Journal of the Neurological Sciences 349 (2015) 179–184

Contents lists available at ScienceDirect

Journal of the Neurological Sciences

j ourna l homepage: www.e lsev ie r .com/ locate / jns

Olfactory identification in amnestic and non-amnestic mild cognitiveimpairment and its neuropsychological correlates

Martin Vyhnalek a,b,⁎, Hana Magerova b, Ross Andel c, Tomas Nikolai a,b, Alexandra Kadlecova a,b,Jan Laczo a,b, Jakub Hort a,b

a International Clinical Research Center, St. Anne's University Hospital Brno, Brno, Czech Republicb Memory Clinic, Department of Neurology, Charles University in Prague, 2nd Faculty of Medicine and Motol University Hospital, Prague, Czech Republicc School of Aging Studies, University of South Florida, Tampa, FL, USA

Abbreviations:MHST,Motol Hospital Smell Test.⁎ Corresponding author at: Memory Clinic, Depar

University in Prague, 2nd Faculty of Medicine and Motol150 06 Prague 5, Czech Republic. Tel.: +420 22443 6801;

E-mail address: martin.vyhnalek@fnmotol.cz (M. Vyhn

http://dx.doi.org/10.1016/j.jns.2015.01.0140022-510X/© 2015 Published by Elsevier B.V.

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 30 June 2014Received in revised form 7 January 2015Accepted 8 January 2015Available online 14 January 2015

Keywords:Mild cognitive impairmentOlfaction disordersSmellAlzheimer's diseaseCognitionMemory

Background: Olfactory identification impairment in amnestic mild cognitive impairment (aMCI) patients is welldocumented and considered to be caused by underlying Alzheimer's disease (AD) pathology, contrasting withless clear evidence in non-amnestic MCI (naMCI). The aimwas to (a) compare the degree of olfactory identifica-tion dysfunction in aMCI, naMCI, controls and mild AD dementia and (b) assess the relation between olfactoryidentification and cognitive performance in aMCI compared to naMCI.Methods: 75 patients with aMCI and 32 with naMCI, 26 patients with mild AD and 27 controls underwent themultiple choice olfactory identification Motol Hospital Smell Test with 18 different odors together with a com-prehensive neuropsychological examination.Results: Controlling for age and gender, patients with aMCI and naMCI did not differ significantly in olfactoryidentification and both performed significantly worse than controls (p b 0.001), albeit also better than patientswith mild AD (p b .001). In the aMCI group, higher scores on MMSE, verbal and non-verbal memory and visuo-spatial tests were significantly related to better olfactory identification ability. Conversely, no cognitive measure

was significantly related to olfactory performance in naMCI.Conclusion:Olfactory identification is similarly impaired in aMCI and naMCI. Olfactory impairment is proportion-al to cognitive impairment in aMCI but not in naMCI.

© 2015 Published by Elsevier B.V.

1. Introduction

Olfactory impairment has been demonstrated in Alzheimer's disease(AD) [10,20,36], presumably as a consequence of early degeneration of ol-factory bulb, olfactory nerve and olfactory cortex situated predominantlyin themedial temporal lobe [1,2,3]. Among the threemajor types of olfac-tory ability (detection, discrimination and identification), the olfactoryidentification is impaired earlier compared to the olfactory detection inAD patients [50]. So far, most studies with cognitively impaired patientshave investigated only olfactory identification, which strongly correlateswith olfactory threshold and is easier to test [9, 11,12]. The olfactory iden-tification deficit seems to be specifically linked mainly to the temporallobe including the amygdala, hippocampus and parahippocampal gyrus,

tment of Neurology, CharlesUniversity Hospital, V Úvalu 84,fax: +420 22443 6820.alek).

but even anterior temporal damage is sufficient to provoke olfactory iden-tification impairment [21].

Dementia syndrome in AD and in other degenerative disorders is al-most always preceded by mild cognitive impairment (MCI) syndromein which the patients have objective cognitive impairment on neuro-psychological examination but do not show substantial deficits in activ-ities of daily living [43,45].

MCI patients with objective memory impairment are labeledamnestic MCI (aMCI). These patients progress mainly to AD dementia[15,42,44], aMCI patients with isolatedmemory (amnestic) impairmentare labeled as the single domain aMCI (aMCIsd), and aMCI patientswithan additional impairment in the other cognitive domains beyondmem-ory (e.g., executive impairment, language, visuospatial) are called mul-tiple domain aMCI (aMCImd) [42,44].

Among aMCI patients, olfactory identification impairment has beendemonstrated in a number of studies with a cross sectional design [8,14,23,59]. To our best knowledge only one study also investigatedother olfactory modalities beyond olfactory identification in patientswith aMCI and it reported impaired olfactory detection and identifica-tion. There was also an olfactory discrimination deficit but that was

180 M. Vyhnalek et al. / Journal of the Neurological Sciences 349 (2015) 179–184

accounted for by an abnormal olfactory threshold [8]. Finally, olfactoryimpairment in aMCI represents a risk factor for subsequent cognitivedecline and conversion to AD dementia, as was demonstrated by somelongitudinal studies [6,53,54].

MCI patients with normal memory function but with cognitive im-pairment in non-memory domains (e.g. executive functions, visuospa-tial functions, language) are classified as non-amnestic MCI (naMCI).Patientswith naMCImay convertmore frequently to non-ADdementias[42,44], especially to frontotemporal lobar degeneration (FTLD),Parkinson disease and Lewy body disease (LBD) inwhich olfactory iden-tification impairment is frequently found [20,29,36,46].

However, there have been only a few studies investigating olfactoryfunctions in naMCI, which report inconsistent results [7,23,59]. Consid-ering the evidence of olfactory identification impairment in these non-AD dementias, which are typically preceded by naMCI subtype, wewould expect large olfactory impairment in patients with naMCI thatmay resemble that of patients with aMCI.

Association between olfactory and cognitive impairment in AD andMCI patients is not yet fully understood. The association betweenmem-ory and olfactory identification performance was demonstrated only ina mixed cohort of healthy elderly and MCI patients and in a mixed co-hort ofMCI and dementia patients [7,8,23,48,60]. To our best knowledgethe relation between olfactory identification and cognitive performancein MCI, specifically in the amnestic versus non-amnestic MCI subtypes,has not been assessed.

Because of anatomical and functional proximity of brain areas re-sponsible for memory and olfaction (both situated predominantly inthe medial temporal lobe), we would expect proportional degree of ol-faction and memory impairment in pre-dementia and dementia stagesof AD.

To build on previous research, the aim of this study was to:

1) Compare olfactory identification deficit in patients with aMCI vs.naMCI, as well as aMCI and naMCI vs. controls and mild AD.

2) Compare olfactory identification between patients with single vs.multiple domain aMCI.

3) Analyze the association between olfactory identification and cogni-tive performance in aMCI vs. naMCI.

2. Methods

2.1. Subjects

All subjects were recruited from referrals to the Memory DisordersClinic at Motol Hospital, an affiliate of Charles University in Prague,and signed an informed consent approved by the local ethics commit-tee. They underwent standard protocol which consisted of magneticresonance imaging, neurological, medical and laboratory evaluation,questionnaires and complex neuropsychological assessmentmentionedbelow. A total of 160 participants were included in the analyses.

The MCI group subjects met the revised Petersen's criteria for MCI[43]. The participants had cognitive complaints reported by themselvesor by their caregiver, they were impaired on objective cognitive tasks,not dementedwith largely intact functional activities with CDR (ClinicalDementia Rating scale) of 0.5.

The MCI patients (n = 107) were further classified into the follow-ing groups: (a) patients with naMCI (n = 32) or (b) patients with aMCI(n = 75).

All aMCI patients hadmemory complaints and scoredmore than 1.5of standard deviation (SD) lower than age matched controls in verbalmemory tests (Auditory Verbal Learning Test [AVLT] and EnhancedCued Recall [ECR] test).

Patients with naMCI had impairment only in the non-memory cog-nitive domains, manifesting in all patients as attentional-executive def-icit, in addition 5 patients suffered from language deficit and 3 patientsfrom visuospatial deficit.

Of the 75 broadly defined aMCI cases, only 18 had pure amnesia (allother tests were within the normal range— aMCIsd), while the remain-ing participants, labeled as aMCImd, suffered from other subtle seman-tic, visuoconstructive or attention-executive function deficits (morethan 1.5 SD), or both.

Two other groups were used in this study:

–The control group (n = 27) Reported no cognitive problems, whichwas subsequently confirmed by neuro-psychological testing and a CDR scoreof 0.0. They were recruited from staffand patient's relatives andwere selectedto be as similar as possible to the othergroups in terms of age, education andgender.

–Themild probable ADgroup (n=26) Met theDiagnostic and Statistical

Manual of Mental Disorders IV criteria fordementia and the National Institute ofNeurological and Communicative Disor-ders and Stroke and Alzheimer Diseaseand Related Disorders Association criteriafor probable AD [32]. Patients withdementia had an impairment of memoryand another cognitive domain, impairedfunctional activities, and their CDRwas 1.0 or higher. They didn't have signif-icant vascular impairment on brain MRI(Fazekas scale 0 or 1) [16]. All AD patientswere on a stable dose of cholinesteraseinhibitors for at least 3 months.

2.2. Exclusion criteria

Subjects with history of smoking in past 10 years, acute or chronicrhinitis or another ORLdiagnosis causingpotential hyposmia or subjectswith preexisting hyposmia of another etiology (posttraumatic, profes-sional exposure to toxics) and subjects with depression (scoring morethan 5 on Geriatric Depression Scale) were excluded from the study.

2.3. Neuropsychological assessment

Following neuropsychological tests and questionnaires were ad-ministered: Clinical Dementia Rating (CDR) [38], and 15-item GeriatricDepression Scale [62]. Additional neuropsychological testing includedthe Mini Mental State Examination (MMSE) [17], verbal memorytests: Auditory Verbal Learning Test (AVLT) [51], 16-item versionGrober and Buschke's Test with Enhanced Cued Recall (ECR) procedure[19], nonverbal memory test and visuospatial test: Rey–OsterriethComplex Figure Test (ROCF) copy and reproduction [37], workingmemory attention: digit span (DS) forward and reversed [58], TrailMaking Tests (TMT) A [47], executive functions: Category Fluency andInitial Letter Fluency (FAS) tests [26] and TMT B.

2.4. Smell identification assessment

In all subjects, smell identification was assessed using the MotolHospital Smell Test (MHST) — a multiple-choice smell identificationtest developed and evaluated at our memory clinic [29,30]. MHST iscomposed of 18 odors verywell known among elderly Czechpopulation(pine-tree, peach, lemon, rose, cherry, grapefruit, clove, lavender, pep-permint, orange, cinnamon, vanilla, coffee, honey, lilac, strawberry,black currant, rum). Odors are presented as essential oils in specialphials in the amount of 200 μl to both nostrils simultaneously. The es-sential oils are replaced every 2months in the phials in order to preventdegradation of the odor. After sniffing the odor, subjects are asked to

181M. Vyhnalek et al. / Journal of the Neurological Sciences 349 (2015) 179–184

select one right answer from the four choices list. It was previously ap-proved that MHST results correlate with the results of the University ofPennsylvania Smell Identification Test (r= .68, p b .001) [11,30]. To ex-clude non-olfactory components of the smell identification failure, thePicture Identification Test (PIT) was used [56].

2.5. Statistical analysis

All analyses were performed using SAS statistical software (SAS In-stitute, Cary, NC). All variableswere examined for normality of distribu-tion. No variable presented significant deviation from normal scoredistribution. Therefore, values are presented as average ± standard de-viation or as percent. Controls were compared to aMCI, naMCI and ADon the MHST using analysis of covariance (ANCOVA). As the group dif-fered in age, education and gender, these variables were entered as co-variates. The post hoc Tukey's Honestly Significant Difference test wasused for group comparison. In an additional analysis, only aMCI patientswere used, allowing for a direct comparison between aMCIsd andaMCImd subgroups. To address possible concerns regarding statisticalpower in the main ANCOVA models, we estimated statistical powerusing SAS procedure POWER, entering covariate-adjusted means andusing weighting to account for differences in sample size across groups.We found that both ANCOVAs had upwards of 90% power to detect anydifferences as statistically significant. In addition, to help interpretbetween-group differences, we also report effect sizes in the form ofCohen's d (covariate-adjusted mean between-group difference overpooled standard deviation) as this statistic is relatively independent ofsample size. Conventionally, Cohen's d is interpreted as small (about0.2), medium (about 0.5) and large (about 0.8 or larger) [5].

In addition, we used ordinary least squares regression to examinethe relationship between olfactory ability and cognitive performanceboth in the entire sample and separately across diagnostic groups, con-trolling for age and gender. The score for each cognitive domain (verbaland nonverbal memory, executive, attention/working memory, visuo-spatial) was expressed as a unit-weighted composite z-score(mean = 0, SD = 1) from the relevant neuropsychological tests tomore easily compare results across regression models. The valuesfrom the TMT A and B tests, which are expressed in seconds to comple-tion,were reversed before the z-scoreswere generated. The significancelevel was set at .05.

3. Results

Demographic and neuropsychological characteristics of the diagnos-tic groups and results of MHST are summarized in Table 1. All fourgroups differed in age (F [3, 157] = 6.15, p = .001), years of education(F [3, 157]=4.65, p b . 004), gender (χ2 [3]=11.53, p b .05), andMMSE(F [3, 157]= 39.0, p b .0001). The subgroups ofMCI did not differ in age,years of education, or gender (F [1, 76] b 1.2, p N .27 in all analyses) butthey differed inMMSE (F [1, 76]=6.72, p= .012), whereby AD patientswere significantly older compared to controls and both aMCI subgroups.

There was a significantmain effect for group as a function of olfacto-ry identificationMHST scores (F [3, 157]=11.17, p b .001). The post hoctest revealed significant impairment in aMCI (Cohen's d= 0.64), naMCI(Cohen's d=0.78) andAD (Cohen's d=1.37) groups compared to con-trols (all ps b .001). In addition olfactory identification was significantlymore impaired in AD group compared to both aMCI and naMCI sub-groups (p b .001). No significant difference was found between aMCIand naMCI groups. Next, we compared olfactory identification in pa-tients with aMCIsd vs. aMCImd, finding that aMCImd patients per-formed significantly worse than those with aMCIsd (F [1, 76] = 7.21,p = .009, Cohen's d = 0.63).

Finally, we used regression analyses to assess the relation between ol-faction and cognitive performance (see Table 2). In the aMCI group,higher MMSE, better visuospatial function, verbal memory, and non-verbal memory were all significantly related to better olfactory ability

regardless of age and gender. The relationship between better atten-tion/working memory, executive function and olfactory abilityapproached significance. No cognitive measure was significantly relatedto olfactory performance in naMCI.

4. Discussion

We assessed olfactory function in a relatively large sample of partic-ipants and considered analytically both diagnostic entities (normal,aMCI, naMCI, AD) and neuropsychological assessment scores. Control-ling for age and gender, we found that (a) olfactory identification im-pairment was similar in amnestic and non-amnestic MCI subgroups,(b) aMCImd patients performed worse than those with aMCIsd, and(c) olfactory identification impairment was related mainly to memoryand visuospatial function in aMCI but has no cognitive correlate innaMCI.

In addition, the study confirms previous findings of olfactory identi-fication impairment early in the course of AD, already at the stage ofmild dementia [20,36] and in individuals with aMCI, an entity which isoften attributed to prodromal AD [8,14,23,59].

Impairment in aMCI subgroupwasmilder than in patientswithmildAD dementia, showing no floor effect for olfactory identification usingMHST in aMCI patients. Analogous results were found in other studywith UPSIT [8].

In a previous study of olfactory identification impairment in aMCIusing a relatively small sample, the authors reported a statisticallynon-significant tendency for patients with aMCImd to be more im-paired than aMCIsd [23]. To our knowledge, the present study is thefirst to show the significant difference between these 2 subgroups.Our findings ofmore severe olfactory impairment in aMCImd comparedto aMCIsd serve as indirect evidence for aMCIsd corresponding tomilder and more localized neuropathological changes compared toaMCImd and it is in agreement with studies confirming higher conver-sion rate to AD in the latter group [31,55].

We found olfactory identification deficits of similar extent in pa-tients with naMCI compared to controls as in patients with aMCI com-pared to controls. This is in agreement with a study using brief 12items test reporting modest impairment in MCI population as a whole(amnestic and non-amnestic together) compared to control populationwith nodifference betweenMCI subtypes [59]. The direct comparison ofdifferent MCI subtypes with controls was, however, not mentioned inthis paper. On the other hand, another study using UPSIT did not findany considerable impairment in olfactory identification in naMCI pa-tients suggesting sparing of olfactory brain in this MCI subtype. Howev-er, naMCI group used in this study performed relatively well on theMMSE scale; the control groupwas recruited from hospitalized patientsand their cognitive status was not confirmed with comprehensive neu-ropsychological testing [23]. Finally, conflicting results have been foundin a large community based multi-ethnic study, in which MCI patientswere again impaired as a whole group in odor identification and aMCIdiffered from cognitively intact elderly but no difference betweennaMCI and controls nor between aMCI and naMCI was detected. How-ever, this study was also community-based, thus using potentially lessimpaired MCI patients compared to population frommemory clinic [7].

There are several plausible explanations for profound olfactory iden-tification impairment in naMCI found in our study. First, as stated above,previous studies have foundprofoundolfactory impairment in FTLD andLBD [46,61] and inconsistent results in VD [13,18]. These dementia dis-orders are considered to be preceded typically by naMCI, so the olfacto-ry impairment in this subgroup does not seem surprising [42,44].Second, a recent publication revealed that in a population of partici-pants 75 years of age, AD was the most common type of dementia fol-lowing not only aMCI but also naMCI subgroup, although the risk ofdeveloping AD dementia was lower in naMCI subgroup compared toaMCI [22]. Similar results were found in previous studies [4,31]. Thus,it is possible that at least in some of our naMCI patients, their olfactory

Table 1Descriptive characteristics of the sample.

aMCI naMCI Controls Mild AD

aMCI total aMCIsd aMCImd

(n = 75) (n = 18) (n = 57) (n = 32) (n = 27) (n = 26)

Demographic characteristicsGender (male/female) 36/39 6/12 30/27 15/17 6/21 6/26Age 72.67 ± 9.07 70.61 ± 8.40 73.32 ± 9.25 69.25 ± 9.87 69.15 ± 8.50 74.2 ± 7.98Education 13.96 ± 3.40 14.56 ± 3.63 13.73 ± 3.32 15.47 ± 2.73 15.22 ± 2.73 12.05 ± 2.97

Test scoresMMSE 26.64 ± 2.76 28.06 ± 1.89 26.19 ± 2.85 28.13 ± 1.98 29.29 ± 7.13 21.15 ± 4.11ECR-IR 5.30 ± 3.25 7.29 ± 3.35 4.70 ± 2.99 7.96 ± 2.57 10.75 ± 2.18 2.28 ± 2.53ECR-TR 13.30 ± 3.50 14.35 ± 3.20 12.98 ± 3.54 15.17 ± 2.74 15.94 ± .250 7.92 ± 4.42AVLT 1–5 30.51 ± 9.13 36.33 ± 8.58 28.67 ± 8.58 46 ± 7.04 50.06 ± 10.1 19.96 ± 71AVLT 30 3.13 ± 3.38 4.83 ± 3.55 2.60 ± 3.16 8.74 ± 3.05 10.56 ± 3.25 .67 ± 1.27ROCF-R 7.66 ± 6.57 12.96 ± 6.50 5.97 ± 5.68 9.60 ± 6.06 15.25 ± 6.58 3.43 ± 5.73ROCF-C 26.85 ± 7.74 32.07 ± 2.70 25.19 ± 8.10 26.78 ± 8.50 30.00 ± 40.3 19.20 ± 10.3TMT A 27.5 ± 12.86 18.82 ± 5.57 30.19 ± 13.29 28.04 ± 16.8 18.94 ± 4.49 69.04 ± 59.1TMT B 222.3 ± 146 99.65 ± 23.7 258.91 ± 147 211.4 ± 133 75.31 ± 36.0 367.54 ± 155COWAT 33.6 ± 11.70 47.33 ± 9.01 29.19 ± 8.72 34.04 ± 11.9 44.00 ± 8.9 25.46 ± 11.73F-DigitSpan—NM 5.64 ± 1.26 6.11 ± 1.32 5.49 ± 1.21 5.71 ± 1.20 6.69 ± 1.14 5.19 ± 1.27F-DigitSpan—SC 8.59 ± 2.19 9.17 ± 2.33 8.40 ± 2.13 8.50 ± 1.67 10.81 ± 2.17 8.35 ± 4.45R-DigitSpan—NM 4.12 ± 1.10 4.78 ± .94 3.91 ± 1.07 4.08 ± 1.21 4.88 ± 1.09 3.15 ± 1.26R-DigitSpan—SC 5.49 ± 1.91 6.61 ± 1.75 5.14 ± 1.84 5.42 ± 1.91 6.94 ± 2.05 4.23 ± 3.42MHST score 11.69 ± 3.60 13.89 ± 3.05 11.00 ± 3.50 11.58 ± 3.77 15.19 ± 1.47 9.0 ± 3.27

Abbreviations: aMCI, amnesticmild cognitive impairment; naMCI, non-amnesticmild cognitive impairment; sd, single domain;md,multiple domain;MMSE, total score; AVLT 1–5, sumoftrials 1 to 5; AVLT 30, recall after 30min; ECR-FR, Enhanced Cued Recall test— free recall; ECR-TR, Enhanced Cued Recall test— total recall after cueing; ROCF-R, visual reproduction after 3min; ROCF-C, copy score; TMT A, given in seconds; TMT B, time given in seconds; F-DigitSpan—NM, forward Digit Span — numbers; F-DigitSpan—SC, forward Digit Span — score; R-DigitSpan—NM, reversed Digit Span — numbers; R-DigitSpan—SC, reversed Digit Span— score; COWAT, Czech version with “N”, “K”, “P” letters, MHST — Motol Hospital Smell test.

182 M. Vyhnalek et al. / Journal of the Neurological Sciences 349 (2015) 179–184

deficit could be attributable to underlying AD pathology. Third, with thenaMCI group largely overlapping with both the aMCI group and thecontrols in our study, the results confirm the established notion fromlongitudinal studies that naMCI is probably etiologically a very hetero-geneous group [4,22,31].

Finally, the reason formore profoundolfactory identification impair-ment found in our study compared to the previous ones is that, contraryto other studies, we used clinically based groups of patients referred byfamilies and primary care physicians for cognitive impairment and theirimpairment was greater than in population based studies [7].

As expected, the degree of olfactory impairment in aMCI was relatedmainly to memory performance, where the scores in verbal and non-verbalmemory tests explained a large part of variance in olfactory iden-tification scores in our sample (26% and 40% respectively). These resultscorrespond to earlier findings in non-demented elderly associating an-osmia with AD-like memory impairment as assessed by a QualitativeEvaluation of Dementia checklist [48]. Other studies have demonstratedmoderate to strong correlation of olfactory identification impairmentwith memory scores and only weak to moderate correlation withother cognitive domains in non-demented subjects (aMCI, naMCI andcontrols together) [7,23]. Similar results were noted in a pooled sampleof healthy controls, MCI and demented patients [8,60]. The strongest re-lation with memory domains confirms that olfaction and memory inaMCI share the common neuroanatomical substrate and underlies theimportance of medial temporal lobes for olfactory identification in thisgroup.

Table 2Association between cognitive scores and olfactory identification in aMCI.

Cognitivedomain/variable

Estimate Standard error ofthe estimate

P Adjustedr-squarea

MMSE 1.37 0.56 .017 .26Verbal memory 1.44 0.63 .026 .25Non-verbal memory 1.71 0.47 b .001 .40Visuospatial function 1.44 0.45 .002 .37Executive function −0.01 0.01 NS .21Working memory 0.06 0.45 NS .20

a Note that adjusted r-square for age and gender alone was .21.

We found olfactory identification more closely related to nonverbalmemory as opposed to verbal memory. This may be the result of sug-gested lateralization of nonverbalmemory as well as odor identificationto the right hemisphere [27,49].

In our aMCI sample, a significant contribution to olfaction scoreswasfound with visuospatial functions, but not with working memory andexecutive functions. To our knowledge only one study has examinedthe relation between visuospatial functions and olfaction and found anassociation between olfaction and copy of ROCF using a sample of cog-nitively impaired patients similar to ours [60]. The reason for this asso-ciation is not very well understood as visuospatial functions aredependent mainly on the parietal and occipital cortexes which do notplay a crucial role in olfactory identification. There are several possibleexplanations: ROCF task performance reflects the influence of manybrain networks, which may preclude precise interpretation of neuro-psychological test scores [24]. Alternatively, copy of ROCF may reflecthippocampal atrophy, as demonstrated in our recent study innondemented older adults [57]. Finally, the early concomitant degener-ation of posterior part of the brain in aMCI patients could be another un-derlying cause.

Contrary to results in aMCI patients, we did not find any significantcontribution of cognitive function to the variability in olfactory identifica-tion ability in naMCI group. One plausible explanation may be that, asstated earlier, contrary to aMCI, which is causedmainly by AD pathology,naMCI is an etiologically heterogeneous condition caused by differentprocesses leading to dementia including vascular and neurodegenerativediseases like FTLD, DLB and AD [3, 15,42,44].

Pathways to olfactory identification impairment in non-AD demen-tias are not fully understood. For example, olfactory identification im-pairment in FTLD is explained by impairment of brain centers relevantto olfactory identification in frontal and temporal lobes [33,40,46],while other studies speculated about the olfactory identification impair-ment being caused by cognitive impairment [28,41]. Olfactory deficit inLBD has been explained by Lewy body pathology in olfactory brain re-gions as well as by concomitant AD pathology which is very commonin this disease [25,35,39,52]. There is no pathological study with odoridentification in LBD. However, the only autopsy study linked anosmiato Lewy body pathology present in olfactory brain (mainly cingulum),

183M. Vyhnalek et al. / Journal of the Neurological Sciences 349 (2015) 179–184

with no influence of concomitant AD pathology [34]. There is conflictingevidence about the extent and cause of olfactory impairment in VD andthe impairment of olfaction in this group has been consideredmainly asa consequence of cognitive deficits and concomitant AD pathologywhich is common in VD [18,52].

The lack of an association between olfactory performance and cogni-tion could therefore be due to the presence of various factors leading toolfactory identification impairment in naMCI: disease specific neuro-pathological changes in the olfactory brain regions, concomitant ADneuropathological changes in olfactory bulb and nerve and cognitiveimpairment [28,33,40,41,46,18,25,52].

Our results correspond to the earlier findings in non-AD dementias,where no relation between cognition and olfaction was found in FTLDand LBD [29,34].

The limitations of the study are that (asmost other olfactory studies)only olfactory identification test was used and the olfactory thresholdand discrimination were not assessed. Therefore, the detected deficitsin olfactory identification could be at least partly due to a deficit at amore basic level of olfactory processing. The cross-sectional design ofthe study did not allowus tomakeprecise conclusions regarding the eti-ology of MCI in our patients. Furthermore, the lack of an association be-tween olfaction and cognitive performance in the naMCI group suggeststhat olfactory performancemay not predict dementia conversion in thatgroup.

The heterogeneity of olfactory identification results in naMCI group(with some patients scoring as high as controls) and the absence of re-lation to cognitive decline in this subgroup depend probably on the un-derlying pathology leading to conversion into the various types ofdementia which could also explain different results of naMCI in variousstudies.

Olfactory impairment in aMCI together with strong relationshipwith memory scores points to the relative clinical and neuropathologi-cal uniformity of this subgroup proposed in other MCI studies [42].

As postulated in previous studies, olfactory identification impair-ment is more pronounced in aMCImd compared to aMCIsd patientsconsistent with the fact that the former subtype may represent a moreadvanced pathological stage than the latter.

We conclude that olfactory identification impairment is present notonly in aMCI but also in naMCI patients. Unlike in naMCI, we found aproportional impairment between cognition and olfaction in aMCI.This may be explained by probable underlying AD; and the more pro-nounced impairment in aMCImd compared to aMCIsd may reflectmore advanced stage of the AD-related pathological process. On theother hand, the absence of a link between cognition and olfactory iden-tification in naMCI patients may be caused by known clinical and etio-logical heterogeneity of this group.

Longitudinal studies and studies with neuropathological correla-tions are necessary to confirm this hypothesis.

Acknowledgment

Supported by the FNUSA-ICRC project (no. CZ.1.05/1.1.00/02.0123)from the European Regional Development Fund, by the European SocialFund within the project Young Talent Incubator II (reg. no. CZ.1.07/2.3.00/20.0117), and by the MH CZ — DRO, Motol University Hospital,Prague, Czech Republic 00064203.

References

[1] Attems J, Lintner F, Jellinger KA. Olfactory involvement in aging and Alzheimer'sdisease: an autopsy study. J Alzheimers Dis 2005;7(2):149–58.

[2] Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. ActaNeuropathol 1991;82(4):239–59.

[3] Braak H, Braak E. Staging of Alzheimer's disease-related neurofibrillary changes.Neurobiol Aging 1995;16(3):271–8 discussion 278–284.

[4] Busse A, Hensel A, Guhne U, Angermeyer MC, Riedel-Heller SG. Mild cognitive im-pairment: long-term course of four clinical subtypes. Neurology 2006;67(12):2176–85. http://dx.doi.org/10.1212/01.wnl.0000249117.23318.e1.

[5] Cohen JA. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale:Lawrence Erlbaum Associates; 1988.

[6] Devanand DP, Michaels-Marston KS, Liu X, Pelton GH, Padilla M, Marder K, et al. Ol-factory deficits in patients with mild cognitive impairment predict Alzheimer's dis-ease at follow-up. Am J Psychiatry 2000;157(9):1399–405.

[7] Devanand DP, TabertMH, Cuasay K,Manly JJ, Schupf N, Brickman AM, et al. Olfactoryidentification deficits and MCI in a multi-ethnic elderly community sample.Neurobiol Aging 2010;31(9):1593–600. http://dx.doi.org/10.1016/j.neurobiolaging.2008.09.008.

[8] Djordjevic J, Jones-Gotman M, De Sousa K, Chertkow H. Olfaction in patients withmild cognitive impairment and Alzheimer's disease. Neurobiol Aging 2008;29(5):693–706. http://dx.doi.org/10.1016/j.neurobiolaging.2006.11.014.

[9] Doty RL. Office procedures for quantitative assessment of olfactory function. Am JRhinol 2007;21(4):460–73.

[10] Doty RL, Reyes PF, Gregor T. Presence of both odor identification and detection def-icits in Alzheimer's disease. Brain Res Bull 1987;18(5):597–600. http://dx.doi.org/10.1016/0361-9230(87)90129-8.

[11] Doty RL, Shaman P, Dann M. Development of the University-Of-Pennsylvania smellidentification test— a standardizedmicroencapsulated test of olfactory function. Phys-iol Behav 1984;32(3):489–502. http://dx.doi.org/10.1016/0031-9384(84)90269-5.

[12] Doty RL, Shaman P, Kimmelman CP, Dann MS. University of Pennsylvania SmellIdentification Test: a rapid quantitative olfactory function test for the clinic. Laryn-goscope 1984;94(2 Pt 1):176–8.

[13] Duff K, McCaffrey RJ, Solomon GS. The pocket smell test successfully discriminatingprobable Alzheimer's dementia from vascular dementia and major depression. JNeuropsychiatry Clin Neurosci 2002;14(2):197–201.

[14] Eibenstein A, Fioretti AB, Simaskou MN, Sucapane P, Mearelli S, Mina C, et al. Olfac-tory screening test in mild cognitive impairment. Neurol Sci 2005;26(3):156–60.http://dx.doi.org/10.1007/s10072-005-0453-2.

[15] Farias ST, Mungas D, Reed BR, Harvey D, DeCarli C. Progression of mild cognitive im-pairment to dementia in clinic-vs community-based cohorts. Arch Neurol 2009;66(9):1151.

[16] Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalitiesat 1.5 T in Alzheimer's dementia and normal aging. AJR Am J Roentgenol 1987;149(2):351–6. http://dx.doi.org/10.2214/ajr.149.2.351.

[17] Folstein MF, Folstein SE, McHugh PR. “Mini-mental state”. A practical method forgrading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12(3):189–98.

[18] Gray AJ, Staples V, Murren K, Dhariwal A, Bentham P. Olfactory identification is im-paired in clinic-based patients with vascular dementia and senile dementia ofAlzheimer type. Int J Geriatr Psychiatry 2001;16(5):513–7.

[19] Grober E, Buschke H. Genuine memory deficits in dementia. Dev Neuropsychol1987;3(1):13–36.

[20] Hawkes C. Olfaction in neurodegenerative disorder. Mov Disord 2003;18(4):364–72.

[21] Jones-Gotman M, Zatorre RJ, Cendes F, Olivier A, Andermann F, McMackin D, et al.Contribution of medial versus lateral temporal-lobe structures to human odouridentification. Brain 1997;120(10):1845–56.

[22] Jungwirth S, Zehetmayer S, Hinterberger M, Tragl KH, Fischer P. The validity ofamnestic MCI and non-amnestic MCI at age 75 in the prediction of Alzheimer's de-mentia and vascular dementia. Int Psychogeriatr 2012;24(6):959–66. http://dx.doi.org/10.1017/s1041610211002870.

[23] Lehrner J, Pusswald G, Gleiss A, Auff E, Dal-Bianco P. Odor identification and self-reported olfactory functioning in patients with subtypes of mild cognitive im-pairment. Clin Neuropsychol 2009;23(5):818–30. http://dx.doi.org/10.1080/13854040802585030.

[24] Lezak MD, Black HD, W., L. D. Neuropsychological assessment. V ed. Oxford: OxfordUniversity Press; 2004.

[25] Liberini P, Parola S, Spano P, Antonini L. Olfactory dysfunction in dementia associat-ed with Lewy bodies. Parkinsonism Relat Disord 1999;5:30.

[26] Loonstra AS, Tarlow AR, Sellers AH. COWAT metanorms across age, education, andgender. Appl Neuropsychol 2001;8(3):161–6.

[27] Loring DW, Lee GP, Meador KJ. Revising the Rey–Osterrieth: rating right hemisphererecall. Arch Clin Neuropsychol 1988;3(3):239–47.

[28] Luzzi S, Snowden JS, Neary D, Coccia M, Provinciali L, Lambon Ralph MA. Distinctpatterns of olfactory impairment in Alzheimer's disease, semantic dementia,frontotemporal dementia, and corticobasal degeneration. Neuropsychologia 2007;45(8):1823–31.

[29] Magerova H, VyhnalekM, Laczo J, Andel R, Rektorova I, Kadlecova A, et al. Odor iden-tification in frontotemporal lobar degeneration subtypes. Am J Alzheimers Dis OtherDemen 2014. http://dx.doi.org/10.1177/1533317514539033.

[30] Magerova H, Vyhnalek M, Laczo J, Bojar M, Hort J. Smell perception testing in earlydiagnosis of neurodegenerative dementia. Ceska Slov Neurol Neurochir 2008;71(3):298–302.

[31] Manly JJ, Tang MX, Schupf N, Stern Y, Vonsattel JP, Mayeux R. Frequency and courseof mild cognitive impairment in a multiethnic community. Ann Neurol 2008;63(4):494–506. http://dx.doi.org/10.1002/ana.21326.

[32] McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. Clinical diag-nosis of Alzheimer's disease: report of the NINCDS-ADRDA Work Group under theauspices of Department of Health and Human Services Task Force on Alzheimer'sDisease. Neurology 1984;34(7):939–44.

[33] McLaughlin NCR, Westervelt HJ. Odor identification deficits in frontotemporal de-mentia: a preliminary study. Arch Clin Neuropsychol 2008;23(1):119–23.

184 M. Vyhnalek et al. / Journal of the Neurological Sciences 349 (2015) 179–184

[34] McShane RH, Nagy Z, Esiri MM, King E, Joachim C, Sullivan N, et al. Anosmia in de-mentia is associated with Lewy bodies rather than Alzheimer's pathology. J NeurolNeurosurg Psychiatry 2001;70(6):739–43.

[35] Merdes AR, Hansen LA, Jeste DV, Galasko D, Hofstetter CR, Ho GJ, et al. Influence ofAlzheimer pathology on clinical diagnostic accuracy in dementia with Lewy bodies.Neurology 2003;60(10):1586–90.

[36] Mesholam RI, Moberg PJ, Mahr RN, Doty RL. Olfaction in neurodegenerative disease:a meta-analysis of olfactory functioning in Alzheimer's and Parkinson's diseases.Arch Neurol 1998;55(1):84.

[37] Meyers JE, Meyers KR. Rey complex figure test and recognition trial: professionalmanual. Odessa, FL: Psychological Assessment Resources; 1995.

[38] Morris JC. The Clinical Dementia Rating (CDR): current version and scoring rules.Neurol Neurol 1993;43(11):2412–4.

[39] Nedelska Z, Ferman TJ, Boeve BF, Przybelski SA, Lesnick TG, Murray ME, et al. Patternof brain atrophy rates in autopsy-confirmed dementia with Lewy bodies. NeurobiolAging 2014. http://dx.doi.org/10.1016/j.neurobiolaging.2014.07.005.

[40] Omar R, Mahoney CJ, Buckley AH, Warren JD. Flavour identification in frontotemporallobar degeneration. J Neurol Neurosurg Psychiatry 2013;84(1):88–93.

[41] Pardini M, Huey ED, Cavanagh AL, Grafman J. Olfactory function in corticobasal syn-drome and frontotemporal dementia. Arch Neurol 2009;66(1):92–6.

[42] Petersen RC. Mild cognitive impairment. N Engl J Med 2011;364(23):2227–34.[43] Petersen RC. Mild cognitive impairment as a diagnostic entity. J Intern Med 2004;

256(3):183–94.[44] Petersen RC. Clinical practice. Mild cognitive impairment. N Engl J Med 2011;

364(23):2227–34. http://dx.doi.org/10.1056/NEJMcp0910237.[45] Petersen RC, Smith GE, Waring SC, Ivnik RJ, Tangalos EG, Kokmen E. Mild cogni-

tive impairment: clinical characterization and outcome. Arch Neurol 1999;56(3):303.

[46] Rami L, Loy CT, Hailstone J, Warren JD. Odour identification in frontotemporal lobardegeneration. J Neurol 2007;254(4):431–5.

[47] Reitan Rm WD. The Halstead–Reitan neuropsychological test battery. Theory andclinical interpretation. South Tucson: Neuropsychology Press; 1993.

[48] Royall DR, Chiodo LK, Polk MS, Jaramillo CJ. Severe dysosmia is specifically associat-ed with Alzheimer-like memory deficits in nondemented elderly retirees.Neuroepidemiology 2002;21(2):68–73 [doi: 48619].

[49] Royet JP, Plailly J. Lateralization of olfactory processes. Chem Senses 2004;29(8):731–45. http://dx.doi.org/10.1093/chemse/bjh067.

[50] Serby M, Larson P, Kalkstein D. The nature and course of olfactory deficits inAlzheimer's disease. Am J Psychiatry 1991;148(3):357–60.

[51] Schmidt M. Rey auditory verbal learning test: a handbook. Los Angeles, LA: WesternPsychological Services; 1996.

[52] Schneider JA, Arvanitakis Z, Bang W, Bennett DA. Mixed brain pathologies accountfor most dementia cases in community-dwelling older persons. Neurology 2007;69(24):2197–204. http://dx.doi.org/10.1212/01.wnl.0000271090.28148.24.

[53] Swan GE, Carmelli D. Impaired olfaction predicts cognitive decline in nondementedolder adults. Neuroepidemiology 2002;21(2):58–67 [doi: 48618].

[54] Tabert MH, Liu X, Doty RL, Serby M, Zamora D, Pelton GH, et al. A 10-item smellidentification scale related to risk for Alzheimer's disease. Ann Neurol 2005;58(1):155–60. http://dx.doi.org/10.1002/ana.20533.

[55] TabertMH,Manly JJ, Liu X, Pelton GH, Rosenblum S, JacobsM, et al. Neuropsycholog-ical prediction of conversion to Alzheimer disease in patients with mild cognitiveimpairment. Arch Gen Psychiatry 2006;63(8):916–24. http://dx.doi.org/10.1001/archpsyc.63.8.916.

[56] Vollmecke TA, Doty RL. Development of the Picture Identification Test (PIT) — a re-search companion to the University-of-Pennsylvania Smell Identification Test(UPSIT). Chem Senses 1985;10(3):413–4.

[57] Vyhnalek M, Nikolai T, Andel R, Nedelska Z, Rubinova E, Markova H, et al. Neuropsy-chological correlates of hippocampal atrophy in memory testing in nondementedolder adults. J Alzheimers Dis 2014. http://dx.doi.org/10.3233/jad-132642.

[58] Wechsler D. Wechsler memory scale. San Antonio; Toronto: The PsychologicalCorporation; 1997.

[59] Westervelt HJ, Bruce JM, CoonWG, Tremont G. Odor identification in mild cognitiveimpairment subtypes. J Clin Exp Neuropsychol 2008;30(2):151–6. http://dx.doi.org/10.1080/13803390701287408.

[60] Westervelt HJ, Ruffolo JS, Tremont G. Assessing olfaction in the neuropsychologicalexam: the relationship between odor identification and cognition in older adults.Arch Clin Neuropsychol 2005;20(6):761–9.

[61] Westervelt HJ, Stern RA, Tremont G. Odor identification deficits in diffuse Lewy bodydisease. Cogn Behav Neurol 2003;16(2):93.

[62] Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M, et al. Development andvalidation of a geriatric depression screening scale: a preliminary report. J PsychiatrRes 1983;17(1):37–49.