Keywords
smell - taste - olfactory testing - gustatory testing - chemosensory dysfunction
Introduction
Although the sensory input from eating is often described as ‘taste’, the perception
of flavour is truly a multisensory experience. Thus, if any sensory component is reduced
or lost, it can have sweeping effects on perception. Olfactory dysfunction is the
most common sensory loss, but its symptoms are often mistakenly ascribed to an impairment
in the gustatory function, since the two senses are interrelated.[1] This is in part due to the fact that ‘taste of substances’ is mainly driven by aromas,
which are perceived by the olfactory system.
Gustatory function consists of at least five basic taste qualities; sweet, salty,
bitter, sour, and umami.[2] The simultaneous olfactory stimulation of retronasal aromas are often not perceived
as a contribution from the nose, which makes it difficult to distinguish between these
senses.[3]
Olfactory disorders are common, affecting up to 20% of the population.[4] They can lead to decreased quality of life,[5] depression,[6] increased health risk (ingestion of rotten food, and lack of response to fires and
gas leaks, for example), and malnutrition.[7] Self-rating of the olfactory function is often used in the clinical practice.[8] Specifically, subjective reduction or loss of smell constitutes one of the four
diagnostic criteria for chronic rhinosinusitis.[9] Although the updated guidelines on chronic rhinosinusitis advise the performance
of olfactory testing,[10] this is not always done routinely for both taste and smell functions. Correct assessment
of the chemosensory function is an important factor in diagnosis and treatment; thus
it is crucial to examine the quality of the self-assessment.
Diagnostic tests can be applied to determine the olfactory and gustatory functions.
Yet, thorough testing is time-consuming. Therefore, it is of great importance to consider
whether or not self-assessment is a valid predictor of olfactory and gustatory dysfunction
compared with diagnostic testing. Previous studies[11]
[12] have established a positive correlation between self-assessment of olfactory abilities
and olfactory function, while others[13] demonstrated a poor correlation between these parameters. Information on the gustatory
correlation between the measured and subjective functions is sparse. Due to the confusion
between taste and smell, an evaluation of both senses is a prerequisite to untangle
the assessment of these ambiguous sensory assessments.
The aim of the present study was to evaluate the correlation between subjective and
measured olfactory and gustatory dysfunctions in referred patients complaining of
taste and/or smell dysfunction to identify the need of chemosensory testing. Moreover,
the aim was also to assess the need for chemosensory testing depending on the subjective
evaluation.
Materials and Methods
Patients
Patients with complaints of olfactory and or gustatory dysfunction were included in
a Research Electronic Data Capture (REDCap, Vanderbilt University, Nashville, Tennessee,
USA) registry from January 2017 to April 2019. Patients were referred from either
a private ear, nose and throat (ENT) specialist or an ENT department to the Flavour
Clinic (a Danish specialized national out-patient clinic for taste and smell disorder)
for diagnostics and treatment. All referred patients were eligible for inclusion in
the registry, and no patients declined inclusion. A previous[13] study has been published in which the diagnostic workflow is more thoroughly described.
A part of the data of the current study was also included in this previous study;
however, not the details of the association between the subjective and measured functions,
which is the primary focus of the current study.
Prior to referral, the patients had been examined by an ENT specialist and undergone
inspection of the nasal cavity and rhino pharynx, allergy testing, and computed tomography
of the nasal cavity and paranasal sinuses.
Testing Procedure
Prior to testing, patients filled out a questionnaire on demographics and self-rated
olfactory and gustatory functions (good, normal, reduced, or absent). The patients
avoided smoking, eating, and drinking (with the exception of water) one hour before
testing. The Mini-Mental State Examination (MMSE) was applied to all patients to ensure
sufficient cognitive capacity during the testing, and the 22-item Sinonasal Outcome
Test (SNOT22) was applied to quantify sinonasal symptoms.
Olfactory Testing
The olfactory function was examined with the full version of the Sniffin' Sticks olfactory
test.[14] This test includes the assessment of the olfactory threshold (T), discrimination
(D), and identification (I) score, which generates a total TDI score (1–48 points).
A Danish version of this test has been validated,[15] making it applicable for research and clinical work in Denmark.
The total individual TDI score was computed, and the participants were separated into
3 groups; normosmia (TDI ≥ 30), hyposmia (TDI bertween 15 and 29), and anosmia (TDI ≤ 16).[16] The cut-off value for hyposmia is based on normative data on healthy individuals
with subjective normal olfactory function from a Danish population, in accordance
to the test guidelines.[15]
Gustatory Testing
The clinical pipeline for gustatory testing was changed during the course of the present
study. As such, gustatory testing was conducted in two different ways. Initially,
the Taste Strips examination was applied.[17] As some patients had difficulties with this test, a study was conducted to assess
the re-test reliability of the Taste Strips and another gustatory test, the Taste-Drop-Test.
As the Taste-Drop-Test proved more reliable,[18] it was subsequently used.
Test Procedures
Taste Strips Test
Taste strips are filter paper strips impregnated with taste solutions representative
of four basic tastants at four different concentrations (sweet: 0.4 g/mL, 0.2 g/mL,
0.1 g/mL, and 0.05 g/mL of sucrose; sour: 0.3 g/mL, 0.165 g/mL, 0.09 g/mL, and 0.05 g/mL
of citric acid; salty: 0.25 g/mL, 0.1 g/mL, 0.04 g/mL, and 0.016 g/mL of sodium chloride;
bitter: 0.006 g/mL, 0.0024 g/mL, 0.0009 g/mL, and 0.0004 g/mL of quinine-hydrochloride).
The Taste Strips test validated to assess gustatory sensitivity. It is administered
in a pseudo-randomized manner in increasing concentrations, resulting in a total taste
score ranging from 0 to 16 points for each patient. For a more detailed description
of the Taste Strips, see Mueller et al.[17]
The Taste-Drop-Test
The Taste-Drop-Test consists of five different options of tastants (salty, sour, bitter,
neutral, and sweet). One drop of the given tastant was applied on each side of the
tongue with a transfer pipette. Subsequently, the patients had to select which tastant
was applied before rinsing the mouth with water and repeating the procedure. The tastants
were semi-randomized and presented in increasing concentrations until the correct
identification, after which a staircase confirmation of sensitivity level for each
tastant was applied. The total taste score (from 0 to 40 points) was established for
each tested patient. For a more detailed description of the Taste-Drop-Test, see and
Fjaeldstad et. al.[18]
Cut-off values for Gustatory Testing
For the Taste-Drop-Test and Taste Strips test, total taste scores below 25 and 9 points
defined hypogeusia respectively. This cut-off value was defined as the 10th percentile
of normative values. Ageusia, for the Taste-Drop-Test and Taste Strips, was defined
as scores below 18 and 4 respectively.[18]
Statistics
Statistical analyses were completed using JMP (SAS Institute, Cary, NC, US) software,
version 14.0. The mean total TDI scores for subjective olfactory ratings were compared
using a two-tailed t-test. The Spearman rank correlation coefficient (ρ) was used to investigate possible
correlations. The ability to classify loss of taste or smell was assessed by calculating
the area under the curve (AUC) in receiver operating characteristic (ROC) curves.
This was calculated for subjective ratings and sensory loss diagnosed with olfactory
or gustatory testing.
The α level of statistical significance was set at 0.05.
Results
A total of 602 patients were tested with the Sniffin' Sticks olfactory test. A random
sample of these patients (n = 258) also underwent gustatory testing. In total, 55% (n = 143) and 45% (n = 115) of patients were tested with the Taste Strips and Taste-Drop-Test, respectively.
The measured olfactory impairments were frequent, with a mean total TDI score of 18.0
(95% confidence interval [95%CI]: 17.4 to 18.7). A high proportion of patients was
found to be anosmic (45.3%, n = 273) or hyposmic (44.4%, n = 267). Subjective olfactory impairment was also frequent, as patients often rated
themselves as anosmic (57.3%, n = 345) or hyposmic (39.5%, n = 238).
The measured gustatory impairments occurred less frequently, as 24.8% (n = 64) of patients were hypogeusic, while only 10.5% (n = 27) were ageusic. However, subjective ratings of hypogeusia (55.8%, n = 144) and ageusia (29.8%, n = 77) were more frequent than the measured reduced gustatory function, as shown in
[Table 1].
Table 1
Demographics of the study sample
|
Olfaction and subjective olfactory function
|
|
Subjective olfactory function
|
All (n = 602)
|
Absent (n = 345)
|
Reduced (n = 238)
|
Normal/Good (n = 19)
|
|
Age, years (IQR)[†]
|
59 (49–68)
|
61 (49–70)
|
58 (50–66)
|
53 (42–61)
|
|
Gender: male, n (%)
|
239 (40)
|
130 (38)
|
103 (43)
|
6 (32)
|
|
Total TDI score, mean (SD)
|
18.0 (7.9)
|
14.5 (6.1)
|
22.0 (7.4)
|
32.2 (2.6)
|
|
Normosmic score, n (%)
|
62 (10.3)
|
6 (1.7)
|
40 (16.8)
|
16 (84.2)
|
|
Hyposmic score, n (%)
|
267 (44.4)
|
117 (33.9)
|
147 (61.7)
|
3 (15.8)
|
|
Anosmic score, n (%)
|
273 (45.3)
|
222 (64.4)
|
51 (21.4)
|
0 (0.0)
|
|
Taste Strips test, mean (SD); n = 143
|
11.8 (4.1)
|
12.5 (3.9)
|
11.3 (4.0)
|
6.0 (2.5)
|
|
Taste-Drop-Test, mean (SD); n = 115
|
23.4 (6.6)
|
23.2 (6.9)
|
24.2 (6.2)
|
21.4 (5.8)
|
|
Normogeusic score, n (%)
|
167 (64.7)
|
105 (69.1)
|
60 (65.9)
|
2 (13.3)
|
|
Hypogeusic score ,n (%)
|
64 (24.8)
|
28 (18.4)
|
25 (27.5)
|
11 (73.4)
|
|
Ageusic score, n (%)
|
27 (10.5)
|
19 (12.5)
|
6 (6.6)
|
2 (13.3)
|
|
Gustation and subjective gustatory function
|
|
Subjective gustatory function
|
All (n = 258)
|
Absent (n = 77)
|
Reduced (n = 144)
|
Normal/Good (n = 37)
|
|
Age, years (IQR)[†]
|
60 (49–70)
|
65 (55–74)
|
58 (48–68)
|
58 (38–65)
|
|
Gender: male/female, n (%)
|
99 (38)
|
25 (32)
|
57 (40)
|
17 (46)
|
|
Taste Strips test, mean (SD)
|
11.8 (4.9)
|
11.1 (4.8)
|
11.7 (4.0)
|
13.7 (2.1)
|
|
Taste-Drop-Test, mean (SD)
|
23.4 (6.6)
|
22.0 (7.1)
|
24.2 (6.4)
|
23.9 (6.3)
|
|
Normogeusic score, n (%)
|
167 (64.7)
|
42 (54.5)
|
96 (66.7)
|
29 (78.4)
|
|
Hypogeusic score, n (%)
|
64 (24.8)
|
20 (26.0)
|
38 (26.4)
|
6 (16.2)
|
|
Ageusic score, n (%)
|
27 (10.5)
|
15 (19.5)
|
10 (6.9)
|
2 (5.4)
|
|
Total TDI score, mean (SD)
|
18.6 (8.4)
|
18.1 (7.5)
|
19.3 (9.1)
|
17.1 (7.3)
|
|
Normosmic score, n (%)
|
36 (14.0)
|
8 (10.4)
|
27 (18.8)
|
1 (2.7)
|
|
Hyposmic score, n (%)
|
108 (41.9)
|
36 (46.7)
|
56 (38.9)
|
16 (43.3)
|
|
Anosmic score, n (%)
|
114 (44.2)
|
33 (42.9)
|
61 (42.4)
|
20 (54.0)
|
Abbreviations: IQR, interquartile range; SD, standard deviation; TDI, threshold, discrimination
and identification.
Notes: Normosmia, hyposmia and anosmia were computed from the total TDI score. Normogeusic,
hypogeusic, and ageusic scores were computed from the total score of the Taste Strips
test or Taste-Drop-Test.
† Median age.
The mean MMSE score of the patients was of 28.4 (95%CI: 28.3 to 28.6). Only 4 patients
scored below 24 points, indicating a cognitive deficit.
Patients had a mean SNOT-22 score of 21.8 (95%CI: 20.6 to 23.0), which was not statistically
correlated to the olfactory TDI score (ρ = 0.0251; p = 0.5401).
Association between Olfactory and Gustatory Self-ratings and Measured Test Results
Patients with a absent or reduced subjective sense of taste (85.7%, n = 221) often had a normal taste test score (n = 138; 62.4%). Most patients with subjective loss of taste and normal taste tests
had a measurable olfactory deficit (n = 121; 87.7%). Contrarily, among the patients with a reduced or absent self-rated
olfactory function and a normal olfactory test, only 2.7% (n = 7) were identified as hypogeusic or ageusic when tested.
In total, 81.0% (n = 209) of the patients who had undergone olfactory and gustatory testing had a combined
subjective smell and taste deficit (either reduced or absent), as shown in [Fig. 1a]. However, only 28.3% (n = 73) had a combined measurable loss of taste and smell, as shown in [Table 1]. Of the 258 (100%) patients who had undergone taste and smell tests, only 18 (7.0%)
had an isolated loss of taste, while 149 (57.8%) had an isolated loss of smell, as
shown in [Fig. 1b].
Fig. 1 (a) Distribution of the subjective olfactory and gustatory ratings (n = 258). The patient group underwent both smell and taste tests (Taste-Strips test
or Taste-Drop-Test). Impairment of smell and taste reflects subjective self-rating
as reduced or absent. (b) Distribution of the olfactory and gustatory test results (n = 258). The patient group underwent both smell and taste tests (Taste-Strips test
or Taste-Drop-Test). Loss of smell reflects hyposmic and anosmic patients. Loss of
taste reflects hypogeusic and ageusic patients.
Accuracy of the Subjective Olfactory Loss
The subjective assessment of the olfactory function was a moderate measurement for
olfactory function based on the olfactory Sniffin' Sticks score (ROC AUC: 0.7951),
as shown in [Fig. 2a]. Patients with subjective anosmia had a statistically significant lower average
mean TDI score compared with the subjective hyposmic group (mean difference: 7.5;
p < 0.001; two-tailed t-test).
Fig. 2 Receiver operating characteristic curves that illustrate the decreased or absent
subjective smell/taste function as a classifier for measured taste/smell function.
(a) Subjective assessment of smell function versus the threshold, discrimination, and
identification (TDI) score. (b) Subjective assessment of taste function versus the Taste-Drop-Test score. (c) Subjective assessment of taste versus Taste Strips test score.
In total, 98.3% of the patients with an olfactory self-assessment rated as absent
did have a measurable reduction in olfactory function.
Only 19 patients were subjectively normosmic, which, for 84.2% of patients corresponded
with a normal TDI score. However, out of the 345 (57.3%) patients who rated their
olfactory function as absent, 123 (35.6%) scored within the normosmic or hyposmic
range. As such, a third of the patients with a subjective belief of an inability to
perceive odour were in fact able to detect smell.
Accuracy of Subjective Gustatory Function in Patients
In total, 54.5% (n = 42) of the patients with subjective ageusia and 66.7% (n = 96) of the patients with subjective hypogeusia scored within the normogeusic range.
The subjective assessment of gustatory function was a poor measurement for gustatory
function based on the Taste-Drop-Test (ROC AUC: 0.5802) and Taste Strips test (ROC
AUC: 0.5436), as shown in [Fig. 2b,c].
Discussion
Key Findings
We found that the subjective olfactory function was a moderate predictor of the measured
olfactory functionthat the , and subjective gustatory function did not reflect the
measured gustatory abilities. Furthermore, the subjective gustatory dysfunction was
associatied with the measured olfactory dysfunction.
Moreover, patients complaining of complete anosmia often had a measurable olfactory
impairment.
Comparisons with Other Studies
Accuracy of Subjective Olfaction in Patients
Patients with an absent subjective sense of smell had a statistically significant
lower average TDI score compared with patients with a reduced subjective sense of
smell. The subjective assessment of olfactory function was associated with the Sniffin'
Sticks score (ROC AUC: 0.7951). In agreement with those findings,[11]
[12] other studies a reported similar moderate correlation (ρ = 0.55 to 0.57) between
self-assessment and measured olfaction in patients. These findings suggest that patients
with a subjective loss of smell are, to some extent, aware of the degree of olfactory
impairment.
We also found that 98.3% (n = 339) of patients who rated their olfactory function as absent scored within the
anosmic or hyposmic range. This result clearly indicates that patients categorizing
themselves as having an absent sense of smell often appear to suffer from some degree
of olfactory impairment. Contrarily,[20] a recent study found that 28.9% of their anosmic patient group rated their olfactory
function as at least average. This suggests that a noticeable share of anosmic patients
are in fact unaware of such a deficit. These findings are in line with those of Shu
et al.,[21] who found that 36% of their patient group scored within the anosmic or hyposmic
range, while only 12% reported olfactory dysfunction.
In the present study, the report by the patients of subjective anosmia was a strong
indicator of measured anosmia or hyposmia. However, subjective olfactory complaints
should be followed up with a measurement of the olfactory function for several reasons.
First, it is of great importance to ensure an accurate diagnosis, since self-assessed
chemosensory impairment may be overestimated or misclassified. As found in our study,
16.8% of patients with a reduced subjective sense of smell had a normal measured olfactory
function, and only 64.4% (n = 222) of patients with an absent subjective sense of smell were found to have lost
the sense completely. Second, accurate results from a validated olfactory test are
indispensable to monitor the efficacy of an applied treatment such as olfactory training.
Third, there is an age-dependent loss of smell that is not reflected in the definition
of hyposmia/anosmia threshold values.[22] This suggests that especially elderly subjects who present themselves with loss
of smell may as well perform better than expected relative to their age group. Thus,
olfactory testing serves as an informative instrument to clarify whether subjective
impairment reflects an actual loss of smell or a benign age-dependent condition.
Accuracy of Subjective Gustatory Function in Patients
We found that the subjective assessment of gustatory function was a good measurement
of gustatory function based on test scores (Taste-Drop-Test, ROC AUC: 0.5802; Taste
Strips test, ROC AUC: 0.5436). In total, 55% of the patients with subjective ageusia
and 67% of the patients with subjective hypogeusia scored within the normogeusic range.
There are several possible explanations for these results. Gustatory impairment is
relatively rare in patients with reported chemosensory dysfunction. We found that
only 7% (n = 18) of patients with chemosensory complaints suffered from an isolated loss of
taste. This low number is consistent with previous results ranging from 1% to 4%.[23]
[24] Another factor is that patients may have mistaken an actual olfactory deficit for
a subjective loss of taste, since olfaction and gustation are interrelated, and both
are vital for the perception of flavour. Also, smell dysfunction is highly frequent
compared with taste disorders.[25]
We found that 78% (n = 29) of the patients with a normal subjective taste function had a normal measured
taste function had a normal measured taste function. In accordance with our results,
a previous study by Soter et al.[26] found that the accuracy of questionnaire statements in detecting loss of taste was
poor, while patients with no gustatory complaints were often correct in their statements.
However, the present study differed from the study by Soter et al.[26] in several aspects. Instead of basing the diagnosis of olfactory impairment on an
olfactory identification test, olfactory threshold and discrimination scores were
also assessed. The questionnaire method to assess subjective chemosensory loss was
different: while we asked the patients to rate the general taste and smell functions,
Soter et al.[26] asked the patients to rate specific scenarios of flavour perception. Furthermore,
the patients in the present study were all referred from ENT specialists, while information
regarding referral was not included by Soter et al.[26] One could hypothesize that the risk of chemosensory misclassification is lower in
patients assessed by an ENT specialist who has increased awareness of chemosensory
function and related organs. With our selected cohort, these findings contribute to
the findings made by Soter et al.[26] In general, this suggests that patients are rather inaccurate in reporting a genuine
taste problem, irrespective of being asked to rate the general taste function or regarding
questions on specific tastants. Contrarily, patients reporting the non-existence of
a taste problem are typically correct in their assessment, and gustatory testing is
in that case unnecessary.
Subjective Loss of Taste Usually Reflects Loss of Smell
We found that patients with a self-assessed impaired gustatory function (n = 221) often had a normal taste test (n =1 38) results and a reduced measured olfactory function (n = 186/221, 84%). This finding may suggest that the patients have a tendency to classify
an olfactory impairment as a subjective taste disorder. This corresponds with the
findings of a previous study, in which patients complaining of isolated taste loss
were three times more prone to demonstrate an olfactory deficit than a gustatory deficit.[24] Furthermore, in a recent study[27] on 358 patients with subjective loss of taste, only 9.5% of patients with a combined
taste and smell complaint had a measurable gustatory impairment, whereas 87% had an
impaired olfactory function. However, in this study,[27] the suprathreshold test for gustatory function was applied, and no olfactory discrimination
test was included. This limited the interpretation of the relationship between the
two senses, and highlighted the need for the present study.
The aforementioned findings may reflect that patients confuse smell impairment, taste
impairment, and combined chemosensory impairment. Overall, olfactory impairment should
be considered as an underlying cause when patients present with a subjective loss
of taste.
Limitations and Strengths of the Study
As with prior studies, the population of patients included in the present study is
a selected cohort. All patients in the present study were referred to a specialized
taste and smell clinic by an ENT specialist. However, as the results are comparable
with those of previous studies, the current findings highlight the argument that chemosensory
testing is imperative for accurate diagnostics, irrespective of whether the patients
have previously been assessed by an ENT specialist.
In the present study, cognitive function or lack of confrontation with the nature
of their chemosensory deficit are not likely factors that can explain the current
results. All patients were confronted with their claimed subjective chemosensory loss
prior to the referral by an ENT specialist, and 99.3% (n = 598) of all patients had a normal MMSE score at the time of their visit.
The study uses two different taste tests to estimate gustatory function in the patient
group. Optimally, such testing should be performed with one test only to make the
data more homogeneous. However, both tests are validated with good reproducibility,
and are used in clinical settings.
Quantitative measures were used for the olfactory and gustatory test scores. However,
such measures do not entirely cover the subjective disturbances for these senses.
Distortions or phantom taste or smell sensations could also be experienced by patients,
which is not always directly reflected in the olfactory test scores.[28]
[29] This adds to the complexity of chemosensory dysfunction and calls for a combination
of chemosensory-specific patient history and thorough chemosensory testing.
Conclusion
Subjective gustatory dysfunction was poorly correlated with measured gustatory dysfunction.
Instead, it was often found to reflect olfactory dysfunction. Contrarily, subjective
olfactory dysfunction was a more reliable measurement of olfactory dysfunction. Although
subjective anosmia was a strong indicator of measured anosmia or hyposmia, the existence
of remaining olfactory function was frequently found in these patients. This finding
is relevant for prognosis and effects of olfactory training.
As such, validated chemosensory testing should be performed in patients with a perceived
olfactory or gustatory deficit, as this could help ensure increased diagnostic precision
and relevant treatment.
