Cannabinoids

Human Studies of Cannabinoids and Medicinal Cannabis 743

Fig. 1.Diary Card intoxication scores.BL, baseline.DB, double-blind. Patients self-titrated active medication
(THC:CBD) or placebo against symptom relief or intolerable unwanted effects. Doses reached a plateau after
4 weeks. Open patients from both arms re-titrated onto active medication. (Reproduced with kind permission
from Arnold Publishers)

3.3

Effects on Mental Health

All the following considerations refer to information derived from recreational
cannabissmokers,andtheimplicationsformedicinalusersareunknown.However,
an obvious difference exists between the two groups: the primary intention of the
former is to experience intoxication, whilst the vast majority of the latter seek to
avoid it.
Cannabis smoking is known to produce anxiety, dysphoria, panic, paranoia,
tiredness and low motivation in a proportion of users, particularly younger people
and those with unusual personalities or social disadvantage (Hall et al. 1994). Large
doses can produce a transient “toxic psychosis” with hallucinations and delusions
that generally resolves within a week or so of abstinence (Johns 2001). Although
there are exceptions to this, a consensus view among psychiatrists would be that
recreational cannabis use is likely to aggravate the symptoms and behavioural con-
sequences of pre-existing psychiatric illness (Johns 2001). This would suggest that
patients with existing psychiatric illness or a strong family history should avoid
cannabis-based medicines. Intriguingly, raised concentrations of endocannabi-
noids were discovered in the cerebrospinal fluid of schizophrenia patients in com-
parison with normal controls (Leweke et al. 1999), leading the authors to speculate
that an imbalance in endocannabinoid signalling may contribute to the pathogen-
esis of schizophrenia.
A much more controversial question is whether cannabis might actually be an
independent risk factor for schizophrenia in previously healthy subjects. Undoubt-
edly, cannabis smoking is more prevalent in psychiatric populations (Regier et al.
1990), but retrospective or cross-sectional studies are of no help in evaluating the