Content uploaded by Eva Drnovšek
Author content
All content in this area was uploaded by Eva Drnovšek on Aug 12, 2024
Content may be subject to copyright.
Content uploaded by Eva Drnovšek
Author content
All content in this area was uploaded by Eva Drnovšek on Aug 12, 2024
Content may be subject to copyright.
Parkinsonism and Related Disorders 127 (2024) 107091
Available online 6 August 2024
1353-8020/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Short communication
Human perception of Parkinson’s disease body odor in comparison to the
volatile organic compounds of Parkinson’s disease
Eva Drnovsek
a
,
*
, Alexandra Parichenko
b
, Nicole Power Guerra
a
, Julia Pabst
a
,
Kristof Wunderlich
c
, Bj¨
orn Falkenburger
c
, Shirong Huang
b
, Gianaurelio Cuniberti
b
,
Antje Haehner
a
, Thomas Hummel
a
a
Smell &Taste Clinic, Department of Otorhinolaryngology, Faculty of Medicine Carl Gustav Carus, Technische Universit¨
at Dresden, Dresden, Germany
b
Institute for Materials Science and Max Bergmann Center for Biomaterials, Technische Universit¨
at Dresden, Dresden, Germany
c
Department of Neurology, Faculty of Medicine Carl Gustav Carus, Technische Universit¨
at Dresden, Dresden, Germany
ARTICLE INFO
Keywords:
Body odor
Parkinson’s disease
Neurodegenerative disease
Sebum
Olfaction
Smell
ABSTRACT
Introduction: Patients with Parkinson’s Disease (PD) have a distinctive body odor, which was rst described by a
patient’s wife as musky and strong. Later analysis of sebum of patients with PD revealed four volatile organic
compounds (VOC) (perillic aldehyde, hippuric acid, eicosane, octadecanal), that differed from healthy subjects,
and the patient’s wife conrmed that three of them smelled like patients with PD. However, it is unclear whether
other people can also perceive this PD body odor and whether it can be articially recreated. Hence, we aimed to
systematically assess whether young women can perceive the PD body odor and whether they can discriminate
between the PD body odor and the “articial PD odor”composed of the four VOCs mentioned above.
Methods: T-shirts were collected from 19 people with idiopathic PD and 15 age- and gender-matched healthy
participants to represent the PD body odor and the healthy body odor, respectively. The four VOCs were diluted
in 1,2-propanediol to prepare the articial PD body odor. Body odors were rated by 26 young women.
Results: PD body odor was perceived as more musty, strong, smelly, and unpleasant compared to healthy and
articial PD body odor. Furthermore, around 80 % of women were able to discriminate PD body odor from
articial PD body odor.
Conclusion: Overall, this study conrmed a distinctive body odor quality of patients with PD, which can be
perceived by young women. However, the four VOCs, composing the articial PD body odor, were insufcient to
reproduce the body odor from PD patients.
1. Introduction
Parkinson’s Disease (PD) is the second most common neurodegen-
erative disease, which is diagnosed clinically after development of
motor symptoms such as resting tremor, rigidity, bradykinesia, and
postural instability. However, PD develops years before and early
detection could be benecial [1]. A change in body odor (“PD BO”)
could be of use as an early symptom of imminent PD because it
reportedly changes years before the motor symptoms develop. This
phenomenon has been rst observed by a patient’s wife, who was able to
discriminate between 6 samples from PD patients and 6 healthy
individuals with 100 % accuracy. Interestingly, she identied one con-
trol as a PD patient, however, he was later also diagnosed with PD [2].
According to their preliminary experiments, the PD BO seems to be most
prominent in areas with high sebum production, namely in the upper
back [2,3]. PD BO could therefore be attributed to the altered sebum
composition. Sebum analysis in 2019 found four volatile organic com-
pounds (VOC) (perillic aldehyde, hippuric acid, eicosane, octadecanal),
which differed among sebum samples of PD patients and healthy people
[3]. Perillic aldehyde was signicantly lower, whereas eicosane was
signicantly higher in PD patients compared to healthy people. Hippuric
acid and octadecanal were not signicantly different yet trended to be
* Corresponding author. Smell &Taste Clinic, Department of Otorhinolaryngology, Faculty of Medicine Carl Gustav Carus, Technische Universit¨
at Dresden,
Dresden, Germany.
E-mail addresses: eva.drnovsek@tu-dresden.de (E. Drnovsek), alexandra.parichenko@tu-dresden.de (A. Parichenko), nicole.powerguerra@ukdd.de (N. Power
Guerra), julia.pabst@ukdd.de (J. Pabst), kristofmarian.Wunderlich@ukdd.de (K. Wunderlich), bjoern.falkenburger@dzne.de (B. Falkenburger), shirong.huang@
tu-dresden.de (S. Huang), gianaurelio.cuniberti@tu-dresden.de (G. Cuniberti), antje.haehner@ukdd.de (A. Haehner), thomas.hummel@tu-dresden.de (T. Hummel).
Contents lists available at ScienceDirect
Parkinsonism and Related Disorders
journal homepage: www.elsevier.com/locate/parkreldis
https://doi.org/10.1016/j.parkreldis.2024.107091
Received 18 March 2024; Received in revised form 1 August 2024; Accepted 5 August 2024
Parkinsonism and Related Disorders 127 (2024) 107091
2
higher. To conrm the importance of this nding, the experimentee
from Morgan et al. was asked to smell the VOCs from an odor port of a
gas chromatography-mass spectrometry. She identied the three VOCs
(hippuric acid, eicosane, and octadecanal) as similar to the musky PD BO
[3]. However, not much is known about human PD BO perception in
general.
Overall, the present study aimed to systematically assess, whether
young women can perceive PD BO and whether they can discriminate
between PD BO and articial PD BO composed of the four VOCs (perillic
aldehyde, hippuric acid, eicosane, octadecanal).
2. Methods
2.1. Body odor donors and human raters
Body odors were collected from PD patients and healthy age- and
gender-matched BO donors. Furthermore, young women were recruited
as human raters. Participants gave informed written consent. The study
was performed in accordance with the Declaration of Helsinki and
approved by the Ethics Committee at the University Clinic of the TU
Dresden (application number: BO-EK-548112021).
PD patients were recruited at the Department of Neurology (Faculty
of Medicine Carl Gustav Carus). Main inclusion criterion was idiopathic
parkinsonism (diagnostic criteria in the Supplement). Exclusion criteria
were a COVID-19 infection at the time of recruitment, renal diseases,
and inability to participate in the study, e.g. due to movement
restriction.
Healthy BO donors were age- and gender-matched and recruited via
verbal propaganda. Exclusion criteria were pregnancy, smoking, and
signicant health restrictions (e.g. pronounced renal insufciency, PD).
Human raters were recruited via verbal propaganda. To reduce the
variance among the raters female participants aged 18–40 years were
included. Additionally, only participants, who were familiarized with
the PD BO as a part of another BO study were included. In the end of this
previous study, they were told which bottle represented the PD BO.
Exclusion criteria were pregnancy, smoking, and signicant health dis-
orders that may interfere with olfactory function (e.g., diabetes mellitus,
renal insufciency, neurodegenerative diseases such as PD).
2.2. Body odors
BO donors were asked to wear a pre-washed cotton T-shirt for one
night following a strict protocol as explained in the Supplement. Next,
the back collar of the T-shirts was cut out, and stored individually in
plastic bags in a freezer at –20 ◦C [4]. BOs were presented in six 500 ml
brown bottles. All 15 T-Shirt pieces of healthy participants were put in
one bottle (healthy BO) and all 19 T-Shirt pieces of PD patients in the
second bottle (PD BO) to reduce the inuence of individual BO and to
enhance BO stimulus. Next, articial PD BO was prepared by mixing
10uM of each of the VOCs (perillic aldehyde, hippuric acid, eicosane,
and octadecanal, Supplement Table S1) with 1,2-propanediol. Four
bottles of articial PD BO were prepared, each containing 15 ml of the
mixture on a gauze pad. Pieces of a blank odorless pre-washed T-shirt
were added. To equalize the background odors, a gauze pad with 15 ml
of 1,2-propanediol was added to the healthy BO and PD BO as well. A
detailed protocol for BO collection and sample preparation can be found
in the Supplement.
2.3. Human ratings
For human ratings there were two appointments to reduce olfactory
fatigue. At the rst appointment, olfactory function was assessed using
the Snifn’Sticks battery with threshold (T), discrimination (D), and
identication (I) summed in the TDI score [5]. Additionally, raters were
familiarzied with the PD BO by smelling BO from PD patients and
healthy age- and gender-matched controls as part of another study.
At the second appointment, raters underwent qualitative and quan-
titative BO evaluation. For the quantitative part, two discriminations
tasks were performed. In the qualitative part, each BO was described
using the description matrix and perceptual descriptors. In the end,
some of the participants underwent the quantitative part again.
2.4. “Qualitative part”
In the qualitative part, participants were individually presented with
three bottles in the same order. They started with the healthy BO, fol-
lowed by the PD BO, and the articial PD BO. They were asked to select
the words that best described the presented BO from a description ma-
trix [6]: biting, owery, deodorized, subtle, foul, fresh, moist, individ-
ual, cold, musty, natural, neutral, salty, clean, sour, sweaty, strong,
pungent, smelly, sweet, unpleasant, warm. These twenty words are the
most common words used to describe human BOs among German people
[6]. In the end, they were asked to rate the odor on a visual analog scale
(VAS) from »not at all «to »very «using the following descriptors:
pleasant, intense, and familiar. Furthermore, they had to rate how much
distance they wanted from the person whose BO was presented on a
scale from »very much «to »none«.
2.5. “Quantitative part”
In the quantitative part, participants had to discriminate the different
BO in two 3-alternative forced choice (3-AFC) tasks. In both tasks, three
bottles were presented. Each bottle was presented 2 cm from the nose for
around 2 s. Participants were blinded. In the end, they were asked to
choose the different odor from the three samples in a forced choice
manner. There were three rounds per task. In the rst discrimination
task, two bottles contained articial PD BO and one contained healthy
BO. In the second discrimination task, two bottles contained articial PD
BO and one contained PD BO.
2.6. Statistical analysis
Statistical analysis was performed using R Statistical Software [7]
(version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria).
Detailed information can be found in the Supplement.
3. Results
BO was collected from 15 healthy people and 19 people with PD.
There was no difference regarding age or gender among the healthy
people and people with PD (Supplement Table S2). For BO ratings, 26
young women were included (Supplement Table S3).
3.1. “Qualitative part”
In the qualitative part, participants were asked to rate each sample.
Firstly, they had to choose words that best describe a certain odor using
a description matrix. As can be seen on Fig. 1 and Supplement Table S4,
PD BO was described as more musty, strong, smelly, and unpleasant
compared to the articial PD BO and healthy BO, indicating that young
women were able to perceive the distinctive quality of the PD BO. On the
other hand, the articial PD BO was described as more neutral, and clean
compared to the PD BO and healthy BO. Secondly, perceptual descriptor
ratings on a VAS for the three samples are shown in Supplement
Table S5. PD BO was perceived as less pleasant compared to the articial
PD BO and the healthy BO, whereas there was no difference in odor
intensity, familiarity, or preferred distance.
3.2. “Quantitative part”
In the quantitative part, participants underwent two discrimination
tasks. Here, 2 or 3 out of 3 correct answers were considered a cut-off for
E. Drnovsek et al.
Parkinsonism and Related Disorders 127 (2024) 107091
3
participants, who were able to distinguish the different sample. Inter-
estingly, more than 80 % of participants were able to distinguish PD BO
from articial PD BO (Table 1). On the other hand, only around 50 % of
participants were able to distinguish healthy BO from articial PD BO.
Participants who could discriminate the BOs had a signicantly better
olfactory function measured using the TDI score (Supplement Fig. S1).
Next, the expected (random) binominal distribution was calculated. This
is the distribution one would expect if participants randomly chose the
“different”bottle. Then our distribution was compared to the expected
binominal distribution (Supplement Table S6). The distribution of
healthy BO discrimination was not signicantly different from the ex-
pected distribution in both rounds. On the other hand, the distribution of
PD BO discrimination differed compared to the expected distribution in
both rounds indicating that our group of young females was able to
distinguish PD BO from the articial PD BO.
4. Discussion
We systematically assessed whether young women can perceive the
PD BO and whether they are able to discriminate PD BO from articial
PD BO composed of four VOCs (perillic aldehyde, hippuric acid, eico-
sane, octadecanal). This study conrmed a distinctive BO in PD patients
perceived by young women. However, the four VOCs, composing the
articial PD BO, were insufcient to reproduce it.
Patients with PD have a distinctive BO, which the experimentee from
Morgan et al. described as strong and musky [2,3]. Here, PD BO was
perceived as more unpleasant, strong, smelly, and musty than healthy or
articial PD BO. Musky was not included in the description matrix
because it is not part of the validated assessment of BO [6], whereas
strong was also used by our raters. Noteworthy, only young women, who
were previously familiarized with the PD BO, were included.
The articial PD BO included four VOCs, which were signicantly (or
at least trending to be) different in sebum of PD patients compared to
healthy people [3]. Here, the PD BO was easily discriminated from the
articial PD BO indicating that using only these four VOCs is not suf-
cient to reproduce the PD BO - although they might contribute to the
unique PD BO [3]. BOs are complex mixtures, and further VOCs are
likely needed to reproduce the PD signature. This is in line with the
experimentee from Morgan et al., who rated the VOC samples with the
addition of the healthy sebum as more familiar to the PD BO compared
to the VOC samples without the addition of the healthy sebum [2,3].
Another potential explanation may be the low concentrations of the four
VOCs within the mixture. Additionally, more recent studies compared
Fig. 1. Word clouds presenting the words chosen from the description matrix for healthy (Healthy) body odor (BO) in purple, Parkinson’s Disease (PD) BO in orange,
and articial PD BO (ArtPD) in blue. All words that were chosen at least once are shown. The word cloud in the middle shows the rst 10 words that differed in their
chosen frequencies, colors corresponding to the BO. Those that were signicantly different among the three BOs are marked with *. The exact frequencies and
comparison is shown in the Supplement Table S4.
For the three word clouds presenting the three BOs, the size of the text corresponds to the number of times the descriptor was used for this BO and does not
correspond to the signicant differences among the three samples.
Table 1
The two discrimination tasks. Discrimination of the different sample using two
bottles with articial PD body odor and one with healthy body odor (healthy) or
PD body odor (PD). The olfactory function of participants, who could discrim-
inate, and participants, who could not, is also reported.
Round^Participants
who
Healthy
N (%)
TDI
Median (25
%–75 %)
PD
N
(%)
TDI
Median (25
%–75 %)
1 (N =
26)
Discriminated 12 (46
%)
39.3
(36.7–40.4)
b
21
(81
%)
a
37.3
(34.8–39.5)
a
Did not
discriminate
14 (54
%)
33.9
(31.5–35.6)
b
5
(19
%)
a
33.5
(30.5–34.3)
a
2 (N =
19)
Discriminated 10 (53
%)
36.1
(34.3–39.9)
17
(89
%)
b
35.0
(34.3–39.5)
Did not
discriminate
9 (47
%)
35.0
(33.5–37.5)
2
(11
%)
b
34.1
(32.3–35.9)
1&2
(N =
19)
Discriminated 6 (32
%)
39.8
(37.8–41.3)
a
13
(68
%)
a
37.3
(34.8–40.0)
Did not
discriminate
6 (32
%)
34.3
(31.3–35.6)
a
1 (5
%)
a
30.5
Participants who answered 2 or 3 out of 3 correctly, could discriminate the
different sample. Others could not.
a
p value <0.05.
b
p value <0.001.
^
The discrimination task was performed by all 26 participants (rst round)
and additionally by 19 out of 26 participants (second round). The 1&2 shows the
number of participants who could discriminate or could not discriminate in both
rounds.
E. Drnovsek et al.
Parkinsonism and Related Disorders 127 (2024) 107091
4
sebum of PD patients and healthy people. Overall, studies agree that
there is a difference in the sebum composition. However, which VOCs
are responsible for this change is less clear [8–10]. Interestingly, one
study was able to predict PD from the odor prole of sebum with 79 %
accuracy using an articial intelligent olfactory system [10].
One limitation of this study is that it is unclear which VOCs are
responsible for the difference between sebum of PD patients and healthy
people [3,8,9]. Of note, three of the VOCs were higher and one was
lower in PD sebum compared to healthy sebum, therefore it would be
interesting to also include a sample using only these three VOCs.
Furthermore, although the patient’s wife reported that these three VOCs
have a similar smell to the PD BO [3], additional VOCs appear to be of
signicance. Another limitation is the solubility of the VOCs. In our
study, 1,2-propanediol, a solvent commonly used in human olfaction
research, was used. However, it may not be the best one for all four
VOCs. In future studies the possibility of using different solvents for each
compound might be explored. Furthermore, highly sensitive
nanomaterials-based electronic nose may be explored to discriminate
the articial PD BO from the healthy BO for futher use in the early
diagnosis of PD [11,12]. Another possible limitation is that healthy BO
donors did not undergo olfactory testing which might have resulted in
the inclusion of a person in early phases of PD. Furthermore, direct
comparison of the healthy BO and PD BO should be investigated and
futher studies on a bigger sample are needed to conrm the results from
this pilot study.
In conclusion, this study showed that young women perceive the BO
of PD patients as different in quality compared to healthy BO or articial
PD BO. However, the four VOCs, used for the articial PD BO, were
insufcient to reproduce it. These results support the previous nding
that the sebum of PD patients differs from the sebum of healthy people
[3,8–10], indicating that the phenotype of PD goes beyond the dopa-
minergic and motor systems. In the future, this difference could poten-
tially be used to screen for PD patients using highly sensitive
nanomaterials-based electronic nose (e-nose) [11,12].
Data availability
Data and code are available upon request on: https://doi.org/
10.7303/syn53278108.
Funding
This work was supported by the European Union for the project
smellodi (smart electronic olfaction for BO diagnostics) [grant number:
101046369] and by Volkswagen Stiftung as a part of the project
“Olfactorial Perceptronics”[grant number Az 9B396].
CRediT authorship contribution statement
Eva Drnovsek: Writing –original draft, Visualization, Methodology,
Investigation, Formal analysis, Conceptualization. Alexandra Par-
ichenko: Writing –review &editing, Methodology, Conceptualization.
Nicole Power Guerra: Writing –review &editing, Methodology,
Investigation, Conceptualization. Julia Pabst: Methodology, Investiga-
tion. Kristof Wunderlich: Writing –review &editing, Methodology,
Investigation. Bj¨
orn Falkenburger: Writing –review &editing, Meth-
odology, Investigation. Shirong Huang: Supervision, Methodology,
Investigation. Gianaurelio Cuniberti: Writing –review &editing, Su-
pervision, Methodology, Funding acquisition, Conceptualization. Antje
Haehner: Writing –review &editing, Methodology, Conceptualization.
Thomas Hummel: Writing –review &editing, Supervision, Investiga-
tion, Funding acquisition, Conceptualization.
Declaration of competing interest
The authors declare the following nancial interests/personal re-
lationships which may be considered as potential competing interests:
Since 2021 TH has collaborated on research projects with Sony, Stutt-
gart, Germany; Smell and Taste Lab, Geneva, Switzerland; Takasago,
Paris, France; aspuraclip, Berlin, Germany. He received consultancy fees
from Baia Foods, Madrid, Spain; Burghart, Holm, Germany; air-up,
Munich, Germany. Other authors declare that they have no conict of
interest.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.parkreldis.2024.107091.
References
[1] T. Prell, F. Siebecker, M. Lorrain, C. Eggers, S. Lorenzl, J. Klucken, T. Warnecke,
C. Buhmann, L. Tonges, R. Ehret, I. Wellach, M. Wolz, Recommendations for
standards of network care for patients with Parkinson’s disease in Germany, J. Clin.
Med. 9 (5) (2020).
[2] J. Morgan, Joy of super smeller: sebum clues for PD diagnostics, Lancet Neurol. 15
(2) (2016) 138–139.
[3] D.K. Trivedi, E. Sinclair, Y. Xu, D. Sarkar, C. Walton-Doyle, C. Liscio, P. Banks,
J. Milne, M. Silverdale, T. Kunath, R. Goodacre, P. Barran, Discovery of volatile
biomarkers of Parkinson’s disease from sebum, ACS Cent. Sci. 5 (4) (2019)
599–606.
[4] P. Lenochova, S.C. Roberts, J. Havlicek, Methods of human body odor sampling:
the effect of freezing, Chem. Senses 34 (2) (2009) 127–138.
[5] A. Oleszkiewicz, V.A. Schriever, I. Croy, A. Hahner, T. Hummel, Updated Snifn’
Sticks normative data based on an extended sample of 9139 subjects, Eur. Arch.
Oto-Rhino-Laryngol. 276 (3) (2019) 719–728.
[6] A. Bierling, I. Croy, Intermediate Dataset of Clustered Body Odor Descriptors,
Techniche Universit¨
at Dresden, 2023.
[7] R, Core Team, R: A Language and Environment for Statistical Computing, R
Foundation for Statistical Computing, Vienna, Austria, 2017.
[8] D. Sarkar, E. Sinclair, S.H. Lim, C. Walton-Doyle, K. Jafri, J. Milne, J.P.C. Vissers,
K. Richardson, D.K. Trivedi, M. Silverdale, P. Barran, Paper spray ionization ion
mobility mass spectrometry of sebum classies biomarker classes for the diagnosis
of Parkinson’s disease, JACS Au 2 (9) (2022) 2013–2022.
[9] E. Sinclair, C. Walton-Doyle, D. Sarkar, K.A. Hollywood, J. Milne, S.H. Lim,
T. Kunath, A.M. Rijs, R.M.A. de Bie, M. Silverdale, D.K. Trivedi, P. Barran,
Validating differential volatilome proles in Parkinson’s disease, ACS Cent. Sci. 7
(2) (2021) 300–306.
[10] W. Fu, L. Xu, Q. Yu, J. Fang, G. Zhao, Y. Li, C. Pan, H. Dong, D. Wang, H. Ren,
Y. Guo, Q. Liu, J. Liu, X. Chen, Articial intelligent olfactory system for the
diagnosis of Parkinson’s disease, ACS Omega 7 (5) (2022) 4001–4010.
[11] S. Huang, A. Croy, A.L. Bierling, V. Khavrus, L.A. Panes-Ruiz, A. Dianat,
B. Ibarlucea, G. Cuniberti, Machine learning-enabled graphene-based electronic
olfaction sensors and their olfactory performance assessment, Appl. Phys. Rev. 10
(2) (2023).
[12] A. Parichenko, S. Huang, J. Pang, B. Ibarlucea, G. Cuniberti, Recent advances in
technologies toward the development of 2D materials-based electronic noses,
TrAC, Trends Anal. Chem. 166 (2023).
E. Drnovsek et al.
