Journal of Neuroscience Methods 243 (2015) 111–114
Contents lists available at ScienceDirect
Journal of Neuroscience Methods journal homepage: www.elsevier.com/locate/jneumeth
Clinical Neuroscience Short communication
Extended version of the “Sniffin’ Sticks” identification test: Test–retest reliability and validity A. Sorokowska a,b,∗,1 , E. Albrecht a,1 , A. Haehner a , T. Hummel a a b
Department of Otorhinolaryngology, TU Dresden, Germany Institute of Psychology, University of Wroclaw, Poland
h i g h l i g h t s • • • •
We examined the reliability of the revised, extended Sniffin’ Sticks test. We reconfirmed the high test–retest reliability and validity of this test. Additionally, we presented normative values for this test. The extended identification test is a useful and detailed diagnostic tool.
a r t i c l e
i n f o
Article history: Received 28 October 2014 Received in revised form 28 January 2015 Accepted 30 January 2015 Available online 10 February 2015 Keywords: Odor identification Sniffin’ Sticks Olfactory testing Smell Validity Reliability
a b s t r a c t Background: The extended, 32-item version of the Sniffin’ Sticks identification test was developed in order to create a precise tool enabling repeated, longitudinal testing of individual olfactory subfunctions. New method: Odors of the previous test version had to be changed for technical reasons, and the odor identification test needed re-investigation in terms of reliability, validity, and normative values. Results: In our study we investigated olfactory abilities of a group of 100 patients with olfactory dysfunction and 100 controls. We reconfirmed the high test–retest reliability of the extended version of the Sniffin’ Sticks identification test and high correlations between the new and the original part of this tool. In addition, we confirmed the validity of the test as it discriminated clearly between controls and patients with olfactory loss. Comparison with existing method(s): The additional set of 16 odor identification sticks can be either included in the current olfactory test, thus creating a more detailed diagnosis tool, or it can be used separately, enabling to follow olfactory function over time. Additionally, the normative values presented in our paper might provide useful guidelines for interpretation of the extended identification test results. Conclusions: The revised version of the Sniffin’ Sticks 32-item odor identification test is a reliable and valid tool for the assessment of olfactory function. © 2015 Elsevier B.V. All rights reserved.
1. Introduction People are often inaccurate in terms of the ratings of their own olfactory abilities (e.g., Landis et al., 2003). Accordingly, the sense of smell is typically assessed with specialized psychophysical tests (Hummel and Welge-Lüssen, 2006), for example with the “Sniffin’ Sticks” which is a well-established tool (Hummel et al., 1997, 2007).
∗ Corresponding author at: Smell and Taste Clinic, Department of Otorhinolaryngology, TU Dresden, Fetscherstrasse 74, 01307 Dresden, Germany. Tel.: +49 351 458 4189; fax: +49 351 458 7370. E-mail address: [email protected] (A. Sorokowska). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.jneumeth.2015.01.034 0165-0270/© 2015 Elsevier B.V. All rights reserved.
The “Sniffin’ Sticks” test consists of 3 subtests allowing to diagnose different elements of olfactory function – it enables testing odor threshold (OT), odor discrimination (OD), and odor identification (OI). In addition to the “classical” 16-item version of the identification and discrimination subtests, there is also the extended, 32-item version of these two tests (Haehner et al., 2009). They were developed in order to create more precise tools enabling repeated, longitudinal testing of individual olfactory subfunctions. Previous work has shown the test–retest reliability of these extended subtests (Haehner et al., 2009). However, because the manufacturing source of odors of the 2009 version (items 17–32) had to be changed for technical reasons, the odor identification test needed re-investigation in terms of reliability, validity, and normative values.
112
A. Sorokowska et al. / Journal of Neuroscience Methods 243 (2015) 111–114
2. Materials and methods Investigations were performed according to the Guidelines for Biomedical Studies Involving Human Subjects (Helsinki Declaration). The protocol was approved by the Ethics Committee of the Medical Faculty of the TU Dresden (application number 156052012). All subjects provided written informed consent prior to their inclusion to the study. 2.1. Participants In the study participated 100 healthy controls (49 women and 51 men) aged 22–70 years (M = 36.3; SD = 15.9) and 100 patients with olfactory loss (61 women and 39 men) aged 24–85 years (M = 59.2; SD = 13.4). The participants underwent diagnostic evaluation; they received a detailed otorhinolaryngological investigation including structured medical interview and nasal endoscopy. 2.2. Procedure The participants were examined with the extended, 32-item “Sniffin’ Sticks” identification test (compare Haehner et al., 2009). By means of a multiple-choice task, identification of odors was performed from a list of four descriptors each. In the data analysis, we used three test scores for each subject: (a) the first 16 items of the task (i.e., the “classic” Sniffin’ Sticks; score 0-no odor identified to 16-all odors identified), (b) the 16 “new” items of the task (score 0–16), and (c) the total of 32 items (i.e., the extended version of the task; score 0–32). In controls, testing was performed again after a mean interval of one week to investigate test–retest reliability. Statistical analyses were performed by means of Statistica v.10 (StatSoft, Inc., www.statsoft.com, Tulsa, OK, USA). MANOVA analyses with Bonferroni post hoc tests were employed for comparison between scores of (i) patients and controls and (ii) healthy men and women. Correlation analyses were performed using Pearson statistics.
Table 1 Percentages of correct identifications of every item in controls and patients. Item
Odor
Controls
Patients
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Orange Leather Cinnamon Peppermint Banana Lemon Liquorice Turpentine Garlic Coffee Apple Cloves Ananas Rose Anise Fish Pear Coke Lilac Grapefruit Grass Raspberry Honey Ginger Coconut Lavender Melon Peach Mushroom Smoked ham Chocolate Onion
97 91 73 98 92 66 87 42 95 88 42 95 85 90 90 94 84 87 98 72 79 88 95 75 87 72 73 84 98 69 98 77
68 64 43 61 57 41 41 37 61 64 27 71 45 66 47 69 56 45 69 38 40 60 71 57 57 52 47 49 61 56 79 57
Mean percentage total Mean percentage items 1–16 Mean percentage items 17–32
83.2 82.8 83.5
54.9 53.9 55.9
Table 2 Differences in scores obtained by controls and patients with olfactory disorders. Controls
3. Results 3.1. Comparison between healthy subjects and patients with olfactory dysfunction Table 1 presents the percentages of correct identifications of each item for the patients and the healthy controls. On average, the patients identified the items correctly in 54.9% of the cases, and healthy participants in 83.2% of the cases. As indicated by difference in proportions tests (two-tailed tests), in both groups the percentages of correct identifications were not different in the “classic”, the “new” and the extended versions of the test. In the extended test, the scores of the patients ranged between 3 and 30 and the scores of the controls were between 10 and 31. The patients scored on average 9.05 lower than controls. Table 2 presents the comparison of scores obtained by controls and patients with olfactory dysfunctions. Generally, in all the tests the scores of controls were significantly different than the scores of the patients, as indicated by MANOVA analysis [Wilks’ Lambda = .62, F(2, 197) = 61.49, p < .001, 2 = .38]. Bonferroni post hoc tests revealed that in all three conditions, patients scored lower than controls (all ps < 05). With regard to normative data, in the extended, 32-item test the controls aged 22–35 years obtained scores between 22 and 31, with the mean value of M = 27.6 (SD = 2.0). The 10th percentile for the healthy subjects equaled 25, and this value was taken as the cutoff point separating hyposmia from normosmia. With regard to the patients taking part in our study, 80% of people with smell
Percentage of correct identifications
“Classic” 16-item test “New” 16-item test Extended 32-item test
Patients
M
SD
M
SD
13.25 13.36 26.61
1.73 2.12 3.34
8.62 8.94 17.56
4.16 3.68 7.43
loss obtained scores equal or lower than this value. Since generally the probability of providing more than 14 correct answers by chance in the 32-item identification test is less than 5%, this result cannot be distinguished from a purely random result and must be interpreted as being consistent with functional anosmia (see e.g., Wolfensberger et al., 2000). It therefore follows that this value should be taken as a cutoff point separating hyposmia from functional anosmia. In our sample, 39% of the patients and none of the controls aged less than 36 years obtained scores equal to or less than this value. 3.2. Effect of sex Women scored on average 13.2 in the “classic” 16 items, 13.65 in the “new” 16 items, and 26.86 in the full, extended test. Men scored on average 13.29 in the “classic” 16 items, 13.08 in the “new” 16 items, and 26.37 in the full, extended test. Men and women in our sample did not perform differently in any of the tests as indicated by MANOVA analysis [Wilks’ Lambda = .98, F(2, 197) = 1.70, p = .19, 2 = .02].
A. Sorokowska et al. / Journal of Neuroscience Methods 243 (2015) 111–114
113
6.27
Difference between two sessions
4.18
2 SD
2.09
0
-2.09
-2 SD
-4.18
-6.27 0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
Idenficaon test result Fig. 1. Bland–Altman diagram for test–retest reliability of the extended, 32-item Sniffin’ Sticks identification test in healthy controls (SD – standard deviation of the differences in scores obtained during the first and the second testing).
3.3. Test–retest reliability and the “classic” vs. the new test–correlation analyses We correlated the scores of all tests across all the subjects, the controls and the patients, respectively (Pearsons r correlations). The results of “classic” test and the “new” test were significantly correlated across all the subjects (r = .83; p ≤ .001), the controls (r = .49; p ≤ .001) and the patients (r = .80; p ≤ .001). The test–retest reliability in controls was very high – the results of the repeated testing were significantly correlated with the scores obtained during the first meeting in the case of the “classic” test (r = .60; p ≤ .001), the “new” test (r = .67; p ≤ .001) and, most importantly, the full, extended test (r = .76; p ≤ .001). Additionally, the controls’ scores obtained during the first testing with the “classic” test significantly correlated with the scores they obtained during the second testing with the “new” test, and vice versa – the scores from the first testing with the “new” test correlated with the scores of the second testing with the “classic” test (r = .46; p ≤ .001 and r = .38; p ≤ .001, respectively). 4. Discussion Olfactory identification tests are useful, both in regular otorhinolaryngological practice and also in scientific studies. Additionally, smell loss might be associated with a number of different diseases (Landis et al., 2004), which highlights the importance of proper olfactory testing for specialists from various disciplines. In our study, for technical reasons related to the change of manufacturer, we reinvestigated the extended, 32-item Sniffin’ Sticks identification subtest (Haehner et al., 2009). We reconfirmed that the full, extended test and its two parts are reliable and highly correlated with each other. The extended test was also highly discriminative in terms of a difference between scores of controls and patients with olfactory disorders. Generally, the described test can be used in its full version as a simple tool allowing detailed diagnosis of odor identification, or it could be split onto two individual tests, used for repeated measurements of odor identification over time (Fig. 1).
Olfactory disorders are relatively frequent with about 5% of the general population exhibiting functional anosmia and approximately 15% exhibiting reduced olfactory function (Landis et al., 2004). Olfactory disorders have different causes. Generally, age is a main factor leading to chemosensory dysfunction (Sorokowska et al., 2014), but often the disorders result from upper respiratory tract infections (URTI), or chronic rhinosinusitis (CRS) (WelgeLüssen and Wolfensberger, 2006). Such olfactory dysfunctions may be transitory (Reden et al., 2006), requiring frequent testing for documentation. It is possible to repeatedly use the same tool to monitor the changes in olfactory acuity (Gudziol et al., 2006), however in the case of a 16-item identification test it might be relatively easy for subjects to memorize the correct answers (or answer options). This, in turn, might result in an increase in score without real improvement of olfactory abilities. Reduced olfactory function can also be observed in various disorders, like early stages of idiopathic Parkinson’s disease (Haehner et al., 2007), or in depression (Pause et al., 2001). If a subject is repeatedly screened with the same identification test, it might be easy to miss a possible change in olfactory acuity, simply because the person knows the test. In both of these cases, the problem might be solved by examination with the additional 16 items, which can be used as a separate tool. At the same time the tests are highly correlated and methodologically identical to the “classic” Sniffin’ Sticks. Simple, reliable and validated olfactory tests – like the extended Sniffin’ Sticks identification test – seem to be useful in everyday medical and scientific practice. Smell tests are becoming more popular also in non-specialized units, for example in order to investigate possible side effects of drugs (Lötsch et al., 2012) or testing smell loss as an accompanying symptom of other disorders (Wolfensberger et al., 2000). Usage of the tool presented in the current study seems to be a good solution for such tasks, because it is a relatively quick method that, at the same time, is highly discriminative in terms of a difference between scores of controls and patients with olfactory disorders. As expected, in the current study the scores of controls were significantly higher than the scores of patients. We were able to
114
A. Sorokowska et al. / Journal of Neuroscience Methods 243 (2015) 111–114
mark a clear cutoff point separating hyposmia and normosmia (although some overlap remains reflecting the continuum between normosmia and hyposmia) – of the patients taking part in our study, 80% of people with smell loss obtained scores equal or lower than this value. Also, percentages of correct identifications were much lower for the patients than for the controls. The only two problematic items were “turpentine” and “apple”, for which the correct response rate was rather low also for the controls. Low identification rate for turpentine might have resulted from familiarity problem. As for the apple – we suggest that there might have been two possible sources of high error rate. First – the distractors could have been too difficult, i.e., too many smells were easily confused with the target odorant (see also: Negoias et al., 2010); second – the odor of apple used in Sniffin’ Sticks smells like a special species of apple – Granny Smith. As we found in Poland (Sorokowska and Hummel, 2014), some people might find it difficult to identify that this odor is actually “an apple”. Still, the two discussed items constitute a significant part of the well-investigated and commonly applied testing procedure with normative values in many countries, and they should not be changed in further versions of the Sniffin’ Sticks. In addition to the issues related to validity and reliability of the investigated test, we would like to highlight the result regarding sex differences (or rather no such differences) in our sample. Although the extended test enabled a detailed diagnosis of our participants’ sense of smell, we did not observe any significant sex differences in the olfactory identification scores. This result is consistent with the fact that sex-related differences in olfactory function are not always present, but if they are, then typically women would outperform men (e.g., Doty and Cameron, 2009). In conclusion, the current study demonstrates that the revised version of the Sniffin Sticks 32-item odor identification test is a reliable and valid tool for the assessment of olfactory function. Our findings seem to be of particular importance not only in the research involving psychophysical olfactory identification tests or in a clinical context, but also in other experiments investigating human olfaction and cognition. Acknowledgements We would like to thank Burghart, Wedel, Germany, for providing us with 3 sets of the new extended odor identification test. Data from this study were also used for a different publication on the effects of stimulus presentation order on performance in the 32-item version of the Sniffin’ Sticks odor identification test which
will be published elsewhere. During the project AS was supported by funds of Polish Ministry of Science and Higher Education (scholarship for years 2013–2016) and Polish National Science Centre (ETIUDA scholarship #2013/08/T/HS6/00408). The funding sources had no involvement in study design; in the collection, analysis and interpretation of data; in the writing of the report; nor in the decision to submit the article for publication. References Doty RL, Cameron EL. Sex differences and reproductive hormone influences on human odor perception. Physiol Behav 2009;97:213–28. Gudziol V, Lötsch J, Hähner A, Zahnert T, Hummel T. Clinical significance of results from olfactory testing. Laryngoscope 2006;116:1858–63. Haehner A, Hummel T, Hummel C, Sommer U, Junghanns S, Reichmann H. Olfactory loss may be a first sign of idiopathic Parkinson’s disease. Mov Disord 2007;22:839–42. Haehner A, Mayer AM, Landis BN, Pournaras I, Lill K, Gudziol V, et al. High test–retest reliability of the extended version of the Sniffin’Sticks test. Chem Senses 2009;34:705–11. Hummel T, Kobal G, Gudziol H, Mackay-Sim A. Normative data for the Sniffin’Sticks including tests of odor identification, odor discrimination, and olfactory thresholds: an upgrade based on a group of more than 3,000 subjects. Eur Arch Otorhinolaryngol 2007;264:237–43. Hummel T, Sekinger B, Wolf SR, Pauli E, Kobal G. ‘Sniffin’sticks’: olfactory performance assessed by the combined testing of odor identification, odor discrimination and olfactory threshold. Chem Senses 1997;22:39–52. Hummel T, Welge-Lüssen A. Assessment of olfactory function. Adv Otorhinolaryngol 2006;63:84–98. Landis BN, Hummel T, Hugentobler M, Giger R, Lacroix JS. Ratings of overall olfactory function. Chem Senses 2003;28:691–4. Landis BN, Konnerth CG, Hummel T. A study on the frequency of olfactory dysfunction. Laryngoscope 2004;114:1764–9. Lötsch J, Geisslinger G, Hummel T. Sniffing out pharmacology: interactions of drugs with human olfaction. Trends Pharmacol Sci 2012;33:193–9. Negoias S, Troeger C, Rombaux P, Halewyck S, Hummel T. Number of descriptors in cued odor identification tests. Arch Otolaryngol Head Neck Surg 2010;136:296–300. Pause BM, Miranda A, Göder R, Aldenhoff JB, Ferstl R. Reduced olfactory performance in patients with major depression. J Psychiatr Res 2001;35:271–7. Reden J, Mueller A, Mueller C, Konstantinidis I, Frasnelli J, Landis BN, et al. Recovery of olfactory function following closed head injury or infections of the upper respiratory tract. Arch Otolaryngol Head Neck Surg 2006;132:265–9. Sorokowska A, Schriever VA, Gudziol V, Hummel C, Hähner A, Iannilli E, et al. Changes of olfactory abilities in relation to age: odor identification in more than 1400 people aged 4 to 80 years. Eur Arch Otorhinolaryngol 2014., http://dx.doi.org/10.1007/s00405-014-3263-4. Sorokowska A, Hummel T. Polska wersja testu Sniffin’Sticks – adaptacja i normalizacja. Otolaryngol Pol 2014;68:308–14. Welge-Lüssen A, Wolfensberger M. Olfactory disorders following upper respiratory tract infections. In: Hummel T, Welge-Lüssen A, editors. Taste and smell. An update. Adv Otorhinolaryngol, vol. 63. Basel: Karger; 2006. p. 125–32. Wolfensberger M, Schnieper I, Welge-Lüssen A. Sniffin’Sticks: a new olfactory test battery. Acta Otolaryngol 2000;120:303–6.
Contents lists available at ScienceDirect
Journal of Neuroscience Methods journal homepage: www.elsevier.com/locate/jneumeth
Clinical Neuroscience Short communication
Extended version of the “Sniffin’ Sticks” identification test: Test–retest reliability and validity A. Sorokowska a,b,∗,1 , E. Albrecht a,1 , A. Haehner a , T. Hummel a a b
Department of Otorhinolaryngology, TU Dresden, Germany Institute of Psychology, University of Wroclaw, Poland
h i g h l i g h t s • • • •
We examined the reliability of the revised, extended Sniffin’ Sticks test. We reconfirmed the high test–retest reliability and validity of this test. Additionally, we presented normative values for this test. The extended identification test is a useful and detailed diagnostic tool.
a r t i c l e
i n f o
Article history: Received 28 October 2014 Received in revised form 28 January 2015 Accepted 30 January 2015 Available online 10 February 2015 Keywords: Odor identification Sniffin’ Sticks Olfactory testing Smell Validity Reliability
a b s t r a c t Background: The extended, 32-item version of the Sniffin’ Sticks identification test was developed in order to create a precise tool enabling repeated, longitudinal testing of individual olfactory subfunctions. New method: Odors of the previous test version had to be changed for technical reasons, and the odor identification test needed re-investigation in terms of reliability, validity, and normative values. Results: In our study we investigated olfactory abilities of a group of 100 patients with olfactory dysfunction and 100 controls. We reconfirmed the high test–retest reliability of the extended version of the Sniffin’ Sticks identification test and high correlations between the new and the original part of this tool. In addition, we confirmed the validity of the test as it discriminated clearly between controls and patients with olfactory loss. Comparison with existing method(s): The additional set of 16 odor identification sticks can be either included in the current olfactory test, thus creating a more detailed diagnosis tool, or it can be used separately, enabling to follow olfactory function over time. Additionally, the normative values presented in our paper might provide useful guidelines for interpretation of the extended identification test results. Conclusions: The revised version of the Sniffin’ Sticks 32-item odor identification test is a reliable and valid tool for the assessment of olfactory function. © 2015 Elsevier B.V. All rights reserved.
1. Introduction People are often inaccurate in terms of the ratings of their own olfactory abilities (e.g., Landis et al., 2003). Accordingly, the sense of smell is typically assessed with specialized psychophysical tests (Hummel and Welge-Lüssen, 2006), for example with the “Sniffin’ Sticks” which is a well-established tool (Hummel et al., 1997, 2007).
∗ Corresponding author at: Smell and Taste Clinic, Department of Otorhinolaryngology, TU Dresden, Fetscherstrasse 74, 01307 Dresden, Germany. Tel.: +49 351 458 4189; fax: +49 351 458 7370. E-mail address: [email protected] (A. Sorokowska). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.jneumeth.2015.01.034 0165-0270/© 2015 Elsevier B.V. All rights reserved.
The “Sniffin’ Sticks” test consists of 3 subtests allowing to diagnose different elements of olfactory function – it enables testing odor threshold (OT), odor discrimination (OD), and odor identification (OI). In addition to the “classical” 16-item version of the identification and discrimination subtests, there is also the extended, 32-item version of these two tests (Haehner et al., 2009). They were developed in order to create more precise tools enabling repeated, longitudinal testing of individual olfactory subfunctions. Previous work has shown the test–retest reliability of these extended subtests (Haehner et al., 2009). However, because the manufacturing source of odors of the 2009 version (items 17–32) had to be changed for technical reasons, the odor identification test needed re-investigation in terms of reliability, validity, and normative values.
112
A. Sorokowska et al. / Journal of Neuroscience Methods 243 (2015) 111–114
2. Materials and methods Investigations were performed according to the Guidelines for Biomedical Studies Involving Human Subjects (Helsinki Declaration). The protocol was approved by the Ethics Committee of the Medical Faculty of the TU Dresden (application number 156052012). All subjects provided written informed consent prior to their inclusion to the study. 2.1. Participants In the study participated 100 healthy controls (49 women and 51 men) aged 22–70 years (M = 36.3; SD = 15.9) and 100 patients with olfactory loss (61 women and 39 men) aged 24–85 years (M = 59.2; SD = 13.4). The participants underwent diagnostic evaluation; they received a detailed otorhinolaryngological investigation including structured medical interview and nasal endoscopy. 2.2. Procedure The participants were examined with the extended, 32-item “Sniffin’ Sticks” identification test (compare Haehner et al., 2009). By means of a multiple-choice task, identification of odors was performed from a list of four descriptors each. In the data analysis, we used three test scores for each subject: (a) the first 16 items of the task (i.e., the “classic” Sniffin’ Sticks; score 0-no odor identified to 16-all odors identified), (b) the 16 “new” items of the task (score 0–16), and (c) the total of 32 items (i.e., the extended version of the task; score 0–32). In controls, testing was performed again after a mean interval of one week to investigate test–retest reliability. Statistical analyses were performed by means of Statistica v.10 (StatSoft, Inc., www.statsoft.com, Tulsa, OK, USA). MANOVA analyses with Bonferroni post hoc tests were employed for comparison between scores of (i) patients and controls and (ii) healthy men and women. Correlation analyses were performed using Pearson statistics.
Table 1 Percentages of correct identifications of every item in controls and patients. Item
Odor
Controls
Patients
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Orange Leather Cinnamon Peppermint Banana Lemon Liquorice Turpentine Garlic Coffee Apple Cloves Ananas Rose Anise Fish Pear Coke Lilac Grapefruit Grass Raspberry Honey Ginger Coconut Lavender Melon Peach Mushroom Smoked ham Chocolate Onion
97 91 73 98 92 66 87 42 95 88 42 95 85 90 90 94 84 87 98 72 79 88 95 75 87 72 73 84 98 69 98 77
68 64 43 61 57 41 41 37 61 64 27 71 45 66 47 69 56 45 69 38 40 60 71 57 57 52 47 49 61 56 79 57
Mean percentage total Mean percentage items 1–16 Mean percentage items 17–32
83.2 82.8 83.5
54.9 53.9 55.9
Table 2 Differences in scores obtained by controls and patients with olfactory disorders. Controls
3. Results 3.1. Comparison between healthy subjects and patients with olfactory dysfunction Table 1 presents the percentages of correct identifications of each item for the patients and the healthy controls. On average, the patients identified the items correctly in 54.9% of the cases, and healthy participants in 83.2% of the cases. As indicated by difference in proportions tests (two-tailed tests), in both groups the percentages of correct identifications were not different in the “classic”, the “new” and the extended versions of the test. In the extended test, the scores of the patients ranged between 3 and 30 and the scores of the controls were between 10 and 31. The patients scored on average 9.05 lower than controls. Table 2 presents the comparison of scores obtained by controls and patients with olfactory dysfunctions. Generally, in all the tests the scores of controls were significantly different than the scores of the patients, as indicated by MANOVA analysis [Wilks’ Lambda = .62, F(2, 197) = 61.49, p < .001, 2 = .38]. Bonferroni post hoc tests revealed that in all three conditions, patients scored lower than controls (all ps < 05). With regard to normative data, in the extended, 32-item test the controls aged 22–35 years obtained scores between 22 and 31, with the mean value of M = 27.6 (SD = 2.0). The 10th percentile for the healthy subjects equaled 25, and this value was taken as the cutoff point separating hyposmia from normosmia. With regard to the patients taking part in our study, 80% of people with smell
Percentage of correct identifications
“Classic” 16-item test “New” 16-item test Extended 32-item test
Patients
M
SD
M
SD
13.25 13.36 26.61
1.73 2.12 3.34
8.62 8.94 17.56
4.16 3.68 7.43
loss obtained scores equal or lower than this value. Since generally the probability of providing more than 14 correct answers by chance in the 32-item identification test is less than 5%, this result cannot be distinguished from a purely random result and must be interpreted as being consistent with functional anosmia (see e.g., Wolfensberger et al., 2000). It therefore follows that this value should be taken as a cutoff point separating hyposmia from functional anosmia. In our sample, 39% of the patients and none of the controls aged less than 36 years obtained scores equal to or less than this value. 3.2. Effect of sex Women scored on average 13.2 in the “classic” 16 items, 13.65 in the “new” 16 items, and 26.86 in the full, extended test. Men scored on average 13.29 in the “classic” 16 items, 13.08 in the “new” 16 items, and 26.37 in the full, extended test. Men and women in our sample did not perform differently in any of the tests as indicated by MANOVA analysis [Wilks’ Lambda = .98, F(2, 197) = 1.70, p = .19, 2 = .02].
A. Sorokowska et al. / Journal of Neuroscience Methods 243 (2015) 111–114
113
6.27
Difference between two sessions
4.18
2 SD
2.09
0
-2.09
-2 SD
-4.18
-6.27 0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
Idenficaon test result Fig. 1. Bland–Altman diagram for test–retest reliability of the extended, 32-item Sniffin’ Sticks identification test in healthy controls (SD – standard deviation of the differences in scores obtained during the first and the second testing).
3.3. Test–retest reliability and the “classic” vs. the new test–correlation analyses We correlated the scores of all tests across all the subjects, the controls and the patients, respectively (Pearsons r correlations). The results of “classic” test and the “new” test were significantly correlated across all the subjects (r = .83; p ≤ .001), the controls (r = .49; p ≤ .001) and the patients (r = .80; p ≤ .001). The test–retest reliability in controls was very high – the results of the repeated testing were significantly correlated with the scores obtained during the first meeting in the case of the “classic” test (r = .60; p ≤ .001), the “new” test (r = .67; p ≤ .001) and, most importantly, the full, extended test (r = .76; p ≤ .001). Additionally, the controls’ scores obtained during the first testing with the “classic” test significantly correlated with the scores they obtained during the second testing with the “new” test, and vice versa – the scores from the first testing with the “new” test correlated with the scores of the second testing with the “classic” test (r = .46; p ≤ .001 and r = .38; p ≤ .001, respectively). 4. Discussion Olfactory identification tests are useful, both in regular otorhinolaryngological practice and also in scientific studies. Additionally, smell loss might be associated with a number of different diseases (Landis et al., 2004), which highlights the importance of proper olfactory testing for specialists from various disciplines. In our study, for technical reasons related to the change of manufacturer, we reinvestigated the extended, 32-item Sniffin’ Sticks identification subtest (Haehner et al., 2009). We reconfirmed that the full, extended test and its two parts are reliable and highly correlated with each other. The extended test was also highly discriminative in terms of a difference between scores of controls and patients with olfactory disorders. Generally, the described test can be used in its full version as a simple tool allowing detailed diagnosis of odor identification, or it could be split onto two individual tests, used for repeated measurements of odor identification over time (Fig. 1).
Olfactory disorders are relatively frequent with about 5% of the general population exhibiting functional anosmia and approximately 15% exhibiting reduced olfactory function (Landis et al., 2004). Olfactory disorders have different causes. Generally, age is a main factor leading to chemosensory dysfunction (Sorokowska et al., 2014), but often the disorders result from upper respiratory tract infections (URTI), or chronic rhinosinusitis (CRS) (WelgeLüssen and Wolfensberger, 2006). Such olfactory dysfunctions may be transitory (Reden et al., 2006), requiring frequent testing for documentation. It is possible to repeatedly use the same tool to monitor the changes in olfactory acuity (Gudziol et al., 2006), however in the case of a 16-item identification test it might be relatively easy for subjects to memorize the correct answers (or answer options). This, in turn, might result in an increase in score without real improvement of olfactory abilities. Reduced olfactory function can also be observed in various disorders, like early stages of idiopathic Parkinson’s disease (Haehner et al., 2007), or in depression (Pause et al., 2001). If a subject is repeatedly screened with the same identification test, it might be easy to miss a possible change in olfactory acuity, simply because the person knows the test. In both of these cases, the problem might be solved by examination with the additional 16 items, which can be used as a separate tool. At the same time the tests are highly correlated and methodologically identical to the “classic” Sniffin’ Sticks. Simple, reliable and validated olfactory tests – like the extended Sniffin’ Sticks identification test – seem to be useful in everyday medical and scientific practice. Smell tests are becoming more popular also in non-specialized units, for example in order to investigate possible side effects of drugs (Lötsch et al., 2012) or testing smell loss as an accompanying symptom of other disorders (Wolfensberger et al., 2000). Usage of the tool presented in the current study seems to be a good solution for such tasks, because it is a relatively quick method that, at the same time, is highly discriminative in terms of a difference between scores of controls and patients with olfactory disorders. As expected, in the current study the scores of controls were significantly higher than the scores of patients. We were able to
114
A. Sorokowska et al. / Journal of Neuroscience Methods 243 (2015) 111–114
mark a clear cutoff point separating hyposmia and normosmia (although some overlap remains reflecting the continuum between normosmia and hyposmia) – of the patients taking part in our study, 80% of people with smell loss obtained scores equal or lower than this value. Also, percentages of correct identifications were much lower for the patients than for the controls. The only two problematic items were “turpentine” and “apple”, for which the correct response rate was rather low also for the controls. Low identification rate for turpentine might have resulted from familiarity problem. As for the apple – we suggest that there might have been two possible sources of high error rate. First – the distractors could have been too difficult, i.e., too many smells were easily confused with the target odorant (see also: Negoias et al., 2010); second – the odor of apple used in Sniffin’ Sticks smells like a special species of apple – Granny Smith. As we found in Poland (Sorokowska and Hummel, 2014), some people might find it difficult to identify that this odor is actually “an apple”. Still, the two discussed items constitute a significant part of the well-investigated and commonly applied testing procedure with normative values in many countries, and they should not be changed in further versions of the Sniffin’ Sticks. In addition to the issues related to validity and reliability of the investigated test, we would like to highlight the result regarding sex differences (or rather no such differences) in our sample. Although the extended test enabled a detailed diagnosis of our participants’ sense of smell, we did not observe any significant sex differences in the olfactory identification scores. This result is consistent with the fact that sex-related differences in olfactory function are not always present, but if they are, then typically women would outperform men (e.g., Doty and Cameron, 2009). In conclusion, the current study demonstrates that the revised version of the Sniffin Sticks 32-item odor identification test is a reliable and valid tool for the assessment of olfactory function. Our findings seem to be of particular importance not only in the research involving psychophysical olfactory identification tests or in a clinical context, but also in other experiments investigating human olfaction and cognition. Acknowledgements We would like to thank Burghart, Wedel, Germany, for providing us with 3 sets of the new extended odor identification test. Data from this study were also used for a different publication on the effects of stimulus presentation order on performance in the 32-item version of the Sniffin’ Sticks odor identification test which
will be published elsewhere. During the project AS was supported by funds of Polish Ministry of Science and Higher Education (scholarship for years 2013–2016) and Polish National Science Centre (ETIUDA scholarship #2013/08/T/HS6/00408). The funding sources had no involvement in study design; in the collection, analysis and interpretation of data; in the writing of the report; nor in the decision to submit the article for publication. References Doty RL, Cameron EL. Sex differences and reproductive hormone influences on human odor perception. Physiol Behav 2009;97:213–28. Gudziol V, Lötsch J, Hähner A, Zahnert T, Hummel T. Clinical significance of results from olfactory testing. Laryngoscope 2006;116:1858–63. Haehner A, Hummel T, Hummel C, Sommer U, Junghanns S, Reichmann H. Olfactory loss may be a first sign of idiopathic Parkinson’s disease. Mov Disord 2007;22:839–42. Haehner A, Mayer AM, Landis BN, Pournaras I, Lill K, Gudziol V, et al. High test–retest reliability of the extended version of the Sniffin’Sticks test. Chem Senses 2009;34:705–11. Hummel T, Kobal G, Gudziol H, Mackay-Sim A. Normative data for the Sniffin’Sticks including tests of odor identification, odor discrimination, and olfactory thresholds: an upgrade based on a group of more than 3,000 subjects. Eur Arch Otorhinolaryngol 2007;264:237–43. Hummel T, Sekinger B, Wolf SR, Pauli E, Kobal G. ‘Sniffin’sticks’: olfactory performance assessed by the combined testing of odor identification, odor discrimination and olfactory threshold. Chem Senses 1997;22:39–52. Hummel T, Welge-Lüssen A. Assessment of olfactory function. Adv Otorhinolaryngol 2006;63:84–98. Landis BN, Hummel T, Hugentobler M, Giger R, Lacroix JS. Ratings of overall olfactory function. Chem Senses 2003;28:691–4. Landis BN, Konnerth CG, Hummel T. A study on the frequency of olfactory dysfunction. Laryngoscope 2004;114:1764–9. Lötsch J, Geisslinger G, Hummel T. Sniffing out pharmacology: interactions of drugs with human olfaction. Trends Pharmacol Sci 2012;33:193–9. Negoias S, Troeger C, Rombaux P, Halewyck S, Hummel T. Number of descriptors in cued odor identification tests. Arch Otolaryngol Head Neck Surg 2010;136:296–300. Pause BM, Miranda A, Göder R, Aldenhoff JB, Ferstl R. Reduced olfactory performance in patients with major depression. J Psychiatr Res 2001;35:271–7. Reden J, Mueller A, Mueller C, Konstantinidis I, Frasnelli J, Landis BN, et al. Recovery of olfactory function following closed head injury or infections of the upper respiratory tract. Arch Otolaryngol Head Neck Surg 2006;132:265–9. Sorokowska A, Schriever VA, Gudziol V, Hummel C, Hähner A, Iannilli E, et al. Changes of olfactory abilities in relation to age: odor identification in more than 1400 people aged 4 to 80 years. Eur Arch Otorhinolaryngol 2014., http://dx.doi.org/10.1007/s00405-014-3263-4. Sorokowska A, Hummel T. Polska wersja testu Sniffin’Sticks – adaptacja i normalizacja. Otolaryngol Pol 2014;68:308–14. Welge-Lüssen A, Wolfensberger M. Olfactory disorders following upper respiratory tract infections. In: Hummel T, Welge-Lüssen A, editors. Taste and smell. An update. Adv Otorhinolaryngol, vol. 63. Basel: Karger; 2006. p. 125–32. Wolfensberger M, Schnieper I, Welge-Lüssen A. Sniffin’Sticks: a new olfactory test battery. Acta Otolaryngol 2000;120:303–6.
