Prognostic Factors of Olfactory Dysfunction

ORIGINAL ARTICLE Prognostic Factors of Olfactory Dysfunction

Thomas Hummel, MD; Jörn Lötsch, MD

Objective: To determine appropriate counseling of ond assessment, 278 (31.1%) had functional anosmia, 496 patients with olfactory dysfunction. (55.5%) had hyposmia, and 120 (13.4%) had normal olfactory function (PϽ.001). Changes in smell scores de- Design: Retrospective analysis. pended positively on the initial score and negatively on age and smoking habits. Normosmia was more likely to Setting: Interdisciplinary Center for Smell and Taste, be restored in females and when residual olfactory func- University of Dresden Medical School, Technical Uni- tion was relatively high. In contrast, the origin of dys- versity of Dresden. function had no direct predictive value because it was mostly reflected by initial smell scores. However, in a sub- Patients: A total of 361 males and 533 females, aged analysis omitting the initial olfactory performance as a 11 to 84 years, who twice reported to the Interdiscipli- potential predictor, the initial presence of parosmia was nary Center for Smell and Taste. associated with a lower probability of anosmia as the final outcome. Main Outcome Measures: Residual olfactory performance, duration of olfactory loss until first assessment, presence or absence of parosmia or phantosmia, origin Conclusion: The prognosis of olfactory dysfunction of olfactory loss, interval between assessments, age, sex, mainly depends on residual function, sex, parosmia, smok- and smoking habits. ing habits, and age, whereas in this statistical model, origin plays only a second-line role, reflected in different Results: Although 431 patients (48.2%) had functional degrees of initial olfactory loss. anosmia at the first olfactory assessment, 444 (49.7%) had hyposmia, and 19 (2.1%) had normosmia; at the sec- Arch Otolaryngol Head Neck Surg. 2010;136(4):347-351

OSS OF OLFACTORY ACUITY TO dysfunction, such as origin of olfactory loss, the degree of anosmia is pre- specific demographic factors, and the ini- sent in approximately 5% of tial severity of olfactory dysfunction. the population.1-4 In fact, it has been estimated for Aus- METHODS tria,L Switzerland, and Germany that approximately 80 000 people consult an oto- PATIENTS rhinolaryngologist every year because of problems related to sense of smell.5 Study participants (Table) were 361 males and Today’s treatment options for olfactory 533 females, aged 11 to 84 years, who had re- dysfunction are limited. Most improve- ported to the Interdisciplinary Center for Smell ments appear to be owing to spontaneous and Taste of the University of Dresden Medical recovery. However, many physicians may School, Technical University of Dresden, with concerns related to the sense of taste or smell. Author Affiliations: not provide the patient with a prognosis of The origin of the concerns was determined on Interdisciplinary Center for his or her olfactory loss. To increase the in- the basis of an exhaustive, standardized inter- Smell and Taste, Department of formational basis for appropriate counsel- view, including questions for smoking status. Otorhinolaryngology, University ing of patients, we retrospectively ana- In addition, all participants underwent thor- of Dresden Medical School, lyzed prospectively collected olfactory ough otorhinolaryngologic examinations, in- Technical University of function data of patients who sought treat- cluding nasal endoscopy. All participants un- Dresden, Dresden, Germany derwent detailed tests of orthonasal olfactory (Dr Hummel); and Institute for ment at a specialized otorhinolaryngologic function and retronasal or gustatory function Clinical Pharmacology, Goethe center at least 2 times. We focused our whenever deemed necessary. Electrophysi- University Frankfurt, Frankfurt, analysis on factors that would provide in- ologic investigations using olfactory event- Germany (Dr Lötsch). formation about the prognosis of olfactory related potentials and imaging of the head (com-

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 136 (NO. 4), APR 2010 WWW.ARCHOTO.COM 347

©2010 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/03/2021 tive forced-choice task with presentation of a list of 4 descrip- Table. Characteristics of Patients’ Demographics tors for each pen (reference range, Ն12 correct identifica- and Olfactory Diagnoses tions). The clinical evaluation of olfactory performance was based on the composite threshold discrimination identification (TDI) No. (%) score represented by the sum of the scores from the 3 sub- Characteristic of 894 Patientsa tests.7 Pathological olfactory function was indicated by a TDI Sex score of 30.5 or less, with the separation of hyposmia from func- 6 Male 361 (40.4) tional anosmia at a TDI score of15.5. Female 533 (59.6) Age, mean (range), y 55.6 (19.0-97.0) STATISTICAL ANALYSIS Weight, mean (range), kg 74.7 (40.0-160.0) Current smokers 76 (8.5) Olfaction-related condition A patient with deteriorated olfactory function will probably URT infection 463 (51.8) be interested in the future change in olfactory performance Trauma 211 (23.6) and in the final olfactory diagnosis. Thus, target factors were Sinunasal disease 220 (24.6) (1) the numerical change in the TDI score from the first to the Diagnoses second assessment (ie, second minus first TDI scores); (2) a At first assessment clinically significant change in olfactory diagnosis by 6 points, Anosmia 431 (48.2) coded as −1 for worsening, 0 for changes of fewer than 6 points, Hyposmia 444 (49.7) and 1 for TDI score improved by 6 or more points8; and Normosmia 19 (2.1) (3) the final diagnosis number, coded as 1, 2, or 3, respec- Parosmia 301 (33.7) tively, for the 3 diagnoses. Candidate factors were (1) mean (SD) Phantosmia 151 (16.9) TDI score at the first assessment, (2) duration of olfactory loss At second assessment until the first assessment, (3) presence or absence of paros- Anosmia 278 (31.1) mia,9 (4) presence or absence of phantosmia,9 (5) origin of ol- Hyposmia 496 (55.5) factory loss (eg, sinunasal disease[SND], trauma, or upper res- Normosmia 120 (13.4) piratory tract [URT] infection), (6) interval between first and second assessments; and demographic factors, including (7) age Abbreviation: URT, upper respiratory tract. in years at first assessment, (8) sex, and (9) smoking habits. a Data are presented as number (percentage) of patients unless otherwise indicated. Baseline values of olfactory function were compared for the candidate factors by means of univariate analysis of variance with post hoc ␣-corrected (Bonferroni) t tests, simple t tests, puted tomography or magnetic resonance imaging) were rank correlation (Spearman), or ␹2 statistics, as appropriate for performed whenever it was thought necessary to clarify the di- the respective data. Candidate factors were submitted to- agnosis. The retrospective study complied with the Declaration gether to linear or binary logistic regression analysis (Stata/IC of Helsinki on Biomedical Research Involving Human Subjects 10.1 for Linux; StataCorp LP, College Station, Texas). To in- and had been approved by the Ethics Committee of the Univer- crease the validity of the results, we repeated regression analy- sity of Dresden Medical School, Technical University of Dresden. ses for 1000 data sets created from the original data set by means of bootstrap resampling.10 Subsequently, the influence of factors identified by regression analysis as exerting significant ef- OLFACTORY TESTING fects on the target factors were assessed by means of analysis of variance, ␹2 statistics, or rank correlation. To adjust for dif- The first olfactory assessment took place when the patient re- ferent baseline TDI scores, the percent changes from the first ported to the Smell and Taste Clinic for the first time. The sec- to the second TDI assessments were analyzed in addition to the ond assessment followed at a mean (SD) interval of 1.3 (1.29) differences in TDI scores (ie, [second − first score]/[first score years (range, 0.083-8.83 years). Olfactory function was as- ϫ 100]). sessed birhinally with the extended version of the Sniffin’ Sticks test (Heinrich Burghart GMBh, Wedel, Germany).6 In this vali- dated test, odors are presented in felt-tip pens. For odor pre- RESULTS sentation, 1 pen at a time (cap removed) is placed in front of the nostrils at a distance of approximately 1 to 2 cm. BASELINE OLFACTORY FUNCTION Odor thresholds were obtained for the roselike odor phenyl ethyl alcohol, presented in 16 successive 1:2 dilutions in liquid form, starting with a concentration of 4%. Using a 3-al- Reasons for olfactory loss were SND, head trauma, and ternative forced-choice task and a staircase paradigm starting URT infections. The olfactory function at the first as- at low phenyl ethyl alcohol concentrations, 1 pen containing a sessment (Table) differed among origin groups (PϽ.001, certain odorant and 2 without odorant were presented at each ␹2 test). That is, although anosmia was equally associ- dilution step. Two successive correct identifications or 1 in- ated with SND (30.6% of patients with anosmia had SND correct identification triggered a reversal of the staircase. Odor as the origin of their condition), trauma (34.3%) and URT threshold was defined as the mean of the last 4 of 7 staircase infection (35.0%), hyposmia or normosmia was most reversals (reference ranges, Ͼ6 for males and Ͼ6.5 for fe- 6 frequently associated with URT infection (66.9% of males ). Odor discrimination was determined with 16 triple sets patients with hyposmia and 78.9% of those with norm- of pens, 2 containing the same odorant and 1 containing a dif- osmia had this origin of olfactory dysfunction). Because ferent one (for detailed listing of the odorants used, see Hum- meletal6), using a 3-alternative forced-choice task (reference they were closely related to the olfactory diagnosis, the range, Ն11 correct discriminations). Odor identification was mean (SD) initial TDI scores differed between origin determined with 16 odors (orange, peppermint, turpentine, groups (SND, 15.1 [6.0]; trauma, 14.1 [6.0]; URT infec- clove, leather, banana, garlic, rose, fish, lemon, coffee, anise, tion, 19.0 [6.1]; analysis of variance: df=2893, F=2220.4, cinnamon, licorice, apple, and pineapple), using a 4-alterna- PϽ.001; post hoc ␣-corrected t tests: significant differ-

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 136 (NO. 4), APR 2010 WWW.ARCHOTO.COM 348

©2010 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/03/2021 A Smokers, URT infection Nonsmokers, URT infection B Smokers, trauma Nonsmokers, trauma 50 Smokers, SND Nonsmokers, SND 50 Identity Regression line

40 40 Normosmia

30 30 Hyposmia

TDI Sum Score 20 20

10 10 Anosmia

0 First Test Second Test First Test Second Test 0102030 40 50 Men Women Initial TDI Sum Score Gender and Olfactory Test Number

Figure. Threshold discrimination identification (TDI) scores. A, Distribution of the TDI scores at the first and second olfactory assessments, separately for sex. Dots indicate individual data points; red indicates an origin of olfactory loss as sinunasal; green, upper respiratory tract (URT) infection; and blue, trauma. SND indicates sinunasal disease. The boxes span the 25th to 75th percentiles, with the median crossing the box as a horizontal line and the whiskers spanning values within 1.5 times the 25th to 75th percentiles. The limits of olfactory diagnoses are shown as horizontal dotted lines. B, Scatterplot of the final (ordinate) vs the initial (abscissa) TDI scores. The dotted diagonal lines indicate equality between TDI scores at the second and the first assessments. The greater the distance from the identity line to the upper left, the greater the improvement of olfactory acuity observed in the respective patient.

ences between infection and the other origins, PϽ.001, TDI sum scores between the second and first olfactory but not between SND and trauma, P=.24; Figure). Fe- assessments. For the percent changes in TDI score, re- males had significantly, but only slightly better, mean (SD) gression analysis found the initial TDI score (PϽ.001) initial TDI scores than males (17.4[6.4] vs 16.1[6.4]; and age (P=.02) to be significant predictors, and when P=.004). Sex did not significantly differ with respect to omitting the first of the candidate factors, age (P=.003) age (mean [SD] age: males, 54.9[13.4] years; females, and smoking habits (P=.04) were also found to be sig- 56.4[11.8] years; t test for age, P=.08), and age was not nificant predictors. The predictions were in the same di- correlated with the initial TDI score (Spearman ␳=0.0571, rection as for the difference changes in TDI scores. P=.09). Finally, smokers (n=76) had the same TDI sores as nonsmokers (n=818; P=.41). Improvement in Olfactory Function

TEMPORAL CHANGES Clinically relevant improvement of olfactory function, de- OF OLFACTORY FUNCTION fined as an increase in TDI score by at least 6 points, was observed in 263 patients (29.4%). Only 47 patients (5.3%) Changes in TDI Score worsened significantly in their olfactory function, whereas no significant change was observed in 584 patients Linear regression analysis identified the initial TDI score (65.3%). (PϽ.001; Figure), the presence of parosmia (P=.03), age Linear regression analysis identified that changes in (P=.04), and smoking habits (P=.04) as significant fac- TDI scores by a relevant degree (ie, by Ն6 points) de- tors modulating the numerical difference in TDI sum pended on the initial TDI score (PϽ.001), smoking hab- scores between the second and first olfactory assess- its (P=.02), the presence of parosmia (P=.04), and the ments. However, parosmia failed to pass the post hoc duration of olfactory loss until first assessment (P=.047). t test (P=.20). Changes in the TDI score depended nega- However, the duration of the olfactory loss until first as- tively, although weakly, on the initial TDI score (Spear- sessment and the presence of parosmia failed to pass the man ␳=−0.2236, PϽ.001). In addition, changes in mean post hoc analysis (P=.20 and .49, respectively; ␹2 tests). (SD) TDI score were significantly smaller in smokers Of the remaining factors, significant change in TDI was (1.88[5.67]) than in nonsmokers (3.72[6.32]; t test, negatively associated with initial TDI scores; that is, pa- P=.02). Age was weakly but significantly negatively cor- tients whose olfactory acuity worsened had initially higher related with the increase in TDI score between assess- TDI scores, whereas those who improved had the low- ments (Spearman ␳=−0.067, P=.04). When omitting the est initial scores (change of Յ6 points: mean [SD] ini- initial TDI score from the regression analysis, age (P=.01) tial TDI score, 22.3[5.8]; change of Ͻ6 absolute points: and smoking habits (P=.03) were again identified as sig- TDI score, 17.2[6.6]; change of Ն6 points: TDI score, nificant factors modulating the numerical difference in 15.1[5.4]; analysis of variance: df=2893, F=29.7, PϽ.001;

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 136 (NO. 4), APR 2010 WWW.ARCHOTO.COM 349

©2010 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/03/2021 post hoc t tests: significant differences between all groups ity is of interest, higher initial scores predict lower im- at PϽ.001). Of smokers, 9.2% worsened in their olfac- provement. By contrast, when normal olfactory func- tory function and only 19.7% improved, whereas of non- tion is of interest, higher initial scores are associated with smokers, only 4.4% worsened but 31.7% improved higher probability of normosmia. (P=.04; ␹2 test). When omitting the initial TDI score from It is widely recognized that women outperform men the regression analysis, only smoking habits (P=.01) were in olfactory tests.17 In this study, sex did not result in a identified as a significant factor modulating the numeri- difference at the initial diagnosis of olfactory loss, but fe- cal difference in TDI sum scores between the second and males improved more. Also, the fact that age and smok- first olfactory assessments. ing status are negative predictors of olfactory function is not surprising.4 However, the lack of importance of the Final Olfactory Diagnosis origin of olfactory loss for its prognosis disagrees with the common perception that olfactory function is more The initial distribution of olfactory diagnosis moved to- likely to be recovered when it was lost because of a URT ward improvement at the second assessment (PϽ.001; infection than because of head trauma. Indeed, separate ␹2 test for the diagnosis at the first vs that at the second analysis of improvement in olfactory function for its ori- assessment; Table). Linear regression analysis identi- gin resulted in significant effects. That is, the numerical fied the initial TDI score (PϽ.001), sex (P=.006; Fig- changes in TDI score depended on the origin of olfac- ure), and the duration of olfactory loss until the first as- tory loss (analysis of variance of the mean [SD] differ- sessment (P=.02) as significant factors affecting the final ences between the second and first TDI scores: df=2893, olfactory diagnosis. However, the duration of olfactory F=34.9, PϽ.001; SND, 5.8[7.6] points; trauma, 0.9[4.9] loss until the first assessment failed post hoc analysis of points; infection, 3.3[5.8] points; percent changes from variance (P=.92). With respect to the final significant fac- the first TDI score: F=32.2, PϽ.001; SND, 52.3%[75.8%]; tors, the better the TDI score at the start of the assess- trauma, 11.8%[43.6%]; infection, 25%[45.8%]; ␣- ment, the better the diagnosis at the end of the assess- corrected post hoc t tests: PϽ.05 for all groups). The ori- ment (analysis of variance: df=2893, F=199.1, PϽ.001; gin of olfactory loss was also significantly related to clini- mean [SD] initial TDI scores: final anosmia, 11.7[4.1]; cally relevant improvement or worsening of olfactory final hyposmia, 18.4[5.4]; final normosmia, 22.8[7.6]; function (PϽ.001, ␹2 test). It was furthermore signifi- post hoc t tests: significant differences between all groups cantly different between groups of final diagnosis at PϽ.001). Males had slightly worse final diagnoses than (PϽ.001, ␹2 test). Although 61.6% of the patients with females as indicated by 38.1%, 54.0%, and 7.9% of the trauma had anosmia at the final diagnosis, this was the males but 23.3%, 65.1%, and 11.6% of the females hav- case in only 30.4% and 17.5% of the patients with SND ing anosmia, hyposmia, or normosmia, respectively, as or URT infection, respectively. On the other hand, nor- his or her final diagnosis (PϽ.001, ␹2 test). When omit- mal olfactory function at the second assessment was seen ting the initial TDI score from the regression analysis, in 19.5% of the patients with SND, 15.3% of the patients sex (P=.001) and the duration of olfactory loss until the with URT infection, but only 2.8% of the patients with first assessment (P=.001) were again identified as sig- trauma as the origin of their olfactory loss. Neverthe- nificant factors modulating the numerical difference in less, multivariate statistical analysis failed to identify ori- TDI sum scores between the second and first olfactory gin as a significant predictor of the development of ol- assessments. In addition, the presence of parosmia was factory function after it had been decreased or lost. A a further factor identified in the reduced regression analy- possible reason for this apparent discrepancy was the in- sis (PϽ.001). Patients who had parosmia at the initial terdependence of initial olfactory function and its ori- assessment had slightly better final diagnoses. This was gin. However, when omitting the initial diagnosis from indicated by 36.3%, 54.0%, and 9.7% of the patients with the candidate factors, origin still failed to provide a sta- parosmia, but 11.0%, 79.1%, and 9.9% of the patients tistically significant prognostic factor. Thus, consistent without parosmia having anosmia, hyposmia, or norm- with the literature,11 the origin of olfactory loss appears osmia, respectively, as their final diagnosis (PϽ.001, ␹2 to play a smaller role in recovery of olfactory function test). than intuitively perceived. Sex, age, smoking, and pos- sibly the presence of parosmia, which could indicate re- generative processes,9 are more relevant prognostic COMMENT predictors. In conclusion, we show that the prognosis of olfac- Improvement of olfactory acuity was most likely in young tory dysfunction primarily depends on residual func- nonsmokers with severely restricted olfactory perfor- tion and secondarily on sex, smoking habits, age, and par- mance. Clinically relevant improvement of olfactory loss osmia, whereas in this statistical model, origin of olfactory was most likely in nonsmokers with a low TDI score at loss plays only a second-line role, reflected in different the time of reporting to the clinic. However, normal ol- degrees of initial olfactory loss. factory function at the second assessment was most likely in females with initially better TDI scores. Olfactory func- Submitted for Publication: April 19, 2009; final revi- tion at the first visit was the strongest predictor of de- sion received July 9, 2009; accepted July 19, 2009. velopment of olfactory function, as previously re- Correspondence: Thomas Hummel, MD, Interdiscipli- ported,11 with a dependence, however, on the kind of nary Center for Smell and Taste, Department of Otorhi- prognosis.12-16 When simple improvement in olfactory acu- nolaryngology, University of Dresden Medical School,

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 136 (NO. 4), APR 2010 WWW.ARCHOTO.COM 350

©2010 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 10/03/2021 Technical University of Dresden, Fetscherstr 74, 01307 dysfunction: the Skövde population-based study. Laryngoscope. 2004;114 Dresden, Germany ([email protected]). (4):733-737. 3. Landis BN, Konnerth CG, Hummel T. A study on the frequency of olfactory Author Contributions: Both authors had full access to dysfunction. Laryngoscope. 2004;114(10):1764-1769. all the data in the study and take responsibility for the 4. Vennemann MM, Hummel T, Berger K. The association between smoking and integrity of the data and the accuracy of the data analy- smell and taste impairment in the general population. J Neurol. 2008;255(8): sis. Study concept and design: Hummel. Analysis and in- 1121-1126. 5. Damm M, Temmel A, Welge-Lüssen A, et al. Olfactory dysfunctions: epidemi- terpretation of data: Hummel and Lötsch. Drafting of the ology and therapy in German, Austria, and Switzerland [in German]. HNO. 2004; manuscript: Hummel and Lötsch. Critical revision of the 52(2):112-120. manuscript for important intellectual content: Hummel 6. Hummel T, Sekinger B, Wolf S, Pauli E, Kobal G. “Sniffin’ Sticks”: olfactory per- and Lötsch. Statistical analysis: Lötsch. Study supervi- formance assessed by the combined testing of odor identification, odor discrimi- sion: Hummel. nation and olfactory threshold. Chem Senses. 1997;22(1):39-52. 7. Wolfensberger M, Schnieper I, Welge-Lüssen A. “Sniffin’ Sticks”: a new olfac- Financial Disclosure: None reported. tory test battery. Acta Otolaryngol. 2000;120(2):303-306. Funding/Support: The work was supported by the 8. Gudziol V, Lötsch J, Hähner A, Zahnert T, Hummel T. Clinical significance of re- Interdisciplinary Center for Neuroscience Frankfurt sults from olfactory testing. Laryngoscope. 2006;116(10):1858-1863. (Dr Lötsch). 9. Reden J, Maroldt H, Fritz A, Zahnert T, Hummel T. A study on the prognostic significance of qualitative olfactory dysfunction. Eur Arch Otorhinolaryngol. 2007; Previous Presentation: A subset of these data, includ- 264(2):139-144. ing less than half the number of patients, has been pub- 10. Efron B, Tibshirani RJ. An Introduction to the Bootstrap. San Francisco, CA: Chap- lished previously by Reden J, Mueller A, Mueller C, man & Hall; 1995. et al. Arch Otolaryngol Head Neck Surg. 2006;132(3):265- 11. London B, Nabet B, Fisher AR, White B, Sammel MD, Doty RL. Predictors of prog- 269. nosis in patients with olfactory disturbance. Ann Neurol. 2008;63(2):159-166. 12. Aiba T, Sugiura M, Mori J, et al. Effect of zinc sulfate on sensorineural olfactory Additional Contributions: Bettina Hauswald, MD, Volker disorder. Acta Otolaryngol Suppl. 1998;118(538):202-204. Gudziol, MD, Mareike Trittin, MD, Simona Negoias, MD, 13. Mori J, Aiba T, Sugiura M, et al. Clinical study of olfactory disturbance. Acta Oto- Franziska Krone, and Monika Roesner helped with ac- laryngol Suppl. 1998;538(6):197-201. quisition and preparation of the data. 14. Costanzo RM, Zasler ND. Head trauma. In: Getchell TV, Doty RL, Bartoshuk LM, Snow JBJ, eds. Smell and Taste in Health and Disease. New York, NY: Raven Press; 1991:711-730. 15. Seiden AM. Olfactory loss secondary to nasal and sinus pathology. In: Seiden REFERENCES AM, ed. Taste and Smell Disorders. New York, NY: Thieme Medical Publishers; 1997:52-71. 16. Leopold D. Distortion of olfactory perception: diagnosis and treatment. Chem 1. Murphy C, Schubert CR, Cruickshanks KJ, Klein BE, Klein R, Nondahl DM. Preva- Senses. 2002;27(7):611-615. lence of olfactory impairment in older adults. JAMA. 2002;288(18):2307-2312. 17. Doty RL, Cameron EL. Sex differences and reproductive hormone influences on 2. Brämerson A, Johansson L, Ek L, Nordin S, Bende M. Prevalence of olfactory human odor perception. Physiol Behav. 2009;97(2):213-228.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 136 (NO. 4), APR 2010 WWW.ARCHOTO.COM 351