Abstract
Background: Osmophobia is an under-investigated associated symptom in migraine.
Objective: To evaluate the clinical correlates and diagnostic utility of osmophobia.
Methods: Adult patients with migraine (with or without aura), probable migraine (PM), tension-type headache (TTH), and cluster headache (CH) were recruited retrospectively from our headache registration system. Migraine patients with and without osmophobia were compared. The newly proposed criteria for migraine requiring at least two of photo-, phono-, or osmophobia instead of both photo- and phonophobia were validated.
Results: In total, 2883 patients were included: 1809 migraine, 792 PM, 138 TTH, and 144 CH. Osmophobia was more common in migraine (62.2%) than in PM (33.8%), TTH (14.5%), and CH (31.3%) (all p < 0.001). Migraine patients with osmophobia scored higher on the Hospital Anxiety and Depression Scale (HADS) than those without (15.6 ± 7.6 vs. 13.3 ± 7.8, p < 0.001). The concordance in headache diagnoses between ICHD-2 (International Classification of Headache Disorders, Second Edition) and the newly proposed criteria was “almost perfect” (κ = 0.886, p < 0.001). Additionally, 16.9% of ICHD-2 PM patients were reclassified as migraine using the newly proposed criteria.
Conclusions: Osmophobia was prevalent in migraineurs, and was associated with higher HADS scores. The newly proposed criteria appear comparable to the ICHD-2 criteria, and may increase the diagnostic yield for Asian migraineurs, among whom photophobia was less common.
Keywords
Introduction
Osmophobia refers to an increased sensitivity to certain odors during migraine attacks, which could cause avoidance of those odors or aggravation of migrainous symptoms. It occurs in 24.7–47.7% of adult migraine patients according to previous epidemiological studies, and has been reported to be quite specific for the diagnosis of migraine, especially in the absence of nausea or vomiting (1–6). Despite such confirmed associations, osmophobia remains under-recognized as a migraine-associated symptom, and its clinical implications are under-investigated. There are only a limited number of studies available to date (1–8), and only one small-scale study has been conducted in Asia (8).
Previous studies have demonstrated an increased risk for anxiety and depressive disorders among migraineurs, and such psychiatric co-morbidities can have a great impact on headache-related quality of life (9–14). The underlying mechanisms for such co-occurrence remain undetermined. Olfactory transduction in the brain involves the amygdala, hippocampus, insula, anterior cingulate cortex, and orbitofrontal cortex, which are also involved in emotion processing. These structures are reported to have functional alterations in affective disorders (15). Osmophobia, in fact, consists of an emotional component, namely aversion. Despite the apparent overlaps in neuroanatomical structures involved, the correlation between osmophobia and the affective aspects of migraine has never been explored.
Osmophobia is included in the appendix criteria but not the formal criteria for migraine without aura in the International Classification of Headache Disorders, Second Edition (ICHD-2) (16). The item of associated symptoms (criterion D) in the ICHD-2 formal criteria requires the presence of either (a) nausea or vomiting or (b) photophobia and phonophobia. In comparison, any two associated symptoms (nausea, vomiting, photophobia, phonophobia, and osmophobia) will suffice in the appendix criteria, which have been validated as a reasonable alternative in a Western sample (7). On the other hand, the prevalence of migraine in Asian studies generally lies in the lower range of those in the Western populations, and photophobia is also reported to be less common among Asian migraineurs (17–21). Therefore, we have proposed another set of criteria taking regional differences into account when diagnosing migraine. The validity of the aforementioned criteria could have a substantial impact on how clinicians diagnose, and subsequently treat, their headache patients.
The aims of the present study were threefold: (1) to compare the percentages of osmophobia among different headache diagnoses, (2) to assess differences in clinical manifestations between migraine patients with and without osmophobia, and (3) to verify the clinical utility of the newly proposed criteria for migraine without aura.
Methods
Patients
This was a retrospective hospital-based study involving consecutive new headache patients between the ages of 20 and 80 years who made their first visit to the Headache Clinic of Taipei Veterans General Hospital (TVGH) between November 2005 and October 2009.The headache diagnoses were made according to the diagnostic criteria of the ICHD-2 (16) by headache specialists. The study intended to include patients with migraine (code 1.1, 1.2.1, 1.5.1, and 1.5.2), probable migraine (PM) (code 1.6), tension-type headache (TTH) (code 2.1, 2.2, and 2.3), and cluster headache (CH) (code 3.1). Concomitant medication overuse was allowed, whereas other secondary headache disorders and acute headaches, which began less than one month before the evaluation, were excluded. Patients were diagnosed as migraine if they fully fulfilled the criteria for migraine with or without aura, and a diagnosis of PM was given if they met all but one of the criteria. Patients were classified as having migraine/PM if they had both migraine/PM and TTH. Questionnaire-based interviews were carried out after detailed history taking and neurologic examination. Our headache registration system was established in 1996, and the questionnaire for structured interviews had been validated for clinical and research purposes. The intake form questionnaire for each patient at the first visit includes demographic and clinical characteristics. Osmophobia during headache attacks was specifically queried. Each patient also completed the Migraine Disability Assessment (MIDAS) (22–24) and Hospital Anxiety and Depression Scale (HADS) (25) as part of the routine evaluations. The study protocol was approved by the Institutional Review Board of TVGH.
Questions pertaining to osmophobia
Osmophobia was defined as the presence of an increased sensitivity to certain odors during headache attacks, and the frequency of such a symptom was questioned. Patients were classified as having frequent osmophobia if their headache attacks were almost always or usually accompanied by osmophobia, and infrequent osmophobia was defined as an occurrence of “rare” or “sometimes.” In patients with osmophobia, enquiries were also made as to whether the relevant odor(s) caused aggravation of their symptoms during headache attacks.
Migraine Disability Assessment
The MIDAS questionnaire contains five items, and assesses headache-related disability over a three-month period. It records the amount of time lost from employment or school, household work, or family and social activities due to headache. The MIDAS score ranges from 0 to 270 (22–24).
Hospital Anxiety and Depression Scale
The HADS is a self-administered instrument designed to evaluate anxiety and depression in the setting of a hospital medical outpatient clinic. As physical symptoms are not included in the evaluation, the confounding influences from somatic complaints can be eliminated. This instrument contains 14 items; half of each pertains to anxiety and depression. The HADS scores lie between 0 and 42 (25).
Validation of the newly proposed criteria for migraine
The only difference between the newly proposed criteria and the ICHD-2 formal criteria is in Item 2 of criterion D, the associated symptoms. The ICHD-2 formal criteria require the presence of both photophobia and phonophobia; whereas in our newly proposed criteria, any two of photophobia, phonophobia, and osmophobia will suffice. Comparisons on headache diagnoses by different sets of criteria were also made. The validity of the newly proposed criteria and the ICHD-2 appendix criteria was verified in the entire study population after excluding patients with CH.
Statistical analysis
Descriptive data were expressed as mean ± standard deviation or percentages. Categorical data were analyzed using the χ2 test. When comparing migraine patients with and without osmophobia, statistically significant variables identified by the χ2 test or the Student’s t-test were selected for subsequent multivariable logistic regression analysis with the forward (Wald) method to determine their odds ratios (OR) and 95% confidence intervals (CI) for the presence of osmophobia. The Student’s t-test was used to compare age, MIDAS scores, average headache severity (0–10 scale), average days of use of headache abortive treatment per month, and the HADS scores between two groups of patients. The differences in the HADS scores among patients with frequent, infrequent, and no osmophobia were determined using one-way analysis of variance (ANOVA), followed by post hoc analysis using least significant difference (LSD) test. Kappa (κ) statistics was used to determine the concordance in headache diagnoses based on different sets of criteria. A κ value of 0.41 to 0.60 indicated “moderate” concordance, 0.61 to 0.80 indicated “substantial” concordance, and 0.81 and above indicated “almost perfect” concordance (26). The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of osmophobia for establishing a diagnosis of migraine were calculated after excluding patients with CH. Positive predictive value is defined as the percentage of migraine patients among patients with osmophobia (not including CH). Negative predictive value is defined as the percentage of patients with PM or TTH among patients without osmophobia (not including CH). The ICHD-2 formal criteria were used as the standard for the diagnosis of migraine. All statistical analyses were carried out using PASW Statistics 18 for Windows (SPSS Inc., Chicago, IL, USA). All calculated p values were two-tailed, and statistical significance was defined as p < 0.01 for balancing the type I and type II errors.
Results
Study participants
Demographics and clinical features of study participants.
ICHD-2: International Classification of Headache Disorders, Second Edition.
Statistical significance vs. migraine (χ2 test): ap ≦ 0.001.
Percentages of osmophobia in different headache groups
Among the 2883 study participants, 1458 patients (50.6%) reported osmophobia, and osmophobia was associated with a younger mean age (40.8 ± 13.1 vs. 45.7 ± 14.4 years, p < 0.001) and a higher proportion of female patients (79.1% vs. 67.4%, p < 0.001). Patients with migraine were more likely to have osmophobia (62.2%) than were those with PM (33.8%), TTH (14.5%), and CH (31.3%) (Table 1) (all p < 0.001), and the percentage of osmophobia was higher in PM than in TTH (p < 0.001).
Osmophobia and migrainous symptoms in study participants
Eight migrainous symptoms included in the ICHD-2 formal criteria were selected for further analyses: unilateral location, pulsating quality, moderate to severe intensity, aggravation by physical activities, nausea, vomiting, photophobia, and phonophobia. The percentage of osmophobia rose with increasing numbers of migrainous symptoms in the study participants (Figure 1). In those who had more migrainous symptoms, the percentages of osmophobia were higher.
The relation between percentage of osmophobia and number of the other eight migrainous symptoms (unilateral location, pulsating quality, moderate to severe intensity, aggravation by physical activities, nausea, vomiting, photophobia, and phonophobia) in the entire study population (consisting of migraine, probable migraine, tension-type headache, and cluster headache patients) (n = 2883) (p < 0.001, linear-by-linear association) The percentages of osmophobia increased in patients with headache as they had more migrainous symptoms.
Clinical correlates of osmophobia in patients with migraine
Of the 1809 migraine patients, 1125 (62.2%) had osmophobia, including 423 (37.6%) with frequent and 702 (62.4%) with infrequent osmophobia during their migraine attacks. Among patients with osmophobia, 789 (70.1%) reported association with aggravation of headache. Migraine patients with osmophobia were younger (age 39.9 ± 12.4 vs. 43.1 ± 13.0 years, p < 0.001), and had a higher proportion of female patients (83.6% vs. 78.1%, p = 0.004). By using χ2 tests, osmophobia was associated with higher proportions of headaches with a pulsating quality (p < 0.001), aggravation by physical activities (p < 0.001), vomiting (p = 0.005), photophobia (p < 0.001), phonophobia (p < 0.001), and photophonophobia (p < 0.001) in migraine patients, but not with aura, chronic migraine, or medication overuse. Subsequent analysis with multivariable logistic regression revealed that female gender (OR 1.60, 95% CI 1.24–2.07, p < 0.001), headaches with a pulsating quality (OR 1.45, 95% CI 1.17–1.79, p = 0.001), aggravation by physical activities (OR 2.13, 95% CI 1.33–3.42, p = 0.002), photophobia (1.46, 95% CI 1.18–1.81, p < 0.001), phonophobia (OR 1.40, 95% CI 1.14–1.73, p = 0.001), and higher HADS scores (OR 1.04, 95% CI 1.03–1.06, p < 0.001) were more likely to be associated with osmophobia, whereas the OR for age was 0.98 (95% CI 0.97–0.99, p < 0.001).
Osmophobia and psychological disturbances in migraine
Demographics and clinical features between migraine patients with and without osmophobia.
HADS: Hospital Anxiety and Depression Scale; MIDAS: Migraine Disability Assessment.
Diagnostic utility of osmophobia alone in migraine
Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of different symptoms for the diagnosis of migraine.
Comparisons among different sets of criteria among study participants after excluding patients with cluster headache
In addition to the 1809 migraine patients diagnosed based on the ICHD-2 formal criteria, another 134 patients (16.9% of PM patients according to the ICHD-2 formal criteria) could be diagnosed as having migraine using our new criteria. However, fewer patients received the diagnosis of migraine according to the appendix criteria (n = 1770) (Table 4). None of the patients diagnosed as having TTH using the ICHD-2 formal criteria were reclassified as migraine by the appendix or the newly proposed criteria, whereas nine of them were re-allocated to the category of PM by each of the other two sets of criteria. The concordance in headache diagnoses was “almost perfect” between the newly proposed criteria and the ICHD-2 formal criteria (κ = 0.886), and “substantial” between the ICHD-2 appendix and formal criteria (κ = 0.749) (both p < 0.001). Only patients diagnosed as having migraine by either the ICHD-2 formal or appendix criteria would be categorized as migraine by the new criteria (Figure 2); whereas patients classified as PM by both the ICHD-2 formal and appendix criteria remained in the category of PM based on the new criteria. As compared with the 1809 migraine patients diagnosed using the ICHD-2 formal criteria, the additional 134 migraine patients diagnosed using the new criteria were older (age 45.1 ± 15.2 vs. 41.1 ± 12.7 years, p = 0.004) and had higher HADS scores (17.1 ± 7.9 vs. 14.7 ± 7.8, p < 0.001). Nevertheless, there was no difference in terms of gender distribution and headache features and severity (data not shown).
The number of patients diagnosed as migraine by different sets of criteria. Distribution of headache diagnoses based on different criteria for migraine. ICHD-2: International Classification of Headache Disorders, Second Edition.
Discussion
Our study showed that osmophobia was quite prevalent in patients with migraine, but was less common in those with PM and other headache disorders. The association of osmophobia with other migrainous symptoms supports its role as part of the craniofacial sensory hypersensitivity in migraine, as headache pulsatility, aggravation by physical activities, photophobia and phonophobia could reflect an increased sensitivity to vascular pulsation, movement, light, and sound. Also, the percentage of osmophobia was positively correlated with the number of migrainous symptoms, which might indicate that osmophobic patients were more “migrainous” than their non-osmophobic counterparts. Although the percentage of migraineurs with osmophobia seemed to be higher than figures reported in prior epidemiological studies (62.2% vs. 24.7–47.7%) (1–6), such disparity could be accounted for by the discrepancies in classification strategies and ethnic differences. In fact, even though a South American study reported a prevalence of 47.7%, it should be noted that only 37.3% of their patients never experienced osmophobia (4). In addition, the result was consistent with that of another small-scale study conducted in Japan (63%) (8), in which the ethnicity is closer to that of the present study. Further studies are needed to clarify whether ethnic or cultural differences play a role. In our study, osmophobia was also quite common in patients with CH (31.3%). Such an association has also been reported in the literature (5,27), although the underlying mechanisms remain to be elucidated. It seems plausible that CH might share part of the pathophysiology with migraine. However, more studies are needed to determine the exact prevalence of osmophobia in CH.
Migraine patients with osmophobia were found to have higher levels of depression and anxiety, as suggested by higher HADS scores. This was not the case for the other eight migrainous symptoms. The exact mechanisms of this are unknown, but it could be hypothesized that psychological disturbances might modulate the negative or aversive emotional content of olfaction, namely osmophobia. Prior studies showed that emotional stimulation could alter olfactory processing (28), and depression was associated with olfactory negative alliesthesia, that is unpleasant odorants were perceived as more unpleasant in patients with depression (29). Both osmophobia and psychological disturbances might possibly result from dysfunction of common brain structures, such as the amygdala, hippocampus, insula, etc., which have been reported to be involved in both olfactory and emotion processing (15,30–32). According to a recent study, ictal migraineurs exhibited increased activities in limbic structures following olfactory stimulation (33). It is yet to be determined whether such functional alterations would be more prominent in osmophobic migraineurs. However, the differences in HADS scores between migraine patients with and without osmophobia were, in fact, not huge, and further studies are needed to confirm such a correlation.
Our study showed that diagnostic utility of osmophobia alone was far from promising for migraine. Even in the absence of nausea or vomiting, osmophobia might better serve as an indicator to exclude the diagnosis of migraine for a NPV of 91.9%. This was in sharp contrast to a study reporting a high specificity of osmophobia (2). The discrepancies could be ascribed to differences in study populations (migraine, PM, and TTH only vs. virtually all kinds of headache disorders in the previous study), as well as the strategy used for labeling osmophobia. Of note, our newly proposed criteria appear to be a reasonable alternative to the ICHD-2 formal criteria for migraine. The newly proposed criteria differ from the ICHD-2 formal criteria only in the associated symptoms, namely Item 2 of criterion D. Both photophobia and phonophobia are required in the formal criteria. In the newly proposed criteria, some flexibility is allowed, and any two of photophobia, phonophobia, and osmophobia will suffice. Diagnoses based on the ICHD-2 formal criteria, appendix criteria, and our newly proposed criteria generally overlapped, and concordance with the formal criteria was even better for the newly proposed criteria (κ = 0.886, p < 0.001). The newly proposed criteria would give a diagnosis of migraine only to those who were diagnosed as migraine by either of the other two sets of criteria, but not to those who received a diagnosis of “non-migraine headache” using both the ICHD-2 formal and appendix criteria. The risk for over-diagnosis is therefore limited. Although the ICHD-2 appendix criteria were assumed to be easier to administer (7), this has never been systemically verified. More importantly, as was reported previously (7), fewer patients could be diagnosed as migraine using the appendix criteria in our study as ICHD-2 migraine patients with either nausea or vomiting alone would be categorized as PM. This can be circumvented with our newly proposed criteria as patients diagnosed as migraine using either the ICHD-2 formal or appendix criteria would remain in the same category. Furthermore, the newly proposed criteria increased the diagnostic yield of migraine by 7.4% (134/1809). Incorporation of osmophobia might avoid under-diagnosis of migraine in Asian populations, as photophobia is less prevalent in Asian migraineurs (17–21). Also, the advantages of the ICHD-2 formal criteria in the Western populations remain unaltered. Although the additionally diagnosed patients were slightly older than the conventionally diagnosed patients, their headache profiles were comparable. Further studies are needed to determine the underlying disease processes and response to conventional migraine treatment of these additionally diagnosed patients. Moreover, a recent study pointed out that osmophobia in juvenile patients was associated with a higher conversion rate from TTH to migraine during a three-year follow-up (34). Although migraine was reportedly less prevalent in adolescents than in adults (35–37), the prevalence of osmophobia in migraine patients of both age groups was comparable (38,39). Whether or not the application of the newly proposed criteria in adolescent headache patients could also increase the diagnostic yield by identifying those with a potential for conversion to migraine remains to be determined.
Our study has limitations. First, even though this is a large-scale hospital-based study, whether or not the findings could be applied in the community-based settings is yet to be determined. For example, the prevalence of migraine varies in different settings, and sensitivity, specificity, PPV, and NPV could also vary with disease prevalence (40). Second, because of the retrospective nature of this study, the reliability of data collected depended solely on the recollection of the patients, and thus could be compromised by recall bias. Third, ethnic differences could also limit the generalizability of our results given the presence of disparities in the prevalence of migraine-associated symptoms, such as photophobia.
In conclusion, our study confirmed that osmophobia is one of the most common associated symptoms of migraine. Osmophobia was associated with a higher level of depression and anxiety as measured by the HADS. Our newly proposed criteria for migraine without aura, which integrate osmophobia, that is at least two of photophobia, phonophobia, and osmophobia, proved to be a useful alternative to the formal ICHD-2 criteria in Taiwanese patients.
Footnotes
Funding
The study was supported by Taiwan National Science Council (100-2314-B-010-019-MY2, 100-2314-B-010-018-MY3), Taipei Veterans General Hospital (VGHUST101-G7-1-1, V101C-106,V101E7-003, V100B-009, V101B-022), NSC support for the Center for Dynamical Biomarkers and Translational Medicine, National Central University, Taiwan (NSC100-2911-I-008-001), Brain Research Center, National Yang-Ming University and a grant from Ministry of Education, Aim for the Top University Plan.
Conflict of interest
None declared.