Olfactory dysfunction revisited: a reappraisal of work-related … · 2018. 9. 4. · paper on functional testing and dysfunction of olfaction, Herberhold [23] assumed several decades

REVIEW Open Access

Olfactory dysfunction revisited: areappraisal of work-related olfactorydysfunction caused by chemicalsSabine Werner* and Eberhard Nies

Abstract

Occupational exposure to numerous individual chemicals has been associated with olfactory dysfunction, mainly inindividual case descriptions. Comprehensive epidemiological investigations into the olfactotoxic effect of workingsubstances show that the human sense of smell may be impaired by exposure to metal compounds involvingcadmium, chromium and nickel, and to formaldehyde. This conclusion is supported by the results of animalexperiments. The level of evidence for a relationship between olfactory dysfunction and workplace exposure toother substances is relatively weak.

Keywords: Chemically induced anosmia, Formaldehyde, Metals, Occupational exposure, Olfaction disorders

BackgroundMany occupational groups are reliant upon intact olfac-tory function in order to perform their work and fortheir safety. Examples are chefs, gas fitters, firefighters,perfumers, sommeliers, coffee and tea tasters, grocers,workers in the chemical industry, and domestic helpers.The importance of the olfactory function for early detec-tion of hazardous substances with an odour is illustratedby the specific case of an anosmic who lit a cigarettewhilst in close proximity to a leaking petrol pipe, therebycausing an explosion [1]. Muttray et al. [2] report thecase of a patient who did not become aware of his olfac-tory dysfunction until his colleagues fled their work-place, to him for no apparent reason, owing to anintense solvent smell. In Germany, an assessment of ol-factory function is a requirement for persons applyingfor certification of their fitness to perform fumigation[3], and loss of olfactory function constitutes groundsfor example for the discharge of members of the USmilitary, including reservists, and of coastguard em-ployees [4].

Diagnosis and assessment of olfactorydysfunctionAssessment of olfactory function and diagnosis of olfac-tory dysfunction requires, firstly, a detailed medical historyand examination by an otolaryngologist [5, 6]. The med-ical history should include information on the triggeringevents, development, complementary symptoms, surgicaloperations, medication and toxicants. The ENT diagnosiscomprises medical status, endoscopy of nose and naso-pharyngeal space and evaluation of the olfactory cleft. If aneurological disorder is suspected, an examination by aneurologist including tests of cognition and memory couldbe necessary. Secondly, a validated test method is neededthat enables subjective sensory perceptions to be quanti-fied objectively. This is essential for a standardised distinc-tion between normosmia (normal olfactory function),hyposmia and anosmia (impaired olfactory function andits complete loss, respectively), and hyperosmia (olfactoryoversensitivity). A screening of global taste function (ret-ronasal smelling) is also advantageous owing to the closeconnection between smell and taste (patients complainingof a dysfunctional sense of taste are in fact often sufferingfrom olfactory impairment).Only in the last three decades have standardised and

practicable psychophysical tests for humans been devel-oped. Of these, UPSIT (University of Pennsylvania Smell

* Correspondence: [email protected] for Occupational Safety and Health of the German Social AccidentInsurance, Unit Toxicology of Industrial Chemicals, Alte Heerstrasse 111,53757 Sankt Augustin, Germany

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Werner and Nies Journal of Occupational Medicine and Toxicology (2018) 13:28 https://doi.org/10.1186/s12995-018-0209-6

http://crossmark.crossref.org/dialog/?doi=10.1186/s12995-018-0209-6&domain=pdfhttp://orcid.org/0000-0002-7531-902Xmailto:[email protected]://creativecommons.org/licenses/by/4.0/http://creativecommons.org/publicdomain/zero/1.0/

Identification Test [7]) and the “Sniffin’ Sticks” testwidely used in Europe [8, 9], are important examples.“Sniffin’ Sticks” are felt sticks that release aromatic

substances when the cap of the stick is removed. In con-ventional form, they permit threshold, discriminationand identification tests; the last two of these areabove-threshold tests. The threshold test indicates theconcentration above which an odour is sensed (the sen-sory threshold). As standard, n-butanol or phenyl ethylalcohol (rose scent) are used for testing. The non-verbaldiscrimination test examines the ability to distinguishbetween odours. In the identification test, 16 odours aretested for recognition [10]. This is a structured, reliableand validated test system that is widely used in Europeand is readily available. Extensive validation studies anddefined standard values exist for this test [11, 12]. Regu-larly updated standardised values are published for ex-ample on the website of the Interdisciplinary CenterSmell & Taste (University Clinic Dresden) [13].A similar identification test used in America is the

UPSIT method, in which up to 40 odorants, microen-capsulated on a sheet of paper, are released by scratchingwith the point of a pencil. In this test, the various odor-ants must be identified with reference to a list of fourterms per substance. The test kit has a long shelf life, isvery well validated and is widely used. The UPSIT testdoes not require clinician supervision and is thereforevery convenient. International versions are also available;they have however rarely been validated specifically forindividual countries. A drawback of the test is that itstudies only the identification of odours.Where patients might not be able to comply with psycho-

physical testing, or in medico-legal assessments, the olfac-tory dysfunction can be assessed objectively by recordingelectrical activity of the brain (olfactory event-related poten-tials, OERPs) following presentation of odours. Thismethod requires virtually no active participation on thepart of the test subject, and has been used since the1970s. It involves the application of olfactory stimuli tothe nose of the test subject with the aid of an olfactom-eter. The stimuli trigger corresponding OERPs, whichcan then be registered on electrodes applied to the testsubject’s head. The use of an electro-olfactograph (therecording of generator potential via an electrode incontact with the olfactory epithelium) is limited to re-search. Olfactory functional imaging methods such asPET (positron emission tomography) and fMRI (func-tional magnetic resonance imaging) are also largely lim-ited to research applications [5, 6].Important information on the toxic properties of

inhaled substances that may affect olfaction has beenobtained from experiments on animals. Besides histo-logical analyses of the olfactory epithelium, the resultsof behavioural tests are also relevant. Since the

laboratory animal is not able to communicate activelyto the researcher whether or not it senses an odour,an operant conditioning test is usually performed. Insuch a test, mice for example are first taught toexpect a reward (such as water following restrictedaccess to water) after sensing a certain olfactorystimulus. The animals are then presented with otherodours that are not followed by a reward, in additionto the odour that they have learnt to associate withthe reward. Where the animal has recognised the cor-rect odour and looks for its reward, this can be regis-tered, for example by means of a light barrier (for adetailed description, see Kuner and Schaefer 2011[14]; an up-to-date overview with detailed test proto-cols can be found in Zou et al. 2016 [15]).

Work-related olfactory dysfunctionThe prevalence of olfactory dysfunction in the wider popu-lation is estimated at around 5% (for functional anosmia)[16]. It is considerably more common among older people,and around a quarter of the population aged over 50 exhibitan impaired sense of smell. The results of recent studiessuggest that specific anosmias, the failure to sense a specificodour, are far more prevalent than was previously assumedand are the norm rather than the exception [16, 17]. Theproportion of olfactory dysfunction caused by occupationalexposure remains unclear. Figures for olfactory dysfunctioncaused by exposure to the effects of chemicals and pharma-ceuticals fluctuate between 0.5 and 5% of all cases [18]. Ac-cording to a large-scale survey of all ear, nose and throatclinics in the German-speaking world, covering 79,000 pa-tients, 2% of the cases had toxic causes [19]. A recent Bel-gian publication covering a substantially smaller collectiveof 496 patients with exclusively non-sinusoidal complaintsfrom a specialist clinical centre assumed that 3.4% weretoxic in origin [20]. “Idiopathic” cases of olfactory dysfunc-tion, which are put at between 10 and 25%, may howeverinclude cases of chemically induced damage caused byworkplace exposure not classified as such [21, 22]. In hispaper on functional testing and dysfunction of olfaction,Herberhold [23] assumed several decades ago that patho-logically elevated olfactory thresholds were present inaround 30% of workers in the metals and chemical indus-try, possibly rising to around 50% with increasing age andincreasing duration of exposure to the hazardous sub-stances. Today’s working conditions and exposures areclearly not comparable with those of the 1970s; in her re-view however, Dalton 2010 [24] also cites a questionnaireand survey conducted in 1995 among 712,000 individualsin Canada and the USA which revealed that factoryworkers of all ages reported a weaker sense of smell andperformed significantly worse in an olfactory test thanmembers of other occupational groups.

Werner and Nies Journal of Occupational Medicine and Toxicology (2018) 13:28 Page 2 of 26

According to Herberhold [23], thermal, mechanicaland chemical noxae may lead to olfactory dysfunction,the effects triggered by chemicals being more pro-nounced the more active the chemical substance, thesmaller the particles, and the longer the duration of ex-posure of the sensory apparatus to them. Occupationalexposure to numerous industrial chemicals, notablythose that are irritative and corrosive to the mucousmembranes or harmful to the nerves, is associated withthe incidence of olfactory dysfunction (see for exampleKlimek et al. 1999 [25] and Muttray et al. 2006 [26]). Inhis overview, Amoore [27] lists over 100 substances pre-sumed capable of causing olfactory dysfunction. This in-formation is based almost entirely upon case reportsrather than on large-scale studies. The discussion belowtherefore gives consideration to the substances associatedwith occupational olfactory dysfunction that have beenstudied under standardised conditions in epidemiologicalstudies and studies of test subjects. The results of relevantanimal experiments concerning the substances identifiedin this way that to some extent permit conclusions regard-ing the possible mechanism of action are also presented.

Industrial chemicals with a potential impact uponolfactionIn order to identify industrial chemicals exposure to whichmay potentially lead to olfactory dysfunction, a literature sur-vey was conducted in the Pubmed database in order to iden-tify substances characterised as olfactotoxic primarily on thebasis of epidemiological studies and studies of test subjectsconducted under standardised conditions. The terms “anos-mia”, “hyposmia”, “dysosmia”, “smell disorders”, “olfactoryfunction”, “olfactory dysfunction” and “olfaction disorders”were each combined in the search with “occupational”, “pro-fessional” and “workplace”. In the second step, the sub-stances identified in this way were used as search terms incombination with the relevant clinical pictures (see above) inorder to identify animal experiments. Relevant studies in thebibliographies of the identified literature were considered.Pharmacological and environmental studies associated withthe identified industrial chemicals were also included. Casereports, for example concerning accident-type events involv-ing very high exposures, were disregarded.A comprehensive overview of long-term effects of oc-

cupational exposures to metals and olfactory toxicity canbe found in the reviews by Gobba 2006 and Sunderman2001 [21, 28]. One aspect addressed by the recent reviewin Doty 2015 is likewise the influence of exposure toneurotoxic substances in the environment or at theworkplace upon the sense of smell [4]. Accordingly, thestudies referred to in these publications will not beconsidered in further detail in the present paper. Thefocus here lies on the epidemiological and animal studiesnot stated there. The latter will be described in detail.

Human studies on olfactotoxic effects caused by chemi-cals are summarised in Table 1.

Cadmium and nickelA considerable number of epidemiological studies demon-strate an association between exposure to metals in theform of dusts and vapours, and occupational olfactorydysfunction. Anosmia and hyposmia were diagnosed forexample among workers exposed to dust containing nickelor cadmium in plants for the production of alkaline bat-teries, nickel refineries, and the cadmium industry.

Human studies

Cadmium exposure Nine epidemiological studies pub-lished in the period from 1950 to 2003 [29–37] addressthe association between workplace cadmium/nickel ex-posure and olfactory impairment. The older studies lackclear information on the test methods and comparisonswith non-exposed subjects [29–32]. Anosmia/hyposmiawas detected in a significantly high proportion of the ex-posed workers in all human studies. Differentiated sen-sory and identification tests were conducted in threestudies [34–37]; Rydzewski et al. and Sulkowski et al.[35, 36] describe the same data, which can therefore alsobe counted as one. Whereas Rose et al. and Mascagni etal. [34, 37] demonstrated that the values obtained in thesensory threshold test were considerably higher amongthe workers than in the control group, the identificationtest revealed no significant differences. The sensorythreshold test for n-butanol or phenyl ethyl alcohol isregarded as an instrument for diagnosing the function ofthe peripheral olfactory receptor neurons. According tothis interpretation, the ability to identify odours is basedprimarily upon the processing of olfactory information inthe cerebral cortex. This absolute distinction is problem-atic however, since the sensory threshold test also encom-passes complex functions of the central nervous system,and the identification of odours is dependent upon the ac-tivity of the olfactory receptor neurons.Many of the epidemiological studies that describe

the incidence of olfactory dysfunction followingexposure to cadmium are older. In the 1950s and1960s, occupational exposure to cadmium was appreciablyhigher than it is today, for example 0.6–236 mg/m3 in abattery factory [31]. The threshold limit values (TLVs)proposed by the American Conference of GovernmentalIndustrial Hygienists (ACGIH) stated a limit concentra-tion of 50 μg/m3 in 1975 and 10 μg/m3 as of 1995. Con-versely, more recently published measured exposurevalues of 0.004–0.187 mg/m3 [32], 0.3 mg/m3 [34] and1.53 mg/m3 (1975) and 0.0171 mg/m3 (1995) [37] demon-strate that damage to olfactory function may arise even atlow concentrations.


Table

1Work-relatedolfactorydysfun

ction:Hum

anstud

ies(in

chrono

logicalo

rder)

Hum

anstud

iesReference

Occup

ation

Expo

sure

(mg/m

3 )Expo

sedworkers,

reside

nts,patients

Con

trols

Testmetho

dResults

Cadmium/nickel

a Frib

erg1950

[29]

Batteryfactory

3–15

Cd

10–150

Ni

43(group

1:long

erem

ploymen

t9–34

years)

15(group

2:shorter

employmen

t1–4

years)

–Testwith

odou

rsamples

(coffee,

perfu

me,

pepp

ermintand

petrol)

Group

1:37%

impairedsenseof

smell(32.6%:ano

smia).

Group

2:6.7%

(one

case)partially

impairedsenseof

smell.

a Baade

r1951

[30])

Batteryfactory

N.i.

8–

Testwith

odou

rsamples

(nofurthe

rde

tails)

4hypo

smicor

anosmic.

a Potts1965

[31]

Batteryfactory

Cddu

standfumes:

0.6–236(1949)

<0.5(1950)

<0.1(1956)

70–

N.i.

64%

anosmic.

a Liu

etal.1985[32]

Cadmium

smelters

Cdoxide:0.004–0.187

65–

N.i.

21.5%

anosmic.

a AdamsandCrabtree

1961

[33]

Batteryfactory

Cd:

0.028–2.76

Ni:0.0016–0.056

106

84Thresholdtestwith

phen

ol27.3%

anosmic(Co:4.8%

anosmic)

a Roseet

al.1992[34]

Factory

prod

ucing

refrige

ratin

gcoils

Cdfumes

upto

0.3

5516

Thresholdtestwith

n-bu

tano

l.Iden

tificationtest

with

10different

olfactorystim

uli

13%

severe

hypo

smia(Co:0%

),44%

weaklyhypo

smic(Co:31%).

Iden

tificationtest:nosign

ificant

difference.

a Rydzewskietal.1998[35]

a Sulkowskietal.2000[36]

Batteryfactory

Atm

osph

ericcadm

ium

concen

trations:0.05–2.1

7343

forestry

workers

Threshold

iden

tificationtest

with

mod

ified

Elsberg-Levys

metho

d

Sign

ificantlyim

pairedability

tode

tect

andiden

tifyod

ours.

Overallresult:26%

hypo

smia,

17.8%

parosm

ia,1.4%

anosmia

(Co:30.2%

parosm

ia,69.8

norm

osmia).

a Mascagn

ietal.2003[37]

Workersin

acadm

ium

foun

dry

andsinteringplant

Cadmium

concen

trations

(max.):

1.530(1975)

-0.0171

(1995);b

y1978,the

measuremen

tresults

hadalreadydrop

ped

substantially

below

1(at0.207)

3339

driversand

storekeepe

rs,

23welde

rs

Thresholdtestwith

phen

ylethyl

alcoho

lIden

tificationtest

with

Wrig

ht’s

confusionmatrices

Olfactorythresholdsign

ificantly

high

erin

Cdworkersthan

incontrols.

Iden

tificationtest:nosign

ificant

difference(a

non-sign

ificant

impairm

entin

cadm

ium

workers).

Overallresults:

3.1%

anosmic,3.1%

severely

hypo

smic,24.1%

mildly/m

oderately

hypo

smic(welde

rs:0%

anosmic,

4.3%

severelyhypo

smic,8.7%

mildly/

mod

eratelyhypo

smic;C

o:0%

anosmic,0%

severelyhypo

smic,

7.7%

mildly/m

oderatelyhypo

smic).


Table


ction:Hum

anstud

ies(in

chrono

logicalo

rder)(Con

tinued)

Hum

anstud

iesReference

Occup

ation

Expo

sure

(mg/m

3 )Expo

sedworkers,

reside

nts,patients

Con

trols

Testmetho

dResults

a 2stud

iesfro

mUSSRcitedin

Sund

erman

[28]:

(Tatarskaya1960)

(Kucharin

1970)

ElectrolyticNi

refinery

ElectrolyticNi

refinery

N.i.

N.i.

N.i.

458

N.i.

N.i.

N.i.

N.i.

Freq

uent

olfactoryim

pairm

ent,

atroph

icnasalm

ucosa,nasalsep

tal

ulceratio

nandsinu

sitis.

Ano

smiain

114/251workers(46%

)with

chronicsinu

sitis;lesssevere

loss

ofsm

ellinotherNi-exposed

workers.

Chrom

ium

Seeb

eret

al.1976[53]

Chrom

epaint

plant(m

anufacture

ofbasiczinc

chromate(zinc

yellow))

Values

exceed

ed0.1

CrO

3air,androse

atcertainpo

intsto

upto

20CrO

3.Moreprecise

values

areno

tstated

.

5chronically

expo

sed

workers,14interm

itten

tlyexpo

sedworkers(not

long

erthan

2hpe

rday),

5mask-wearin

gworkers

(wearin

gfine-du

stfilter

masks

forthefull

duratio

n)

9officeworkersat

thesamecompany,

23em

ployeesat

hospital

Testwith

odou

rstrip

s(6

different

stim

uliin8

concen

trations)

Olfactorysensitivity

ofcontrolg

roup

andno

n-expo

sedindividu

als

substantially

high

erthan

that

ofexpo

sedindividu

als.

“Significant”relatio

nshipbe

tween

chromium

expo

sure

andloss

ofolfaction.

Chron

icallyexpo

sedindividu

als:

damageto

thenasalm

ucou

smem

brane.

Interm

itten

tlyexpo

sedgrou

pand

workerswith

amask:damageto

the

nasalm

ucou

smem

braneon

lyinapart.

Seeb

erandFikentsche

r1980

[54]

Seeabove

Occup

ationalexposure

“sub

stantially

redu

ced”

bysuitablemeasures

(not

specified

)

3chronically

expo

sed

workers,9

interm

itten

tlyexpo

sedworkers(not

long

erthan

2hpe

rday)

4mask-wearin

gworkers

(wearin

gfine-du

stfilter

masks

forthefull

duratio

n)

7officeworkersat

thesamecompany,

23em

ployeesat

hospital

Seeabove

Con

firmationof

therelatio

nship

betw

eenthepatholog

icalnasal

mucou

smem

branefinding

sand

olfactorydysfun

ctionin

16workers

expo

sedto

different

levels.

Noim

provem

entin

themucou

smem

braneor

olfaction.

a WatanabeandFukuchi

(1981)

[55]

Chrom

ate

prod

uctio

nplant

AirCrconcen

tration

of20.17according

to[56]

33–

T&Tolfactom

eter

(odo

urde

tection

thresholdand

odou

rrecogn

ition

)

“Middleandhigh

gradede

crease

ofod

orrecogn

ition

faculty”in

18workers(54.5%

)includ

ing2anosmiac,

oneof

who

malso

complaine

dof

atastedisorder.

51%

exhibitednasalsep

tum

perfo

ratio

ns.

Relatio

nbe

tweende

gree

ofolfaction

loss

anddu

ratio

nof

employmen

tin

thechromateprod

ucingfactory.

a Kitamuraet

al.2003[56]

Crplatingfactory

Average

atmosph

eric

concen

trationwas

0.0228

2734

T&Tolfactom

eter

(odo

urde

tection

thresholdand

odou

rrecogn

ition

)Olfactory

percep

tion

thresholdtest

Nosign

ificant

differences

forsensory

andpe

rcep

tionthreshold.

Sign

ificantlyhigh

ervalues

forthe

odou

rrecogn

ition

testthan

thosefor

controls,p

ositive

correlationwith

duratio

nof

expo

sure.

Non

eof

Crworkersshow

ednasal

septum

perfo

ratio

nor

ulceratio

n.


Table


ction:Hum

anstud

ies(in

chrono

logicalo

rder)(Con

tinued)

Hum

anstud

iesReference

Occup

ation

Expo

sure

(mg/m

3 )Expo

sedworkers,

reside

nts,patients

Con

trols

Testmetho

dResults

Aiyer

etal.2003

Chrom

ium

plating

indu

stry

N.i.

28–

N.i.

11workersanosmic(nasalseptal

perfo

ratio

nof

different

magnitude

son

allexposed

workers,m

ajority

with

initialsymptom

ofnasalirritatio

n).

Mangane

sea Lucchinietal.1997[58]

Ferroalloy

prod

uctio

nplant

Mangane

sedu

stexpo

sure:0.026–0.750

(geo

metric

mean:0.193)

3537

Olfactory

percep

tion

thresholdto

PM-carbino

l(3-m

ethyl-1-

phen

ylpe

ntan-3-ol

dilutio

nseries)

Nosign

ificant

differences

(alth

ough

itwas

negativelyassociated

with

Mn

levelsin

urinein

theexpo

sedgrou

p)

a Mergler

etal.1994[59]

Ferrom

angane

seandsilicom

angane

sealloyplant

Mangane

sedu

stexpo

sure:0.014–11.48

TotalM

nlevelsin

dust:

0.89

(geo

metric

mean)

Mangane

seconten

tof

respirabledu

stfraction:

0.001and1.273

(geo

metric

mean:0.04)

74(m

atched

pairs)

74Olfactory

percep

tion

thresholdto

PM-

carbinol

Sign

ificantlyincreasedolfactory

percep

tionam

ongworkerscompared

tocontrols.

a Antun

eset

al.2007[60]

a Bow

leret

al.2007[61]

SanFrancisco/

Oakland

BayBridge

welde

rs

Atm

osph

eric

mangane

selevelslay

betw

een0.11

and0.46

(55%

>0.20)

43welde

rs43

(matched

byage,

sex,ed

ucationand

smokingstatus

from

thedatabase

ofthe

University

ofPenn

sylvaniaSm

ell

andTasteCen

ter)

UPSITiden

tification

test

Sign

ificantlyweakerolfactoryfunctio

nof

welde

rsthan

incontrols.

Scoreof

88%

oftheworkersbe

low

theirindividu

allymatched

controls.

3%of

thewelde

rswereanosmic.

Better

olfactoryfunctio

nin

workers

with

thehigh

estMnbloo

dlevelsthan

thosewith

thelowestMnbloo

dlevels.

Bowleret

al.2011[62]

Follow-upstud

ythreeandahalf

yearslater

–26

welde

rsfro

mthe

SanFrancisco/Oakland

BayBridge

welde

rstud

y(13stud

yparticipantswereno

long

erworking

aswelde

rs)

–UPSITiden

tification

test

Nosign

ificant

differences

from

earlier

finding

s.Thebloo

d-mangane

selevelsof

the

form

erworkersweresign

ificantly

lower

than

thoseof

theircolleagues.

Senet

al.2011[63]

Welde

rsCum

ulativeMn

expo

sure

(mg/m

3 xyears):

Welde

rs0.881(±

0.567)

Con

trols0.002(±

0.0003)

7welde

rs7

MRI

Increasedmangane

sede

positio

nin

theolfactorybu

lbandin

othe

rregion

sof

thebrain.

Nostatisticallysign

ificant

differences

intheolfactorytestbe

tweenthetw

ogrou

ps(datano

tshow

n).

Guarneros

etal.(2013)

Person

slivingin

proxim

ityto

amangane

seplant

Elevated

mangane

sehairconcen

tration

(mean0.00973vs.

0.00101)

30pe

rson

sliving

with

inaon

e-kilometre

radius

ofaMexican

mangane

semine

30controlsliving

morethan

50km

away

Sniffin’Sticks

olfactorytestseries

(thresho

ld,

discrim

inationand

iden

tificationtests)

Sign

ificantlydiminishe

dolfactory

functio

nin

thoselivingcloseto

the

mangane

semine(in

threshold,

discrim

inationandiden

tificationtests).


Table


ction:Hum

anstud

ies(in

chrono

logicalo

rder)(Con

tinued)

Hum

anstud

iesReference

Occup

ation

Expo

sure

(mg/m

3 )Expo

sedworkers,

reside

nts,patients

Con

trols

Testmetho

dResults

Lucchini

etal.2012[65]

Reside

ntsof

Valcam

onica:

Italianregion

,markedby

ferrou

salloyplantsun

til2001

Average

Mnatmosph

eric

andsoilvalues

atthe

timeof

thestud

y:0.0495,958

154youn

gpe

ople

aged

betw

een11

and14

157youn

gpe

ople

from

Lake

Garda

region

Sniffin’Sticks

(iden

tificationtest)

Sign

ificantlypo

orer

olfactoryfunctio

nassociated

with

Mnconten

tin

thesoil.

Iann

illietal.2016

Valcam

onica(see

above)

andBo

gnolo

Mella:reg

ions

with

ahistoryof

high

Mn

contam

ination

–9youn

gpe

oplefro

mValcam

onicaand

Bagn

oloMella

expo

sedto

mangane

se

4youn

gpe

ople

from

theLake

Garda

region

Sniffin’Sticks

(iden

tificationtest),

fMRT

(respon

seactivity

toolfactory

tasks)

Nosign

ificantlydifferent

results

ofthe

iden

tificationtestbe

tweenthetw

ogrou

ps.

Redu

ctionin

activity

inrelevant

olfactorybrainregion

s.In

comparison

with

alarger

control

grou

pfro

madatabase,significant

differences

wereob

served

with

regard

tothesize

oftheolfactory

bulb

andin

theolfactorytest

involvingSniffin’Sticks.

Casjenet

al.2017[67]

Blue

collarworkers

Med

ian:58.3μg

/m3x

years

(interquartile

rang

e19–185

μg/m

3xyears)

354

1031

Sniffin’Sticks

(iden

tificationtest)

Norelevant

associationof

form

erMn

expo

sure

atrelativelylow

levelswith

impairedolfaction.

Zinc

Pyatayev

etal.1971[73]

Zinc

prod

uctio

nplant

Zinc

oxide,zinc

sulfate

andmetaldu

sts.

Co-expo

sure

tofurthe

rstrong

lycorrosiveand

irritant

substances

301

63Olfactom

eter

employingmint

anddilute

acetic

acid

Sign

ificant

elevations

ofsensory

thresholds.

Elevations

werehigh

estam

ong

workersrespon

sibleforroastin

gthe

zinc

ore.

Ano

nymou

s1938,aTisdall

etal.1938[74,75]

(Treatmen

tin

specialclinicswith

zinc

innasalspray

forpreven

tionof

poliomyelitis

infection)

Nasalspraysolutio

ncontaining

1%zinc

sulfate

and0.5%

tetracaine

5233

children(4713

(received

two

spraying

s)+520

(received

one

spraying

))

6300

N.i.

Nomorethan

aqu

arterexhibited

tempo

rary

anosmia.

a Tisdallet

al.1938[75]

(Patientstreatedby

thesame

otolaryngo

logists,

butin

theirprivate

practices)

Seeabove

5000

childrenand

adults

–N.i.

6mon

thsfollowingtreatm

ent:44

ofthepatientswerepe

rmanen

tlyanosmic(52with

disturbances

ofsm

elland

taste).

Davidsonet

al.2010[76]

(NasalDysfunctio

nClinicof

the

University

ofCaliforniain

SanDiego

)

Intranasalzinc

glucon

atege

l10

patients

–n-Bu

tano

lthreshold

UPSIT

(iden

tificationtest)

3subjectswith

anosmiaand7with

hypo

smia.

Alexand

eret

al.2006[77]

(See

above)

Intranasalzinc

glucon

atege

l17

patients

–n-Bu

tano

lthreshold,

Impairedolfactionin

allp

atients:

7patientsanosmic,10hypo

smic.


Table


ction:Hum

anstud

ies(in

chrono

logicalo

rder)(Con

tinued)

Hum

anstud

iesReference

Occup

ation

Expo

sure

(mg/m

3 )Expo

sedworkers,

reside

nts,patients

Con

trols

Testmetho

dResults

iden

tificationtest

employingseven

know

nod

orants

andon

esubstance

fortestingthe

trigem

inalfunctio

n,UPSIT(9

patients)

Traumaandinfectionwereruledou

tas

causes

ofhypo

smiaandanosmia

diagno

sedin

15of

thesepatientsin

therelevant

tests.

Jafeket

al.2004[78]

(Spe

cial“Taste

and

SmellC

enter”at

the

University

ofColoradoScho

olof

Med

icine)

Use

ofzinc

glucon

atege

l10

patients

–N.i.

Sufferin

gof

severe

hypo

smiain

conjun

ctionwith

parosm

iaor

anosmiafollowinguseof

zinc

glucon

atege

l.

“Pesticides”

a Calvertet

al.1998[102]

Structural

fumigationworkers

Lifetim

edu

ratio

nof

methylb

romideand

sulfurylfluoride

expo

sure:1.2and

2.85

years

Sulfurylfluoridevalues

ofearlier

NIOSH

measuremen

ts:b

elow

20

123

120

UPSIT

(iden

tificationtest)

Sign

ificantlyweakerolfactoryfunctio

nin

workerswith

high

sulfurylfluoride

expo

sure

over

theyear

preced

ing

exam

ination.

Quand

tet

al.2016[104]

Latin

ofarm

workers

Lifetim

eexpo

sure.

Totalyearsin

jobs

involvingpe

sticide

expo

sure:farmworkers

13.11(m

ean),n

on-

farm

workers3.84

(mean)

304

247

Sniffin’Sticks

Kit

(iden

tificationtest

andthresholdtest

with

n-bu

tano

l)

Nodifferencein

odou

riden

tification

perfo

rmance

butsign

ificantlyhigh

erod

ourthresholds.

Form

alde

hyde

a Holmström

andWilhelmsson

1988

[111]

Workersat

achem

icalplant

whe

reform

alde

hyde

andprod

uctsbased

onform

alde

hyde

wereprod

uced

0.05–0.5form

alde

hyde

(forthegrou

pof

workersim

preg

natin

gpape

rup

to1)

7036

(office

workers

with

0.9mean

expo

sure

toform

alde

hyde

)

Sensorythreshold

testem

ploying

pyrid

ine

Sign

ificantlyredu

cedolfactory

functio

n.

a Holmström

andWilhelmsson

1988

[111]

Workwith

glued

woo

din

the

prod

uctio

nof

furnitu

re

0.2–0.3(fo

rmalde

hyde

)1–2(woo

ddu

st)

100

36Seeabove

Sign

ificantlyredu

cedolfactory

functio

n(but

nodifferencebe

tween

theform

alde

hyde

andthe

form

alde

hyde

-woo

ddu

stgrou

ps).

a Hisam

itsuet

al.2011[112]

Med

icalstud

ents

durin

gcadaver

dissectio

n

0.64–1.2(m

iddleof

the

labo

ratory),0.28–0.88

(inthecorners)

41–

Nagashimajet

nebu

lising

olfactiontestwith

brom

ine(detectio

nthreshold)

Sign

ificantlydiminishe

dolfactory

functio

n(32%

,tem

porary).


Table


ction:Hum

anstud

ies(in

chrono

logicalo

rder)(Con

tinued)

Hum

anstud

iesReference

Occup

ation

Expo

sure

(mg/m

3 )Expo

sedworkers,

reside

nts,patients

Con

trols

Testmetho

dResults

Kilburnet

al.1985[113]

Histology

technicians

0.25–2.38(exposureto

othe

rsolven

tssuch

asxylene

,toluene

andalso

chloroform

)

7656

Questionn

aire

Sign

ificantlymoreself-repo

rted

frequ

entredu

cedsenseof

smellin

histolog

ytechnicians

(32%

ofthewom

enwho

were

expo

sedto

form

alde

hyde

for1to

3handlikew

ise32%

ofthewom

enwho

wereexpo

sedforover

4h

stated

that

theirolfactoryfunctio

nwas

diminishe

d,5%

Co).

Edlinget

al.1988[114]

Workersat

different

plants(laminate

plants,p

article

boardplants)

0.1–1.1with

expo

sure

peaksof

upto

5(fo

rmalde

hyde

),0.6–1.1

(woo

ddu

st)

7525

Histological

exam

ination

Patholog

icalchange

sof

thenasal

mucou

smem

branein

72individu

als.

Acrylates

a Schwartzet

al.1989[119]

Workersat

achem

ical

facilitymanufacturin

gacrylatesand

methacrylates

0.0416–232.96ethyl

acrylate

andacrylic

acid

(mostly

arou

nd30

for

acrylic

acid

and20.8for

ethylacrylate)

55(high)

164(low)

512

UPSIT

(iden

tificationtest)

Noassociationin

thefirstinstance

betw

eenexpo

sure

andresults

ofolfactorytests.

a Schwartzet

al.1989[119]

embe

dded

case-con

trol

stud

ySeeabove

Seeabove

7777

Dose-effect

relatio

nshipbe

tween

olfactorydysfun

ctionandcumulative

expo

sure;the

effect

appe

ared

tobe

reversible.

Workerswho

hadne

versm

oked

had

thehigh

estrelativeriskof

olfactory

dysfun

ction.

Muttray

etal.1997[120]

Che

micalworkersin

acrylic

sheetprod

uctio

nMethylm

ethacrylate

(MMA):104–416(1988),

41.6–208

(1989–1994),

9.6±7.1years(m

ean

duratio

nof

MMA

expo

sure)

175

88Rh

inoIden

tification

Test(6

tested

arom

as,very

similarin

design

totheUPSITtest)

Nosign

ificant

differencebe

tween

expo

sedworkersandcontrolg

roup

.

Muttray

etal.2007[121]

Health

yvolunteers

208MMAandroom

airin

anexpo

sure

cham

berat

aninterval

ofon

eweek,in

each

case

for4h

20–

Olfactorythreshold

forn-bu

tano

l(Sniffin’Sticks)

Olfactorythreshold:

nochange

s.

Styren

ea Che

nget

al.2004[123]

Injection-mou

lding

workersexpo

sedto

acrylonitrile-butadiene

-styren

e(m

ainly

manufacture

ofcompu

tershells)

–52

72Olfactorythreshold

testem

ploying

1-bu

tano

l,CCCRC

olfactory

test(id

entification)

Slight

butsign

ificant

redu

ctionin

olfactoryfunctio

n(thresho

ld)at

the

endof

theshift

(whe

reas

theinitial

situationat

thebe

ginn

ingof

theshift

hadbe

enthesame).

Theiden

tificationtestrevealed

nodifferences

before

andaftertheshift.


Table


ction:Hum

anstud

ies(in

chrono

logicalo

rder)(Con

tinued)

Hum

anstud

iesReference

Occup

ation

Expo

sure

(mg/m

3 )Expo

sedworkers,

reside

nts,patients

Con

trols

Testmetho

dResults

a Daltonet

al.2003[124]

Workersin

the

reinforced

-plastics

indu

stry

Meanairborne

styren

econcen

trations:89.2

(day

1),106.1(day

2)Means

andrang

esof

historicexpo

suresto

airborne

styren

efor

individu

alworkers

(n=52):57.37

(15.16–134.23)

Cum

ulativemean

expo

sure:675.48

(59.75–1420.24)

Peak

year

expo

sure:

112.58

(22.52–331.25)

5252

Olfactorythreshold

forph

enylethyl

alcoho

land

styren

e,retron

asalod

our

iden

tificationtest

with

fivestim

uli,

iden

tificationtest

with

20od

orants

Nosign

ificant

differences

betw

een

styren

e-expo

sedworkersand

matched

controlsin

theresults

ofthe

phen

ylethylalcoh

olthresholdtest,

retron

asaltest,orod

ouriden

tification

test.

Sign

ificantlydifferent

odou

rde

tection

thresholds

forstyren

eam

ongexpo

sed

andun

expo

sedgrou

ps.

a Daltonet

al.2007[125]

Workersin

areinforced

plasticsbo

at-

manufacturin

gfacility

Calculatedeffective

meanconcen

tration

ofstyren

ein

air:43.3–108.25

(measuredairborne

styren

econcen

trationforthisgrou

p:303.1–346.4)

1515

Olfactorythreshold

forph

enylethyl

alcoho

l(PEA)and

styren

e,iden

tificationtest

with

18od

orants

Nosign

ificant

differencein

olfactory

thresholdforPEA.Significantdifference

inthethresholdtestinvolvingstyren

eam

ongexpo

sedandun

expo

sed

grou

ps.

Sign

ificant

differencein

the

iden

tificationtestbe

tweenworkers

with

high

vs.low

expo

sure

(possible

explanations:m

oreindividu

alswith

poor

values

inbo

thgrou

psthan

inno

rmalpo

pulatio

n,cultu

raland

educationald

ifferen

cesbe

tweenthe

twogrou

ps).

Organicsolven

tsandmineral

oilp

rodu

cts

a Schwartzet

al.1990[130]

Workersin

paint

manufacturin

gfacilities

Solven

ts:toluene

,xylen

eand

methylethylketone

Lifetim

ehydrocarbo

ndo

sein

ppm

years(average

sat

the

twoprod

uctio

nplants):180

(±128),97(±70)

187

–UPSITiden

tification

test

Sign

ificant

dose-dep

ende

ntde

terio

ratio

nin

olfactoryfunctio

nwith

risinglifetim

eexpo

sure

amon

gworkerswho

hadne

versm

oked

.

a Sandm

arket

al.1989[131]

Painters

Noqu

antitativeexpo

sure

measuremen

t(described

as“lo

wto

mod

erate”)

5442

UPSITiden

tification

test

Nostatisticallysign

ificant

impairm

ent

ofolfactoryfunctio

nafteradjustmen

tforageandsm

okinghabits.

Stud

iescitedin

Muttray

etal.

1998

[133]:(Drago

mitretzky

etal.1970)

(Kmita

1953)

(Latkowskietal.1981[134])

Workersat

ashoe

factory,

petroleum

chem

istry

workers

Ape

troleum

mixture

(“Galoscha”),ethylacetate

andbu

tylacetate

(solvent

concen

tration:220–300),

rubb

eradhe

sive

containing

petroleum

(upto

3000)

N.i.

216

205

547

N.i.

N.i.

100(cottonindu

stry

workers)

N.i.

N.i.

Smelland

taste

tests

31%

sufferedtheloss

ofolfactory

functio

n,andam

ongtheremaind

er,

olfactoryfunctio

nwas

impairedin

comparison

toacontrolg

roup

.Olfactorydysfun

ctionas

aresultof

rhinitisaroseafterafew

years.

Hyposmiain

238(43.5%

)andanosmia

in50

(9.2%)subjects.

Dysge

usiain

319(58.3%

).


Table


ction:Hum

anstud

ies(in

chrono

logicalo

rder)(Con

tinued)

Hum

anstud

iesReference

Occup

ation

Expo

sure

(mg/m

3 )Expo

sedworkers,

reside

nts,patients

Con

trols

Testmetho

dResults

(Co:dysgeusiain

69%,ano

smiain

24%

ofsubjects).

a Ahlstrom

etal.1986[132]

Tank

cleane

rsMineraloilp

rodu

cts(heavy

andlight

oils,hydrocarbon

conten

t:240–1615)

2040

Thresholdtest,

perceivedod

our

intensity

test

(pyridine,dimethyl

disulfide

,n-butanol

andhe

ated

oil

vapo

ur)

Elevated

olfactorythresholdvalues

for

heated

oilvapou

randn-bu

tano

lin

comparison

tocontrols(fo

rn-bu

tano

lwith

intheno

rmalrang

e,oilvapou

rhadno

tbe

enstud

iedin

thisrespect

before).

Highe

rthresholdvalues

foralltested

substances

exhibitedby

tank

cleane

rsexpo

sedon

edaypriorto

thetestin

comparison

toindividu

alsexpo

sed

earlier.

Lower

odou

rintensity

values

atthe

loweststim

ulus

concen

trations

forall

4substances

incomparison

tothose

ofcontrolsub

jects(“o

dour

intensity

recruitm

ent”).

N.i.no

tindicated,

Cocontrolg

roup

,astud

ieswhich

have

alread

ybe

endiscussedin

thereview

sby

Sund

erman

,Gob

baan

dDoty[4,2

1,28

]


Nickel exposure With the exception of the studiesstated by Sunderman, we are not aware of any studiesinto the association between workplace nickel exposureand olfactory dysfunction. Sunderman [28] cites twostudies from the former Soviet Union from 1960 and1970, according to which workers exposed to nickel inelectrolytic nickel refineries exhibited olfactory dysfunc-tion/anosmia in addition to atrophy of the nasal mucousmembrane, chronic sinusitis and ulceration of the nasalseptum.In the studies cited concerning cadmium, the workers

were often exposed not only to cadmium, but also tonickel (see above). Whereas in the 1940s, occupationalnickel exposure of 10–150 mg/m3 for example was pos-sible [29], considerably lower values of up to 0.056 mg/m3 were measured in the publication by Adams andCrabtree in 1961 [33]. ACGIH currently sets a TLV of1.5 mg/m3 for the inhalable fraction of nickel.

Animal experimental studies

Cadmium exposure Following inhalation tests on rats(250 and 500 μg/m3 CdO, 5 h per day, 5 days per week for20 weeks), an elevated cadmium level was determined inthe olfactory bulb. This was accompanied neither by sig-nificant histopathological changes in the mucous mem-brane, nor by a reduction in the olfactory function [38].In another animal experiment, administration of 400 μg

of CdCl2 to mice by intranasal instillation resulted in par-tial damage to the olfactory epithelium, reversible loss ofolfactory discrimination, and specific cadmium depositionin the olfactory bulb but not in other parts of the centralnervous system [39]. Czarnecki et al. [40] also observedanosmia in a behavioural test following intranasal instilla-tion of a cadmium chloride solution in mice. They furtherdemonstrated a dose-dependent reduction in theodour-induced release of neurotransmitters from the ol-factory nerve into the olfactory bulb. Moreover, a 20%drop in the dendrite density of the olfactory epitheliumwas described at the highest dose (20 μg CdCl2). In furtherexperiments, Czarnecki et al. demonstrated a clear cad-mium accumulation on mice specifically in the olfactorybulb by bilateral instillation of a buffer solution of 20 μg ofCdCl2 per nostril. The accumulation was still measurable4 weeks after exposure. A reduction in the axonal termi-nals of the olfactory receptor neurons was also demon-strated histologically. A decrease in neurotransmitterrelease in response to olfactory stimulation was detectedin vivo on the mice exposed to cadmium (intranasal instil-lation with 0.2, 2 and 20 μg CdCl2) 2, 7 and 28 days afterexposure. After the laboratory animals treated with cad-mium had exhibited significant olfactory deficits in a be-havioural test, these deficits disappeared after two weeksof olfactory training; however, the mice with restored

olfaction continued to exhibit damaged projections of theolfactory receptor neurons in the results of opticalimaging. Czarnecki et al. conclude from this that restor-ation of olfactory function is attributable to neuroplasti-city: the brain, they assume, has learnt to reinterpret thereduced stimuli appropriately. Such processes of neuronalplasticity could mask severe damage by neurotoxic sub-stances [41].Cadmium-induced olfactory impairment was also con-

firmed in fish [42]: after 8-h exposure to 347 ppb of Cdin fresh seawater, coho salmon exhibited not only histo-logical changes to the olfactory epithelium and dimin-ished olfaction in the behavioural test (for example lossof the tonic immobility response to olfactory alarm sig-nals), but also significantly reduced expression of olfac-tory receptors and increased expression of enzymesinvolved in the antioxidant reaction in relation tometals. During 48-h exposure to 3.7 ppb, tonic immo-bility responses were diminished and histologicalchanges to the olfactory epithelium likewise occurredthat were not as pronounced as in the highly exposedfish group.

Nickel exposure Inhalation of NiSO4 (0.635 mg, 6 h perday, 16 days) caused atrophy of the olfactory epitheliumin rats, α-Ni3S2 additional chronic inflammation of thenasal tissue. Significant impairment of olfaction was notrecorded [43–47]. Studies on rats and apes confirm thetransport of nickel into the olfactory bulb following in-halation of soluble NiSO4 [47]. Following intranasal in-stillation of 63Ni2+ in rats, the uptake pathway wastracked from the olfactory epithelium, via the axons ofthe primary olfactory neurons, into the glomeruli in theolfactory bulb and into further parts of the brain [48]. Amaximum 63Ni2+ transport rate of 0.13 mm/h was mea-sured in the olfactory neurons of pike [49]. In a recentexamination by Jia et al. [50], intranasal instillation ofnickel sulfate (0.5 and 2.5 mg/kg) in mice led todose-dependent and time-dependent atrophy of the ol-factory epithelium of the turbinate bone, but not of theseptum. The sustentacular cells were affected first byapoptotic cell loss on the first day post exposure, the ol-factory receptor neurons on the third day; a significantincrease in cell proliferation in the olfactory epitheliumwas detected after 5 to 7 days.Neuronal signal transduction by calcium and apoptosis

is a factor in olfactory impairment by nickel: accordingto Zhao et al. [51], NiSO4 is capable of inducing apop-tosis by activation of the death receptor 3 and caspase-8and subsequent activation of caspase-3; Jia et al. suspectNiSO4-induced apoptosis of the olfactory receptor neu-rons to be attributable to this mechanism. Moreover, ac-cording to Gautam et al. [52], Ni2+ can reduce theodour-induced calcium influx by inhibition of the T-type


Ca2+ channels in the olfactory receptor neurons, therebyimpairing signal transduction.

ChromiumHuman studiesExposure to chromium is frequently encountered incombination with nickel and other metals. According toSeeber et al. [53], ulcers on the skin and mucous mem-brane and perforation of the nasal septum caused bychromium were known as long ago as 1826. Few epi-demiological studies exist of a possible association be-tween chromium exposure and olfactory dysfunction.In all 5 human studies known to us, deficits in the ol-

factory function of the exposed workers were detectedthat were associated significantly with the chromium ex-posure and the duration of employment [53–57]. Thestudy by Seeber et al. and the follow-up research by See-ber and Fikentscher of 1976 and 1980 were not men-tioned in the reviews by Gobba, Sunderman and Doty,and are described accordingly in more detail here: See-ber et al. (1976) and Seeber and Fikentscher (1980) re-ported on damage to the nasal mucous membrane andolfactory dysfunction among workers at a chrome paintplant in the German Democratic Republic (GDR) inwhich basic zinc chromate (zinc yellow) was manufac-tured [53, 54]. In their comparison between chronicallyexposed workers (5), intermittently exposed workers (14,not longer than 2 h per day), mask-wearing workers (5,wearing fine-dust filter masks for the full duration),non-exposed individuals (9 office workers at the samecompany) and a control group (23 employees at a hos-pital), they determined by means of rhinoscopic exami-nations that damage to the nasal mucous membrane wasevident on all five of the chronically exposed individuals,but occurred only in a part of the intermittently exposedgroup and among the workers wearing masks, and notat all in the other groups. Olfactory tests involving odourstrips for ascertaining the sensitivity to certain sub-stances showed the olfactory sensitivity of the controlgroup and the non-exposed individuals to be substan-tially higher than that of the exposed individuals. Theauthors considered this relationship between chromiumexposure and loss of olfaction to be “significant”. Ac-cording to the authors, the dust values measured at thepoints at which zinc chromate dust was produced clearlyexceeded the occupational exposure limit for chromium(VI) applicable in the GDR at this time of 0.1 mg CrO3/m3 air, and rose at certain points to up to 200 times theoccupational exposure limit. More precise values are notstated. Four years later, after violation of the occupa-tional exposure limit in this plant had been “substantiallyreduced” by suitable measures (not specified), the work-force was examined once again. It was found that in 16workers exposed to different levels, the relationship

between the pathological nasal mucous membrane find-ings and olfactory dysfunction was confirmed, and thaton average, no improvement in the mucous membraneor olfaction was detected [54].Besides the results obtained by Watanabe and Fukuchi

(1981) and Kitamura et al. (2003), in which an impairedolfactory function in workers in the chromate and galva-nising industry was detected by means of the T&T ol-factometer and which have already been described indetail in Gobba 2006, workers in galvanising are alsoshown to be affected by olfactory dysfunction in an In-dian publication from 2003: the authors reported on 28workers in the chromium industry aged between 22 and37 and exposed to chromium for between 5 and 14 years.Of these, all 28 employees exhibited nasal septumperforations, and 11 were anosmic. Information on thelevel of exposure and on the test method was not pro-vided [57].Kitamura et al. reported chromium-induced olfactory

dysfunction at low workplace concentrations [56]. Theexposure values for chromium measured in this studywere 0.0047 to 0.059 mg/m3. ACGIH set a TLV of0.05 mg/m3 for water-soluble Cr (VI) compounds and0.01 mg/m3 for insoluble Cr (VI) compounds.

Animal experimental studiesIn his review of the relationship between exposure tometals and nasal toxicity, Sunderman cites an animal ex-periment on rats. Following 40 days’ inhalation of so-dium dichromate (0.2 mg/m3, 6 h per day, 40 days), therats exhibited no morphological nasal changes. Olfactionwas not tested [28].

ManganeseHuman studiesAccording to the results of our searches, 10 human stud-ies are available to date that examine the association be-tween manganese exposure and impairment of the senseof smell [58–67]. The study populations encompass notonly exposed workers, but also persons living close to amanganese mine and young people living in a region inwhich manganese emissions of industrial origin were veryhigh prior to 2001. In the two studies conducted onworkers in the metals industry [58, 59], in which only thesensory test method was employed, the results were eithernot significant [58] or, surprisingly, revealed a significantlyincreased olfaction perception among the workers in mea-surements of the sensory olfactory threshold [59]. By com-parison, use of the olfactory identification test employingSniffin’ Sticks on welders or the inhabitants of an areawith elevated background manganese values revealedsignificantly poorer values [60–62, 64–66]. In furthermagnetic or functional resonance imaging studies, ele-vated manganese deposition in the olfactory bulb was


measured on welders [63] and a reduction in activity inrelevant olfactory brain regions was measured in youngpeople living in a region exhibiting elevated manganesevalues [66].The following recent studies have not yet been dis-

cussed in the reviews by Sunderman, Gobba and Dotyand will therefore be presented in more detail here:In a follow-up survey, 26 welders who had participated

in the San Francisco/Oakland Bay Bridge welder study[60, 61] were examined three and a half years later bythe same methods. Although 13 participants were nolonger working as welders, the results obtained in theUPSIT did not differ significantly from the earlier find-ings. The blood-manganese levels of the workers whowere no longer welding were significantly lower thanthose of their colleagues [62].For seven welders, increased manganese deposition in

the olfactory bulb and in other regions of the brain wasdemonstrated by means of functional magnetic reson-ance tomography [63].Guarneros et al. (2013) conducted a Sniffin’ Sticks ol-

factory test series encompassing threshold, discrimin-ation and identification tests on persons living inproximity to a manganese plant and exhibiting an ele-vated manganese concentration (mean 9.73 μg/g vs.1.01 μg/g) in their hair [64]. Significant differences be-tween the subjects were observed. In this study, 30 per-sons (non-smokers) living within a one-kilometre radiusof a Mexican manganese mine were compared with 30controls living more than 50 km away. The groups werematched by age, sex and school education; none hadpreviously worked in a job involving manganese expos-ure. The results of the Sniffin’ Sticks test revealed sub-stantially diminished olfactory function in those livingclose to the manganese mine.By means of the identification test employing Sniffin’

Sticks, Lucchini et al. also documented significantlypoorer olfactory function associated with the Mn con-tent in the soil on 154 young people aged between 11and 14 in Valcamonica (Italy). Up until 2001, this regionwas marked by ferrous alloy plants and the emissionsfrom them (average Mn atmospheric and soil values atthe time of the examination: 49.5 ng/m3 and 958 ppmrespectively). Young people from the region around LakeGarda were tested as the control group [65]. In a furtherstudy employing functional magnetic resonance tomog-raphy, the activity of the brain in 9 young people fromValcamonica and Bagnolo Mella exposed to manganesewas compared with that of 4 young people from theLake Garda region. In the exposed young people, a re-duction in activity in relevant olfactory brain regionswas detected, for example in the orbitofrontal cortexand piriform cortex and in further brain regions typicallyassociated with olfactory function, such as the middle

frontal gyrus and cerebellum. In comparison with a lar-ger control group from a database, significant differenceswere also monitored with regard to the size of the olfac-tory bulb and in the olfactory test involving Sniffin’Sticks. Reduced activity in comparison with that of thecontrols was also noted in the regions of the limbic sys-tem [66].In a recent prospective cohort study, Casjens et al.

examined the influence of work-related manganese ex-posure upon the olfactory function. The study popula-tion comprised 1385 men, of whom 354 had potentiallybeen exposed to manganese in their earlier occupations.No relevant association was determined between man-ganese exposure and a deterioration in olfactory func-tion [67].

Animal experimental studiesExperiments on pikes and rats showed that following in-tranasal application of a dilute 54MnCl2 solution, manga-nese is absorbed by the olfactory epithelium andtransported on into the brain. In the process, manganeseaccumulates in the olfactory bulb and can be detectedafter 12 weeks throughout the brain and spinal cord[68–71]. Transport of the manganese from the nasalcavities into the brain requires the axonal projections ofthe olfactory epithelium’s receptor neurons to be intact[72]. Foster et al. also evaluated the transport of manga-nese from the olfactory epithelium to the olfactory bulb:a bilateral instillation of 40 μl 200 mM MnCl2 in ratsleads to an increase in manganese levels in both the ol-factory epithelium and the olfactory bulb, and the ratsexposed to manganese exhibit decreased performance inthe olfactory discrimination task. Manganese accumula-tion in the olfactory bulb and in other regions of thebrain was also demonstrated by MRT studies onnon-human primates exposed to aerosolised MnSO4 (≥0.06 mg Mn/m3) [71].

ZincExposure to zinc in the form of fumes and dust fre-quently occurs during the manufacture and processingof metals. With the exception of one study published in1971 in Russian [73], we are not aware of any studies ofa possible relationship between the incidence of olfac-tory dysfunction and occupational zinc exposure. Studiesdo exist of the frequent incidence of anosmia followingmedical intranasal application of sprays or gels contain-ing zinc, as do studies on animals demonstrating an as-sociation between intranasal exposure to zinc salts andadverse influence upon olfactory function [74–94].

Human studiesIn 1971, 301 workers at a zinc production plant were ex-amined with regard to their olfactory and trigeminal


function and compared with a control group comprising63 workers at a machine factory [73]. The olfactory testswere performed by means of an olfactometer employingmint and dilute acetic acid (trigeminal stimulation). Incomparison with the sensory threshold values of thecontrol group the sensory thresholds of the exposedworkers were statistically significantly elevated. The ele-vations were highest among the workers responsible forroasting the zinc ore. High concentrations – accordingto the author several times higher (without closer speci-fication) than the limits in force at the time – of furtherstrongly corrosive and irritant substances such as sulfurdioxide, sulfur anhydride, sulfuric acid, chlorine, hydro-gen fluoride and others were however released in allthree working areas (roasting, leaching, electrolysis) cov-ered by the study, besides zinc oxide, zinc sulfate andmetal dusts. Assessing the impact upon health to a par-ticular substance is therefore difficult.The link postulated by Seeber and Fikentscher between

the effects of olfactory impairment observed amongworkers in a zinc chromate plant and exposure to chro-mium may perhaps equally be associated with zinc. Thispossibility was not examined in these studies [53, 54].The suspicion that zinc in the form of a pharmaceut-

ical component was capable of triggering olfactory dys-function dates back to the 1930s. The suitability of asolution containing 1% zinc sulfate and 0.5% tetracaine(a topical anaesthetic) for use as a nasal spray for pre-vention of poliomyelitis infection was studied in Torontoin 1938 [28]. It was found not only that the desired pro-tection was not achieved, but that some children andadults also developed anosmia. According to reports bythe British Medical Journal and the Journal of Pediatrics,of 5233 children (for the most part aged between 3 and10) treated in special clinics by otolaryngologists, ap-proximately a quarter exhibited temporary anosmia [74,75]. Unfortunately, the documentation contains no de-scription of the diagnostic method, quantitative detailsof temporary and permanent olfactory dysfunction, orany indication whatsoever of systematic olfaction testingin the control group [74, 75]. Information on the controlgroup and the test method were relevant insofar as re-cent studies indicate that olfaction is poorer in childrenthan in adults: in the study by Sorokowska et al. of 1422test subjects (aged between 4 and 80), children agedunder 10 and adults aged over 70 performed worst in anidentification test involving Sniffin’ Sticks with 16 differ-ent odours [95]. Tisdall et al. (1938) were in possessionof data from a collective of an estimated 5000 furtherpatients (children and adults) treated with zinc sulfate ofwhich 44 were permanently anosmic as a consequenceof the treatment. However, in this collective the 44 pa-tients identified as having permanent olfactory dysfunc-tion accounted for fewer than 1%, which is below the

estimated figure for olfactory dysfunction in the widerpopulation [16, 96].Olfactory dysfunction occurring in patients following in-

tranasal use of Zicam gel, claimed by the manufacturer tobe “homeopathic” (according to Mossad, Zicam nasal gelcontains 33 mmol/l zinc gluconate [97]) for prophylacticor therapeutic purposes against the symptoms of colds,was documented by Davidson and Smith, Alexander andDavidson, and Jafek et al. [76–78].In 2009, the US Food and Drug Administration (FDA)

issued warnings to consumers against three intranasal“Zicam” products containing zinc owing to the suspicionthat they “may cause a loss of sense of smell”, possiblypermanent [98]. The products concerned were thentaken off the market in 2009.The possible olfactory effect of a nasal insulin spray

containing zinc as an additive was also the subject of re-cent discussion [99, 100]. Indeed, it is unclear how muchof the zinc spray actually reaches the olfactory cleft. Not-able in this context are the results of experiments byHerranz Gonzalez-Botas and Padin Seara, who examinedthe efficacy of the nasal gel form of application andascertained that pigmented nasal gel is not detectable inthe olfactory cleft following self-application by 16 testsubjects [101].

Animal experimental and in vitro studiesThe cytotoxic effect of Zicam was also demonstrated invivo on mice and in vitro on human nasal tissue. Follow-ing instillation of Zicam by injection in the nasal cavities(15 μl per cavity), the olfactory epithelium was especiallyaffected. The results of the behavioural test showed thattreatment of the mice with Zicam led to olfactory dys-function that still persisted two months after treatment.In the human nasal tissue samples, necrosis of the epi-thelial and subepithelial structures was observed follow-ing the application of Zicam [80].Numerous further histological studies have been

performed that illustrate the degenerative effect ofzinc salts upon the olfactory epithelium in mice andfish [81–86]. Several studies in which ZnSO4 is usedfor experimental induction of anosmia in laboratoryanimals demonstrate that direct treatment of theolfactory mucous membrane with zinc sulfate solutionimpairs olfaction in mice, rats, hamsters and pigeons[81, 84, 87–91]. McBride et al. (2003) provide anoverview of 22 behavioural studies on mice in whicholfactory dysfunction was induced by means of intra-nasal irrigation with ZnSO4 [81].Following intranasal zinc gluconate instillation on mice

(33 mM, 50–100 μl per nostril), Duncan-Lewis et al.(2011) were able to demonstrate, by means of a behav-ioural test, a significant reduction in olfaction comparedto control mice treated only with phosphate-buffered


saline (PBS) [92]. Significantly weaker olfaction was alsoexhibited by mice following treatment with copper gluco-nate. Irrigation with magnesium gluconate yielded no dif-ferences. The olfactory dysfunction was reversible, sinceno differences in olfactory behaviour were observed in afurther behavioural test performed one month aftertreatment.Electroolfactograms and patch-clamp tests on isolated

rat olfactory epithelia showed that exposure to zincmetal nanoparticles in the picomolar range had a signifi-cantly reinforcing effect upon the reaction of the olfac-tory neurons following olfactory induction, whereas Zn2+ ions in the same concentration led to a reduced re-sponse to olfactory stimulation [93].The toxic properties of zinc oxide nanoparticles in

the olfactory system of rats are presented by Gaoet al.: once-off instillation of a suspension of zincoxide nanoparticles led to significant damage of theolfactory epithelium and to inflammation reactions[94]. In addition, the exposed rats exhibited a changein sniffing behaviour and appeared no longer able todistinguish between vanillin diluted with distilledwater and distilled water alone. In an in-vitro assayon primary human olfactory cells, Osmond-McLeodet al. demonstrated that zinc oxide nanoparticles areable to induce cellular stress reactions, inflammationreactions and apoptosis, but do not activate DNArepair mechanisms [79]. They established that thecellular reactions to zinc oxide nanoparticles with acoated surface were weaker.

“Pesticides”Human studiesAs already reported by Doty the neurological functions ofworkers employed in structural fumigation involving thepesticides of methyl bromide and sulfuryl fluoride werecompared in 2015 in a cross-sectional study with those ofcontrol persons [102]. Significantly weaker olfactory func-tion tested with UPSIT was observed among the workerssubject to high sulfuryl fluoride exposure. A correspond-ing observation was not made for the workers with highexposure to methyl bromide. It should be noted here thatwith the exception of 11 individuals, the majority ofworkers were subject to coexposure with methyl bromide.Clear distinction between the effects of methyl bromideand those of sulfuryl fluoride is therefore difficult, notleast since division of the workers by the criteria of “highexposure to sulfuryl fluoride” and “high exposure to me-thyl bromide” is based upon statements made by theworkers themselves in a questionnaire. Methyl bromide ishighly toxic and harmful to the central nervous system. Inanimal experiments it also causes damage to the olfactoryepithelium. According to the authors, further pesticides(including chlorpyrifos, organophosphates, carbamates,

pyrethrins, organochlorine pesticides) with which the testsubjects had come into contact during work or leisure re-vealed no significant association with the results of the ol-factory test and the memory test involving patterns.During work, the fumigators were also exposed to “smallamounts” of chloropicrin, which was used as an irritantwarning substance during fumigation with sulfuryl fluor-ide and methyl bromide (no concentration stated). Chlo-ropicrin is a highly irritant gas [103].In one recently published study an impaired olfactory

function was demonstrated in Latino farmworkers ex-posed to pesticides [104]. 304 farmworkers exposed topesticides were compared with 247 non-farmworkers. Atsignificantly greater self-reported lifetime pesticide ex-posure, the farmworkers required significantly higherconcentrations for odour detection; the odour identifica-tion did not differ between the groups (Sniffin’ Sticks).Unfortunately, it is not specified which types of pesti-cides the farmers worked with. In this context it is inter-esting that an impaired olfactory function could be anearly symptom of Parkinson’s disease (PD), and that pes-ticides are also suspected of inducing symptoms of Par-kinson’s. A strong association between farmers with aPD diagnosis and a reduced sense of smell is shown byShrestha et al. 2017 [105].

Animal experimental studiesWhereas methyl bromide can strongly damage the ol-factory epithelium in animal tests [106], the availableliterature provides no clear indication of the effect ofsulfuryl fluoride upon olfaction. Inflammation of thenasal tissue was reported after high exposure to sulfu-ryl fluoride (inhalation of 300–600 ppm) in rats andrabbits [107, 108].

FormaldehydeOlfactory dysfunction in humans caused by formalde-hyde was described relatively early. Spealman [109] pro-vides an indirect indication in that he cites the medicaldepartment of an airline that had rejected the use of de-odorants containing formaldehyde in airliners on thegrounds that formaldehyde was known to impair olfac-tion. Lehnhardt and Rollin mention the strange case of acompany emergency response officer who developed an-osmia allegedly owing to compulsive sniffing of a thera-peutic agent that released formaldehyde [110].

Human studiesFour human studies involving formaldehyde and its associ-ation with olfactory dysfunction were identified [111–114].In an epidemiological cross-sectional study by Holmström

et al. of workers at a factory producing formaldehyde andformaldehyde-based products and workers exposed to bothformaldehyde and wood dust, significantly reduced olfactory


function was measured (sensory threshold test employingpyridine) [111]. Hisamitsu et al. demonstrated a significantlydiminished olfactory function in medical students who wereexposed to formaldehyde vapours during an anatomy coursein Japan. It must be emphasised that all affected individualsalready had a preexisting history of allergic rhinitis. Olfactoryfunction was fully restored after a year [112]. The publica-tions of Holmström et al. and Hisanitsu et al. [111, 112] arementioned by Doty 2015 [4] and will not be presented ingreater detail here.Kilburn et al. describe the result of a survey involving

76 female histology technicians (average age 40.8) whowere exposed to formaldehyde (0.2–1.9 ppm) and toother substances such as xylene and toluene during theirwork with histological preparations [113]. 22 of thesewomen were exposed to formaldehyde for 1 to 3 h, 47for over 4 h. In a survey, 32% of the women in each sub-group stated that their olfactory function was dimin-ished. Among a control group of 56 non-exposedwomen (average age 39.5), this was stated by only 5%.A study of 75 male workers (average age: 38) with oc-

cupational exposure either to formaldehyde alone (0.1–1.1 mg/m3 with exposure peaks of up to 5 mg/m3) or toformaldehyde and wood dust (0.6–1.1 mg/m3) detectedpathological changes of the nasal mucous membrane in72 individuals [114]. No differences in the histologicalfindings were observed between the workers exposedsolely to formaldehyde and those who had also been ex-posed to wood dust.Based upon the results of long-term toxicological inhal-

ation studies on laboratory animals, an NOAEL for nasaldamage of 1 ppm formaldehyde was determined in the lit-erature [115]. In their cross-sectional study, Holmströmand Wilhelmsson observed a significant worsening of thesensory threshold even at formaldehyde concentrationsbelow 0.37 mg/m3 (= 0.3 ppm, and corresponding to theGerman OEL) [111]. Irreversible damage to the sensorytissue need not however be anticipated below the irrita-tion threshold, upon which the OEL is based.

Animal experimental studiesInhaled formaldehyde (3.2 ppm) and acrylic aldehyde(0.67 ppm) led to degeneration of the respiratory epithe-lium in rats, inhaled acetaldehyde (750 ppm) to degener-ation of the olfactory epithelium. It was demonstratedon the same species of laboratory animal that the effectsof a combined exposure to these aldehydes can increaseexponentially [116].10 rats were exposed for 4 h per day for 7 days to

12.5 mg/m3 formaldehyde by Li et al. [117]. Examinationof the olfactory bulbs and hippocampi of the exposedrats revealed severe morphological damage compared toan untreated control group. Reduced production of

glutamate, gamma-aminobutyric acid and nitrogen oxidesynthase was also detected in this damaged tissue. Sincethe publication was in Chinese, only the English abstractcould be evaluated here.Diminished olfactory function was determined in rats

that had been subject for 30 min twice a day for 14 daysto inhalative exposure to a formaldehyde concentrationof 13.5 ± 1.5 ppm [118].

AcrylatesHuman studiesGobba and Doty have already described the study bySchwartz et al. in their reviews: hundreds of workerswere studied in a cross-sectional analysis at a factoryproducing acrylates and methacrylates [119]. Theworkers were divided into four classes according to theirexposure. No association was established in the first in-stance between exposure and the results of olfactorytests (UPSIT, p = 0.09). By means of an embeddedcase-control study, a dose-effect relationship was ob-served between olfactory dysfunction and cumulative ex-posure; the effect appeared to be reversible. The risk ofdeveloping olfactory dysfunction was greatest in thegroup of non-smokers.Two studies were, however, not mentioned by Gobba

or Doty:In a cross-sectional study of 175 workers exposed to

methyl methacrylate and a control group of 88non-exposed workers employed in the same productionunit for acrylic glass sheet casting, performance of theRhino Identification Test (6 tested aromas, very similar indesign to the UPSIT test) revealed only a single hyposmiccase in the exposed group [120]. At 58.3%, the proportionof smokers was higher in the exposed group than in thecontrol group (34.1%). Over an 8-h shift, exposure lay be-tween 25 and 100 ml/m3 in 1988 and between 10 and50 ml/m3 in the period from 1989 to 1994. The averageexposure duration was 9.6 ±7.1 years. With the exceptionof 2 workers who were additionally exposed briefly to for-maldehyde up to 1990 and 4 workers who additionallyhad contact with acrylonitrile, and a further 2 workerswho were additionally exposed to both formaldehyde andacrylonitrile, all workers were exposed solely to methylmethacrylate. These results permit the conclusion that atexposures of up to 50 ml/m3 methyl methacrylate, olfac-tory function is not harmed.In a more recent exposure study, 20 healthy male vol-

unteers (non-smokers, aged 20–62) were exposed in anexposure chamber once to 49.2 (±1.4) ppm methylmethacrylate for 4 h. Following exposure, no changesoccurred in the olfactory threshold for n-butanol, whichwas measured by means of Sniffin’ Sticks, nor had themeasured mucociliary transit time (time from


introduction of a saccharin particle in the lower nasalvestibule to perception of a sweet taste in the throat)changed. Also unchanged were the measured concentra-tions of protein and mRNA markers of inflammation inthe nasal secretion and respiratory epithelium. Further-more, only minor differences were observed in the men-tal state, which was evaluated by questionnaire. Theauthors concluded from this that acute exposure to50 ppm methyl methacrylate causes no inflammatorychanges to the respiratory nasal mucous membrane, andthat in view of the absence of a rise in the olfactorythreshold following exposure to 50 ppm methyl meth-acrylate, this dose is not sufficient to cause toxic damageto the olfactory epithelium. These results of acute expos-ure cannot be readily transferred to chronic conditions[121].

Animal experimental studiesChronic exposure to 100 ppm methyl methacrylate cancause degeneration and atrophy of the olfactory epithe-lium in rats. It must be considered that the activity ofthe carboxylesterase in the olfactory epithelium of thenasal mucous membrane is considerably higher in ratsthan in humans. The unspecific carboxylesterase hy-drolyses methyl methacrylate to methyl acrylic acid,which is responsible for the local toxicity [122].

StyreneHuman studiesThree studies have already been listed by Doty [4]:Cheng et al. [123] compared injection moulding

workers exposed to styrene with non-exposed workers.At the end of a working day, a slight but significant re-duction in olfactory function was detected in the olfac-tory threshold test employing 1-butanol, whereas theinitial situation at the beginning of the shift had beenthe same. An identification test employing 7 odours(Connecticut Chemosensory Clinical Research Center(CCCRC) olfactory test) revealed no differences beforeand after the shift. According to Doty [4] these resultssupport the “concept that the olfactory thresholds reflectadaptation rather than sustained neurological damage”.Unfortunately, no exposure measurements were performedin this study.Other epidemiological studies of workers in the glass

fibre reinforced plastic industry revealed no relationshipbetween styrene exposure and a general deterioration inolfactory function [124, 125]. In both studies the olfac-tory threshold for styrene was significantly higher amongthe exposed workers. According to Dalton et al., this isalso explained by an adaptation effect, which leads to areversible reduction in sensitivity. This has already beenfrequently observed for volatile substances in industry orthe laboratory and is correspondingly well documented

[126]. Whereas the identification test revealed no differ-ences between exposed and non-exposed individuals inthe study published in 2003, the identification test pub-lished in 2007 resulted in a significant difference be-tween the workers with high vs. low exposure [125].Dalton et al. state that the proportion of individualsshowing poor values in the identification test is substan-tially higher in both groups (40 and 20%) than in thenormal population (10%) [125]. Even with the aid ofmultiple regression analysis, no association was demon-strated between the results of the identification test andthe exposure values measured at present or in the pastin the group of workers subject to high exposure. Theauthors therefore suspect that the high proportion ofimmigrants in the exposed group had influenced the re-sults of the test, and they do not consider the results tobe valid evidence of impairment of human olfactoryfunction by workplace exposure to styrene.

Animal experimental studiesIn two studies on rodents conducted in 1997 and 1998,already cited by Doty 2015, styrene exposure of between20 and 50 ppm led to lesions of the olfactory epithelium[127, 128]. Green et al. assumed that the nasal lesionsinduced in mice by exposure to styrene were caused bystyrene oxide, which cannot be detected in the humannasal epithelium [129].

Organic solvents and mineral oil productsHuman studiesThe publications by Schwartz et al. [130], Sandmark etal. [131] and Ahlstrom [132] are epidemiological studiesevaluating a possible association between solvent expos-ure and impairment of olfaction. They are also cited byGobba and Doty:With consideration for smoker status, the UPSIT iden-

tification test performed by Schwartz et al. [130] yieldeda significant dose-dependent deterioration in olfactoryfunction of workers exposed to solvent in paint manu-facture with rising lifetime exposure among workerswho had never smoked. This effect was not observedamong workers who had always smoked. The resultsremained the same when the confounders of age andcultural background (vocabulary testing) were taken intoaccount. Schwartz et al. suspect that the induction ofcytochrome P450 enzymes by cigarette smoke leads toan increase in the metabolism and thereby to detoxifica-tion of organic olfactotoxins before they reach the olfac-tory epithelium. The exposed workers who had alwayssmoked performed worse in the UPSIT olfactory testcompared to reference values, albeit not with dose de-pendency. The best olfactory function was exhibited bythe workers with the lowest exposure who had neversmoked.


Documents

Olfactory dysfunction revisited: a reappraisal of work-related … · 2018. 9. 4. · paper on functional testing and dysfunction of olfaction, Herberhold [23] assumed several decades