphasis and content of this
editorial.
Nahas & Latour cited
the study of Andreasson et al (ref 34, 1987, also rewritten in ref 2), which detected a six-fold greater
incidence of schizophrenia among heavy cannabis users (use on more than 50
occasions) than in non-users. However, when the authors corrected for other
variables, including previous psychiatric history, the relative risk dropped to
3.1 times (95% confidence interval: 2.1, 4.7). This was not mentioned by Nahas
& Latour.
It should be noted that
such studies cannot establish causality, as pointed out in 33 by Negrette and
other commentaries (see above), because:
These criticisms (ref 33) make Nahas & Latour' s claim
that "The ability of cannabis to induce long lasting mental disturbances in
Western man, now epidemiologically documented . ..", much less compelling than
implied. Summaries of " older anecdotal reports" (ref 35, a monograph edited by Nahas, not
retrieved) add little weight to their claims.
In the final paragraph of
this review Nahas & Latour claim to have demonstrated that the animal
research of the 1970s which described some of the possible areas of cannabis
toxicity have actually been "good predictors of the long term pathophysiological
manifestations observed 20 years later in chronic marijuana smokers". If their
evidence for this statement resides in the arguments presented in their review
in the Medical Journal of Australia in 1992 we cannot accept that they have
proved this claim. We have described here a high incidence of inaccuracies in
their description of the original data that they have cited. All these
inaccuracies operate in the direction of finding an adverse effect of marijuana.
These authors may hold a sincere belief that cannabis is a more dangerous drug
than the scientific literature presents it to be. However, the inaccurate
citation of scientific data as we have described in this analysis of their
review in the Medical Journal of Australia does nothing to advance rational
debate on this issue. The danger is that their review confused the thinking of
very many concerned people.
References cited by
Nahas and Latour
1. Hollister LE.
Health aspects of cannabis. Pharmacol Rev l986;38:1--32.
2. Nahas G, Latour C,
editors. Cannabis. ln: First lnternational Colloquium on Illicit Drugs. Advances
in the biosciences.
3 Nahas GG.
Cannabis: toxicological properties and epidemiological aspects Med J Aust
l986;145:82-7.
7 Cabral GA, Vasquez R.
Marijuana decreases macrophage antiviral and antitumor activities. ln: Nahas GG,
Latour C, editors. First lnternational Colloquium on Illicit Drngs. Advances in
the biosciences.
10 Donald pJ. Marijuana and
upper aerodigestive tract malignancy in young patients. ln: Nahas GG, Latour C,
editors. First lnternational Colloquium on Illicit Drugs. Advances in the
biosciences.
11 Endicott J, Skipper P.
Marihuana and the upper aerodigestive tract malignancy . in young objects. In:
Internationaies Symposium gegen Drogen in der Schweiz.
12
16 Rosenkrantz H. Effects
of cannabis on fetal development of rodents. In: Nahas GG, Paton WDM, editors.
Marihuana, biological effects.
26 Varma VK,
28 Herkenham M,
Lynn AB, Little MD, Cannabinoid receptor localization in brain. Proc Natl Acad
Sci
29
31 Chesher GB, Bird KD,
Stramarcos A, Nikias M. A comparative study of the dose response relationship of
alcohol and cannabis on human skills performance. ln:
32 Department of
Transportation,
33 Negrete JC.
Cannabis and schizophrenia [editorial].Br J Addict
l989;84:349-51.
Informó
Claudio
Izaguirre
Presidente
Asociación Antidrogas de la República
Argentina
Delegado por Argentina ante
Estados Unidos 1312 Capital Federal - 4942-1789 y 155 132
8916
asociacionantidrogas@drwebsa.com.ar
www.drwebsa.com.ar/aara
http://personales.clipsite.com.ar/IZAGUIRRE/IZAGUIRRE.htm