BACKGROUND: To identify which clinical factors can modify the probability o
f the appearance of the psychotic syndromes in patients with idiopathic Par
kinson's disease treated with levodopa.
PATIENTS AND METHODS: 214 patients were retrospectively studied to evaluate
the appearance of hallucinosis, delusions or mental confusion, from the be
gining of the treatment with levodopa to a transversal evaluation along the
course of the disease. To determine which clinical factors were independen
t predictors of psychosis, a multivariate logistic regression model was obt
ained, using the variables for which the univariate studies showed p values
under 0.25.
RESULTS: The multivariate model showed that the probability of developing p
sychosis during levodopa treatment was higher for the patients with interme
diate or advanced stages of the disease (Hoehn and Yahr scale), at the begi
ning of the treatment (OR: 4.5; 95% CI: 1.86-11.23), when amantadine was ad
ministrated as associated drug (OR: 3.31; 95% CI: 1.19-9.23) and for the pa
tients who presented meter fluctuations (OR: 3.08; 95% CI: 1.32-7.16). Univ
ariate studies showed a significant association between levodopa psychosis
and dyskinesias (univariate OR: 2.44; 95% CI: 1.12-5.33). Patients who suff
ered from pshychotic complications had received significantly higher mean l
evodopa daily dose (p = 0.016) and the punctuation reached in the Folstein'
s Mini-Mental Scale was significantly lower (p = 0.0001).
CONCLUSIONS: Levodopa psychosis appears in a "bad pronostic" group of patie
nts, characterized by a greater motor and cognitive impairment and by the o
ccurrence of other levodopa central adverse effects, higher doses of levodo
pa and a more frequent administration of other antiparkinsonian drugs.