Percentage of subjects reporting impairment in selected activities of daily living.
Histogram showing the number of activities of daily living (ADL) affected by olfactory impairment reported by each subject.
Percentage of subjects reporting impairment in selected quality of life issues.
Level of overall satisfaction with life reported by subjects in both groups.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Miwa T, Furukawa M, Tsukatani T, Costanzo RM, DiNardo LJ, Reiter ER. Impact of Olfactory Impairment on Quality of Life and Disability. Arch Otolaryngol Head Neck Surg. 2001;127(5):497–503. doi:10.1001/archotol.127.5.497
To determine whether olfactory loss affects patients' quality of life or level of disability.
Retrospective survey using questionnaire data and clinic database review.
Two university medical center smell and taste clinics.
A total of 1407 patients were tested for smell and taste disturbances from 1984 through 1998. Surveys were mailed to 1093 patients who had abnormal test scores; 420 (38.4%) returned completed surveys. Patients were grouped by self-rated ability to smell as "impaired" (those reporting persisting deficits) or "improved" (those reporting no smell problem when surveyed).
Main Outcome Measures
Response frequencies were compared between the 2 groups for questions regarding ability to perform common activities of daily living and quality-of-life issues.
Mean (±SD) number of activities of daily living affected by olfactory loss was 4.70 ± 3.56 for the impaired group and 0.61 ± 1.58 for the improved group (P<.001). Among specific activities, the most common cited impairments were ability to detect spoiled food (impaired vs improved groups, 75% vs 12%; P<.001), gas leaks (61% vs 8%; P<.001), or smoke (50% vs 1%; P<.001); eating (53% vs 12%; P<.001); and cooking (49% vs 12%; P<.001). Differences in quality-of-life issues were reported primarily in the areas of safety and eating. Overall satisfaction with life was reported by 87% of the improved group but only 50% of the impaired group (P<.001).
Patients reporting persistent olfactory impairment after previously documented olfactory loss indicate a higher level of disability and lower quality of life than those with perceived resolution of olfactory compromise.
A SURVEY BY the National Institute on Deafness and Other Communication Disorders estimated that more than 2.7 million adults in the United States (1.4% of the population) have chronic olfactory impairment.1 Causes of these losses include head injury, aging, paranasal sinus disease, neoplasm, medications, toxic exposure, upper respiratory tract infection, surgical trauma, and congenital defects.2-6 Despite the relatively high prevalence, there is a lack of public awareness regarding this problem. Moreover, limited availability of olfactory testing centers may hinder patients seeking assistance. Unfortunately, even with medical attention, there is often little that can be done to restore lost olfactory function.
In recent years, progress has been made toward elucidating various mechanisms of olfactory losses and their natural course.7-10 Despite an improved understanding of olfactory disorders, little is known about the impact of olfactory impairment on patients' lives. Anecdotal reports suggest that a significant proportion of olfactory-impaired individuals have a diminished quality of life (QOL) and ability to perform normal activities of daily living (ADLs). A few studies have demonstrated this in some patient populations.4,10 The relative scarcity of reports documenting disability or QOL changes relegates olfactory loss to the status of mere inconvenience rather than that of chronic illness, as with other sensory impairments such as vision or hearing loss. In the present study, we used patient databases from 2 large smell and taste centers to examine the nature and extent of functional disabilities and alterations in QOL experienced by olfactory-impaired individuals. This information is particularly important for otolaryngologists, who are frequently involved in the examination and treatment of such patients.
Patient databases at the smell and taste clinics of the Virginia Commonwealth University's Medical College of Virginia (MCV) Campus, Richmond, and the Ear Nose and Throat Clinic of the Kanazawa University Hospital, Kanazawa, Japan, were accessed to identify patients who had undergone olfactory testing during a 15-year period (January 1, 1984, through December 31, 1998). A total of 1407 patients (693 from the MCV clinic and 714 from the Kanazawa University clinic) had been evaluated and were thus eligible to be surveyed. Appropriate approval was obtained from both parent institutions' Committee on the Conduct of Human Research before patient contact.
Patients had been examined for olfactory or gustatory dysfunction. All had undergone quantitative olfactory testing at the time of their clinic visit. At the MCV clinic, testing was performed by means of the butanol detection threshold and odor identification tests as described by the Connecticut Chemosensory Clinical Research Center.11 Testing was performed at the Kanazawa University clinic by means of the T & T Olfactometer (Daiichi Yakuhin Sangyo Co, Ltd, Tokyo, Japan) threshold determination technique.12 Patients' degree of olfactory impairment was rated according to their olfactory function test results as follows: (1) anosmic (MCV test score, 0-14; Kanazawa test score, ≥5.6); (2) severe (MCV, 15-44; Kanazawa, 4.1-5.5); (3) moderate (MCV, 45-64; Kanazawa, 2.6-4.0); (4) slight (MCV, 65-89; Kanazawa, 1.1-2.5); or (5) normal (MCV, 90-100; Kanazawa, ≤ 1.0). Stratification of patients into 5 levels of olfactory function along the spectrum from anosmia to normosmia is routinely used with both systems.11,12
A questionnaire was developed to determine the present status of subjects' health and olfactory function. The questionnaire included demographic and historical information, including general medical history and nature of olfactory impairment (cause, character, change with time, and severity of loss). The questionnaire asked patients to indicate whether they felt impaired by their olfactory loss in the performance of a number of common ADLs. These focused on activities we perceived as "olfaction-related," such as cooking, detecting spoiled foods or gas leaks, eating, and others, such as working, socializing, and playing sports or exercising. The survey also asked questions relating to patients' enjoyment of life. We term these QOL issues. The questionnaire was first prepared in English and then translated into Japanese by one of us (T.M.). Surveys were mailed to patients' last known addresses and were returned to the MCV and Kanazawa University clinics. Data were recorded and organized by means of a spreadsheet (Microsoft Excel 97; Microsoft Corp, Redmond, Wash), and statistical analysis was performed with SPSS (version 9.0; SPSS Inc, Chicago, Ill).
Among 1407 patients seen in the 2 smell and taste clinics from 1984 to 1998, 1093 had impaired olfaction on clinical testing and address information allowing questionnaires to be mailed. Thirty-two questionnaires were returned with notification that the patient was no longer alive and 627 were never returned. Of the 434 patients who returned questionnaires, 14 were excluded for incomplete survey responses. Thus, 420 (38.4%) of the mailed surveys (312 from the Kanazawa University clinic and 108 from the MCV clinic) were complete and available for study.
The demographic characteristics of all study subjects are shown in Table 1. All had some degree of olfactory impairment documented during their clinic visit. The patients were grouped as "improved" (no smell problem at time of survey) or "impaired" on the basis of their response to the survey question "Do you have problems with your sense of smell (yes or no)?" The 2 groups were demographically matched with the exception of race (Table 1). There was no difference between the average ages for the impaired (mean, 59.6 years; range, 9-87 years) and improved (mean, 57.1 years; range, 20-87 years) groups (P = .19, t test).
Subjective and objective differences in olfactory function between the 2 groups were evaluated in several ways. The survey asked patients to subjectively rate their ability to smell on a scale of 0 (cannot smell) to 10 (normal). This number was termed the patient's smell score. Mean smell scores were 8.96 and 2.38, with medians of 10 and 2, for the improved and impaired groups, respectively (P<.001). None of the 345 subjects in the impaired group rated their ability to smell as 10 (normal). The degree of olfactory impairment determined by quantitative testing at the time of clinic evaluation was compared between groups. In the improved group, 52.0% of patients had tested as having slight or moderate impairment, while in the impaired group only 24.6% had this level of impairment (difference significant by Pearson χ2, P<.001). Only 29.3% of the improved group, but 55.4% of the impaired group, tested as being anosmic. Of the anosmic patients, 72.7% (16/22) in the improved group, but only 7.9% (15/191) in the impaired group, reported that their sense of smell had improved between the time of testing and the time of survey. Overall, 58.7% (44/75 patients) in the improved group, but only 14.8% (51/345) in the impaired group, reported improvement between testing and survey.
The subjective smell scores were compared with the objective measurements of olfactory impairment determined by testing in the clinic. The correlation between patient smell scores and olfactory function test results was significant by Pearson χ2 testing (r = −0.47; P<.001). Of 104 patients reporting a smell score of 0 (anosmic), 85.6% tested as being anosmic at the time of the clinic visit.
The characteristics of the olfactory disturbance experienced by subjects in the impaired group are shown in Table 2. Among the 345 subjects in this group, 96.4% (268/278 patients who reported the duration of their smell problem) had had their olfactory disturbance for longer than 1 year. Although 54.2% reported no change in olfaction, 14.8% reported improvement and 9.9% reported worsening since the time of their smell and taste clinic evaluation. Fluctuation in sense of smell was noted by 15.7%. When asked about their type of smell problem, 11.9% reported alteration in the quality of known odors (dysosmia), while 4.1% noted phantom odors (phantosmia). Approximately half reported complete loss of smell (anosmia). The most common causes of olfactory disturbance among this group were nasal or sinus disease (21.4%), head or facial injury (17.1%), documented upper respiratory tract infection (17.1%), and unknown (28.4%).
Figure 1 shows the percentage of subjects in each group who indicated compromise in the ADLs queried by the survey. The most commonly cited activities impaired by olfactory loss for the impaired group were detection of spoiled foods (75%), detection of gas leaks (61%), eating (53%), detection of smoke (50%), cooking (49%), buying fresh food (36%), and using perfume or colognes (33%). Differences between the response frequencies observed in the impaired and improved groups were statistically significant by Pearson χ2 test (P<.05) for all of the listed ADLs, with the exception of sports or exercise (P = .14). The number of ADLs affected by olfactory loss in each patient was termed the ADL disability score. Patients in the impaired group had significantly greater ADL disability scores (mean ± SD, 4.70 ± 3.56) than those in the improved group (0.61 ± 1.58; P<.001 by t test). The distributions of ADL disability scores for both groups are shown in Figure 2.
Figure 3 lists the QOL issues queried by the survey most frequently reported to be affected by olfactory impairment. Safety issues such as worrying about gas leaks, fire, and cleaning solution vapors were frequently cited, with these response rates all differing significantly (t test, P<.05) from the improved group. Personal hygiene issues also generated high response frequencies, although concern about breath and body odor was the only one to achieve statistical significance when compared with the improved group. Eating issues were also frequently noted, the most common being altered taste for food, less enjoyment from eating, and preferring food spicier. More than 25% of those in the impaired group indicated that they enjoyed life less than they used to, although this was not significantly different than in the improved group (t test, P = .90).
We also asked patients to rate their global satisfaction with life as "very dissatisfied," "somewhat dissatisfied," "neither satisfied nor dissatisfied," "somewhat satisfied," or "very satisfied." Eighty-seven percent of those in the improved group, but only about 50% of those in the impaired group, reported being either very or somewhat satisfied with life (Figure 4). Conversely, only 3% in the improved group, but 34% in the impaired group, reported being either very or somewhat dissatisfied with life (Figure 4). These differences in level of satisfaction with life between the improved and impaired groups were significant by Pearson χ2 test (P<.001). An analysis of all patients surveyed showed that there was a clear trend toward greater overall satisfaction with life as the patient-reported smell score increased (Pearson χ2, P<.001). There were, however, no significant differences in the level of dissatisfaction when patients were grouped by age, sex, comorbid illnesses, work status, or level of education. The same finding was obtained when the impaired group was evaluated alone.
Both the medical and lay communities have historically neglected the chemical sense of olfaction. Given the scarcity of successful treatments, the perception exists that the diagnosis of olfactory disorders is of little utility. At present there are fewer than 10 smell and taste research centers in the United States capable of thoroughly diagnosing and characterizing olfactory deficits. This compares with thousands of centers nationwide dedicated to the diagnosis and treatment of disorders of hearing and vision. The lay public remains uninformed, since most medical practitioners have limited experience with olfactory disorders and seldom know where to refer such patients for evaluation of their deficit. As such, people with disorders of olfaction are given little support in dealing with their problem.
The chemical senses have important safety functions such as the early detection of fire, gas leaks, spoiled foods, or dangerous fumes, as well as hedonic functions such as the assessment of the palatability of foods and beverages and the detection of fragrances or aromas. While these functions may be less vital to a person's well-being and functionality than vision or hearing, their loss may still negatively impact a person's QOL and potentially their level of disability. Previous reports have documented olfactory losses from a wide variety of injuries and disease processes1,3-6,13; however, few have systematically evaluated the effects of such losses on patients' QOL and disability.
A retrospective study of 750 patients tested at the University of Pennsylvania concluded that "chemosensory dysfunction influences quality of life."4 In particular, patients with dysosmia or dysgeusia had higher scores on the Beck Depression Inventory, indicating more severe depression, than patients without these conditions. In addition, patients with chemosensory dysfunction reported reduced body weight, appetite, and psychological well-being. Heald and coworkers10 found that the degree of patient-reported chemosensory dysfunction correlated with lower scores on standardized measures of health and QOL in individuals infected with human immunodeficiency virus.10 Human immunodeficiency virus–infected individuals with chemosensory complaints had significantly lower QOL scores in all tested domains, including physical and social function, general health perception, mental health, and energy and fatigue, compared with human immunodeficiency virus–infected counterparts without chemosensory impairment. Other studies have noted relationships between olfaction and health-related issues that likely impact individuals' QOL, such as depression, appetite, and nutritional intake.14-16
The present study sought to determine the impact of olfactory disturbance on patients' perceived level of disability and QOL. Given the hedonic influence of olfaction on human existence, we believed a patient-completed survey was the most relevant instrument to evaluate patients' disability and QOL. Use of a patient-completed survey as the sole instrument in assessing QOL and disability potentially introduces bias. However, by comparing 2 demographically similar groups differing primarily by self-reported olfactory status at the time of survey, the impact of this effect should be minimized.
There are several means whereby a selection bias might have been introduced into this study. First, patients presenting to a smell and taste clinic might represent olfactory-impaired individuals who feel more compromised by their deficit. The data might then overstate the true impact of olfactory loss. However, the 2 study groups were both composed of subjects who had previously sought help. By comparing disability and QOL between these groups, this effect should be minimized. Second, 627 (57.4%) of the 1093 mailed surveys were never returned, suggesting a possible sampling bias. It is possible that those not returning surveys were either not affected by their olfactory loss, leading to overstatement of our results, or severely affected by their olfactory loss, leading to understatement of our results. As the 2 study groups were both composed of subjects who returned surveys, we believe this effect is also minimized. It is possible that a significant proportion of the 627 unreturned surveys were not returned because the subject had moved and the receiving party did not forward the survey. This is likely, as the average time between clinic visit and survey for our patient population was more than 8 years.
The 2 study groups differed by patient-reported level of olfactory function at the time of survey. Previous studies have reported poor correlation between patient-reported and objectively measured degrees of olfactory impairment.15 However, our results show a strong correlation between the patient-reported smell score and the results of olfactory testing performed at the time of evaluation at the smell and taste clinic. This is despite the long average (±SD) time interval (8.08 ± 5.01 years) between testing and survey completion. This correlation was likely achieved because (1) patients in the improved group had better olfactory function at the time of clinic visit (52.0% had a slight or moderate loss determined by testing) compared with patients in the impaired group (only 24.6% had slight or moderate loss) and (2) more patients (58.7%) in the improved group reported improved sense of smell since the time of testing than patients in the impaired group (14.8%). Thus, on average, subjects in the improved group started with better olfactory function at the time of testing and were also more likely to have had improvement between the time of testing and survey than subjects in the impaired group. Therefore, the impaired group represents a population of olfactory-impaired individuals, while the improved group represents a group of mildly or non–olfactory-impaired individuals who had previously experienced olfactory loss.
The study groups also differed significantly in terms of racial distribution. We cannot discount the possibility that this might introduce a cultural bias into the nature of subject responses, such that whites are more likely to complain of olfactory losses than Asians/Pacific Islanders or vice versa. Potential differences in impact of olfactory losses between the Japanese and American subjects will be the subject of a forthcoming study. However, we believe the disparity in racial distribution between the groups is most likely the result of the differing histories and referral patterns of the 2 smell and taste clinics involved in the study. At the MCV clinic, where no study patients were Asians/Pacific Islanders, very few patients are referred to rule out olfactory losses, because of the restrictive nature of the US health care system in terms of patient referrals. Also, this clinic was previously associated with the institution's head injury unit, and thus 27% of patients had head or facial injury as the cause of olfactory loss, while only 23% indicated nasal or sinus disease or respiratory infection. Traumatic olfactory losses generally tend to be more severe or permanent than those caused by nasal or sinus disease.4 Thus, 92% of the subjects from the MCV clinic were included in the impaired group, and only 8% were included in the improved group, as most study subjects from this site either had more severe olfactory losses at the time of testing or did not improve with time. At the Kanazawa University clinic, where all study patients were Asians/Pacific Islanders, far more patients are referred to rule out olfactory losses, as the Japanese health care system is less restrictive in terms of patient referrals. Also, as this clinic was formed within the institution's Department of Otolaryngology–Head and Neck Surgery, there were more patients at this site with olfactory losses from nasal or sinus disease or documented upper respiratory tract infection (45%) than from head injury (14%). Consequently, more patients had less severe or transient olfactory losses. As a result of both of these factors, 21% of the subjects from this site were included in the improved group and 79% in the impaired group.
In this study, subjects reported that their ability to perform or participate in a number of daily activities was hindered by their olfactory loss. Although our survey made no effort to have subjects quantify their degree of disability, we believe that the difference between the impaired and improved groups shown in Figure 1 and Figure 2 supports this statement. Most striking was the high percentage of subjects indicating difficulty detecting spoiled foods, gas leaks, or smoke—activities considered "safety functions" of olfaction. These deficits are significant given the widespread use of gas grills, water heaters, furnaces, and appliances, and the resultant risk of fire in the home. Thirty-seven percent of survey respondents reported at least 1 smoker in the house, thus further increasing the risk of injury or mortality to these olfactory-impaired individuals.17 Subjects also frequently reported difficulties with food-related activities (eating, cooking, buying fresh food, and going out to eat) and hygiene-related activities (using perfumes, colognes, or scented detergents or soaps, detecting soiled diapers, and housecleaning). These activities are not critical in terms of safety or survival but are important to most people in maintaining a desirable lifestyle. Finally, few subjects reported an effect of their olfactory loss on work or hobbies such as gardening or socializing. However, the effect of olfactory loss on subjects' ability to work might depend greatly on vocation. It might be inferred from the previous data that workers in vocations involving any of the above activities, such as food service workers, firefighters, home appliance or heating technicians, utility workers, or retailers of perfumes and soaps, might be compromised in their ability to perform their jobs.
The present study clearly indicates a negative impact of olfactory deficits on patient-perceived QOL. The areas that stand out most are safety and eating, as shown in Figure 3. The high rate of reported concern about gas leaks and fire mirrors patients' reported reduced ability to detect gas leaks and smoke. Patients reported altered taste for food and decreased enjoyment of eating. Although more than 50% of subjects in the impaired group reported increased concern about breath or body odor, concerns about other hygiene issues were not significantly different from those of subjects in the improved group. While these concerns might not seem significant, together they may be a factor contributing to the increased rate of depression seen in olfactory-impaired individuals.14
Despite the noted negative impact on multiple QOL issues among subjects with olfactory losses, the percentage of subjects in both the impaired and improved groups reporting decreased "enjoyment of life" overall was the same (approximately 25%). However, when subjects were asked to rate their overall "satisfaction with life," responses between the 2 groups differed significantly (Figure 4). Moreover, among all subjects, the greater the degree of olfactory loss, the lesser the degree of satisfaction with life. This trend could not be accounted for by subject age, sex, work status, level of education, or comorbid illnesses. The source of disparity between these 2 findings is unclear. However, the response frequency for these 2 questions—only 24% for the "enjoyment of life" question but 94% for the "satisfaction with life" question—strongly suggests a selection bias, making the later finding more reliable.
In conclusion, the present study used the patient databases from 2 large university smell and taste centers to quantify the impact of olfactory loss. Olfactory-impaired individuals reported considerable disability and reduced QOL. Evaluation of suspected olfactory deficits should thus be considered. Although treatment options for olfactory disturbances remain limited, thorough assessment is potentially useful for purposes of patient counseling, particularly with regard to safety issues, and for determination of disability and monitoring for interrelated conditions such as depression. The otolaryngologist, who is uniquely capable of evaluating nasal and thus olfactory anatomy and function, must be aware of the impact of olfactory dysfunction on patients' lives.
Accepted for publication September 22, 2000.
This study was supported in part by grant 98-03-034 from the Univers Foundation, Tokyo, Japan.
Presented at the annual meeting of the Association for Chemoreception Sciences, Sarasota, Fla, April 29, 2000.
Corresponding author and reprints: Evan R. Reiter, MD, Department of Otolaryngology–Head and Neck Surgery, PO Box 980146, Medical College of Virginia, Virginia Commonwealth University, Richmond, VA 23298-0146 (e-mail: firstname.lastname@example.org).