Harrison's Neurology in Clinical Medicine, 3rd Edition


Richard L. Doty Image Steven M. Bromley

All environmental chemicals necessary for life enter the body by the nose and mouth. The senses of smell (olfaction) and taste (gustation) monitor those chemicals, determine the flavor and palatability of foods and beverages, and warn of dangerous environmental conditions, including fire, air pollution, leaking natural gas, and bacteria-laden foodstuffs. These senses contribute significantly to quality of life and, when dysfunctional, can have untoward physical and psychological consequences. A basic understanding of these senses in health and disease is critical for the physician, since thousands of patients present to doctors’ offices each year with complaints of chemosensory dysfunction. Among the more important developments in neurology has been the discovery that decreased smell function is perhaps the first sign of neurodegenerative diseases such as Alzheimer’s disease (AD) and Parkinson’s disease (PD), signifying their “presymptomatic” phase.


Olfactory system

Odorous chemicals enter the nose during inhalation and active sniffing as well as during deglutition. After reaching the highest recesses of the nasal cavity, they dissolve in the olfactory mucus and diffuse or are actively transported to receptors on the cilia of olfactory receptor cells. The cilia, dendrites, cell bodies, and proximal axonal segments of these bipolar cells are situated within a specialized neuroepithelium that covers the cribriform plate, the superior nasal septum, the superior turbinate, and sectors of the middle turbinate (Fig. 23-1). Each of the ~6 million bipolar receptor cells expresses only one of ~450 receptor protein types, most of which respond to more than a single chemical. When damaged, the receptor cells can be replaced by stem cells near the basement membrane. Unfortunately, such replacement is often incomplete.



Anatomy of the olfactory neural pathways, showing the distribution of olfactory receptors in the roof of the nasal cavity. (Copyright David Klemm, Faculty and Curriculum Support [FACS], Georgetown University Medical Center; used with permission.)

After coalescing into bundles surrounded by glia-like ensheathing cells (termed fila), the receptor cell axons pass through the cribriform plate to the olfactory bulbs, where they synapse with dendrites of other cell types within the glomeruli (Fig. 23-2). These spherical structures, which make up a distinct layer of the olfactory bulb, are a site of convergence of information, since many more fibers enter than leave them. Receptor cells that express the same type of receptor project to the same glomeruli, effectively making each glomerulus a functional unit. The major projection neurons of the olfactory system—the mitral and tufted cells—send primary dendrites into the glomeruli, connecting not only with the incoming receptor cell axons but with dendrites of periglomerular cells. The activity of the mitral/tufted cells is modulated by the periglomerular cells, secondary dendrites from other mitral/tufted cells, and granule cells, the most numerous cells of the bulb. The latter cells, which are largely GABAergic, receive inputs from central brain structures and modulate the output of the mitral/tufted cells. Interestingly, like the olfactory receptor cells, some cells within the bulb undergo replacement. Thus, neuroblasts formed within the anterior subventricular zone of the brain migrate along the rostral migratory stream, ultimately becoming granule and periglomerular cells.



Schematic of the layers and wiring of the olfactory bulb. Each receptor type (red, green, blue) projects to a common glomerulus. The neural activity within each glomerulus is modulated by periglomerular cells. The activity of the primary projection cells, the mitral and tufted cells, is modulated by granule cells, periglomerular cells, and secondary dendrites from other mitral and tufted cells. (From www.med.yale.edu/neurosurg/treloar/index.html.)

The axons of the mitral and tufted cells synapse within the primary olfactory cortex (POC) (Fig. 23-3). The POC is defined as the cortical structures that receive direct projections from the olfactory bulb, most notably the piriform and entorhinal cortices. Although olfaction is unique in that its initial afferent projections bypass the thalamus, persons with damage to the thalamus can exhibit olfactory deficits, particularly ones of odor identification. Those deficits probably reflect the involvement of thalamic connections between the primary olfactory cortex and the orbitofrontal cortex (OFC), where odor identification occurs. The close anatomic ties between the olfactory system and the amygdala, hippocampus, and hypothalamus help explain the intimate associations between odor perception and cognitive functions such as memory, motivation, arousal, autonomic activity, digestion, and sex.



Anatomy of the base of the brain showing the primary olfactory cortex.

Taste system

Tastants are sensed by specialized receptor cells present within taste buds: small grapefruit-like segmented structures on the lateral margins and dorsum of the tongue, the roof of the mouth, the pharynx, the larynx, and the superior esophagus (Fig. 23-4). Lingual taste buds are embedded in well-defined protuberances termed fungiform, foliate, and circumvallate papillae. After dissolving in a liquid, tastants enter the opening of the taste bud—the taste pore—and bind to receptors on microvilli, small extensions of receptor cells within each taste bud. Such binding changes the electrical potential across the taste cell, resulting in neurotransmitter release onto the first-order taste neurons. Although humans have ~7500 taste buds, not all harbor taste-sensitive cells; some contain only one class of receptor (e.g., cells responsive only to sugars), whereas others contain cells sensitive to more than one class. The number of taste receptor cells per taste bud ranges from zero to well over 100. A small family of three G-protein-coupled receptors (GPCRs)—T1R1, T1R2, and T1R3—mediate sweet and umami taste sensations. Umami (“savory”) refers to the flavors of meat, cheese, and broth due to glutamate and related compounds. Bitter sensations, in contrast, depend on T2R receptors, a family of ~30 GPCRs expressed on cells different from those which express the sweet and umami receptors. T2Rs sense a wide range of bitter substances but do not distinguish among them. Sour tastants are sensed by the PKD2L1 receptor, a member of the transient receptor potential protein (TRP) family. Perception of salty sensations, such as those induced by sodium chloride, arises from the entry of Na+ ions into the cells via specialized membrane channels such as the amiloride-sensitive Na+ channel.



Schematic of the taste bud and its opening (pore), as well as the location of buds on the three major types of papillae: fungiform (anterior), foliate (lateral), and circumvallate (posterior). TRC, taste receptor cell.

Taste information is sent to the brain via three cranial nerves (CNs): CN VII (the facial nerve, which involves the intermediate nerve with its branches, the greater petrosal and chorda tympani nerves); CN IX (the glossopharyngeal nerve); and CN X (the vagus nerve(Fig. 23-5). CN VII innervates the anterior tongue and all of the soft palate, CN IX innervates the posterior tongue, and CN X innervates the laryngeal surface of the epiglottis, the larynx, and the proximal portion of the esophagus. The mandibular branch of CN V (V3) conveys somatosensory information (e.g., touch, burning, cooling, irritation) to the brain. Although not technically a gustatory nerve, CN V shares primary nerve routes with many of the gustatory nerve fibers and adds temperature, texture, pungency, and spiciness to the taste experience. The chorda tympani nerve is notable for taking a recurrent course through the facial canal in the petrosal portion of the temporal bone, passing through the middle ear, then exiting the skull via the petrotympanic fissure, where it joins the lingual nerve (a division of CN V) near the tongue. This nerve also carries parasympathetic fibers to the submandibular and sublingual glands, whereas the greater petrosal nerve supplies the palatine glands, thereby influencing saliva production.



Schematic of the cranial nerves that mediate taste function, including the chorda tympani nerve (CN VII), the glossopharyngeal nerve (CN IX), and the vagus nerve (CN X). (Copyright David Klemm, Faculty and Curriculum Support [FACS], Georgetown University Medical Center; used with permission.)

The axons of the projection cells that synapse with taste buds enter the rostral portion of the nucleus of the solitary tract (NTS) within the medulla of the brainstem (Fig. 23-5). From the NTS, neurons then project to a division of the ventroposteromedial thalamic nucleus (VPM) via the medial lemniscus. From there projections are made to the rostral part of the frontal operculum and adjoining insula, a brain region considered the primary taste cortex (PTC). Projections from the primary taste cortex then go to the secondary taste cortex, namely, the caudolateral OFC. This brain region is involved in the conscious recognition of taste qualities. Moreover, since it contains cells that are activated by several sensory modalities, it is probably a center for establishing “flavor.”


The ability to smell is influenced by factors such as age, sex, general health, nutrition, smoking, and reproductive state. Women typically outperform men on tests of olfactory function and retain normal smell function to a later age. Significant decrements in the ability to smell are present in over 50% of the population between 65 and 80 years of age and in 75% of those 80 years and older (Fig. 23-6). Such presbyosmia helps explain why many elderly persons report that food has little flavor, a problem that can result in nutritional disturbances. It also helps explain why a disproportionate number of the elderly die in accidental gas poisonings. A relatively complete listing of conditions and disorders that have been associated with olfactory dysfunction is presented in Table 23-1.



Scores on the University of Pennsylvania Smell Identification Test (UPSIT) as a function of subject age and sex. Numbers by each data point indicate sample sizes. Note that women identify odorants better than men at all ages. (From Doty et al: Science 226:1421, 1984. Copyright 1984 American Association for the Advancement of Science.)

TABLE 23-1




Aside from aging, the three most common identifiable causes of long-lasting or permanent smell loss seen in the clinic are, in order of frequency, severe upper respiratory infections, head trauma, and chronic rhinosinusitis. The physiologic basis for most head trauma–related losses is the shearing and subsequent scarring of the olfactory fila as they pass from the nasal cavity into the brain cavity. The cribriform plate does not have to be fractured or show pathology for smell loss to be present. Severity of trauma, as indexed by a poor Glasgow Coma Rating on presentation and the length of posttraumatic amnesia, is associated with higher risk of olfactory impairment. Fewer than 10% of posttraumatic anosmic patients recover age-related normal function over time. Upper respiratory infections, such as those associated with the common cold, influenza, pneumonia, or HIV, can directly and permanently harm the olfactory epithelium by decreasing receptor cell numbers, damaging cilia on remaining receptor cells, and inducing the replacement of sensory epithelium with respiratory epithelium. The smell loss associated with chronic rhinosinusitis is related to disease severity, with most loss occurring in cases in which rhinosinusitis and polyposis are both present. Although systemic glucocorticoid therapy usually can induce short-term functional improvement, it does not, on average, return smell test scores to normal, implying that chronic permanent neural loss is present and/or that short-term administration of systemic glucocorticoids does not mitigate the inflammation completely. It is well established that microinflammation in an otherwise seemingly normal epithelium can influence smell function.

A number of neurodegenerative diseases are accompanied by olfactory impairment, including AD, PD, Huntington’s disease, Down syndrome, parkinsonismdementia complex of Guam, dementia with Lewy bodies (DLB), multiple system atrophy, vascular parkinsonism, corticobasal syndrome, frontotemporal dementia, multiple sclerosis (MS), and idiopathic rapid eye movement (REM) behavioral sleep disorder (iRBD). The olfactory disturbance of MS varies as a function of the plaque activity within the frontal and temporal lobes. In postmortem studies of patients with very mild “presymptomatic” signs of AD, poorer smell function has been associated with higher levels of AD-related pathology even after controlling for apolipoprotein E4 alleles and the level of episodic memory function present at the time of olfactory testing. Olfactory impairment in PD often predates the clinical diagnosis by at least 4 years. Studies of the sequence of Lewy body and abnormal α-synuclein development in staged PD cases, along with evidence that the smell loss presents early, is stable over time, and is not affected by PD medications, suggest that the olfactory bulbs may be, along with the dorsomotor nucleus of the vagus, the site of first neural damage in PD. Smell loss is more marked in patients with early clinical manifestations of DLB than in those with mild AD. Interestingly, smell loss is minimal or nonexistent in progressive supranuclear palsy and 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced parkinsonism.

The smell loss seen in iRBD is of the same magnitude as that found in PD. This is of particular interest to clinicians since patients with iRBD frequently develop PD and hyposmia. iRBD may actually represent an early associated condition of PD. REM behavior disorder not only is seen in its idiopathic form but also can be associated with narcolepsy. This led to a study of narcoleptic patients with and without REM behavior disorder that demonstrated that narcolepsy, independent of REM behavior disorder, was associated with significant impairments in olfactory function. Orexin A, also known as hypocretin-1, is dramatically diminished or undetectable in the cerebrospinal fluid of patients with narcolepsy and cataplexy. The orexin-containing neurons in the hypothalamus project throughout the olfactory system (from the olfactory epithelium to the olfactory cortex), and damage to these orexin-containing projections may be one underlying mechanism for impaired olfactory performance in narcoleptic patients. The administration of intranasal orexin A (hypocretin-1) appears to result in improved olfactory function relative to a placebo, supporting the notion that mild olfactory impairment is not only a primary feature of narcolepsy with cataplexies but that CNS orexin deficiency may be a fundamental part of the mechanism for this loss.


The majority of patients who present with complaints of taste dysfunction exhibit olfactory, not taste, loss. This is the case because most flavors attributed to taste actually depend on retronasal stimulation of the olfactory receptors during deglutition. As noted earlier, taste buds only mediate basic tastes such as sweet, sour, bitter, salty, and umami. Significant impairment of whole-mouth gustatory function is rare outside of generalized metabolic disturbances or systemic use of some medications, since taste bud regeneration occurs and peripheral damage alone would require the involvement of multiple cranial nerve pathways. Nonetheless, taste can be influenced by (1) the release of foul-tasting materials from the oral cavity from oral medical conditions and appliances (e.g., gingivitis, purulent sialadenitis), (2) transport problems of tastants to the taste buds (e.g., drying of the orolingual mucosa, infections, inflammatory conditions), (3) damage to the taste buds themselves (e.g., local trauma, invasive carcinomas), (4) damage to the neural pathways innervating the taste buds (e.g., middle ear infections), (5) damage to central structures (e.g., multiple sclerosis, tumor, epilepsy, stroke), and (6) systemic disturbances of metabolism (e.g., diabetes, thyroid disease, medications). Bell’s palsy is among the most common causes of CN VII injury that results in taste disturbance. Unlike CN VII, CN IX is relatively protected along its path, although iatrogenic interventions such as tonsillectomy, bronchos-copy, laryngoscopy, and radiation therapy can result in selective injury. Migraine is associated on rare occasions with a gustatory prodrome or aura, and certain tastes may trigger a migraine. Although a number of disorders can affect CN IX, including tumors, trauma, vascular lesions, and infection, it remains unclear if noticeable taste disturbance can result from such factors.

Although both taste and smell can be adversely influenced by pharmacologic agents, drug-related taste alterations are more common. Indeed, over 250 medications have been reported to alter the ability to taste. Major offenders include antineoplastic agents, antirheumatic drugs, antibiotics, and blood pressure medications. Terbinafine, a commonly used antifungal, has been linked to taste disturbance lasting up to 3 years. In a controlled trial, nearly two-thirds of individuals taking eszopiclone (Lunesta) experienced a bitter dysgeusia which was stronger in women, systematically related to the time since drug administration, and positively correlated with both blood and saliva levels of the drug. Intranasal use of nasal gels and sprays containing zinc—common over-the-counter prophylactics for upper respiratory viral infections—has been implicated in loss of smell function. Whether their efficacy in preventing such infections, which are the most common cause of anosmia and hyposmia, outweighs their potential detriment to smell function requires study.

As with olfaction, a number of systemic disorders can affect taste. They include chronic renal failure, end-stage liver disease, vitamin and mineral deficiencies, diabetes, and hypothyroidism, to name a few. Psychiatric conditions can be associated with chemosensory alterations (e.g., depression, schizophrenia, bulimia). A review of tactile, gustatory, and olfactory hallucinations demonstrated that no one type of hallucinatory experience is pathognomonic to any specific diagnosis.


In most cases, a careful clinical history will establish the probable etiology of a chemosensory problem, including questions about its nature, onset, duration, and pattern of fluctuations. Sudden loss suggests the possibility of head trauma, ischemia, infection, or a psychiatric condition. Gradual loss can reflect the development of a progressive obstructive lesion. Intermittent loss suggests the likelihood of an inflammatory process. The patient should be asked about potential precipitating events, such as cold or flu infections before symptom onset, as they often are underappreciated. Information regarding head trauma, smoking habits, drug and alcohol abuse (e.g., intranasal cocaine, chronic alcoholism in the context of Wernicke’s and Korsakoff’s syndromes), exposures to pesticides and other toxic agents, and medical interventions are also informative. A determination of all the medications the patient was taking before and at the time of symptom onset is important, since many can cause chemosensory disturbances. Comorbid medical conditions associated with smell impairment, such as renal failure, liver disease, hypothyroidism, diabetes, and dementia, should be assessed. Delayed puberty in association with anosmia (with or without midline craniofacial abnormalities, deafness, and renal anomalies) suggests the possibility of Kallmann syndrome. Recollection of epistaxis, discharge (clear, purulent, or bloody), nasal obstruction, allergies, and somatic symptoms, including headache or irritation, may have localizing value. Questions related to memory, parkinsonian signs, and seizure activity (e.g., automatisms, occurrence of blackouts, auras, and déjàvu) should be posed. Pending litigation and the possibility of malingering should be considered.

Neurologic and otorhinolaryngologic (ORL) examinations, along with appropriate brain and nasosinus imaging, aid in the evaluation of patients with olfactory or gustatory complaints. The neural evaluation should focus on cranial nerve function, with particular attention to possible skull base and intracranial lesions. Visual acuity, field, and optic disc examinations aid in the detection of intracranial mass lesions that induce elevations in intracranial pressure (papilledema) and optic atrophy, especially when one is considering Foster Kennedy syndrome (ipsilateral optic nerve atrophy and contralateral papilledema usually due to a meningioma near the olfactory bulb or tract). The ORL examination should thoroughly assess the intranasal architecture and mucosal surfaces. Polyps, masses, and adhesions of the turbinates to the septum may compromise the flow of air to the olfactory receptors, since less than a fifth of the inspired air traverses the olfactory cleft in the unobstructed state. Blood serum tests may be helpful to identify conditions such as diabetes, infection, heavy metal exposure, nutritional deficiency (e.g., vitamins B6 and B12), allergy, and thyroid, liver, and kidney disease.

As with other sensory disorders, quantitative sensory testing is advised. Self-reports of patients can be misleading, and a number who complain of chemosensory dysfunction have normal function for their age and sex. Quantitative smell and taste testing provides valid information for worker’s compensation and other legal claims as well as a way to assess treatment interventions accurately. A number of standardized olfactory and taste tests are commercially available. Most evaluate the ability of patients to detect and identify odors or tastes. For example, the most widely used of these tests, the 40-item University of Pennsylvania Smell Identification Test (UPSIT), employs norms based on nearly 4000 normal subjects. A determination is made of both absolute dysfunction (i.e., mild loss, moderate loss, severe loss, total loss, probable malingering) and relative dysfunction (percentile rank for age and sex). Although electrophysiologic testing is available at some smell and taste centers (e.g., odor event-related potentials), such tests require complex stimulus presentation and recording equipment and rarely provide additional diagnostic information. In addition to electrogustometers, commercial chemical taste tests are now available. Most employ filter paper strips impregnated with tastants, so no stimulus preparation is required. Like the UPSIT, these tests have published norms for establishing the degree of dysfunction.


Because of the various mechanisms by which olfactory and gustatory disturbance can occur, management of patients tends to be condition-specific. For example, patients with hypothyroidism, diabetes, or infections need to be given specific treatments to correct the underlying process adversely influencing chemoreception. For most patients who present primarily with obstructive/transport loss affecting the nasal and paranasal regions (e.g., allergic rhinitis, polyposis, intranasal neoplasms, nasal deviations), medical and/or surgical intervention is often beneficial. Antifungal and antibiotic treatments may reverse taste problems secondary to candidiasis or other oral infections. Chlorohexidine mouthwash mitigates some salty or bitter dysgeusias, conceivably as a result of its strong positive charge. Excessive dryness of the oral mucosa is a problem with many medications and conditions, and artificial saliva (e.g., Xerolube) or oral pilocarpine treatments may prove beneficial. Other methods to improve salivary flow include the use of mints, lozenges, or sugarless gum. Flavor enhancers may make food more palatable (e.g., monosodium glutamate), but caution is advised to avoid overusing ingredients containing sodium or sugar, particularly in circumstances in which a patient also has underlying hypertension or diabetes. Medications that induce distortions of taste often can be discontinued and replaced with other types of medications or modes of therapy. As mentioned earlier, pharmacologic agents result in taste disturbances much more frequently than smell disturbances, and over 250 medications have been reported to alter the sense of taste. Many drug-related effects are long-lasting and are not reversed by short-term drug discontinuance.

A study of endoscopic sinus surgery in patients with chronic rhinosinusitis and hyposmia revealed that patients with severe olfactory dysfunction before the surgery had a more dramatic and sustained improvement over time compared with patients with more mild olfactory dysfunction before intervention. In the case of intranasal and sinus-related inflammatory conditions such as those seen with allergy, viruses, and traumas, the use of intranasal or systemic glucocorticoids may be helpful. One common approach is a short course of oral prednisone, typically 60 mg daily for 4 days and then tapered by 10 mg daily. The utility of restoring olfaction with either topical or systemic glucocorticoids has been studied. Topical intranasal glucocorticoids were less effective in general than systemic glucocorticoids; however, nasal steroid administration techniques were not analyzed. Intranasal glucocorticoids are more effective if administered in Moffett’s position (head in the inverted position such as over the edge of the bed with the bridge of the nose perpendicular to the floor). After head trauma, an initial trial of glucocorticoids may help reduce local edema and the potential deleterious deposition of scar tissue around olfactory fila at the level of the cribriform plate.

Treatments are limited for patients with chemosensory loss or primary injury to neural pathways. Nonetheless, spontaneous recovery can occur. In a follow-up study of 542 patients presenting with smell loss from a variety of causes, modest improvement occurred over an average period of 4 years in about half the participants. However, only 11% of the anosmic and 23% of the hyposmic patients regained normal age-related function. Interestingly, the amount of dysfunction present at the time of presentation, not etiology, was the best predictor of prognosis. Other predictors were the patient’s age and the time between the onset of dysfunction and initial testing.

A nonblinded study reported that patients with hyposmia may benefit from smelling strong odors (e.g., eucalyptol, citronella, eugenol, and phenyl ethyl alcohol) before going to bed and immediately upon awaking each day over the course of several months. The rationale for this approach comes from animal studies demonstrating that prolonged exposure to odorants can induce increased neural activity within the olfactory bulb. α-Lipoic acid (200 mg two or three times daily), an essential cofactor for many enzyme complexes with possible antioxidant effects, has been reported to be beneficial in mitigating smell loss after viral infection of the upper respiratory tract, although double-blind studies are needed to confirm this observation. This agent has also been suggested to be useful in some cases of hypogeusia and burning mouth syndrome.

The use of zinc and vitamin A in treating olfactory disturbances is controversial; not much benefit is obtained beyond replenishing established deficiencies. However, zinc improves taste function secondary to hepatic deficiencies, and retinoids (bioactive vitamin A derivatives) are known to play an essential role in the survival of olfactory neurons. One protocol in which zinc was infused with chemotherapy treatments suggested a possible protective effect against developing taste impairment. Diseases of the alimentary tract can not only influence chemoreceptive function but occasionally influence B12absorption. This can result in a relative deficiency of B12, theoretically contributing to olfactory nerve disturbance. B2 (riboflavin) and magnesium supplements are reported in the alternative medicine literature to aid in the management of migraine headaches that may be associated with smell dysfunction.

A number of medicines have been reported to ameliorate olfactory symptoms, although strong scientific evidence for efficacy is generally lacking. A report that theophylline improved smell function was not double-blinded and lacked a control group, failing to take into account that some meaningful improvement occurs without treatment. Indeed, the percentage of patients reported to be responsive to the treatment was about the same as that noted by others to show spontaneous improvement over a similar time period (~50%). Anti-epileptics and some antidepressants (e.g., amitriptyline) have been used to treat dysosmias and smell distortions, particularly after head trauma. Ironically, amitriptyline is also frequently on the list of medications that can ultimately distort smell and taste function, possibly from its anticholinergic effects. The use of donepezil (an acetylcholinesterase inhibitor) in AD may result in improvements in smell identification measures that correlate with overall clinician-based impressions of change scales (Clinician Interview Based Impression of Severity [CIBIC]-plus). Smell identification function could become a useful measure to assess overall treatment response with this medication.

A major and often overlooked element of therapy comes from chemosensory testing itself. Confirmation or lack of confirmation of loss is beneficial to patients who come to believe, in light of unsupportive family members and medical providers, that they may be “crazy.” In cases in which the loss is minor, patients can be informed of the likelihood of a more positive prognosis. Importantly, quantitative testing places the patient’s problem into overall perspective. Thus, it is often therapeutic for an older person to know that although his or her smell function is not what it used to be, it still falls above the average of his or her peer group. Without testing, many such patients are simply told they are getting old and nothing can be done for them, leading in some cases to depression and decreased self-esteem.