Changes of olfactory performance during the process of aging – Psychophysical testing and its relevance in the fight against malnutrition

Abstract

Objectives

Olfactory performance plays a vital role in several aspects of everyday life. A decrease in olfactory function leading to reduced appetite and inadequate food intake could contribute to higher risk of malnutrition and has a significant impact on quality of life. Early detection of olfactory impairment may help to maintain normal, or — through early training — to improve olfactory function and could thus contribute to the prevention of malnutrition. This review summarizes recent findings on the association between decreased olfactory performance in older adults, its implication for overall health and discusses issues regarding the assessment of olfactory performance in older people.

Methods

A non-systematic literature search using databases (PubMed, ScienceDirect, Google Scholar) was conducted up to March 2016 to review recent findings on the topics of olfactory impairment, age, malnutrition and issues in psychophysical testing.

Results

Although the association of olfactory impairment and malnutrition is widely accepted strong evidence is scarce. This could be because existing psychophysiological olfactory performance tests are not suitable and inefficient for older adults or people with cognitive impairment due to their required time and concentration. Since prevalence of olfactory impairment remains poorly documented in European regions and varies from 13.9% to >60% depending on age and due to considerable methodological diversity in studies there is need to develop rapid and efficient screening tools which are suitable for this target group.

Conclusion

Due to numerous causes and health consequences of olfactory impairment in older people, early detection of olfactory malfunction by rapid, inexpensive but valid tests, which could be used by geriatricians and other healthcare professionals, might be useful in patient counselling by identifying individuals at nutritional risk.

Introduction

Aging is often accompanied with an impairment of sensory perception due to various factors – vision and hearing loss are the most recognised, but impairment of gustatory, olfactory and texture perception also often occur (1). The sensory appeal of food may be one of the key drivers in food selection in older persons (2). Functional changes with age in sensory perception might have an impact on food preferences of older adults (3).

Olfaction plays an important role in food intake, harm avoidance e.g. via detection of spoiled food, and social life (4). The loss or reduction of olfaction could lead to reduced appetite, inadequate food intake (5) and a significant reduction in quality of life (4, 6). These issues may be partly responsible for higher risk of malnutrition and the so called “Anorexia of Ageing” (3) since poor appetite has been reported to be a single determinant with a strong association with malnutrition (7). Malnutrition leads to unintended weight loss and loss of muscle mass and bone mass and results in functional decline and diminished cognitive function. There is growing evidence that a strong association between inadequate food intake in the elders and risk of frailty exists (8). Clegg et al. (9) named frailty as one of the most problematic conditions in aging due to the numerous health consequences like higher mortality rates and the financial burden on health care systems.

Olfactory impairment (OI) is expressed with increase in the threshold for odor detection, lower perceived odor intensity, and a decreased ability to identify food-related odors (10). It has been found that training of the senses may be a feasible strategy against malnutrition maintaining normal function of smell and taste and thus enjoyment of food: olfactory training might improve olfactory function in individuals with hyposmia (11), with Parkinson's disease (12) or help to maintain normal olfactory function in older people (13). Early detection of OI may therefore be crucial for the prevention of malnutrition, due to an earlier onset of training. Many existing olfactory tests might not be suitable for older adults or people with cognitive impairment (e.g. concerning time span and concentration), which may be impaired by disease (such as Alzheimer's) or other factors (14).

The aim of the current review is to discuss the current evidence on the effects of age on olfactory performance and its consequences for human health in particular malnutrition. In addition issues in psychophysical testing of olfactory performance in older adults will be discussed.

Olfactory system - Anatomy and normal physiological function

The olfactory system plays an important role in human behaviour and emotional experience and olfaction is a complex physiologic process concerted by peripheral and central components (15). The olfactory epithelium is located on the roof of the nasal cavity and enables through its specific location odor perception via ortho- and retronasal airflows (16) – according to Rozin (17) smell has therefore a ‘dual nature' which makes it unique among the senses. This ‘duality' means that volatile compounds enter the nose via the nostrils through breathing/sniffing or via the nasopharynx when they are released during eating and drinking (18). Odor molecules generate complex sensory signals by binding to multiple receptors. It is thought that each odor stimulates a specific pattern of receptors building the characteristic smell and as a result the brain can discriminate between specific scents (19). Sheperd (20) described these patterns as ‘odor images'.

As opposed to taste perception which is limited to a few chemical substances which are bond to single receptors, odor perception covers an estimated 10000 odors (19) and the structure of about 1000 olfactory receptors is encoded in the human genome (20), although only 380 of them are expressed and located in the cilia of olfactory receptor neurons in olfactory epithelium (16). Although each olfactory receptor neuron expresses a specific receptor type, it is known that single odor molecules can bind to different receptor types which implicates that some odors are perceived more intense than others (16).

Stimulation of olfactory receptor neurons results in a cascade of neuronal signals which are sent via the olfactory bulb to brain structures involved in processing of odor information: Although cerebral processing is not fully understood, among other structures the orbitofrontal cortex plays a major role in odor perception (14, 20). The involvement of the hippocampus and amygdala in odor information processing partly explains the emotional character of some odors and the odor-related recalling of memories (16). It has been shown that cannabinoid type-1 (CB1) receptors are present in the olfactory bulb and Wang et al. (21, 22) argued that an activation of CB1 receptors may lead to an increase of overall sensitivity of the olfactory bulb to sensory inputs. Recently Soria-Gomez et al. (23) found that CB1 receptors promote food intake in fasted mice by increasing odor detection. This might have some implications for future research, e.g. increasing food intake by improving odor sensitivity.

Age-related changes in olfactory performance

It is widely accepted that olfactory performance decreases during the process of aging (= presbyosmia) (14, 16, 24, 25). Olfactory sensitivity depends on age and gender, whereas women are superior to men in all aspects of olfactory sensitivity (16, 25, 26). The degree of presbyosmia can also be seen as an expression of overall health (16) because it has been reported that people who aged without any medication seem to have normal smell thresholds (25) while the condition of OI is a potential adverse effect of several types of medications and diseases (16). The age-related loss of smell is not always noticed (16) and people who are affected by unrecognized OI often complain of taste loss and do not reflect the importance of smell on overall flavour perception (14). To emphasise this issue, self-reported estimates of OI were lower than objective measurements from psychophysiological tests (see below).

Data on the prevalence of OI in older adults varies greatly. Karpa et al. (27) found OI in 27.0% of a large group of older people (n =1636, age >60years). In a study with patients of a geriatric day hospital (n =191, mean age 79.6 ± 6.3 years) the prevalence of hyposmia was 39.3% and anosmic conditions were found in 31.9% (28). Toussaint et al. (29) recently assessed the olfactory performance in a sample of vital older persons (n =345, mean age 67.1 years) and a geriatric sample (n = 191, mean age 80.9 years) and found that only 2% of the vital older adults were anosmic, compared with 46% of the geriatric participants. Few studies examined the prevalence of OI in larger cohorts of different ages. The large population based study (n =2491, age 53–97 years) from Murphy et al. (30) revealed a prevalence of OI in 24.5% of all participants, and the prevalence increased to 62.5% among those aged 80–97 years. The large cohort study (n =2838, age 21–84 years) from Schubert et al. (31) found that OI occurred in 3.8% of general population whereas prevalence increased to 13.9% among those >65 years. In a large German-based study (n =1312, age 25-75 years) 3.6% of all participants were anosmic and another 18.3% had a considerable OI (32).

Interestingly self-reported estimates of OI were lower than objective measurements from psychophysical tests ranging from 1.4% (33) to 15.3% (34). Doty and Kamath (14) noted that many older people when they are recognizing OI tend to over- or underestimate its magnitude. Doets and Kremer (24) argued that this large incoherence in prevalence estimates between cohorts occurs due to differences in study population (sample size, age distribution) and study design (methodology for assessing quality of olfactory function, Table 1). Table 1 illustrates that olfactory function is assessed by identification tasks. Due to the higher prevalence of cognitive impairment the usual testing may not always be appropriate for this age group (14). In addition people with cognitive impairments are excluded which implicates that prevalence data often may not be representative and underlines the need for a rapid valuable method for olfactory assessment.

Table 1.

Used methodology to assess quality of olfactory function in studies assessing prevalence of OI

Authors/Study	Method
Karpa et al., 2010 (27)	San Diego Odor Identification Test (SDOIT)
Smoliner et al., 2013 (28)	Sniffin’ Sticks—Screening 12 test
Toussaint et al., 2015 (29)	a) Sniffin’ Sticks—Screening 12 test in the geriatric sample
	b) The extended version of the Sniffin’ Sticks (TDI-score) was used for the vital older adults.
Murphy et al, 2002 (30)	SDOIT and self-report
Schubert et al., 2012 (31)	SDOIT
Vennemann et al., 2008 (32)	Sniffin’ Sticks—Screening 12 test
Hoffman et al., 1998 (33)	Self-reported odor sensitivity; questionnaire
Nordin et al., 2004 (34)	Self-reported odor sensitivity; structured interview
Brämerson et al. 2004 (44)	Scandinavian Odor Identification Test (SOIT); similar to the Sniffin Sticks Screening 16 test
Pinto et al., 2015 (46)	5-item version of the Sniffin Sticks - Identification test 16

In general OI may be caused by different anatomical and (patho-) physiological changes which are amongst others changes in olfactory epithelium, decreased selectivity of olfactory receptors and decrease of olfactory bulb volume (24) but also closure of the foramina of the cribriform plate (35). For example olfactory cells have the ability to regenerate continuously due to the fact that olfactory receptor neurons - although protected by a mucous layer - are directly exposed to environmental conditions (e.g. temperature, toxins) (16, 36). This ability to regenerate diminishes during the process of aging which may contribute to presbyosmia (14, 36, 37). But also changes in brain structures involved in odor information processing occur during aging (24). Several studies with functional magnetic resonance imaging (fMRI) showed that the degree of activation in brain regions involved in olfactory processing is significantly lower in older persons than in younger ones (38, 39, 40).

It has been found that the correct identification of pleasant odors significantly decreased in the elders as compared to younger persons, whereas identification of unpleasant odors was similar between the two groups (41) – this could be due to danger avoidance (e.g. unpleasant odors from spoiled food) and the degree of OI depends therefore on the nature of the odor (24). Quality and quantity of smell disorders – regardless of age – ranging from lack of ability to smell to the perception of odors in the absence of one (Table 2) (42, 43).

Table 2.

Qualitative and quantitative types of smell disorders [adapted from 42, 43]

Quantitative smelling disorders		Qualitative smell disorders
Anosmia	Complete anosmia: absolute loss of olfactory function	Parosmia	Qualitative “wrong” perception of odors
	Functional anosmia: severe limitation of olfactory function		‘Cacosmia’ = perception of a bad smell with or without an odorant stimulus
	Partial anosmia: greatly reduced sensitivity to a particular odor; usually not viewed as pathological		‘Euosmia’ = a pleasant parosmia to selected odorants
Hyposmia	Reduced ability to smell	Phantosmia	Perception of odors in the absence of an odour (hallucination)
Hyperosmia	Enhanced ability to smell

For an in depth insight of anatomy and function of olfaction and its age-related changes see reviews by Doty and Kamath (14) and Mobley et al. (36).

Olfaction and human health

The following section will discuss the issue that age-related OI might also be accompanied by disease and non-disease related factors with a direct or indirect impact on human health, such as deleterious effects on flavour perception, enjoyment of food and quality of life (6, 43). Several risk factors associated with OI have been identified by observational studies (Table 3). It can be seen that ‘age' is one of the strongest risk factors besides being ‘male' and having a low socio-economic status.

Table 3.

Factors associated with impaired olfactory functioning in observational studies^a [adapted from (24)]

Authors/Study	Age	Male gender	Low SESb	Smoking	Nasal congestion
Murphy et al., 2002 (30)	x	x		x	x
Brämerson et al., 2004 (44)	x	x			x
Karpa et al., 2010 (27)	x	x		x	x
Boesveldt et al., 2011 (45)	x	x	x
Schubert et al., 2012 (31)	x	x	x	x (only women)	x
Vennemann et al., 2008 (32)			x	x
Pinto et al., 2015 (46)	x	x

^aFactors were significantly associated with impaired olfactory functioning in multivariate models

^bSES = socio-economic status; includes low education

Beside risk factors genetic factors may also contribute to large inter-individual variation of degree of OI (24, 47). Several diseases can cause or promote progression of OI: Chronic rhinosinusitis, nasal polyposis, allergic rhinitis and respiratory infections but also head injuries (42). Recently it has also been found that subclinical atherosclerosis at a younger age may also be a risk factor for the development of OI in older age (48). Another point which should be taken into account is the condition of OI as a potential adverse effect of several types of medications (16) – This is of special interest and might have some implications due to polypharmacy practices in older age (49). For example antibiotics, antihypertonic drugs, antidepressants (16) but also chemotherapeutic drugs (50) can induce smell disorders.

There is some evidence that OI may be an early sign of neurodegenerative diseases such as Parkinson's disease (51) and Alzheimer's disease (27, 52). Olfactory deficits have been observed in 85-95% of tested patients with Alzheimer's disease (52). It has been proposed that OI might be used as a sensitive biomarker for early detection of motor or cognitive disturbances (24). A systematic review by Sun et al. (53) however came to the conclusion that rigorously designed longitudinal cohort studies are necessary to clarify the value of olfactory testing in predicting the onset of Alzheimer. A recent study in Parkinson's Disease (PD) outpatients (n =118, mean age 68.5 ± 9.0 years) showed an association between low olfactory function at time of PD diagnosis and a reduced MMSE (Mini- Mental-State-Examination) score after three years of diagnosis (54). According to Hummel et al. (16), the degree of smell loss seems to be very high in idiopathic Parkinson's disease, Alzheimer's disease, Lewy body dementia, familial Parkinson's disease PARK 8 but somewhat lower in motor neuron diseases. Interestingly it has been found that people with mild cognitive impairment have greater smell loss when they convert to Alzheimer than people who do not convert (14). Velayudhan et al. (55) recently showed in a small population (n =64), that OI might be also a marker for the progression of Alzheimer's disease.

Beside these associations, OI also have some consequences for human health. It has been proposed olfactory function might be a marker of overall health (16, 29). A recent study demonstrated in a sample of home-dwelling adults (n =1436, age 57-85 years) that OI predicts 5-year mortality – the severity of impairment increased the likelihood of death (56). Keller and Malaspina (6) evaluated in a study (n =1000) the influence of OI on different aspects of quality of life. They found that smell loss-induced social isolation and smell loss-induced anhedonia have detrimental effects on the quality of life of these persons. Croy et al. (4) reported that about one-third of people with OI have severe problems in everyday life and a significant reduction in quality of life. They analysed the results from eight studies (n =1394) regarding the influence of OI on different aspects of everyday life. They found that the majority of negative consequences of OI were associated with lowered ability to enjoy food, decreased appetite and food intake and difficulties in preparing food. This is in line with previous reports that OI could lead to reduced appetite and inadequate food intake (5) and might also have an impact on food choices and results in decreased food variety (57).

Only few studies investigated these possible associations in detail. A recent study found that older adults with moderate/ severe impairment (n =89) versus normal olfactory function (n =468) had poorer diet quality over 5 years whereas this association was significant in women but not in men (58). A previous study by Ådén et al. (51) found that persons with impaired olfaction (n =87) had lower intakes of protein, folate, magnesium, and phosphorous than controls. In another study hyposmic and anosmic patients self-reported lower enjoyment of food and a change in eating habits (28). These issues may be partly responsible for higher risk of malnutrition (3) since poor appetite has been reported to be a single determinant with a strong association with malnutrition (7). This fact is widely accepted but Smoliner et al. (28) failed to establish an association between malnutrition and OI in geriatric patients (n =191, mean age 79.6 ± 6.3 years). In contradiction Toussaint et al. (29) demonstrated a significant association (P =0.015) between olfactory function and nutritional status in a geriatric sample (n =138, mean age 80.9 years).

Issues in the assessment of olfactory performance in older people

It is supposed that the presented contradictory findings occurred due to the fact that the existing psychophysical sensory tests (e.g. tests of odour detection, identification, and discrimination) might not be suitable for older adults or people with cognitive impairment (1, 14). As shown in Table 1 many studies assessed olfactory function by odor identification tasks and Doets and Kremer (24) speculated that the observed findings are more likely word- and name-finding problems than actual OI.

Several types of commercially available tests to assess olfactory performance have been developed over the years - all of them designed to detect hyposmia or anosmia (59). Principally they can be divided into three basic groups (60) whereas many of the tests combine 2 or 3 of the following groups, like the Connecticut Chemosensory Clinical Research Center Test - CCCRC (61), or the Sniffin' StickTM test (62):

1) Odor identification tests: The subject has to identify the odor (presented in a concentration much above threshold). Many of them are established as forced-choice tests like the University of Pennsylvania Smell Identification Test (UPSIT) (63) or the Scandinavian Odor Identification Test (SOIT) (64).

2) Odor discrimination tests where the participant has to discriminate different odors. An odor discrimination task is included in the Sniffin' SticksTM test battery (62).

3) Odor detection threshold. These tests assess the smell sensitivity by determining the minimum perceived odor intensity, like in the tests of Eloit and Trotier (65) and Stamps et al. (66) and Kern et al. (67). Olfactory threshold tests are often presented to the participant as a series of different concentrations of an odorant via sniff bottles, squeeze bottles, felt-tip pens or olfactometers (14).

Existing olfactory tests are not only often too complex but also too expensive and too time-consuming to be of use in clinical routine (68). For example the alternative forced choice (AFC) approach, where participants have to select an odorant against blank samples, is one of the most common approaches (69) whereas the ascending 3-AFC is recognised as standard for smell threshold testing (70). This 3-AFC approach with 2 blank samples is used in the Sniffin' SticksTM olfactory test kit (62). This method requires presentation of several samples in each set. Such multiple replications are likely to cause sensory and cognitive fatigue which may be more pronounced in elderly populations (69). The 3-AFC method including 16 dilutions used in the Sniffin' Sticks test requires significant time (up to 20-30 minutes) and training to administer (67).

Due to these aspects older people and/or people with cognitive impairments are often excluded from testing which results in exclusion of a target group who is often at nutritional risk. This underlines the need for a rapid valuable method for olfactory assessment. To sum up it might be one promising alternative to assess olfactory performance in older or cognitive impaired people by assessing olfactory sensitivity in terms of olfactory thresholds. Discrimination or identification tasks or combined tests might not be suitable for this population. Due to these facts and the numerous health consequences of impaired olfaction it has been proposed that an early detection of olfactory malfunction by rapid, inexpensive but valid tests used by dieticians, geriatricians and other healthcare professionals could be useful in patient counselling identifying those who are at nutritional risk (58).

Conclusion

Olfactory performance plays a vital role in several aspects of everyday life. Decrements of olfaction during aging occur due to disease and non-disease-related factors and have a great impact on quality of life and lead to reduced appetite, and inadequate food intake. It has been reported that poor appetite might be a single determinant with a strong association with malnutrition (7). This fact is widely accepted although there is still a lack of strong evidence between the association of OI and malnutrition due to methodological issues in studies and contradictory results.

It has been discussed that training of the senses may be a feasible strategy against malnutrition maintaining normal function of smell and taste and thus enjoyment of food. Due to the numerous health consequences of OI in older people it has been proposed that an early detection of olfactory malfunction by rapid, inexpensive but valid tests used by dieticians, geriatricians and other healthcare professionals could be useful in patient counselling identifying those who are at nutritional risk (58). Early detection of OI may therefore be crucial for the prevention of malnutrition, due to an earlier onset of training. Since existing olfactory performance tests are not suitable for older adults and or people with cognitive impairment, there is need for development of validated screening tools which are suitable for this target group. To sum up it might be one promising alternative to assess olfactory performance in cognitive impaired people by assessing olfactory sensitivity in terms of olfactory thresholds. Discrimination or identification tasks or combined tests might not be suitable for this population. Therefore future studies should focus on establishing a rapid method addressing the above named issues before assessing olfactory performance in older people.

Conflict of Interest: No conflict of interest to declare.

Ethical standard: This article is not a specific study involving human or animal subject research.