Parkinsonism and Related Disorders 16 (2010) 475e477

Contents lists available at ScienceDirect

Parkinsonism and Related Disorders

journal homepage: www.elsevier.com/locate/parkreldis

Short communication

The impact of and the factors associated with drooling in Parkinson's diseaseq
Joshua Leibner a, Amit Ramjit a, Laura Sedig a, Yunfeng Dai b, Samuel S. Wu b, Charles Jacobson IV c,
Michael S. Okun c, Ramon L. Rodriguez c, Irene A. Malaty c, Hubert H. Fernandez c, *
a
College of Medicine, University of Florida, Gainesville, FL, USA
b
Department of Epidemiology and Health Policy Research, University of Florida, Gainesville, FL, USA
c
Department of Neurology, College of Medicine, University of Florida, Gainseville, FL, USA

a r t i c l e i n f o a b s t r a c t

Article history: We administered a 7-question survey on drooling to PD patients and age-matched controls. Each subject
Received 3 July 2009 was assigned a drooling severity score and categorized as a “drooler” or a “non-drooler”. The age, disease
Received in revised form duration, motor scores, quality of life (PDQ-39), and levodopa equivalent daily dosage (LEDD) were
30 November 2009
compared between PD droolers vs. PD non-droolers.
Accepted 3 December 2009
58 PD patients and 51 age-matched controls participated. In PD patients, the mean: disease duration
was 10.96 years (SD 8.66) and UPDRS on motor score was 30.76 (SD 10.57). The drooling severity score was
Keywords:
significantly different between patients vs. controls (3.41 vs. .58; p < .01). 14% of controls vs. 59% of
Parkinson's disease
Drooling patients were droolers (p < .01). PD droolers scored worse on the ADL subscale of the PDQ-39 (p ¼ .031).
Sialorrhea Furthermore, PD droolers had significant difficulty speaking (7.27% vs. 0%; p < .01); eating (3.64% vs. 0%;
Hallucinations p ¼ .01); and socially interacting (12.73% vs. 0%; p < .01) compared to PD non-droolers. Interestingly, the
Quality of life hallucination component of the UPDRS Part I was significantly correlated with being a drooler (p ¼ .016).
None of the other variables have significant effect on drooling severity scores. There is a high prevalence
of drooling among PD patients compared to controls.
PD droolers had worse quality of life and had more difficulty speaking, eating and socially interacting
compared to PD non-droolers. Experiencing hallucinations was the only factor that correlated with being
a drooler and it may be confounded by medications.
Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction 17e77% of patients [3]. Drooling is thus a common inhibiting

Drooling is a frequent symptom of Parkinson's disease (PD) that
is estimated to occur in 75% of all patients at some point during the
evolution of the disease [1]. This disability can strongly impact
quality of life.
Excessive drooling can be due to either autonomic problems
(excessive production of saliva) or due to weak and infrequent
swallowing. It is now widely agreed upon that the difficulty and
decreased frequency in swallowing may be the primary reason for
drooling in PD. Despite the drooling, there is actually a reduced
level of salivary flow in PD compared to controls [2].
As many as 27% experienced severe drooling symptoms. Social
and emotional consequences of drooling have been reported in
symptom of PD.
There is a poverty of information available examining the factors
associated with drooling in PD. While some studies have observed
a relationship between dysphagia severity and disease duration [4],
other studies have not supported these correlations [5,6]. More-
over, the relationship between the overall severity of the disease
and medication effect on the prevalence of drooling remains
unclear. Further exploration into factors such as the Unified Par-
kinson Disease Rating Scale (UPDRS) score and the levodopa
equivalent daily dosage (LEDD) might provide us with insights on
PD patients at greater risk of suffering from significant drooling.
This study examines the prevalence, associated factors, and the
overall impact of drooling in a cohort of PD patients seen and
followed at a tertiary Movement Disorders Center.

q The review of this paper was entirely handled by an Associate Editor, Rober
Rodnitzky.
* Corresponding author at: Department of Neurology, University of Florida, Rm
L3-100, McKnight Brain Institute, 100 S. Newell Drive, Gainesville, FL 32610, USA.
Tel.: þ1 352 273 5550; fax: þ1 352 273 5575.
E-mail address: fernandez@neurology.ufl.edu (H.H. Fernandez).
2. Methods
After obtaining informed consent, consecutive patients with idiopathic Parkin-
son's disease (diagnosed by a fellowship-trained movement disorders specialist
using strict UK PD Brain Bank Criteria [8]) seen at the University of Florida Parkin-
son's Disease and Movement Disorders Center, between June and August of 2008,

1353-8020/$ e see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.parkreldis.2009.12.003
476 J. Leibner et al. / Parkinsonism and Related Disorders 16 (2010) 475e477

and age-matched control subjects were recruited to participate in this study. The Table 1
study was approved by the UF IRB. Demographic data, including age, disease Drooling survey.
duration, levodopa equivalent daily doses (LEDD) were recorded. All subjects
completed a self-administered seven question multiple choice survey on drooling 1. While you are awake,
[7]. The study was explained to the patients via a standardized presentation by 0 e I don’t drool
research assistants immediately before they filled out the questionnaire. Four of the 1 e Saliva wets my lips
questions (with a score range from 0 to 3) were summed and used to assign 2 e Saliva accumulates on my lips, but I don’t drool
a drooling severity score to each patient. All participants (PD patients and non-PD 3 e I drool
controls) were then categorized as “droolers” or “non-droolers”. A participant was 2. When you are asleep, how much saliva is in your mouth?
considered to be a “drooler” if s/he scored a “2” or higher on any of the four drooling 0 e I don’t notice an increase in saliva
severity questions. The three additional questions in the 7-question survey were 1 e I notice increased saliva in my mouth, but my pillow does not get wet
administered to investigate the relationship of drooling to speech, eating, and social 2 e My pillow gets wet
activity. While there was no formal cognitive testing done on the patients at the time 3 e My pillow and other bedclothes get wet
of completion of the questionnaire, the patients were determined to be capable to
consent and complete the drooling questionnaire without the assistance of other 3. During the day, when do you feel there is more saliva in your mouth?
individuals. Patients that were unable to understand the purpose of the study or had 0 e Never
difficulty completing the questionnaires were not included in the results of the 1 e At meal times
study. 2 e Throughout the day, not related to meals
Among PD subjects, the age, disease duration, UPDRS Parts II and III scores, 3 e All the time, even when I am asleep
Parkinson Disease Quality of Life Questionnaire (PDQ-39) scores, and LEDD were
4. How many times do you drool during the daytime?
compared between PD droolers vs. PD non-droolers using t-tests for initial
0 e Never
comparison. The Wilcoxon Rank test was used to evaluate speaking, eating, social
1 e Not more than 3 times
interaction, and the four components of the UPDRS Part I between PD droolers vs. PD
2 e Often. I have to carry a handkerchief with me all the time
non-droolers. Logistic regression was performed to determine the factors associated
3 e Permanently
with “PD droolers”.
5. Does accumulation of saliva in your mouth impair your eating ability?
0 e No
1 e I must swallow frequently to avoid difficulties
3. Results 2 e I have trouble eating
3 e I can’t eat at all

Fifty eight PD patients and 51 non-PD controls participated in 6. Does accumulation of saliva in your mouth impair your speech?
this study. The mean age of the PD patient group was 69.27 years 0 e No
1 e I must swallow frequently to avoid difficulties
(SD 5.17) and the mean age of the control group was 66.45 years (SD
2 e I have trouble speaking
9.17). In our PD group, the mean disease duration was 11 years (SD 3 e I can’t speak at all
8.66). The mean total UPDRS Part II, and III “on” scores were 13.15
7. When you go out or on social occasions, does saliva accumulation bother you?
(SD 5.92), and 30.76 (SD 10.57), respectively. The mean PDQ-39 0 e No
score was 164.05 (SD 110.3) and the subscale means were as 1 e I notice an accumulation, but it does not bother me
follows: mobility 29.69 (SD 26.80), activities of daily living 26.16 2 e I realize that other people notice it, but I can control the situation
(SD 21.54), emotions 18.12 (SD 15.85), stigma 13.26 (SD 18.10), 3 e I have stopped attending social meetings

social 8.04 (SD 15.42), cognitions 21.49 (SD 17.67), communications Lloret SP, Arce GP, Rossi M, Nemet MLC, Salsamendi P, Merello M. Validation of
21.05 (SD 19.55), and discomfort 29.09 (SD 19.35). The average a new scale for the evaluation of sialorrhea in patients with Parkinson's disease.
Movement Disorders 2007;22(1):107e11.
levodopa equivalent daily dose (LEDD) was 692.47 (SD 544.03). In
the patient group, 64.58% had a Hoehn and Yahr score of 2, 10.41%
had an H&Y of 2.5, 18.75% had an H&Y score of 3 and 6.25% had an
the PDQ-39 survey, there was a significant difference (p ¼ .03)
H&Y score of 4. Amongst the PD patients, only one non-drooler was
within the PDQ-39 subscale of activities of daily living.
on an anti-cholinergic medication (trihexyphenidyl).
A logistic regression analysis found the UPDRS Part I score to
The drooling severity score was significantly different between
have a significant effect on being a drooler (p < .01). Because the
patients and controls (3.41 vs. .58; p < .001). Only 14% of controls
UPDRS Part I is comprised of 4 distinct items on mentation,
were droolers while 59% of PD patients were droolers, demon-
behavior, and mood, we then analyzed each of the four components
strating more chances to be a drooler for patients than controls
separately using a logistic regression to follow up on which variable
(p < .001).
within the UPDRS part I was influencing the drooling. The score on
One-third of PD droolers claimed that they must swallow
the hallucinations component of the UPDRS part I was significantly
frequently to avoid difficulties in speaking while only 4.3% of PD
associated with being a drooler (p ¼ .016). For every one unit
non-droolers had to make the same compensation. An additional
increase in the hallucination score, the odds to be a drooler
12% of PD droolers stated that they have trouble speaking. None of
increased by a factor of 4.93. The 3 remaining components of the
the PD non-droolers had the same difficulty (p < .001).
UPDRS part I were not found to predict the chance of being
Among PD droolers 15.6% claimed they must swallow frequently
a drooler. There was no difference in LEDD between droolers and
to avoid difficulties in eating. None of the non-drooling PD patients
non-droolers on a 2-tailed t-test. The patient’s age, disease dura-
had this same difficulty. Furthermore, 6.3% of PD patients that drool
tion, UPDRS II and III, and the LEDD were all non-significant. See
claimed to have trouble eating. There were no non-droolers that
Table 2.
had difficulty eating (p ¼ .009).
Among PD droolers, only 18.8% stated that accumulation of
saliva did not bother them in social situations, while 18.8% claimed 4. Discussion
that they knew other people may notice the saliva but that they
could control it and another 3.1% said that they stopped attending Drooling is a difficulty frequently encountered by patients with
social meetings altogether. Of those PD non-droolers, there were PD. The results illustrate a significant difference in the prevalence of
none that had any of the above problems with social situations due drooling between patients with PD compared to their age-matched
to saliva accumulation (p < .001). See Table 1. controls. While differences in the overall quality of life survey
While there was no significant difference (p ¼ .08) in the overall between PD droolers and PD non-droolers was not found to be
quality of life between droolers and non-droolers as determined by significant, the activities of daily living subset was most affected by
 J. Leibner et al. / Parkinsonism and Related Disorders 16 (2010) 475e477 477

Table 2 disease duration or motor score. A recent study demonstrated

(a): Performance of Parkinson droolers vs. non-droolers in the Drooling Survey. patients with drooling difficulty had a mean duration of disease of 7
(b): Factors associated with drooling in Parkinson's disease.
years with no difference in disease duration between droolers
(a) and non-droolers3. One explanation for finding no relationship
Disabilitya PD drooler (%) PD non-drooler (%) p-value between the duration of disease and drooling is the ambiguity of
Speaking 0 54.54 95.65 p < .001 “duration of disease”.
1 45.45 4.35 Although one also might expect the UPDRS ADL subscale (Part II)
Eating 0 78.12 100 p ¼ .009
to have correlation with the PDQ-39 ADL subscale among PD
1 21.87 0 droolers, this relationship was not significant. While the UPDRS II
and the PDQ-39 ADL subscale are overlapping, there are some
Social interaction 0 59.37 100 p < .001
1 40.63 0 differences: 1) the UPDRS is clinician administered and therefore
rater-interpreted while the PDQ-39 is self-administered; 2) the
(b)
timeline of the ADL assessment in UPDRS II is the ‘previous week’,
Logistic regression for dependent variable drooler while that of the PDQ-39 is the ‘previous month’; 3) the UPDRS II
Variable PD PD Odds p-value generally measures functional severity while the PDQ-39 measures
droolers non-droolers ratio the impact of functional limitation in quality of life; and, 4) the
Age (mean in years) 69.14 69.47 .91 .185 items probed in the UPDRS II and PDQ-39 are not exactly identical.
Duration of Disease (mean) 12.12 9.33 1.02 .642 There are a few weaknesses in this study which must be recog-
UPDRS Part I nized in the analysis and interpretation of the data. While a corre-
Mentation (% 1) 82.15% 57.15% 3.53 .076 lation was found between hallucinations and drooling, there was no
Hallucinations (% 1) 67.85% 28.58% 4.93 .016 temporality established and it is unknown whether patients who
Depression (% 1) 50% 38.10% 1.15 .824
Apathy (% 1) 71.43% 52.39% 1.90 .211
experienced both of these symptoms developed hallucinations or
UPDRS Part II (mean) 15.54 10.29 .95 .594 drooling first. If a temporal relationship exists between these two
UPDRS Part III (mean) 32.22 28.61 1.04 .416 symptoms, it would help elucidate further understanding of their
PDQ-39b (mean) 184.04 132.10 1.00 .527 correlation. A thorough neuropsychological testing battery should
Levodopa daily (mean in mg) 709.67 665.89 1.00 .650
also be considered in any studies following up the relationship of
Equivalent dosage (LEDD)
drooling and PD. It is possible that a large number of patients that
a
Answers were graded from 0 to 3 in severity. 0 ¼ Never a problem. 3 ¼ Severe
were defined as droolers had some cognitive deficits that may be
problem.
b
t-test for the ADL subset of the PDQ-39 scale was significant (p ¼ .031).
clearly illustrated by neuropsychological testing. Another possibility
is that neuropsychological testing could reveal small cognitive
deficits in droolers that may not be apparent from day-to-day
interaction thus identifying drooling as an early sign of decreasing
drooling. Moreover, speaking, eating, and comfort in social situa- cognitive skills in PD patients.
tions were all identified as being impacted by drooling as PD It is evident that drooling is an important symptom of Parkin-
droolers frequently responded with difficulties in these tasks as son's disease which must be recognized by clinicians. Drooling can
compared to PD non-droolers. greatly affect a patient’s day-to-day lifestyle and should not be
Most interestingly, the presence of hallucinations (based on ignored amongst the vast amount of symptoms a PD patient can
UPDRS Part I) was determined to be the only significant component endure.
influencing the “drooler” status of the PD patients in our cohort. This
result is very intriguing and no direct relationship between drooling
and hallucinations in PD patients has previously been reported. One References
potential explanation is that sialorrhea is a side effect to the anti-
psychotic agents and cholinesterase inhibitors used to control [1] Johnston BT, Li Q, Castell JA, Castell DO. Swallowing and esophageal function in
hallucinations. More patients classified as droolers were on medi- Parkinson's disease. American Journal of Gastroenterology 1995;90:1741e6.
[2] Bagheri H, Damase-Michel C, Lapeyre-Mestre M, Cismondo S, O’Connell D,
cations to treat hallucinations and these medications can potentially Senard JM, et al. A study of salivary secretion in Parkinson's disease. Clinical
cause sialorrhea: clozapine (n ¼ 1), donepezil (n ¼ 4), and perhaps Neuropharmacology 1999;22:213e5.
seroquel (n ¼ 3). None of the non-droolers were on any anti-hallu- [3] Kalf JG, Smit AM, Bloem BR, Zwarts MJ, Munneke M. Impact of drooling in
Parkinson's disease. Journal of Neurology 2007;254:1227e32.
cinatory medications. Hypersalivation has been reported as a side [4] Clarke CE, Gullaksen E, Macdonald S, Lowe F. Referral criteria for speech and
effect in patients taking atypical antipsychotics such as clozapine [9] language therapy assessment of dysphagia caused by idiopathic Parkinson's
but has not been reported, to our knowledge, with quetiapine use. disease. Acta Neurologica Scandinavica 1998;97:27e35.
[5] Nilsson H, Ekberg O, Olsson R, Hindfelt B. Quantitative assessment of oral and
There is also the potential of an unknown underlying mechanism
pharyngeal function in Parkinson's disease. Dysphagia 1996;11:144e50.
which can be shared by the neural pathways involved in hallucina- [6] Proulx M, de Courval FP, Wiseman MA, Panisset M. Salivary production in
tions and drooling. Further evaluation of hallucinations and drooling Parkinson's disease. Movement Disorders 2005;20:204e7.
[7] Lloret SP, Arce GP, Rossi M, Nemet MLC, Salsamendi P, Merello M. Validation of
should be done to determine if this is a reproducible finding and
a new scale for the evaluation of sialorrhea in patients with Parkinson's disease.
what the cause of the relationship may be. Movement Disorders 2007;22:107e11.
It is also important to recognize the variables which were not [8] Gibb WRG, Lees AJ. The relevance of the Lewy body to the pathogenesis of
found to be significantly related to drooling. While many physicians idiopathic Parkinson's disease. Journal of Neurology, Neurosurgery & Psychiatry
1988;51:745e52.
may have the conception that drooling comes about in later stages [9] Davydov I, Botts SR. Clozapine-induced hypersalivation. The Annals of Phar-
of PD, no significant correlation was found between drooling and macotherapy 2000;34:662e5.