Measurement precision of an olfactory perception threshold test for use in field studies

Changes in olfactory function have been associated with workplace exposure to a variety of substances. In the workplace, smell can be particularly important, since it is commonly used to detect potentially hazardous situations or as an indicator of mask cartridge breakthrough. Sensitive quantitative measures of olfactory loss would be useful in epidemiological studies and workplace surveillance. The objective of the present study was to determine the reproducibility of an olfactory perception threshold test and variations with age, gender, and smoking status. The test was a standard olfactory kit (Olfactolab No. 11), including 18 serial dilutions of PM-carbinol, with an equal number of blanks. The forced choice method was used, with both tester and subject blinded as to which bottle contained the odorant. Olfactory perception threshold was recorded when the subject identified the same dilution three times. To assess reproducibility, testing was repeated four times over a period of 4 weeks, on the same weekday and the same time of day. Subjects (n = 63) ranged in age from 20 to 60 years (mean age: 39.7 +/- 12.5 years), 47.6% were women, 29.5% currently smoked, and 27.9% were former smokers. Results showed no inter-week differences in olfactory perception threshold (Analysis of variance for repeated measures: F = 0.59; p much greater than 0.05). Inter-class correlation for assessment of agreement of continuous variables was 0.76. Inter-week concordance of hyposmia showed fair to good agreement (0.55 greater than or equal to kappa less than or equal to 0.66). Three-way analysis of variance (ANOVA) revealed significant differences with respect to age category (F = 7.36; p less than 0.001) and current smoking status (F = 4.54; p less than 0.05), but not for gender (F = 2.32; p greater than 0.05). The multiple regression model with age and smoking as independent variables was highly significant (F = 13.03; p less than 0.001), explaining 28% of the variance; olfactory threshold increased 0.47 ds/year (t = 4.01; p less than 0.001) and 0.27 ds/cigarettes/day (t = 2.46; p less than 0.05). The findings of this study indicate that this test is reproducible and sensitive to expected changes with age and smoking status. It corresponds well to criteria for testing in the field and should be considered for studies characterizing olfactory functions and sensory loss among working populations.