Clinical manifestations

Presentation and course

Olfactory dysfunction is a clinical marker for early stages of neurodegenerative diseases (34). In addition, cerebral neurodegenerative and microvascular lesions are common neuropathologies linking anosmia with mild cognitive impairment in older adults (21).

Idiopathic Parkinson disease. Olfactory deficits involving odor detection, identification, and discrimination are very common early in the course of Parkinson disease (52; 23; 07; 125; 109; 30; 12; 43; 81; 112; 29; 66; 62; 20). Occasionally, patients with Parkinson disease may develop pleasant olfactory hallucinations (phantosmias) (64). Impaired olfaction can predate the motor symptoms of Parkinson disease by at least 4 or 5 years (101; 59; 03; 35; 112). Nevertheless, olfactory defects in Parkinson disease don’t progress markedly with development of motor manifestations (23).

Hyposmia in Parkinson disease is associated with various motor and nonmotor symptoms like cognition, depression, anxiety, autonomic dysfunction, constipation, rapid eye movement sleep behavior disorder, and other sleep disturbances, and with the degree of nigrostriatal dopaminergic cell loss (113; 100). In multivariate regression models adjusting for age, sex, and disease duration, all motor and nonmotor symptoms of Parkinson disease were associated with olfactory function (as measured by UPSIT scores) (100). In addition, olfactory test scores were strongly associated with DaT binding in the putamen and caudate nuclei on DaT-SPECT scans (100).

In patients with Parkinson disease, olfactory deficits are associated with cognitive dysfunction, including memory impairment (02). The decline of cognitive function in early Parkinson disease is more rapid in patients with hyposmia at diagnosis than in those with normosmia (44). Olfactory impairment is also a significant predictor of postoperative delirium in patients with Parkinson disease (62). The presence of olfactory dysfunction in newly diagnosed patients with Parkinson disease increases the risk of dementia up to 10 years after Parkinson disease diagnosis regardless of baseline cognitive function, whereas normal olfaction (with normal cognition) at baseline predicts a benign cognitive course up to 10 years after diagnosis (20).

In patients with Parkinson disease, olfactory deficits are associated with autonomic defects (46). Anosmia in Parkinson disease is associated with autonomic failure, including baroreflex failure and noradrenergic denervation of the heart and other organs, independently of parkinsonism or striatal dopaminergic denervation (46).

Genetic forms of Parkinson disease. Olfaction is significantly reduced in some genetic forms of Parkinson disease, but when present, the degree of impairment is usually less than that in Parkinson disease (40; 06). Olfaction is significantly reduced among Parkin mutation heterozygotes with Parkinson disease but not among their heterozygous relatives without Parkinson disease (01). Subjects with LRRK2 G2019S mutations and Parkinson disease also have been reported to have impaired olfactory identification, although their olfactory impairment is much less than subjects with idiopathic Parkinson disease (73; 105; 114; 40; 57), and olfactory impairment seems to be limited to a specific subgroup of patients (104). Olfactory dysfunction in patients with LRRK2 mutation is positively correlated with reduced uptake of (123)I-meta-iodobenzylguanidine (MIBG) on cardiac scintigraphy, a measure of postganglionic sympathetic cardiac innervation (123). Hyposmia is less common in asymptomatic carriers of the LRRK2 G2019S mutation (114). Olfactory impairment among subjects with LRRK2 mutations and Parkinson disease has not been consistently identified in all studies; one study reported that subjects with LRRK2 mutation and Parkinson disease had a similar sense of smell to healthy LRRK2 mutation carriers and healthy family members without the mutation when adjusting for age (57). Glucocerebrosidase gene mutation-positive individuals also show deterioration in olfactory and cognitive assessment scores compared with controls, which is consistent with the prodrome of Parkinson disease (06).

Patients with Parkinson disease scored lower than age- and gender-matched controls on smell detection, definition, recognition, and forced choice (discrimination) (15). Compared with patients with acute respiratory infections, patients with Parkinson disease scored better on forced-choice testing, but no differences were found regarding other olfactory characteristics. Patients with traumatic brain injury scored significantly worse than patients with Parkinson disease on the olfaction domains.

Dementia with Lewy bodies. In dementia with Lewy bodies, like Parkinson disease, olfactory dysfunction is nearly universal, develops early (ie, before any movement or cognitive disorder), and is often severe (50; 04). Nevertheless, although olfactory disturbances have been considered a useful clinical marker of dementia with Lewy body disease (134), the addition of anosmia to the consensus diagnostic criteria for dementia with Lewy bodies did not significantly improve their overall performance (77; 85; 130). Severe hyposmia is nevertheless helpful in distinguishing neuropathologically confirmed dementia with Lewy bodies from Alzheimer disease (04). In patients with dementia with Lewy bodies, olfactory and trigeminal detection thresholds are significantly lower as compared to Parkinson disease dementia patients (37).

REM sleep behavior disorder. Odor identification is also impaired in patients with REM sleep behavior disorder, a common and very early feature of Lewy body alpha-synucleinopathies (118; 33; 80; 79; 96; 95; 56; 112; 71; 55; 69). Olfactory dysfunction also predicts early transition to a Lewy body disease in idiopathic REM sleep behavior disorder (71; 78) but cannot distinguish between Parkinson disease and dementia with Lewy bodies (78). Olfactory impairment is a sensitive and stable diagnostic biomarker of REM sleep behavior disorder and appears to be useful for identifying patients with idiopathic REM sleep behavior disorder at high risk for early conversion to Parkinson disease (69). Olfactory abnormalities are often measurable at least 5 years before disease onset and progress slowly in the preclinical stages but do not progress much, if at all, after onset (95; 56; 69).

Other movement disorders. Although hyposmia is a frequent, early, and often severe abnormality in Parkinson disease and dementia with Lewy bodies, this is not so in other forms of parkinsonism, including multiple system atrophy, vascular parkinsonism, progressive supranuclear palsy, or corticobasal degeneration, nor is hyposmia a feature of essential tremor (126; 60; 110; 90; 31).

Pathologically confirmed Parkinson disease had reduced olfaction compared with progressive supranuclear palsy or controls (111). Indeed, the presence of hyposmia in progressive supranuclear palsy suggests the presence of additional Lewy body pathology.

Most studies of olfaction in corticobasal degeneration have reported relatively mild deficits, but olfactory dysfunction can be moderate or severe in this disorder (90).

A mild olfactory loss develops later in the course of multiple system atrophy (60), but it is not clear that this is of clinical significance (63).

Olfactory assessment using the TDI global olfactory score (ie, comprised of the Threshold, Discrimination, and Identification subtests) from the Sniffin' Sticks test is a rapid, safe, and easily applicable biomarker that can differentiate tremor-dominant Parkinson disease from essential tremor in doubtful cases (31).

Other dementing disorders (eg, Alzheimer disease and frontotemporal dementia). Some degree of olfactory loss has also been reported in various other dementing disorders, including Alzheimer disease and frontotemporal dementia (26; 129; 132; 130; 39; 103; 66; 70; 18; 48; 124; 17; 19; 36; 72; 86; 91; 97; 136; 16). Olfactory impairment is more marked early in the course of the disease in patients with dementia with Lewy bodies than in those with either Alzheimer disease or frontotemporal dementia (130). Nevertheless, olfactory deficits have been reported in Alzheimer disease (26; 116; 75; 67; 103; 66; 18; 48; 124; 17; 19; 36; 86; 91; 97; 136), and these deficits may be detectable before the appearance of overt memory loss (67; 17), increase with severity of dementia (83; 108; 131; 124; 17), are associated with noncognitive neuropsychiatric symptoms (124), and correlate with density of neurofibrillary tangles in the entorhinal cortex and hippocampus (132) and with cortical Lewy body pathology (77). It remains unclear, though, whether Alzheimer disease is associated with clinically meaningful hyposmia in the absence of Lewy body pathology (77). Olfactory dysfunction, if apparent in Alzheimer disease, can sometimes help in the differential diagnosis with depression (116; 75). Frontotemporal dementia is also associated with relatively mild olfactory deficits, comparable to those seen in Alzheimer disease (76; 70; 72).

Familial ataxias. Mild hyposmia, less severe than that seen in Parkinson disease, has been observed in familial ataxias, but it has been suggested that this may be due to general cognitive deficits rather than specific olfactory problems (82). The ability of patients with spinocerebellar ataxia type 7 to discriminate and identify odors is significantly impaired, although their odor detection thresholds are normal, suggesting that olfactory perception is affected even when olfactory sensory capabilities are spared (42). Olfactory identification deficits have also been observed in FMR1 premutation carriers at risk for fragile X tremor/ataxia syndrome (58). The severity of olfactory impairment is comparable to that reported in hereditary ataxias (82) but less than that in Parkinson disease and Alzheimer disease (58). The frequency and severity of these olfactory defects are greater in cognitively impaired compared to cognitively intact carriers (58).

Motor neuron disease. According to some investigators, olfactory deficits may also occur with motor neuron disease (32; 50; 121; 93), but smell testing is not likely to be of clinical value in this condition. In any case, other investigators have not substantiated significant olfactory deficits in motor neuron disease (65).

Prognosis and complications

In some elderly patients, hyposmia may be an early ("preclinical") marker for the subsequent development of overt neurodegenerative disease. Among older persons without manifest cognitive impairment or parkinsonism, impairment of odor identification predicts subsequent cognitive decline, the development of both mild cognitive impairment and Parkinson disease, and the progression of amnestic mild cognitive impairment to Alzheimer disease (133; 101; 39; 25; 115; 99). Impaired olfaction can predate clinical Parkinson disease in men by at least 4 or 5 years and may be a useful screening tool to detect those at high risk for developing Parkinson disease in later life (101). The severity of olfactory impairment early in the course of Parkinson disease may also be a useful marker for the risk of neuropsychiatric complications (117). Impaired olfaction is also a putative predictor of the conversion of amnestic mild cognitive impairment to subsequent Alzheimer disease (39; 119).

In general, for patients with olfactory dysfunction, the prognosis depends primarily on etiology and the degree of residual function but also secondarily on gender, parosmia (distortion of the sense of smell), smoking habits, and age (38; 54). Male gender, initial presence of parosmia, smoking, and older age are negative prognostic factors (54).

Impaired olfaction adversely affects quality of life and the enjoyment of eating and contributes to hypogeusia, impaired appetite, weight loss, frailty, and depression in elderly patients (41). In addition, it may pose a safety risk, as affected individuals may not smell smoke or gas leaks and are less likely to recognize that food has spoiled.

A systematic review of neurodegenerative changes on MRI in patients with olfactory impairment and mild cognitive impairment or dementia found that hippocampal volume correlates with olfactory performance in individuals with cognitive impairment and that olfactory functional MRI may improve early detection of Alzheimer disease (137); however, the predictive utility of these imaging markers was limited in prospective studies.

In a cross-sectional and longitudinal study of olfaction among 364 initially cognitively normal community-dwelling older adults who had baseline olfaction data and subsequent cognitive assessments during an average 2.4-year follow-up period, poorer olfaction predicted incident mild cognitive impairment (122). Additionally, poorer olfaction was associated with overall and regional amyloid beta peptide (Aβ) in a subset who had PET amyloid imaging. Greater olfaction decline is associated with faster Aβ and tau accumulation in olfaction-related regions.

A cohort study from Shanghai, China, included 1811 participants aged 60 years or older who completed both an olfactory identification test and a cognitive assessment at baseline (2010–2011) (11). At the 10-year follow-up, olfactory impairment was associated with long-term mortality in older adults, and the association was even stronger in those with neurodegenerative diseases. Failure to identify coffee or a rose was associated with a higher mortality risk, and the association was mediated by neurodegenerative diseases.

A systematic review and meta-analysis also found that olfactory impairment is associated with all-cause mortality (89). Olfactory loss was associated with a significantly higher pooled hazard (hazard ratio: 1.52) of all-cause mortality.

Biological basis

Etiology and pathogenesis

Olfactory loss in neurodegenerative disorders is thought to result from central nervous system abnormalities involving olfactory pathways.

Olfactory loss in Parkinson disease and dementia with Lewy bodies is not due to damage to the olfactory epithelium but instead results from central nervous system abnormalities (51; 135; 02). Patients with Parkinson disease and severe hyposmia show decreased gray matter volume in the bilateral cuneus, associative visual area, precuneus, and anterior temporal lobes as well as a widespread significant decrease in amygdala functional connectivity (138). The decrease in amygdala functional connectivity with the inferior parietal lobule, lingual gyrus, and fusiform gyrus was significantly correlated with both reduction of cognitive assessment scores and the severity of hyposmia (138).

Pathology of the olfactory bulb and tract occurs prior to motor signs of Parkinson disease (51). In synucleinopathies, olfactory dysfunction relates specifically to Lewy body pathology, and correlates with cardiac sympathetic denervation, independently of striatal dopamine deficiency or parkinsonism (45). Impaired olfaction in Parkinson disease is associated with the presence of Lewy bodies and neuronal loss in the olfactory bulb and tract, with a strong correlation between neuronal loss and disease duration (92). In Parkinson disease, the olfactory bulb contains numerous Lewy bodies, and severe neuronal loss is present in the anterior olfactory nucleus (63). Immunolabeling for alpha-synuclein in the olfactory bulb and tract occurs prior to clinical signs of parkinsonism in dementia with Lewy bodies (51). The presence of Lewy body alpha-synucleinopathy in the olfactory bulb accurately predicts the presence of Lewy body pathology in other brain regions (05). Cortical olfactory receptors also become dysregulated in Parkinson disease (43). In pathological models overexpressing alpha-synuclein, high expression of this protein in neurons is a critical risk factor for neurodegeneration (120).

A cross-sectional study of 541 participants in the Rhineland Study, a population-based cohort study in Bonn, Germany, found that olfactory bulb volume on MRI was independently associated with odor identification function and was a strong mediator of the age-dependent association between volumes of central olfactory structures and olfactory function (68). Consequently, neurodegeneration-associated olfactory dysfunction may primarily originate from the pathology of peripheral olfactory structures.

Environmental toxins may also play a role in the development of Parkinson disease and in the associated olfactory impairment. From epidemiological studies, chronic inhalation exposure to welding fumes containing metal mixtures is associated with, and may play a role in the etiology of, the development of Parkinson disease. In particular, nasal exposure to vanadium pentoxide adversely affects olfactory bulbs, resulting in neurobehavioral and neurochemical impairments (84). Vanadium is present in welding fumes, such as vanadium pentoxide, and vanadium is increasingly incorporated in the production of high-strength steel. Vanadium pentoxide induces dopaminergic neurotoxicity via protein kinase C delta-dependent oxidative signaling mechanisms in dopaminergic neuronal cells (84).

The pathophysiology of hyposmia in Lewy body alpha-synucleinopathies may have multiple components, including those resulting from degenerative changes in the olfactory bulb and primary olfactory cortex, as well as limbic dysfunction and possibly prefrontal dysfunction (52; 08; 09; 127; 02). Functional imaging indicates that reduced neuronal activity in the amygdala, hippocampus, and piriform cortex (uncus) contributes to olfactory dysfunction in Parkinson disease (127; 02). Hyposmia in Parkinson disease is more closely associated with cholinergic denervation of limbic archicortex (ie, hippocampus and amygdala) than with nigrostriatal dopaminergic denervation (09). However, because olfactory threshold and odor identification in Parkinson disease are not related to duration of disease, to current therapy with levodopa or anticholinergic drugs, or to “on” and “off” states, olfactory impairment in Parkinson disease likely involves mechanisms that are not due to dopaminergic or cholinergic denervation and that are not influenced by pharmacologic manipulation of dopaminergic or cholinergic status (98).

Impaired olfaction in Alzheimer disease may reflect neurodegeneration within basal telencephalic structures that mediate olfactory processing, including the anterior olfactory nucleus (103). Similarly, olfactory dysfunction in behavioral variant frontotemporal dementia has been attributed to disruption of association areas involved in olfactory processing, particularly those in the temporal lobe and amygdala (87). In a large, prospective, population-based study, odor identification impairment was associated with a decline in episodic memory but only in carriers of the ɛ4 allele of the apolipoprotein E gene (86). Other studies of olfaction in Alzheimer disease have implicated an early imbalance in splicing factors, the subsequent interruption of the cycling of neurotransmitters, alteration in toxic and protective mechanisms of beta-amyloid, and ultimately mitochondrial dysfunction in conjunction with a disturbance in neuron-neuron adhesion (139).

A mild olfactory loss develops later in the course of multiple system atrophy (60) and is associated with characteristic glial cytoplasmic inclusions in the olfactory bulb and some degree of neuronal loss in the anterior olfactory nucleus (63).

A relatively mild olfactory impairment has also been reported in patients with normal pressure hydrocephalus, but odor identification test scores are unaffected by either extended lumbar drainage or implantation of a ventriculo-peritoneal shunt (91).

The relationship between sensory impairment and cognitive impairment is not unique to the olfactory system, but a similar relationship or comparable magnitude exists for hearing and visual impairments, which implies that sensory health per se may be a marker of brain health or, equivalently, that impairments of the special senses may be an indicator of brain aging or disease-related dysfunction (36). This does not obviate the strong associations between some forms of modality-specific dysfunction of the special senses and certain disease states.

Epidemiology

Olfactory deficits are very common early in the course of Parkinson disease and are present in more than 90% of patients with early-stage Parkinson disease (52; 23; 07; 109; 30; 49). Using standardized olfactory assessment tools (ie, University of Pennsylvania Smell Identification Test), olfactory deficits are nearly universal in patients with Parkinson disease who have confirmed evidence of nigrostriatal denervation by PET imaging (49). Idiopathic olfactory dysfunction in first-degree relatives of Parkinson disease patients is also associated with an increased risk of developing Parkinson disease within 2 to 5 years (94). Olfactory deficits are nearly universal in the early stages of dementia with Lewy bodies (50). Olfactory deficits are less common and less severe in Alzheimer disease and other neurodegenerative disorders (128), though some studies still report a high prevalence of olfactory dysfunction in Alzheimer disease, approaching 90% in some studies (139).

Prodromal Parkinson disease is infrequent among idiopathic hyposmia patients (74).

Differential diagnosis

Disorders of olfaction in the elderly can be conveniently divided into conductive, sensorineural, and central disorders, where (1) conductive disorders involve transmission of the sensory stimuli to the sensory receptors (usually, but not always, by impeding transmission), (2) sensorineural disorders involve dysfunction of the sensory receptors or conduction of signals from the sensory receptors to the brain, and (3) central disorders involve dysfunction of processing of sensory information within the central nervous system, and particularly within the brainstem and cerebrum.

Conductive hyposmia (depending on the author or circumstance, “conductive” is sometimes stated conversely as “obstructive”) occurs if there are impediments to conveyance of odorants to the olfactory neuroepithelium (eg, upper respiratory infection, chronic rhinosinusitis, nasal polyposis). Sensorineural disorders involve the receptors themselves (“sensory”) or the afferent neural pathways involving the respective cranial nerves, tracts, or central processing centers of the brain (collectively “neural”). In addition to presbyosmia (age-related olfactory dysfunction), other common causes of sensorineural hyposmia in the elderly include tobacco smoking, drug toxicity, head injury, hypothyroidism, and post-viral disorders (following an influenza-like infection). Common causes of central hyposmia in the elderly include Parkinson disease and dementia with Lewy bodies, but other neurodegenerative disorders (eg, Alzheimer disease) may also cause olfactory impairment.

Markedly reduced olfaction in a parkinsonian patient is supportive of Parkinson disease or dementia with Lewy bodies (126), whereas normal smell identification is rare with these conditions and should prompt review of the diagnosis (unless possibly if the patient is female with tremor-dominant disease) (50). Nevertheless, adding anosmia to the consensus criteria for dementia with Lewy bodies did not significantly improve their overall performance (77; 85; 130).

Preserved or mildly impaired olfactory function in a parkinsonian patient is more likely to be related to atypical parkinsonism, such as vascular parkinsonism, multiple system atrophy, progressive supranuclear palsy, or corticobasal degeneration (61; 50). Although hyposmia is a frequent and early abnormality in Parkinson disease and dementia with Lewy bodies, this is not so in other forms of parkinsonism (ie, without Lewy bodies), nor is hyposmia a feature of essential tremor (126; 60; 110; 90). However, these heuristic clinical rules are not absolute because occasional patients with corticobasal degeneration have moderate or severe olfactory impairment (90).

Diagnostic workup

Even if disease-modifying therapies were available for neurodegenerative disease, smell tests alone do not have a high enough predictive power to be used in a screening program for neurodegenerative disease (47). However, they could serve as a component of a short test battery or as part of a stepwise process that could be used to more accurately identify early neurodegeneration in large populations.

Elderly patients with hyposmia should undergo a focused history and examination that particularly addresses the sense of smell, the nasal passages, and the nervous system, with attention to recognizing disorders of movement and cognition. Office testing of smell can include well-standardized, commercially available tests (eg, University of Pennsylvania Smell Identification Test or UPSIT) or crude approaches utilizing identification of a few readily available odorants (eg, oil of Wintergreen, oil of cloves) (27; 28; 22; 24; 88). More complicated odor identification and detection tests are also available but are rarely practical outside of specialized diagnostic laboratories (22). The UPSIT is a forced-choice olfactory discrimination test utilizing micro-encapsulated odorants in standardized “scratch 'n sniff” booklets. UPSIT scores are standardized by gender and age and can be used to identify degrees of hyposmia and some malingerers. UPSIT may be sensitive to decline in dementia and is largely unaffected by premorbid cognitive functioning, making it particularly useful in the evaluation of olfaction in patients with cognitive dysfunction (106). Irritant substances, such as ammonia, are sometimes employed when psychogenic or malingered anosmia is a consideration because such substances are perceived via trigeminal afferent pathways rather than through the olfactory system.

A history of head trauma, past and current tobacco smoking, and medications should be reviewed for potential contributing causes to olfactory dysfunction. Thyroid dysfunction should be excluded with thyroid-stimulating hormone (TSH) and free T4 assays. Patients with evidence of a conductive olfactory disturbance may warrant CT imaging of the paranasal sinuses and/or otolaryngological referral. Patients with neurologic signs or symptoms need further evaluation to exclude neurodegenerative disorders and, rarely, structural lesions of the cribriform plate or basal forebrain. Neuroimaging is recommended in patients with dementia and in those with asymmetric or unilateral hyposmia not attributable to conductive hyposmia. Patients without conductive hyposmia and without evidence of neurologic signs or symptoms most likely have presbyosmia, although early neurologic disease is possible. Periodic re-examination may help identify cases of early neurodegenerative disease (eg, Alzheimer disease, Parkinson disease, dementia with Lewy bodies).

Management

There is no established treatment for hyposmia associated with neurodegenerative diseases, and, in general, sensorineural and central causes of hyposmia are seldom correctable.

In patients with Parkinson disease or dementia with Lewy bodies, olfactory threshold and odor identification do not improve with levodopa therapy or with other pharmacologic manipulation of dopaminergic or cholinergic status (98; 53; 102). The lack of efficacy of dopaminergic therapies to correct olfactory disturbances in Parkinson disease is, at least in part, because dopamine inhibits olfactory transmission in the olfactory glomeruli (53). Deep brain stimulation does not improve olfactory function (10).

There is no evidence that acetylcholinesterase inhibitors or memantine help alleviate the olfactory disturbances in Alzheimer disease.

In general, conductive loss of olfaction is more likely to be remediable (ie, to result in significantly improved function) than is a sensorineural or central loss (107). Concomitant sinonasal disease should be treated when present. Smokers should be advised to quit smoking, and further counseling and pharmacologic management should be employed as necessary to facilitate cessation. Patients should be counseled regarding safety issues and the use of proper monitoring devices for smoke and natural gas; this is particularly important for those who are living alone. In addition, the effects of hyposmia on gustation and appetite should be assessed, and, if necessary, a dietary consult should be obtained. In some cases, dietary supplements or flavor enhancers may prove beneficial.

Multiple studies and reports of olfactory training have made unsupported claims about purported benefits in individuals with dementia. Most have been poorly designed and have not separated the effects of increased activity and engagement in the study groups from supposed benefits of improving depression and attention in subjects (13). In subjects with mild cognitive impairment, olfactory training twice a day for 4 months had no significant effect on olfaction and cognitive function (14).