Loss of smell in patients with aspirin-exacerbated respiratory disease impacts mental health and quality of life

Abstract

Background:

The impact of anosmia on quality-of-life (QoL) for patients with aspirin-exacerbated respiratory disease (AERD) is poorly understood. We aimed to investigate how the severity of smell loss and olfactory dysfunction (OD) in patients with AERD affects their QoL, mental health, and physical well-being.

Methods:

Five validated QoL questionnaires (Sinonasal Outcome Test -22, Asthma Control Test, Healthy Days Core Module-4, Short Form-36, and Patient Health Questionnaire-4) and two newly developed questionnaires assessing severity and consequences of OD were electronically sent to all 2,913 patients in the Brigham and Women’s Hospital AERD registry. Responses were received from 853 participants for analysis.

Results:

85% of participants reported present diminished sense of smell and/or taste and 30% categorized their OD severity was, “as bad as it can be.” There were significant relationships between the severity of self-reported OD and both psychological distress and general health scores, even after adjusting for asthma control. Additionally, incidence rates for physically and mentally unhealthy days in the prior month were higher for patients with moderate or severe OD than for normosmic patients. Patients with diminished smell responded that they could not identify spoiled food (86%), did not enjoy food (71%), felt unsafe (63%), and had encountered dangerous situations (51%) as consequences of their OD.

Conclusions:

Anosmia and hyposmia severely impact the physical, emotional, and mental health of AERD patients, and leads to safety concerns in their daily lives The importance of olfaction and the relevance of OD to patients’ QoL should be acknowledged and evaluated by clinicians caring for these patients.

Introduction:

Aspirin-exacerbated respiratory disease (AERD) is a syndrome characterized by chronic rhinosinusitis with nasal polyps (CRSwNP), asthma, and respiratory reactions to aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs) that inhibit cyclooxygenase-1. AERD affects approximately 0.3–0.9% of the general population and about 7% of adults with asthma.^{1, 2} Patients with AERD often have a high symptom burden, including more severe and difficult-to-treat asthma and upper respiratory symptoms than patients with aspirin-tolerant asthma or CRSwNP. Anosmia affects >90% of patients with AERD and a significant proportion of patients with aspirin-tolerant CRSwNP as well.³ Anosmia and hyposmia are common impairments in the general population with reported incidences of 4–6 % and 13–18%, respectively.^{4, 5} The exact causes of olfactory dysfunction (OD) in AERD are not fully understood and are likely multifactorial. High nasal polyp burden and swelling of the sinonasal mucosa may cause a physical obstruction that prevents odorants from reaching the olfactory epithelium in the superior nasal cavity. Chronic inflammation within the olfactory cleft epithelium may also be a causative mechanism for anosmia in these patients, as elevated nasal mucous levels of IL-2, IL-5, and IL-13 have been associated with abnormal sense of smell, ^{6, 7} suggesting underlying immunologic causes of OD.

Both the severe asthma and chronic sinus congestion that are commonly experienced by patients with AERD can have negative impacts on quality of life (QoL).⁸ Nonetheless, anosmia can also negatively impact wellbeing, though the importance of olfaction is understudied and often overlooked in the medical field. Independent of the underlying cause(s) of OD, there are known associations between anosmia and multiple negative sequalae. In the era of COVID-19, these associations are even more relevant, as up to 50% of patients with COVID-19 experience an acute change in smell, and of those patients, 10% may continue to experience chronic OD.^9–11 Anosmia is associated with self-reported depression, decreased QoL, disruption in relationships, feelings of vulnerability, changes in weight, and cognitive decline, among other symptoms.^12–15 The specific consequences of chronic anosmia experienced by patients with AERD has not been thoroughly studied. To better treat patients with anosmia, regardless of its cause, it is important to fully understand the multi-faceted impact of anosmia on patients.

Here, we report on the relationship between severity of OD in patients with AERD and patients’ self-report of mental and physical well-being, based on results from previously standardized olfaction and QoL surveys, as well as parallel results from our newly developed surveys characterizing the consequences and severity of OD in these patients. A better understanding of the relationship between loss of smell and QoL is the first step toward being able to push forward with seeking new treatments.

Methods:

Survey Questionnaire Design

Seven questionnaires, along with questions regarding medical history, medication use, and demographic information were sent out electronically to patients enrolled in the Brigham and Women’s Hospital (BWH) AERD Registry. The study was approved by the Mass General Brigham Institutional Review Board and all participants provided informed consent. Five questionnaires were previously validated elsewhere, including:

1. The Sino-Nasal Outcome Test (SNOT-22), a disease-specific QoL survey for use in chronic rhinosinusitis.¹⁶

2. The Asthma Control Test (ACT), a five-question survey that stratifies asthma control into three levels (well-controlled, poorly-controlled, and very poorly-controlled).¹⁷

3. The Healthy Days Core Module-4 (HRQOL-4), a four-question survey used by the Centers of Disease Control and Prevention (CDC) to assess health-related quality of life by measuring self-rated general health, physical health, mental health, and activity limitations.¹⁸

4. The general health questions within the 36-item Short Form (SF-36) questionnaire, which assess self-reported general health and level of function (19).

5. The Patient Health Questionnaire-4 (PHQ-4), a screening scale for anxiety and depression with total scores ranging from 0–12 and the following categories of psychological distress: None (0–2), Mild (3–5), Moderate (6–8), and Severe (9–12).¹⁹

The remaining two questionnaires were designed specifically for this study. The Consequences of Anosmia questionnaire assesses the presence or absence of problems in daily life as potential consequences of anosmia. In developing the survey, we considered our own clinical experiences working with anosmic AERD patients, as well as the potential influence of the COVID-19 pandemic on patient experiences (for example, we avoided using questions that asked about frequency of visiting restaurants or retail industries). To create a generalizable survey with cross-cultural relevance, we aimed to use concise, objective wording with binary answer choices. The introductory question asks participants if they currently have a “diminished, decreased, or abnormal sense of smell.” If “yes” is selected, the participant is prompted to answer the next thirteen questions, whereas a response of “no,” terminates the questionnaire. All thirteen questions begin with “because of my change in smell…” and end with a potential consequence of anosmia. The potential consequences include aspects of mental, physiologic, social, and emotional health as well as safety, eating habits, and self-realization (see Supplemental Figure E1 for complete questionnaire). Additionally, we developed a Severity of Anosmia questionnaire that categorizes smell and taste symptoms objectively using language that is mutually exclusive between and within questions. Questions regarding symptoms are all asked in reference to the preceding month. The first question specifically asks how well the participant could smell within the past 1 month, with subsequent questions regarding changes in eating habits, presence of phantom odors, and changes in one’s smell over time (see Supplemental Figure E2 for complete questionnaire).

Patient Characterization and Data Collection

The full set of surveys were uploaded into REDCap® and a link was emailed to all 2,913 participants enrolled in the BWH AERD Registry in April 2021. The registry collects longitudinal information on AERD patients and includes patients followed at BWH and at other non-affiliated institutions. Participants were sent an email reminder one week later. Within the two-week data collection period, 853 patients had completed the survey.

Statistical Analysis

Descriptive statistics were used to present participants’ characteristics. Chi-square and Fishers exact tests were used to assess a relationship between Consequences of Anosmia and Severity of Anosmia questionnaire questions and SNOT-22 question 21 (sense of smell/taste). HRQOL-4 data were summarized descriptively. The summary index of unhealthy days was calculated based on the CDC methodology.²⁰ As HRQOL-4 data were over dispersed and included excess zeros, negative binomial regression was used to analyze a relationship between SNOT-22 question 21 and HRQOL-4. Incident rate ratio and its 95% confidence intervals were reported. Linear regression was used to analyze a relationship between SNOT-22 question 21 and SF-36 modified general health score. PHQ-4 scores were also overdispersed and included excess zeros and were categorized using published cutoffs.¹⁹ Moderate and severe groups of PHQ-4 scores were combined, as there were only 4.7% in the severe group. Ordinal logistic regression was used to analyze a relationship between SNOT-22 question 21 and psychological distress, created using PHQ-4 scores as an ordered outcome. Odds ratio and 95% confidence intervals were reported. All statistical analyses were performed using SAS version 9.4 (Cary, NC).

Results:

Study Population and Demographics

Of all emailed participants, 29.2% responded. The demographics of the respondents largely reflect our institution’s overall AERD demographics (Table I). Most participants completed all seven questionnaires, with completion rates from 96.8–99.4%.

TABLE I.

Patient characteristics

Variables		N = 853 (%)
Age (years)		50.3 ± 13.1
Gender	Female	626 (73.4)
	Male	225 (26.4)
	Nonbinary	2 (0.2)
Race	African american / Black	14 (1.6)
	Asian	9 (1.1)
	Caucasian	783 (91.8)
	Multiple races	12 (1.4)
	Other or unknown	35 (4.1)
Ethnicity	Hispanic	32 (3.8)
Have ever undergone aspirin desensitization	Yes	451 (53.0)
	No	400 (47.0)
Current medications	Aspirin	282 (33.1)
	Inhaled corticosteroids	700 (82.1)
	Nasal corticosteroids	648 (76.0)
	Oral corticosteroids	64 (7.5)
	Antihistamines	456 (53.5)
	Antileukotrienes	482 (56.5)
	Respiratory biologics	309 (36.2)

Data is presented as n (%) for categorical parameters and as mean ± standard deviation for continuous variables.

Self-Reported Olfactory Dysfunction

SNOT-22 Question 21

To determine the portion of respondents who reported olfactory dysfunction at the time of the survey, the answers to two separate questions were analyzed, including question 21 of the SNOT-22, which asks patients to rate their “Sense of taste/smell”, on a scale of 0–5 from “No problem” to “Problem as bad as it can be.” Of the 848 participants who completed the SNOT-22, 724 (85.4%) considered their sense of taste/smell to be at least a very mild problem and 255 (30.1%) reported that the problem was as bad as it could be (Figure 1).

Figure 1.

Answers to Question 1, “Sense of taste/smell”, within the Sinonasal Outcome Test -22 questionnaire (SNOT-22). N=848.

Consequences and Severity of Anosmia Questionnaires

The second question used to evaluate olfactory dysfunction was the initial Yes/No question of the Consequences of Anosmia questionnaire, which 826 participants completed. Six hundred and thirty-two (76.5%) participants reported “Yes,” to indicate current “diminished, decreased, or abnormal sense of smell.” Of the 632 participants, 489 (77.4%) reported in question 21 of the SNOT-22 that their sense of smell/taste was at least a moderate problem. There was a significant association between responses to the initial question of the Consequences of Anosmia questionnaire and question 21 of the SNOT-22 (p<0.0001, Chi-square).

Participants who answered “Yes” to the initial question of the Consequences of Anosmia questionnaire were subsequently asked 13 questions with “Yes/No” responses regarding potential consequences of their abnormal sense of smell (Supplemental Figure E1). Among these 632 participants, the most common reported consequence was the inability to identify spoiled food (n=548, 86.7%), followed by not enjoying food or drink as much as they used to (n=452, 71.5%), and 225 (35.6%) responded that because of their change in smell, their weight had changed in an unhealthy way. Three hundred and ninety-six (62.7%) participants reported that they had felt unsafe at times because of their change in smell and 325 (51.4%) reported having caused a dangerous situation that would have been preventable if they had been able to smell normally (e.g. burning food or leaving a gas stove on). Additionally, 52.5% of participants reported that their change and smell had led to concerns about their own personal hygiene, and 32.8% reported feeling isolated as a consequence. The standardized Cronbach’s alpha, which is a measure used to assess the reliability of a set of test items, or how closely related in concept a set of test items are as a group, was 0.83 among these 13 questions, indicating strong internal consistency. Data from these 13 questions are summarized in Figure 2.

Figure 2.

Percentage of participants who reported experiencing specific consequences as a result of their anosmia. N=632.

Within the Severity of Anosmia questionnaire participants were asked, “has your sense of smell affected your eating habits in the past month?” A majority of the 831 participants answered, “my sense of smell has not affected my eating habits” or “I eat the same as before” (n=378, 62.0% combined). Remaining responses were evenly distributed between those who reported their changed sense of smell caused them to eat significantly less (n=27, 4.3%), a little less (n=65, 10.7%), a little more (n=78, 12.8%) and significantly more (n=62, 10.2%). We also asked whether participants thought they “smelled an odor despite nothing being there (a ‘phantom odor’)” within the last month. Four hundred and ten (49.3%) participants reported experiencing a phantom odor “some days” and 25 (3.0%) participants experienced a phantom odor daily.

Relationship Between Olfactory Dysfunction and QoL, Health Status, and Mental Health

To determine the relationship between self-reported decreased sense of smell and patient-reported measures of QoL, general health status, and mental health, we compared answers to question 21 of the SNOT-22 to QoL-related survey responses. Though there was no significant collinearity between asthma control as determined by the ACT scores and anosmia as determined by responses to question 21 of the SNOT-22, participants’ perceived severity of asthma control could affect their perceived overall QoL and health status. Therefore, we adjusted our results to account for varying levels of asthma control

Healthy Days Core Module

The HRQOL-4 questionnaire asked participants to estimate the number of days out of the previous 30 during which they felt physically unhealthy, mentally unhealthy, in poor health, and/or unable to perform their normal daily routine secondary to poor health (Figure 3).

Figure 3.

Responses to Health-Related Quality of Life (median with interquartile error bars) in relation to olfactory dysfunction, assessed by responses to question 21 (severity of problem with “Sense of taste/smell”) in Sinonasal Outcome Test -22 (SNOT-22). There was a significant relationship between SNOT-22 question 21 response and physically unhealthy days (p<0.0002), mentally unhealthy days (p=0.003), poor health days (p<0.0001), and derived unhealthy days scores in the past 30 days (p<0.0001).

When adjusted for asthma control, the incident rate of physically unhealthy days in the past 30 days for patients who had the most severe OD (defined here as a response to question 21 of SNOT-22 of either “Problem as bad as it can be” or “Severe problem”) was 1.79 times (95% CI: 1.28–2.52, p=0.0008) and 1.48 times (95% CI: 1.03–2.13, p=0.035) the incident rate of the normosmic group who indicated “No problem” with smell on question 21. Moreover, even participants who reported only a “Mild or slight problem” with their smell/taste on question 21 of the SNOT-22 had an adjusted incident rate of physically unhealthy days that was 1.65 times (95% CI: 1.07–2.56, p=0.024) that for the normosmic participants. Similarly, for participants who answered “Problem as bad as it can be,” the adjusted incident rate of mentally unhealthy days was 1.60 times (95% CI: 1.15–2.25, p=0.006) that for the normosmic group and the adjusted incident rate of poor health days was 2.14 times (95% CI: 1.46–3.14, p=0.0001) that for the normosmic group.

“Derived Unhealthy Days” are the number of days in the past 30 days in which a person’s physical or mental health prevented them from doing their usual activities (i.e. self-care, work, or recreation.) The adjusted incident rates of derived unhealthy days for participants with severe OD who answered “Problem as bad as it can be” or “Severe problem” on the SNOT-22 question 21 was 1.74 times (95% CI: 1.33–2.28, p<0.0001) and 1.52 times (95% CI: 1.14–2.03, p=0.005), respectively, the incident rate of derived unhealthy days for the normosmic group. Even the group of participants who reported only a “Mild or slight problem” with their smell/taste on SNOT-22 question 21 had an adjusted incident rate of derived unhealthy days that was 1.52 times (95% CI: 1.08–2.16, p=0.018) that for the normosmic participants.

Patient Health Questionnaire 4

When adjusted for asthma control, there was a significant relationship between severity of OD based on SNOT-22 question 21 response and psychological distress, measured by the PHQ-4, an anxiety/depression screening test (p<0.0036). All patient groups who reported any problem with their sense of smell/taste had significantly higher odds of having moderate and severe psychological distress compared to the reference normosmic group with “No problem” smelling (Table 2). Further, the proportion of participants who reported psychological distress increased with worsening OD (Figure 4), such that 27.7% of the anosmic participants in the “Problem as bad as it can be” group had PHQ-4 scores in the categories of moderate or severe psychological distress.

TABLE II. Relationship between psychological distress assessed by Patient Health Questionnaire-4 (PHQ-4) and olfactory dysfunction as assessed by question 21 on the Sinonasal Outcome Test -22 (SNOT-22) survey.

Participants were grouped by their response to question 21 on the SNOT-22 questionnaire. When adjusted for asthma control, all participant groups who reported any problem with their sense of smell/taste had higher odds of having moderate and severe psychological distress when compared to the normosmic reference group with “No problem” smelling. Moderate and severe psychological distress groups were combined as there were only 39 out of 832 subjects (4.7%) in the severe group.

Question 21 ”Sense of Taste/Smell”	Adjusted OR of moderate and severe psychological distress	Adjusted p-value
Problem as bad as it can be vs. No problem	2.89 (1.72–4.86)	<0.0001
Severe problem vs. No problem	2.60 (1.51–4.49)	0.0006
Moderate Problem vs. No problem	2.24 (1.24–4.05)	0.0079
Mild or slight problem vs. No problem	1.93 (1.01–3.69)	0.0463
Very mild problem vs. No problem	2.01 (1.11–3.67)	0.0222

Figure 4.

Percentage of participants with psychological distress, assessed by Patient Health Questionnaire-4 (PHQ-4) responses in relation to olfactory dysfunction, assessed by responses to question 21 (severity of problem with “Sense of taste/smell”) in Sinonasal Outcome Test -22 (SNOT-22). PHQ-4 total score ranges from 0 to 12, with categories of psychological distress being: None 0–2, Mild 3–5, Moderate 6–8, and Severe 9–12.

36-Item Short Form Survey – General Health

There was a significant relationship between the severity of OD based on SNOT-22 question 21 response and the SF-36 modified general health score (p=0.0001) when adjusted for asthma control. The SF-36 uses a 100-point scale with 0 representing the least favorable health state and 100 representing the most favorable health state. The SF-36 score was 11.9 points lower in the anosmic “Problem as bad as it can be” group compared to the group who reported “No problem” with smell (p<0.0001). All patient groups who reported any problem with their sense of smell/taste had significantly lower SF-36 scores when compared to those of the reference normosmic group (data not shown).

COVID-19 Infection

We also queried participants on their history of COVID-19 infection. Sixty-seven of the 827 participants (8.1%) who answered this part of the survey indicated that they had been diagnosed with COVID-19 at some point, though 65 of them had been free from the acute symptoms of COVID-19 for at least 7 days at the time of the survey.

Discussion:

Prior to the COVID-19 pandemic, with the infection leading to OD in a previously normosmic population,²¹ surprisingly few studies on the effect of OD on QoL had been published. This is the largest investigation into the impact of anosmia on QoL in patients with AERD, and the first to investigate the severity of anosmia in relation to this breadth of QoL measures. The results demonstrate that QoL is severely negatively impacted in patients with AERD loss of smell. AERD patients with anosmia experience sequelae that are directly related to the severity of their perceived OD, demonstrating the significance of olfaction in daily life. Croy et al. proposed two pathways on how OD can affect mental health.²² First, restrictions on daily life, such as decreased enjoyment of food and concerns about personal safety, can impair QoL and lead to depression. Second, olfaction is strongly linked with emotions and memory, which are both impaired in depression. Reduced input from the olfactory bulb to the limbic circuit may affect brain function,^{23, 24} and in fact olfactory training in depressed patients can decrease symptoms of depression.²⁵ Around 30% of patients with OD have been reported to show signs of depression.²⁶ In this cohort of AERD patients, who generally have had longstanding hyposmia or anosmia, we found a significant relationship between subjective OD and psychological distress (PHQ-4), as well as between subjective OD and mental health (HRQOL-4). The number of reported mentally unhealthy days increased significantly with increasing OD, independent of asthma severity, indicating that worsening anosmia in these patients may be linked with worsening mental health. However, only 4% reported that they had a new mental health diagnosis due to their anosmia, suggesting that this is an under-evaluated consequence of smell loss.

In patients with new onset of OD due to COVID-19, Elkholi et al. showed that patients felt significant reduction in health-related QoL,²¹ with 76% of patients reporting an OD-related decline in general QoL. However, some patients develop coping strategies, mostly problem-focused or emotion-focused, and according to a study by Liu et al., patients with COVID-19-induced hyposmia adjusted to their OD over time.²⁷ If this is also true for anosmic AERD patients, and over time they developed coping strategies, their current anosmia-related poor QoL could at some point earlier in their diagnosis have been even worse.

Anosmia is closely associated with diminished ability to distinguish between flavors and can lead to subsequent challenges enjoying food.¹³ In our study, associations between anosmia and food consumption were prominent, as 71.5% did not enjoy food and drink as much as before, one third reported that their weight had changed in an unhealthy way, and 86.7% felt they were not able to identify spoiled food, creating a potential safety concern. Importantly, a notable proportion of our patients felt less safe as a consequence of their anosmia and more than half of the participants had found themselves in an unsafe situation, such as not detecting burnt food or leaving the gas on, which is in line with previous reports on heterogenous groups of patients with OD.^{8, 28–30} A feeling of overall safety is fundamental for humans and a lack thereof could lead to increased emotional distress. Worries about inability to detect one’s own body odor/hygiene was reported by 52.5% of our patients, and some studies have even suggested that this is the most negative perceived consequence patients experience due to anosmia.^{8, 21, 30} It has been speculated that insecurities related to personal hygiene might be reflected negatively in social relations,²⁶ and one fourth of our patients believed their personal relationships were impacted by their OD and a third reported feeling isolated in social situations due to OD.

Our study has some limitations. First, collected data were self-reported and thus patients with higher impairment might have been more likely or motivated to participate in the survey study. Further, the severity of smell impairment was self-reported and could not be compared to objective measurements such as formal olfactory testing, and our patients’ self-report of prior COVID-19 diagnosis also was not objectively verified. However, the breadth and depth of survey types used in this study is a strength, as is the multitude of questionnaires showing similarly correlative results.

OD is a major problem affecting QoL in AERD, including the physical, emotional, and mental health of patients. This study provides insight regarding the frequency of a variety of specific anosmia-related consequences experienced by our patients. We identified the striking effect of loss of smell on patients’ perception of their general, physical, and mental health, as well as a significant relationship between the severity of self-reported OD and levels of psychological distress. These data further the field’s understanding of the struggles experienced by patients with AERD and OD, and should encourage doctors to ask patients about their concerns related to smell, emphasizing the potential impact of smell loss on perceived safety and mental health.

Key messages:

• Most AERD patients report diminished sense of smell/taste and 30% categorize it as a severe problem.

• Smell loss in AERD patients negatively affects quality of life and their mental and physical health.

• Clinicians should ask about safety issues and the mental well-being of patients with olfactory dysfunction.

Abbreviations:

ACT

Asthma Control Test™

ACQ-6

Asthma Control Questionnaire-6

AERD

Aspirin-exacerbated respiratory disease

CRSwNP

Chronic rhinosinusitis with nasal polyps

SNOT-22

Sinonasal outcome test-22

HRQOL-4

Healthy Days Core Module (4 questions)

NSAIDs

Nonsteroidal anti-inflammatory drugs

OD

Olfactory dysfunction

PHQ-4

Patient Health Questionnaire-4

SF-36

36-item Short-Form Survey

QoL

Quality of Life