Self‐Perception of Drooling Impact in People with Parkinson's Disease: A Case–Control and Cross‐Sectional Study

Abstract

Background

Drooling, defined as the involuntary loss of saliva from the anterior oral cavity, is potentially problematic for people with Parkinson's disease (PwP). However, there is little research on how PwP perceive the impact of drooling and what factors contribute to it.

Objectives

The objective was to evaluate the self‐perceived impact of drooling in people with and without Parkinson's disease and the contributing clinical factors in PwP.

Methods

We conducted a cross‐sectional and case‐control study. Participants were clinically examined, and the primary outcome was the Sialorrhea Clinical Scale for Parkinson's disease. Clinical variables were compared between PwP and control subjects using the Mann‐Whitney test, correlations between drooling impact and clinical factors in PwP were analyzed using Spearman's test, and predictors were identified using linear regression.

Results

The study included 101 PwP and 101 sex‐ and age‐matched controls. PwP experienced significantly more severe impact of drooling compared to controls across all domains: diurnal and nocturnal drooling, drooling severity and frequency, social discomfort, speech, and eating impairments. The greater impact of drooling in PwP was significantly associated with drooling severity, disease duration, levodopa equivalent daily dose, clinical global impression of saliva accumulation (CGI‐S), chewing, swallowing, speech, oromotor, motor and non‐motor impairments. Significant predictors of greater impact of drooling in PwP include drooling severity, higher CGI‐S, facial expression, and swallowing impairments.

Conclusions

PwP have a significantly greater impact of drooling compared to controls, affecting several domains. Drooling impact and its contributing clinical factors should be investigated in a Parkinson's consultation.

Drooling, the involuntary loss of saliva from the anterior oral cavity, ¹ is common in people with Parkinson's disease (PwP). ² Despite its high pooled prevalence of 56%, ³ it is often neglected in clinical practice. ⁴

The consequences of drooling include perioral skin damage, ⁵ , ⁶ , ⁷ an increased risk of infection, ⁷ and potential impact on food intake and speech. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² Drooling also affects psychosocial well‐being, ¹¹ , ¹² leading to embarrassment, ¹³ reduced self‐esteem, and social isolation. ¹⁴ Between 17% and 77% of PwP with drooling report social and emotional consequences. ¹³ As Parkinson's disease (PD) progresses, drooling‐related discomfort worsens. ¹⁵

However, research on drooling impact and contributing factors remains limited. ¹¹ A better understanding of self‐perceived impact is needed to guide assessment and interventions, improving quality of life.

The primary aim of this study was to investigate the self‐perception of the impact of drooling in PwP and matched controls without PD. The second and third aims were to explore correlations between clinical factors and the self‐perceived impact of drooling in PwP, and to identify predictors of this impact.

Methods

Study Design

The study was a case‐control, cross‐sectional study, including a group of PwP and a control group of people without PD.

Participants

Participants were recruited through convenience sampling by neurologists at Hospital de Egas Moniz. The inclusion criterion for PwP required a PD diagnosis for at least 3 years according to the UK PD Brain Bank criteria. ¹⁶ Controls were matched to PwP by age (±3 years) and sex. Exclusion criteria for both groups included a diagnosis of Sjögren's syndrome, neurological conditions other than PD, head and neck surgery (except deep brain stimulation), and congenital malformations that could affect saliva control.

Data Collection

Data were collected individually between February and July 2023, with each session lasting approximately 2 hours. A non‐blinded speech‐language pathologist with more than 5 years of experience in swallowing and movement disorders conducted all procedures. The standardized sequence of data collection was as follows: (1) sociodemographic and clinical data, (2) Montreal Cognitive Assessment (MoCA), (3) Movement Disorder Society‐Unified Parkinson's Disease Rating Scale (MDS‐UPDRS), (4) Sialorrhea Clinical Scale for Parkinson's Disease (SCS‐PD), (5) clinical global impression of saliva accumulation (CGI‐S), (6) Frenchay Dysarthria Assessment, second edition (FDA‐2), (7) unstimulated saliva collection, (8) stimulated saliva collection, and (9) Swallowing Clinical Assessment Score in Parkinson's disease (SCAS‐PD). Participants completed the SCS‐PD, as well as the non‐motor aspects of experiences of daily living (Part I) and the motor aspects of experiences of daily living (Part II) of the MDS‐UPDRS, independently or with the assistance of caregivers or spouses.

Primary Outcome Measure

The primary outcome was the SCS‐PD, a 7‐item scale assessing self‐perceived discomfort related to drooling in the following domains: diurnal drooling, nocturnal drooling, drooling severity, speech impairment, eating impairment, frequency of drooling, and social discomfort. Each question contains four response options with a score ranging from 0 to 3. The total score ranges from 0 to 21, with higher scores indicating greater self‐perceived drooling impact. ¹⁷ , ¹⁸

The participants completed the scale based on symptoms from the previous week. Spouses or caregivers could assist, but the SCS‐PD primarily aimed to capture participants' thoughts and feelings. ¹⁸

Secondary Outcome Measures

Structured interviews were conducted to collect sociodemographic data and medication. The levodopa equivalent daily dose (LEDD) was calculated using conversion factors proposed by Tomlinson et al. ¹⁹ Drugs with anticholinergic effects were based on a universal list using objective pharmacological data, according to the Anatomical Therapeutic Chemical classification. ²⁰ These drugs bind to five muscarinic receptors, with their antagonism serving as the primary mechanism of action (eg, oxybutynin, trihexyphenidyl, and ipratropium bromide), as well as drugs whose anticholinergic activity is not connected with their primary therapeutic target or mechanism (eg, antidepressants, antipsychotics, and antihistamines). ²⁰

The MoCA ²¹ was used to assess cognitive abilities, with a total score ranging from 0 to 30 points and higher scores indicating better cognitive performance. ²¹ , ²²

The MDS‐UPDRS was used to assess features associated with PD, including nonmotor (Part I) and motor aspects (Part II) of daily living, motor examination (Part III), and motor complications (Part IV). The total score ranges from 0 to 260, with a higher score suggesting more significant impairments. ²³

As no validated tool exists to assess drooling severity in PwP, we used item “2.2 Saliva and drooling” from the MDS‐UPDRS. This item asks the patient or caregiver the following: “Over the past week, have you usually had too much saliva during when you are awake or when you sleep?” Response options range from normal (0) to severe (4), reflecting the progression of disability or impairment. ²⁴

Total axial signs were calculated by summing MDS‐UPDRS items 3.1 (speech), 3.3 (neck rigidity), 3.10 (gait), 3.11 (freezing), 3.12 (postural stability), and 3.13 (posture). ²⁵ , ²⁶ The total score ranges from 0 to 24, with higher scores indicating more severe axial signs.

The Hoehn and Yahr (H&Y) scale was used to grade the PD severity, ²⁷ ranging from 0 to 5, with higher scores reflecting more advanced stages. ²³

The FDA‐2 was used to evaluate oromotor and speech functions, ²⁸ with a total score ranging from 0 to 104, where higher scores indicate better performance.

The SCAS‐PD was used to assess swallowing function with water (20 mL), yogurt (10 mL), and a biscuit. Scores range from 0 to 354, with higher values indicating greater swallowing impairments. ²⁹

Saliva collection involved stimulated and unstimulated samples. Participants were instructed to abstain from smoking, eating, drinking, oral hygiene, or chewing gum for at least 1 hour before collection. Because circadian variations can influence saliva flow rate, ³⁰ , ³¹ collection was conducted between 8 am and 1 pm.

For unstimulated saliva collection, participants first rinsed their mouths with deionized water and relaxed for 5 minutes. ³² Before collection, participants emptied their mouths of saliva. They were asked to minimize orofacial movements during collection, sitting comfortably for 5 minutes with eyes open, head tilted down, and mouth slightly open. A funnel was placed under the lower lip to direct the saliva into a measuring tube. ³²

To collect stimulated saliva, participants chewed tasteless paraffin wax (Saliva‐Check buffer, GC, reference: GC720000) for 5 minutes, spitting saliva into a funnel every 60 seconds. A metronome set to 70 chews per minute was used to maintain a steady chewing rate. The samples collected for the first 2 minutes were discarded to allow participants to familiarize with the procedure. ³²

For both procedures, the total amount of saliva collected was converted into a mean rate in milliliter per minute.

The CGI‐S was specifically developed by the authors for this study to assess the severity of saliva accumulation in the oral cavity. It is a 10‐point visual analog scale, with 0 representing no accumulation and 10 indicating the worst possible accumulation. The clinician assigns a score by observing saliva accumulation after the participants open their mouths.

Statistical Analysis

The sample size was calculated using G*Power 3.1 ³³ software. The estimation was performed a priori based on a study comparing SCS‐PD total score between PwP (M = 3.5; standard deviation [SD] = 3.8) and controls (M = 0.7; SD = 1.2). ³⁴ Although an effect size of d = 0.99 was obtained, a more conservative effect of d = 0.5 was considered. Assuming a moderate effect size of interest and a significance level of α = 0.05 (type I error), a total sample size of 176 participants (88 per group) was required to achieve enough power (1–β = 0.95; type II error).

Statistical analysis was performed using IBM Statistical Package for Social Sciences SPSS (version 29.0.1.0) and R (version 4.2.2). Descriptive analysis was conducted for sociodemographic and clinical characteristics. For inferential analysis, the Mann‐Whitney test was used due to the non‐normal data distribution. It was performed to compare differences between PwP and controls across primary and secondary outcome measures: self‐perceived impact of drooling (SCS‐PD), disease (duration, H&Y stage, LEDD), cognition (MoCA), motor and non‐motor domains (MDS‐UPDRS—total score, Part III, item 3.13 posture, axial signs), oromotor and speech function (MDS‐UPDRS—item 3.1 speech, item 3.2 facial expression, FDA‐2), chewing and swallowing (MDS‐UPDRS—item 2.3 chewing and swallowing, SCAS‐PD), and saliva and drooling (MDS‐UPDRS—item 2.2 saliva and drooling, unstimulated and stimulated saliva flow rate, CGI‐S). To account for potential confounders, an ANCOVA with robust standard errors was performed to compare SCS‐PD scores between PwP and controls, adjusting for age, cognition, use of anticholinergic drugs and botulinum toxin, and mode of completion (independent or assisted). Unstimulated and stimulated salivary flow rates were also analyzed using analysis of covariance (ANCOVA) with robust standard errors, controlling for the use of anticholinergic drugs and botulinum toxin. The Benjamini‐Hochberg procedure was used to correct for multiple comparisons and reduce the false discovery rate, thereby avoiding type I errors. A P‐value ≤ 0.05 was considered statistically significant.

The Spearman test was conducted to explore correlations between the self‐perceived impact of drooling in PwP and factors such as age and the secondary outcome measures. To control for anticholinergic drugs and botulinum toxin, Spearman's partial correlations were used to assess the relationships between salivary flow rates and SCS‐PD. Furthermore, to account for the effect of disease duration, partial correlations using Spearman's test were also performed to analyze the relationship between the SCS‐PD score and the severity of drooling (MDS‐UPDRS—item 2.2). Correlations were categorized as strong (≥0.70), moderate (0.40–0.69), or low (≤0.39). ³⁵ P‐values were corrected for multiple testing using the Benjamini‐Hochberg procedure, with a significance level of P ≤ 0.05.

A linear regression model was used to estimate the relationship between the severity of the self‐perceived impact of drooling (SCS‐PD) in PwP and potential predictors, including age, sex, disease duration, cognition (MoCA), motor domain (MDS‐UPDRS—Part III), facial expression (MDS‐UPDRS—item 3.2), oromotor and speech function (FDA‐2), swallowing function (SCAS‐PD), drooling severity (MDS‐UPDRS—item 2.2), and CGI‐S. The selection method for the linear regression was stepwise forward Akaike Information Criterion‐based, using the step function from stats R package. A significance threshold of P ≤ 0.10 was set to reduce the risk of type II error (failing to detect a true effect), particularly given the exploratory nature of the study and the limited sample size. Odds ratios (OR), 95% confidence intervals (CI), standard errors, and P‐values of the variables considered in the model were reported.

Results

This study included 101 PwP and 101 controls, accurately matched by sex and age. Sociodemographic characteristics are presented in Table 1.

TABLE 1.

Sociodemographic characteristics

	PwP (n = 101)	Controls (n = 101)
Sex n | %
Male	60 | 59.4%	60 | 59.4%
Female	41 | 40.6%	41 | 40.6%
Age (years)
Mean ± standard deviation	72.19 ± 9.60	71.77 ± 9.76
Range	45–90	45–91
Level of education (ISCED classification) n | %
Primary school education (0–4)	41 | 40.6%	53 | 52.5%
Secondary school education (5–12)	34 | 33.7%	31 | 30.7%
Postsecondary nontertiary education	0 | 0%	4 | 4%
Short‐cycle tertiary education	1 | 1%	2 | 2%
Tertiary education (graduation and postgraduation)	25 | 24.8%	11 | 10.9%
Occupational status n | %
Working	9 | 8.9%	19 | 18.8%
Unemployed	3 | 3%	0 | 0%
Retired	89 | 88.2%	82 | 81.2%
Residence n | %
Own home	97 | 96%	99 | 98%
Family member's home	3 | 3%	0 | 0%
Retirement home	1 | 1%	2 | 2%
Cohabitation n | %
Family	90 | 89.1%	82 | 81.2%
Caregiver	5 | 5%	3 | 3%
Alone	6 | 5.97%	16 | 15.8%

Abbreviations: ISCED, International Standard Classification of Education; PwP, people with Parkinson's disease.

Self‐Perceived Impact of Drooling among PwP and Controls

The self‐perceived impact of drooling was significantly higher in PwP compared to controls, as indicated by the SCS‐PD total score (P = 0.000; Table 2). After adjusting for confounders, such as, age, cognition, use of anticholinergic drugs and botulinum toxin, and mode of completion (independent or assisted), SCS‐PD total score remains significantly different between PwP and controls (p < 0.05).

TABLE 2.

Comparison of Sialorrhea Clinical Scale for Parkinson's Disease (SCS‐PD) total score and domains between people with Parkinson's disease and controls

	PwP (n = 101)			Controls (n = 101)
	Median	Mean ± SD	Range	Median	Mean ± SD	Range	Significance
Diurnal drooling	1	0.97 ± 1.04	0–3	0	0.18 ± 0.52	0–2	U = 2943, z = −6.291, P = 0.000
Nocturnal drooling	0	0.87 ± 1.07	0–3	0	0.22 ± 0.59	0–3	U = 3402.5, z = −5.075, P = 0.000
Drooling severity	0	0.85 ± 1.14	0–3	0	0.03 ± 0.17	0–1	U = 2991, z = −6.869, P = 0.000
Speech impairment	0	0.36 ± 0.73	0–3	0	0 ± 0	0–0	U = 3939, z = −5.072, P = 0.000
Eating impairment	0	0.15 ± 0.46	0–2	0	0 ± 0	0–0	U = 4545, z = −3.401, P = 0.001
Frequency of drooling	0	0.45 ± 0.78	0–3	0	0.0 ± 0.1	0–1	U = 3679, z = −5.535, P = 0.000
Social discomfort	0	0.32 ± 0.71	0–3	0	0 ± 0	0–0	U = 4090.5, z = −4.693, P = 0.000
SCS‐PD—total score	2	3.94 ± 4.80	0–20	0	0.44 ± 0.83	0–4	U = 2300, z = −7.271, P = 0.000

Abbreviations: PwP, people with Parkinson's disease; SD, standard deviation.

Concerning SCS‐PD domains, PwP reported significantly more self‐perceived diurnal and nocturnal drooling, severe drooling, speech and eating impairment, frequency of drooling, and social discomfort compared to controls (Table 2).

Clinical Characteristics among PwP and Controls

PwP had significantly worse results compared to controls in cognitive function (MoCA, P = 0.002), motor and non‐motor domains (MDS‐UDPRS—total score, P = 0.002; Part III, P = 0.000; item 3.13, P = 0.002; axial signs, P = 0.000), oromotor and speech functions (MDS‐UPDRS—item 3.1, P = 0.002; item 3.2, P = 0.002; FDA‐2, P = 0.000), chewing and swallowing functions (MDS‐UPDRS—item 2.3, P = 0.000; SCAS‐PD, P = 0.001), drooling severity (MDS‐UPDRS—item 2.2, P = 0.000; CGI‐S; P = 0.001), and lower saliva flow rate (unstimulated, P = 0.001; stimulated, P = 0.011; Table 3 and Fig. 1). After adjusting for anticholinergic drugs and botulinum toxin, the difference in unstimulated salivary flow rate between PwP and controls remained significant (P = 0.040), whereas the difference in stimulated flow rate was no longer significant (P = 0.858).

TABLE 3.

Clinical characteristics of people with Parkinson's disease and controls

	PwP (n = 101)			Controls (n = 101)
	Median	Mean ± SD	Range	Median	Mean ± SD	Range	Significance
Disease
Duration (years)	7	8.25 ± 5.13	3–28	–	–	–	–
Hoehn and Yahr stage	2	2.24 ± 0.74	1–5	–	–	–	–
Daily levodopa equivalent dose (mg)	700	813.83 ± 572.66	80–3300	–	–	–	–
Cognition
Montreal Cognitive Assessment—total score	20	18.91 ± 5.96	4–30	22	21.52 ± 4.02	8–29	U = 3839.5, z = −3.042, P = 0.002
Motor and non‐motor domains
MDS‐UPDRS—total score	64	66.47 ± 32.23	14–167	5	6.49 ± 5.77	0–37	U = 55.5, z = −12.152, P = 0.002
MDS‐UPDRS—Part III	35	36.55 ± 18.13	9–99	1	1.53 ± 2.52	0–13	U = 5177.5, z = −12.294, P = 0.000
MDS‐UPDRS—item 3.13 posture	1	1.36 ± 1.16	0–4	0	0.27 ± 0.55	0–3	U = 2196, z = −7.644, P = 0.002
Axial signs	7	7.19 ± 4.37	1–23	0	0.21 ± 0.91	0–6	U = 263.5, z = −11.996, P = 0.000
Oromotor and speech functions
MDS‐UPDRS—item 3.1 speech	1	1.50 ± 0.97	0–4	0	0.01 ± 0.10	0–1	U = 734, z = −11.698, P = 0.002
MDS‐UPDRS—item 3.2 facial expression	1	1.53 ± 0.91	0–4	0	0 ± 0	0–0	U = 454.5, z = −12.334, P = 0.002
FDA‐2–total score	87	82.85 ± 15.09	12–103.5	102.5	102.04 ± 2.19	93.5–104	U = 226, z = −11.754, P = 0.000
Chewing and swallowing
MDS‐UPDRS—item 2.3 chewing and swallowing	1	1.13 ± 1.07	0–3	0	0.49 ± 1.40	0–10	U = 1787, z = −9.363, P = 0.000
SCAS‐PD—total score	2	11.87 ± 16.78	0–80	0	1.41 ± 3.96	0–25	U = 2890, z = −5.877, P = 0.001
Saliva and drooling
MDS‐UPDRS—item 2.2 saliva and drooling	1	1.33 ± 1.48	0–4	0	0.12 ± 0.45	0–2	U = 2652.5, z = −7.803.5, P = 0.000
Unstimulated total salivary flow rate (mL/min)	0.5	0.14 ± 0.19	0–1.10	0.25	0.32 ± 0.25	0–2	U = 2800, z = −5.611, P = 0.001
Stimulated total salivary flow rate (mL/min)	0.80	1.01 ± 0.71	0–5	1.2	1.18 ± 0.59	0–3.6	U = 3836, z = −2.538, P = 0.011
Clinical global impression of saliva	0	0.80 ± 1.71	0–7	0	0.02 ± 0.20	0–2	U = 3884.5, z = −5.031, P = 0.001

Abbreviations: FDA‐2, Frenchay Dysarthria Assessment, second edition; MDS‐UPDRS, Movement Disorder Society‐Unified Parkinson's Disease Rating Scale; PwP, people with Parkinson's disease; SCAS‐PD, swallowing clinical assessment score in Parkinson's disease; SD, standard deviation.

FIG. 1.

Key clinical characteristics of people with Parkinson's disease and controls. CGI‐S, clinical global impression of saliva accumulation; FDA‐2, Frenchay Dysarthria Assessment, second edition; MDS‐UPDRS, Movement Disorder Society‐Unified Parkinson's Disease Rating Scale; MoCA, Montreal Cognitive Assessment; ns, nonsignificant; SCAS‐PD, Swallowing Clinical Assessment Score in Parkinson's Disease. **P ≤ 0.001.

Correlations between Self‐Perceived Impact of Drooling and Clinical Characteristics in PwP

In PwP, significant strong positive correlations were observed between the SCS‐PD and the severity of drooling (MDS‐UPDRS—item 2.2; P = 0.000), as well as with the CGI‐S (P = 0.000). Additionally, significant moderate positive correlations were observed between the SCS‐PD and the following variables: global motor and non‐motor domains (MDS‐UDPRS—total score, P = 0.000; axial signs, P = 0.000), speech (MDS‐UPDRS—item 3.1, P = 0.000), and facial expression (MDS‐UPDRS—item 3.2, P = 0.000). The SCS‐PD showed a significant moderate negative correlation with oromotor and speech functions (FDA‐2, P = 0.000). Significant, low positive correlations were obtained between the SCS‐PD and the following variables: disease duration (P = 0.024), LEDD (P = 0.020), motor domain (MDS‐UPDRS—Part III, P = 0.002), posture (MDS‐UPDRS—item 3.13, P = 0.000), chewing and swallowing impairments (MDS‐UPDRS—item 2.3, P = 0.024; SCAS‐PD, P = 0.004), and unstimulated salivary flow rate (P = 0.030). No significant correlations were observed between the SCS‐PD and age, disease severity (H&Y stage), cognition (MoCA), and stimulated salivary flow rate (Table 4).

TABLE 4.

Spearman's correlation values between Sialorrhea Clinical Scale for Parkinson's Disease and clinical variables

	PwP (n = 101)
	r	p
Age	−0.063	0.299
Disease
Duration (years)	0.211	0.024
Hoehn and Yahr stage	0.162	0.064
Daily levodopa equivalent dose (mg)	0.224	0.020
Cognition
Montreal Cognitive Assessment—total score	0.029	0.388
Motor and non‐motor domains
MDS‐UPDRS—total score	0.411	0.000
MDS‐UPDRS—part III	0.293	0.002
MDS‐UPDRS—item 3.13 posture	0.338	0.000
Axial signs	0.415	0.000
Oromotor and speech function
MDS‐UPDRS—item 3.1 speech	0.460	0.000
MDS‐UPDRS—item 3.2 facial expression	0.508	0.000
FDA‐2—total score	−0.435	0.000
Chewing and swallowing
MDS‐UPDRS—item 2.3 chewing and swallowing	0.215	0.024
SCAS‐PD—total score	0.283	0.004
Saliva and drooling
MDS‐UPDRS—item 2.2 saliva and drooling	0.749	0.000
Unstimulated total salivary flow rate (mL/min)	0.199	0.030
Stimulated total salivary flow rate (mL/min)	0.036	0.385
Clinical global impression of saliva	0.711	0.000

Abbreviations: PwP, people with Parkinson's disease; FDA‐2, Frenchay Dysarthria Assessment, second edition; MDS‐UPDRS, Movement Disorder Society‐Unified Parkinson's Disease Rating Scale; SCAS‐PD, swallowing clinical assessment score in Parkinson's disease.

The correlation between the SCS‐PD score and the severity of drooling (MDS‐UPDRS—item 2.2) remained significant and strong (r = 0.820; P < 0.001) even after adjusting for disease duration.

After controlling for the use of anticholinergics drugs and botulinum toxin, the partial correlation between SCS‐PD and unstimulated salivary flow rate lost significance (r = 144, p = 0.166), whereas the partial correlation with stimulated salivary flow rate remained nonsignificant (r = 0.175, p = 0.088).

Predictors of Self‐Perceived Drooling Impact in PwP

The linear regression analysis revealed that higher drooling severity (MDS‐UPDRS—item 2.2) (OR = 4.26 [95% CI: 2.89–6.28], P = 0.000), greater accumulation of saliva (CGI‐S) (OR = 3.51 [95% CI: 2.50–4.93], P = 0.000), higher facial expression impairments (MDS‐UPDRS—item 3.2) (OR = 1.60 [95% CI: 0.96–2.67], P = 0.077), and higher swallowing impairments (SCAS‐PD) (OR = 1.03 [95% CI: 1.00–1.05], P = 0.078) were predictive variables for higher severity of self‐perceived impact of drooling (SCS‐PD). However, age (OR = 0.97 [95% CI: 0.93–1.01], P = 0.164) was not a significant predictor of the self‐perceived impact of drooling (Table 5).

TABLE 5.

Linear regression with the values between Sialorrhea Clinical Scale for Parkinson's disease and clinical variables

Variables	β	SE	P‐value
Age	−0.03	0.02	0.164
MDS‐UPDRS—item 3.2 facial expression	0.47	0.26	0.077
SCAS‐PD—total score	0.03	0.01	0.078
MDS‐UPDRS—item 2.2 saliva and drooling	1.45	0.20	0.000
Clinical global impression of saliva accumulation	1.26	0.17	0.000

Abbreviations: MDS‐UPDRS, Movement Disorder Society‐Unified Parkinson's Disease Rating Scale; SCAS‐PD, Swallowing Clinical Assessment Score in Parkinson's Disease; SE, standard error.

Sex, disease duration, cognition (MoCA), motor domain (MDS‐UPDRS—Part III), and oromotor and speech domains (FDA‐2) were excluded from the model as they were irrelevant for prediction.

Discussion

This study aimed to assess the self‐perception of drooling impact in PwP and controls. PwP reported a significantly higher impact across various domains, including diurnal and nocturnal drooling, drooling severity, speech impairment, eating difficulties, frequency of drooling, and social discomfort. These findings corroborate previous studies on the physical and psychosocial impact of drooling. ¹¹ , ¹² , ¹³

PwP demonstrated greater impairments in cognitive, motor, and non‐motor domains, as well as in oromotor, speech, chewing, and swallowing functions, compared to controls. They also experienced more drooling. These findings are consistent with other studies that have found similar impairments among PwP. ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³

Interestingly, compared to controls, PwP showed significantly lower saliva flow rates, in both unstimulated and stimulated conditions, supporting the hypothesis that saliva flow may not be the primary cause of drooling in PD. ³⁷ Autonomic dysfunction and reduced parasympathetic activity in PD can directly affect saliva flow. ⁴⁴ The cause of this dysregulation is unknown but may be related to Lewy body involvement in the salivary parasympathetic ganglia. ⁴⁵ , ⁴⁶ Another possible explanation is a dopamine deficit, ⁴⁷ as dopamine seems to modulate salivary secretion in both vertebrates and invertebrates. ⁴⁸ , ⁴⁹ Central and peripheral dopamine receptor activation also influences salivary secretion in rats, possibly reflecting a similar mechanism in PwP. ⁴⁹ This highlights dopamine's role in saliva production. After controlling for anticholinergic drugs and botulinum toxin, the difference in unstimulated saliva flow rate between groups remained significant. However, the difference in stimulated saliva flow was no longer significant. Studies indicate that salivary glands express various muscarinic receptor types, with M₃ being the most abundant and most strongly associated with saliva secretion volume. ⁵⁰ , ⁵¹ This receptor is predominantly found in the parotid glands, ⁵² , ⁵³ which are more active during mechanical stimulation. ⁵⁴ Anticholinergic drugs inhibit muscarinic receptors, with M₃ blockade being particularly responsible for reducing salivation. ⁴⁶ This may explain why differences in stimulated saliva flow are no longer significant, supporting the role of anticholinergic drugs in mechanically induced saliva production.

In our study, drooling impact in PwP was correlated with various clinical factors, including drooling severity, saliva accumulation, facial expression, global motor and non‐motor domains, posture, LEDD, disease duration, oromotor, speech, chewing and swallowing functions, and unstimulated saliva flow rate.

Despite reduced saliva flow rate in PwP, a significant positive correlation was observed between the self‐perceived drooling impact and higher unstimulated saliva flow rate. PwP with higher saliva flow may experience increased challenges in managing saliva, potentially contributing to a greater perceived impact. In contrast, no significant correlation was observed with stimulated salivary flow rate. This may be because stimulated saliva flow involves mechanical stimulation and a higher activation of the parotid gland. ⁵⁵ Stimulated saliva production occurs during specific activities, such as speaking or eating, and the SCS‐PD items related to these domains were rated as having less impact on daily life. Unstimulated saliva, on the contrary, is the predominant mechanism during longer periods of the day and night and is, therefore, associated with a higher impact. After accounting for the effects of anticholinergic drugs and botulinum toxin, the correlation between the SCS‐PD score and unstimulated saliva flow rate was no longer significant. Accordingly, although predominantly found in the parotid glands, M₃ receptors are also present in the submandibular glands, which play a major role in saliva production at rest. ⁵⁴ Then, by controlling for the effect of anticholinergic drugs in unstimulated saliva flow, one source of variation in salivary production may be reduced. This might make the group more homogeneous regarding unstimulated saliva and the associated self‐perceived impact, potentially contributing to the loss of significance. These findings suggest that these pharmacological interventions also play a key role in modulating the unstimulated saliva and, in turn, the drooling impact.

Oromotor function is essential for saliva control, ⁵⁶ with lip function crucial for preventing anterior saliva spillage and generating pressure for swallowing. The tongue also plays a key role in the efficiency of saliva swallowing, as it facilitates the removal of saliva from the oral cavity and prevents accumulation. Consequently, impairments in oromotor function contribute to saliva accumulation and spillage, which may exacerbate the perceived impact of drooling. This can explain the significant association between greater oromotor impairment and higher self‐perceived drooling impact.

Speech impairments were associated with higher drooling impact, likely due to shared anatomical structures involved in speech production and saliva control. Some PwP reported that saliva accumulation negatively affects speech, explaining the link between speech function and drooling impact.

Posture was also correlated with drooling impact. PwP tend to adopt a more immobile posture and bend their heads, ⁵⁷ which may cause saliva to drip due to gravity, potentially worsening drooling severity and its perception. ⁵⁷

Longer duration was associated with greater impact, as drooling tends to worsen over time. ¹⁵ This may be due to more pronounced impairments in saliva control as the disease progresses. Despite this association, after adjusting for disease duration, the relationship between self‐perceived impact and the severity of drooling remained strong and significant. This suggests that such relationship is influenced by other factors beyond disease duration, highlighting the multifactorial nature of drooling impact in PD.

LEDD was also correlated with drooling impact, as it tends to be higher in PwP with longer disease duration ⁵⁸ , ⁵⁹ and motor impairments. ⁵⁹ Therefore, LEDD may be related to various factors underlying the severity and drooling impact.

Self‐perceived drooling impact was not significantly correlated with PD severity, measured by H&Y stages, although it was correlated with the MDS‐UPDRS score. Although the H&Y stages account for the overall severity, they may not capture more subtle variations in non‐motor problems. In contrast, the MDS‐UPDRS score provides a more comprehensive assessment of motor and non‐motor domains, which may explain why a significant correlation was observed with the MDS‐UPDRS score but not with the H&Y stages. This suggests that PwP with a greater impact of drooling tend to have higher motor and non‐motor problems, as found in prior studies. ⁵⁶ , ⁶⁰ , ⁶¹

Interestingly, no significant correlations were observed between the self‐perceived drooling impact and cognitive impairment, consistent with some studies ⁶¹ , ⁶² , ⁶³ but in contrast to others. ⁴ , ⁶⁴ , ⁶⁵ These inconsistencies may arise from differences in the definition of cognitive impairment, outcome measures, and participant samples. In our study, we used a screening test that may not fully capture the range of cognitive functioning in PwP. A more detailed neuropsychological assessment could provide a clearer understanding of how cognitive impairment influences drooling and its impact.

Beyond these correlations, higher levels of impairment in swallowing and facial expression, drooling severity, and saliva accumulation significantly predicted the self‐perceived drooling impact in PwP.

As expected, drooling severity was the strongest predictor of drooling impact. More severe drooling was associated with greater impact in physical and psychological domains and contributed to more severe restriction in daily activities. Severe drooling not only impairs speech and eating, which are crucial for social integration, but also increases social discomfort, possibly leading to isolation. ⁶⁶

Saliva accumulation was also a significant predictor of drooling impact. The frequency and severity of drooling may increase if an excessive amount of saliva has already accumulated in the oral cavity, ⁷ contributing to a heightened perception of drooling's effects.

Swallowing impairment predicted drooling impact. Impairment in swallowing liquids and solid foods, as observed in this study, may plausibly extend to saliva. Reduced swallowing efficiency could hinder the adequate transportation of saliva to the pharynx, causing it to accumulate in the oral cavity. This pooling may, in turn, increase the frequency and severity of drooling episodes and thus intensify the impact.

Swallowing impairments in PD ⁶⁷ can be attributed to reduced strength of the muscles involved in swallowing (eg, lips, tongue). ⁶⁸ , ⁶⁹ It can decrease the negative intraoral pressure during swallowing, limiting the salivary clearance in the oral cavity. Moreover, muscle rigidity and bradykinesia associated with PD have been linked to swallowing impairments. ⁷⁰ In particular, tongue bradykinesia seems to play a major role in swallowing impairment and drooling. ⁵⁷ , ⁷¹

Facial expression impairment was another significant predictor. Reduced lip closure may impair saliva retention in the oral cavity, increasing the likelihood of forward leakage. This problem reinforces the connection between drooling and bradykinesia. ⁶⁰ , ⁷²

Limitations and Future Research

Some limitations of this study should be highlighted. The representativeness of the PD cohort was restricted by a limited number of participants with severe motor impairments, as some were unable to travel to the hospital. The ethics committee determined that the risks of home‐based data collection outweighed the benefits. Future studies should consider measures to include these participants, such as providing transportation to the data collection site.

Some PwP participants (3.9%) were at high risk of airway obstruction and/or aspiration of food and/or paraffin wax. In these cases, procedures related to swallowing assessment and stimulated saliva collection were not performed.

The speech‐language pathologist responsible for data collection was not blinded, which is a limitation. In addition, due to the lack of standardized instruments to assess saliva accumulation in the oral cavity, the CGI‐S was developed but has not undergone clinimetric evaluation. Future studies should investigate its clinimetric properties. Moreover, drooling severity was assessed using the item 2.2 of the MDS‐UPDRS, as no validated instrument exists specifically for this purpose in PD. ¹⁷ This assessment relied on participant and caregiver perceptions, which may not fully reflect daily severity. Given the variability of drooling in PD throughout the day and night, assessing its severity within a short data collection period would have been particularly challenging for the clinician.

In conclusion, self‐perceived drooling is a significant concern for PwP. Its impact can occur both during the day and at night, and negative effects are reported in speech, eating, and social participation. The severity of drooling, saliva accumulation, swallowing, and facial expression impairments were significant predictors of the drooling impact. Clinicians should pay more attention to this often‐overlooked problem. Given its multifactorial nature, drooling may benefit from a multidisciplinary approach to assessment and management. It is, therefore, recommended that drooling and its impact be considered during routine Parkinson's consultations.

Author Roles

(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical analysis: A. Design, B. Execution, C. Review and critique; (3) Manuscript preparation: A. Writing of the first draft, B. Review and critique.

D.N.: 1A, 1B, 1C, 2A, 2B, 3A

B.M.: 1B, 2C, 3B

L.G.: 2A, 2B, 2C, 3B

D.A.: 2A, 2C, 3B

T.F.O.: 3B

I.G.: 1A, 1B, 2A, 2C, 3B

J.J.F.: 1A, 1B, 2A, 2C, 3B

Disclosures

Ethical Compliance Statement: This study was reviewed and approved by the Health Care Ethics Committee of the Centro Hospitalar de Lisboa Ocidental, Lisbon, Portugal (reference number 2284). Subjects who were eligible for participation gave written informed consent after receiving detailed information about the objectives and procedures of the study. The authors confirm that they have read the journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflicts of Interest: Funding was received for this study from the Laboratory of Clinical Pharmacology and Therapeutics, Faculdade de Medicina, Universidade de Lisboa. The authors declare no conflicts of interest relevant to this work.

Financial Disclosures for the Previous 12 Months: J.J.F. has provided consultancy and received speaker fees from BIAL, Biogen, AbbVie, Sunovion Pharmaceuticals, Infucure, Zambon, Roche, Stada, ONO Pharma, Britannia, Neuroderm, and SK Chemicals and has received grants from AbbVie, BIAL, Medtronic, and Angelini. D.N., B.M., L.G., D.A., T.F.O., and I.G. declare that there are no additional disclosures to report.

Data Availability Statement

The data that support the findings of this study are openly available in Repositório da Universidade de Lisboa at https://repositorio.ul.pt.