Discussion

The present study comprised 245 patients with a predominance of anosmia and severe hyposmia. Although the composite scores were low in these patients, subcomponent analysis of threshold and suprathreshold scores showed a relative preservation of olfactory identification, especially in those with moderate and severe hyposmia. Pepper was the least perceived smell among patients with anosmia, severe hyposmia and moderate hyposmia. Patients with allergic rhinitis, which was the most common sinonasal cause of olfactory dysfunction, showed a relative preservation of identification scores, unlike those with the other sinonasal diseases, who showed low threshold and identification scores. Similar to the findings in the cohort, pepper was the least perceived smell among patients with various sinonasal diseases.

Most patients with all four types of olfactory dysfunction were in the 18–45 years age group, with a male predominance. Because women are more involved in cooking and care giving, they are thought to appreciate smell disturbances more than men (38.3 vs 27.6 per cent).^{Reference Doty8} The male predominance (61.63 per cent) in our cohort may arise because this is a hospital-based study and men seek medical help faster in Indian population.^{Reference Kapoor, Agrawal, Ravi, Roy, Subramanian and Guleria9} The comparatively lower mean age (42 years) in our study population could indicate that people of prime working age seek medical help, whereas elderly people tend to live with the problem.

Unlike hearing or vision, olfactory losses can go unrecognised for a long time, especially when there is unilateral or less than complete loss. Less than 25 per cent of people with demonstrable smell loss are aware of their loss until they are tested quantitatively. In conditions with gradual deterioration in smell sensation, a certain threshold must be reached before the loss becomes noticeable and the person seeks medical help.^{Reference Wehling, Naess, Wollschlaeger, Hofstad, Bramerson and Bende10} This could explain why almost 68 per cent of patients in the study cohort had anosmia or severe hyposmia. Sudden loss of smell may be recognised and reported more quickly.

The most commonly experienced associated condition was allergic rhinitis (44.4 per cent), followed by chronic rhinosinusitis with nasal polyposis (20.1 per cent) and chronic rhinosinusitis without nasal polyposis (6.7 per cent). A quarter of patients did not state an identifiable cause for the loss of smell.

Different aetiologies affect the olfactory pathway in different ways. Age-related smell decline can be attributed to olfactory cell damage due to repeated exposure to air pollutants, cigarette smoke, viruses, bacteria and other airborne xenobiotics along with reduced olfactory receptor cell regeneration from the basal cells. As age advances, the number and size of foramina in the cribriform plate reduces, resulting in fewer olfactory receptor cell axons from the nasal cavity reaching the brain and inducing olfactory receptor cell necrosis.^{Reference Doty11}

Chronic rhinosinusitis, which accounts for 14–30 per cent of cases with olfactory dysfunction, produces a gradual loss of olfaction. While respiratory mucosal oedema and polyps impede the access of odorants to the olfactory cleft, alterations in the composition and transport of the mucus layer can impair access to or removal from the receptor sites.^{Reference Dalton12} A sensorineural component has also been implicated based on the olfactory neuroepithelial inflammation seen on histopathology and the clinical response to systemic corticosteroids.^{Reference Litvack, Fong, Mace, James and Smith13} Hence, in chronic rhinosinusitis the aetiology could be multifactorial.^{Reference Kern14}

In post-infectious hyposmia and anosmia, the sensory deficit is probably due to a disproportionately high rate of olfactory neuronal apoptosis when compared with the compensatory increased rate of neurogenesis.^{Reference Reden, Herting, Lill, Kern and Hummel15} The lack of basal cells in some post-viral disorders explains the loss of regenerative capacity. However, in post-traumatic olfactory dysfunction the damage resulting from the external force is inflicted on the olfactory filament, and the basal cells with their regenerative capacity are retained in the olfactory epithelium.^{Reference Aiba, Sugiura, Mori, Matsumoto, Tomiyama and Okuda16}

Olfactory abilities assessed by various psychophysical tests measure three main domains, namely, threshold (lowest detectable concentration of odours), discrimination (ability to differentiate between odours) and identification (ability to identify odours). The criteria for olfactory dysfunction based on the measurement of odour detection thresholds for one or more chemicals (e.g. butanol, pyridine, phenylethyl alcohol) and performance on a multiple-item odour identification task are superior to scaling measures or questionnaires because they are less prone to subjective biases and variability.

Compared with higher-order olfactory tasks, such as identification and discrimination, threshold tests place few demands on cognitive function and are independent of cultural adaptation.^{Reference Sorokowska, Sorokowski, Hummel and Huanca17} It has been suggested that the olfactory threshold is strongly related to sensory capability because it is often impaired in peripheral sinonasal disorders, while olfactory discrimination and identification require higher cognitive functions, including working memory, judgment and decision making, and their dysfunction may represent generalised cognitive deterioration.

Soler et al. reported that olfactory dysfunction in patients with chronic rhinosinusitis showed the greatest loss for threshold levels, with only 60 per cent having a loss in the identification domain, and a much greater disparity between the chronic rhinosinusitis without nasal polyposis (38 per cent) and chronic rhinosinusitis with nasal polyposis (75 per cent) subgroups.^{Reference Soler, Kohli, Storck and Schlosser18} Furthermore, 20–40 per cent of patients with allergic rhinitis have olfactory dysfunction, with the severity of dysfunction increasing with the duration, type and severity of allergic rhinitis.

This study showed that patients with chronic rhinosinusitis without nasal polyposis had worse odour threshold scores and showed more anosmia and severe hyposmia than those with chronic rhinosinusitis with nasal polyposis. Although 75 per cent of patients with chronic rhinosinusitis without nasal polyposis and 61 per cent of patients with chronic rhinosinusitis with nasal polyposis had olfactory threshold values of 0, only 43.75 and 48.96 per cent, respectively, of chronic rhinosinusitis with nasal polyposis and chronic rhinosinusitis without nasal polyposis patients had an olfactory identification score of 0. Among allergic rhinitis patients, even though 42.86 per cent of patients had threshold values of 0, 39.62 per cent of patients had identification scores of 7.

Because most of the patients in our cohort had allergic rhinitis and chronic rhinosinusitis, which in turn affected the peripheral olfactory system, preferential involvement of the olfactory threshold with preservation of identification scores even in patients with anosmia and severe hyposmia seems plausible.^{Reference Corwin, Serby and Rotrosen19} Apter et al. reported a higher incidence and severity of olfactory loss among allergic rhinitis patients than rhinosinusitis patients because allergic rhinitis patients are prone to respiratory infections, which can cause olfactory epithelial loss.^{Reference Apter, Gent and Frank20}

When analysing the specific smell loss in various nasal diseases, pepper was found to be the least perceived smell among the allergic rhinitis, chronic rhinosinusitis without nasal polyposis, chronic rhinosinusitis with nasal polyposis and idiopathic anosmia patients. In patients with atrophic rhinitis, the asafoetida and baby powder smells were affected the most. Because this study examined a retrospective cohort, the onset and sequence of loss of different odours was not assessed.

The novel finding from our study is that in most inflammatory sinonasal conditions the odour identification component is relatively preserved, even when the threshold is maximally affected. In patients with comparable severity of olfactory dysfunction based on the composite score of Connecticut Chemosensory Clinical Research Centre testing, a relatively preserved suprathreshold odour identification score may predict better prognosis and recovery. Pepper was the least perceived and most probably the earliest odour to be lost in patients with olfactory dysfunction. Our study is the first to arrive at these conclusions and there are no similar studies with which it can be compared.

The limitations of our study are that it was retrospective in nature and the study subjects were a self-selected group of individuals who had sought medical help. However, many individuals with sinonasal and neurological disorders may not be aware of their olfactory dysfunction unless specific olfactory tests are performed, therefore the conclusions of our study may not extrapolate to the distribution of general olfactory dysfunction in these conditions. A well-designed prospective study of various diseases causing olfactory loss will help to determine the pattern of olfactory dysfunction they cause and the sequence of loss of different odours. This may also help in the early diagnosis and prognostication of neurodegenerative diseases.