One Heal h 13 (2021) 100324
A ailable online 9 Sep embe 2021
2352-7714/© 2021 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY-NC-ND license
(h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/).
Mapping he scien i ic knowledge o an imic obial esis ance in
ood-p oducing animals
Ri a Tinoco To es
a
,
*
, Jo˜
ao Ca alho
a
, Joana Fe nandes
a
, Josman D. Palmei a
a
,
M´
onica V. Cunha
b
,
c
, Ca los Fonseca
a
,
d
a
Depa amen o de Biologia & CESAM, Uni e sidade de A ei o, Campus de San iago, 3810-193, A ei o, Po ugal
b
Cen e o Ecology, E olu ion and En i onmen al Changes (cE3c), Faculdade de Ciˆ
encias, Uni e sidade de Lisboa, 1749-016 Lisbon, Po ugal
c
Biosys ems & In eg a i e Sciences Ins i u e (BioISI), Faculdade de Ciˆ
encias, Uni e sidade de Lisboa, 1749-016 Lisbon, Po ugal
d
Fo es WISE - Collabo a i e Labo a o y o In eg a ed Fo es & Fi e Managemen , Quin a de P ados, 5001-801 Vila Real, Po ugal
ARTICLE INFO
Keywo ds:
AMR
Food -p oducing animals
Global ends
Heal h
Mapping
ABSTRACT
An imic obial esis ance (AMR) can be highligh ed as one o he mos signi ican heal h conce ns among he las
decades, o which an imic obial d ug use in ood-p oducing animals has con ibu ed as one o he majo d i e s.
Food-p oducing animals a e one o he mos impo an and apidly expanding comme cial ag icul u al sec o s
wo ldwide bu he e is cu en ly limi ed knowledge on he empo al and geog aphical dis ibu ion o scien i ic
esea ch on an imic obial esis ance in ood-p oducing animals. We p o ide a global o e iew o he spa ial and
empo al ends o scien i ic knowledge on AMR in ood-p oducing animals. Pee - e iewed pape s o AMR on
ood-p oducing animals we e e ie ed om he Web o Science, sys emized and dissec ed. The inal alida ed
da ase con ained 1341 occu ences obse a ions co e ing he 1957–2018 pe iod. The e has been a shi o
esea ch e o s, bo h geog aphically and empo ally, emphasizing egional di e ences in ood animal p oduc ion
and changing p ac ices in he ood p oduc ion indus y. I becomes e iden ha many egions ha e been poo ly
su eyed, whe ein in ensi ied sampling and es ing e o s should be mos aluable. This sys ema iza ion o
knowledge will be c ucial in helping o de e mine how o op imally alloca e limi ed esou ces a ailable o AMR
moni o and con ol, aiding in he p edic ion whe e he h ea o new esis an in ec ions will be g ea es . AMR
esea ch in ood-p oducing animals in de eloping coun ies is ma kedly g owing, e lec ing changes in ood
animals p oduc ion sys ems bu also posing a pa icula ly signi ican h ea , no only due o in ensi e animal
p oduc ion, bu also exace ba ed by poo sani a ion. We highligh ha he use o an ibio ics in ood p oducing
animals is pe asi e, calling o u gen ac ion. These indings aise he possibili y o ine uning key p io i ies on
AMR global issues.
1. In oduc ion
“The ime may come when penicillin can be bough by anyone in he
shops. Then he e is he dange ha he igno an man may easily unde dose
himsel and by exposing his mic obes o non-le hal quan i ies o he d ug
make hem esis an . (…) I you use penicillin, use enough.”
This ex ac om he 1945 Nobel lec u e o Si Alexande Fleming is
ema kable, as he scien is who disco e ed penicillin a icina ed an e a
whe ein an ibio ics would become less e ec i e. Mo e han 70 yea s
a e Fleming’s Nobel speech, an imic obial esis ance (AMR) is a
ala ming le els, posing se ious global heal h p oblems [1]. Despi e
Fleming’s wa nings, i was no long a e he disco e y o penicillin, ha
esis ance became a se ious clinical p oblem jeopa dizing he conques s
o he p io decade [2]. Today, bac e ial esis ance is epo ed ac oss
nea ly all classes o an ibio ics comme cially a ailable [3,4]. In 2014,
he Wo ld Heal h O ganiza ion (WHO) p o ided a epo in which, o
he i s ime, he ex en o an imic obial esis ance was globally
assessed. The esul s we e d ama ic, showing ha in many a eas o he
globe (e.g., Eas e n Medi e anean, Sou h-Eas Asia Region, among
o he s) esis ance le els displayed by common bac e ia had eached
wo ying le els (WHO, 2014).
An ibio ics a e essen ial o he ea men o bac e ial in ec ions in
humans; hey ha e changed medical p ac ices and sa ed millions o li es
[5]. An ibio ics we e also a s epping s one suppo ing ou ine medical
* Co esponding au ho a : CESAM and Depa men o Biology, Uni e si y o A ei o, Campus de San iago, 3810-193 A ei o, Po ugal.
E-mail add ess: [email p o ec ed] (R.T. To es).
Con en s lis s a ailable a ScienceDi ec
One Heal h
jou nal homepage: www.else ie .com/loca e/onehl
h ps://doi.o g/10.1016/j.onehl .2021.100324
Recei ed 6 Ap il 2021; Recei ed in e ised o m 6 Sep embe 2021; Accep ed 6 Sep embe 2021
One Heal h 13 (2021) 100324
2
p ocedu es a ying om su ge ies o chemo he apy, o gan ans-
plan a ion, among o he s [6]. Whils an ibio ic consump ion and o e -
use is conside ed he p ima y d i e o an ibio ic esis ance [7], se e al
ac o s ha e di ec ly and/o indi ec ly p omo ed esis ance eme gence
and consolida ion along he yea s, including demog aphic changes
associa ed wi h u baniza ion and poo sani a ion, ising li e expec ancy,
discha ge o an ibio ic esidues h ough en i onmen al was and use o
animal manu e, immunosupp ession and inc eased oppo unis ic
in ec ion [8]. In ac , he in ensi ica ion o esis ance in ecen decades is
now jeopa dizing he con ol o in ec ious diseases [5,9]. Vancomycin-
esis an en e ococci (VRE) and me hicillin- esis an S aphylococcus
au eus (MRSA) a e nowadays global p oblems [10,11]. Fo ins ance,
MRSA is esponsible o mo e dea hs, yea ly, in U.S.A. (~19,000) han
he combina ion o HIV/AIDS, Pa kinson’s disease, emphysema and
homicide (Spellbe g e al., 2016; Ven ola, 2015). Resis an bac e ial
pa hogens a e c edi ed o mo e han 700,000 dea hs pe yea , including
214,000 neona al sepsis dea hs [12]. The loss o e icacy o an ibio ics
agains common pa hogens led o inc easing heal h cos s in high-income
coun ies [12], bu also o an inc ease o mo bidi y and mo ali y in
de eloping coun ies [8,13].
In he Eu opean Union (EU), economic losses associa ed wi h esis-
an bac e ial in ec ions a e on he o de o
€
1.5 billion annually and in
he U.S.A. hese losses a e es ima ed o be US$20 billion a yea
(T ansa lan ic Task o ce on An imic obial Resis ance). In ec ions wi h
bac e ia acqui ing mul id ug- esis an pheno ypes a e esponsible,
yea ly, o he dea h o 25,000 pa ien s in he EU (ECDC/EMEA Join
Wo king G oup, 2009) and o e 63,000 people in he U.S.A.. Recen
es ima es a e ala ming, sugges ing ha nea ly 10 million people will die
due o an imic obial esis ance by 2050 [14], howe e such “b oad
b ush es ima es” equi e cau ion o p ojec u u e scena ios [15]. Un-
doub edly, he bu den o an imic obial esis ance on human heal h is
inc easing bu i is di icul o p ecisely quan i y, as de ailed and eliable
da a a e missing [16], bu see [17]. Boeckel e al. [18] mapped, o he
i s - ime, an ibio ic consump ion by humans, showing ha an ibio ics
demand has inc eased wo ldwide by 36%, be ween 2000 and 2010.
BRICS coun ies (B azil, Russia, India, China and Sou h A ica) a e
esponsible o 76% o ha inc ease. These a e imp essi e igu es,
pa icula ly when hese coun ies ep esen 40% o he wo ld’s popu-
la ion [18]. The same ends we e epo ed by Klein e al. (2018) in an
expanded empo al scale (2000–2015), whe e an ibio ic consump ion
inc eased by 65%, mainly d i en by low- and middle-income coun ies.
As global human popula ion is expec ed o inc ease om app oxima ely
7.7 billion in 2019 o app oxima ely 9.2 billion by 2050 (Uni ed Na ions
Depa men o Economic and Social A ai s, 2015), an ibio ic con-
sump ion is expec ed o inc ease by 50% in 2030 [19]. Such human
popula ion inc ease will claim mo e animal-based p o ein and will de-
mand mo e indus ialized me hods o ood-animal p oduc ion, including
managemen ools such as sub he apeu ic dose an ibio ics o g ow h
p omo ion and disease p e en ion [20]. Fu he mo e, 34% o he p o-
jec ed global consump ion o an ibio ics in li es ock by 2030 is due o a
shi in a ming p ac ices as in ensi e animal husband y will inc ease
[13]. A subs an ial pa o he esis ance bu den in humans is a ibu -
able o an imic obial use in li es ock p oduc ion, p ima ily o disease
p e en ion and g ow h p omo ion [21–23], e en i he mechanisms
unde lying g ow h p omo ion emains unce ain [23]. Fo example,
an imic obials used in ood-p oducing animals a e expec ed o accoun
o ci ca 80% o he U.S.A. annual an imic obials consump ion [6,23]
and o ci ca 84% in China [24]. The e is also expanding awa eness ha
he use o an imic obials in ood-p oducing animals may add o he
eme gence o esis ance o an ibio ics egula ly used by humans, mainly
due o he o e lap o molecules om he same an ibio ic classes ha a e
used bo h in human and e e ina y medicine [25]. The an ibio ics used
in ood-p oducing animals can sp ead o humans h ough a ious di ec
and/o indi ec ou es. I can occu di ec ly h ough ood consump ion
and handling, − he ood chain is ac ually conside ed he main ou e o
ansmission o esis ance genes be ween animal and human
popula ions [26], o indi ec exposu e h ough he en i onmen , as 90%
o he an ibio ics gi en o ood-p oducing animals a e exc e ed in hei
ac i e o m in he u ine and aeces o ood-p oducing animals [27].
Rema kably, 90% o he an ibio ics adminis e ed o ood-p oducing
animals a e ul ima ely dispe sed h ough e ilize , g oundwa e , and
su ace uno in he en i onmen [28–30]. Chemical and physical
ea men s o wa e o human consump ion do no e ec i ely emo e
all an ibio ic aces. [30]. Exposu e o en i onmen al niches o s esso s,
such as bac e ial commensals o pa hogens ha bo ing mul id ug esis-
an genes, c ea es a complex scena io ha p omo es adap a ion
mechanisms o commensal o en i onmen al bac e ia and ho izon al
ans e o esis ance de e minan s. An imic obial use in ood-p oducing
animals is he e o e esponsible o an imic obial esis ance wo ldwide
[29,31] and, e en hough con o e sially, consump ion o ood ha was
ca ying an imic obial esis an bac e ia has de eloped in acquisi ion o
an ibio ic esis an in ec ions [32]. An ibio ics used in ood-p oducing
animals a e he e o e a se ious h ea o human heal h. Using global
da ase s o an ibio ic e e ina y use, li es ock densi ies and socio-
economic p ojec ions o mea demand, Boeckel e al. [18] p ojec ed
ha an imic obial use in ood-p oducing animals will inc ease by 67%
by 2030. This s udy was he i s o se a global pic u e o an ibio ic use
in li es ock p oduc ion on a wo ldwide scale. Al hough hese a e
admi able exe cises o ackle di icul ques ions, hey ely on limi ed
da ase s o a ailable da a as mos coun ies a e eluc an o p o ide o
do no ha e an ibio ic consump ion da a he e o e hampe ing he
de elopmen o e icien s a egies. S ill, he [18] p ojec ions we e
de i ed om da a conce ning 32 coun ies [33]. As he li es ock sec o
is one o he mos swi ly g owing comme cial ag icul u al sys em
globally [34], i is undamen al o e alua e he s a e and he
geog aphical dis ibu ion o scien i ic knowledge on an imic obial
esis ance in ood-p oducing animals. Such a global map is essen ial o
swi ly inspec ends, o unde s and he con ibu ion o ood-p oducing
animals o global heal h p oblems, o iden i y egions in which
commi ed esea ch is in ense o apidly inc easing and hose ha u -
gen ly need o gene a e new da a. This sys ema iza ion o knowledge
will be c ucial in helping o de e mine how o op imally alloca e limi ed
esou ces a ailable o AMR moni o and con ol, aiding in he p edic-
ion whe e he h ea o new esis an in ec ions will be g ea es (CDDEP
2018). Cu en unde s anding abou he global dis ibu ion o AMR
bu den is su p isingly limi ed. Mo eo e , he equency wi h which new
bac e ia wi h (mul i) esis ance a e eme ging unde sco es he ine i a-
bili y o shi ing om a eac iona y o a p o-ac i e p e en i e app oach
o mi iga ing AMR.
To mee hese demands, his s udy aimed a mapping he spa ial and
empo al ends o scien i ic knowledge on an ibio ic esis ance in ood-
p oducing animals. Speci ically, we wan ed o assemble and analyze
published pee - e iewed e idence on AMR in he main li es ock axa a
di e en ime poin s and geog aphical egions. This will allow o iden-
i y ho spo s o AMR esea ch, om which impo an esea ch gaps can
be iden i ied. As a as we know, his is he i s s udy mapping, a a
global scale, he scien i ic knowledge on an imic obial esis ance in ood
animals.
2. Ma e ial and me hods
2.1. Da a assembly
2.1.1. O e iew
The e iew ques ion was de ined as “Wha is he global dis ibu ion
o published scien i ic li e a u e o an imic obial esis ance in li e-
s ock?” We complied an ex ensi e da abase comp ising spa ial and
empo al occu ence da a de i ed om he pee - e iewed li e a u e. To
ga he his da abase, a sys ema ic li e a u e e iew was pe o med using
online sea ch engines and he esul an a icles we e manually e iewed.
Repo s by na ional o in e na ional au ho i ies conce ning o icial da a
submission by each coun y we e no conside ed.
R.T. To es e al.
One Heal h 13 (2021) 100324
3
2.2. Pee - e iewed li e a u e sea ch
2.2.1. Da a collec ion
A sys ema ic li e a u e e iew was pe o med using a igo ous sea ch
s a egy in he online e sion o he Science Ci a ion Index Expanded
(SCI-EXPANDED) om he Web o Science (WoS) da abase (h p://www
.isiknowledge.com). No ime and geog aphical loca ion es ic ions
we e placed on hese sea ches, and only hose published in English
e ie ed wi h ull i le and abs ac . The sea ches we e las upda ed on
19 h Janua y 2018. Sea ch esul s we e es ic ed o he ollowing
Boolean que y, execu ed wi hin a single sea ch o assembling h ee
di e en sea ch s ings, one o each ca ego y (an imic obial esis ance
and li es ock) and combining hese by he Boolean ope a o “AND” o
ob ain only he in e sec ion. Speci ically, we used he ollowing Boolean
sea ch s a emen : #1 “an imic obial esis ance”: “ANTIMICROBIAL” OR
“ANTIBIOTIC” AND “RESISTANT” OR “RESISTANCE” and #2 “li e-
s ock”: “PIG” “SUID” “BOVINE” “CATTLE” “POULTRY” “BROILERS”
and he in e cep ion consis ed in #1 AND #2. Ou sea ch was only
ocused ood-p oducing animals ha we e ei he e es ial mammals
and bi ds; hence o h o he g oups ( ish, amphibians, ep iles, and in-
e eb a es) we e excluded. All he esul ed publica ion eco ds we e
impo ed in o a bibliog aphic e e encing ool and manually assessed o
ele ance, emo ing a icles ha did no con ain in o ma ion ela ing o
AMR in li es ock (e.g., me hodology, en i onmen al AMR, clinical
AMR). All esul s we e manually checked be o e excluding duplica es.
To e i ied duplica es, que ies a ge ing iden ical i les and au ho s we e
pe o med and he ea e emo ed. Incomple e publica ions e.g., wi hou
an abs ac and publica ions clea ly indexed as e iew, edi o ial, o
e a a we e excluded. To c ea e he inal da ase , pape s we e en e ed
in o an Excel sp eadshee o u he analysis. Key de ail a iables in he
Excel da abase included he publica ion da e, coun y whe e he s udy
was pe o med, ood-p oducing animal species, bac e ial species, class
o an ibio ic, ci a ion. All he selec ion p ocess was made by wo inde-
penden eade s using p ede ined exclusion c i e ia.
2.2.2. Linking loca ions o da a om he li e a u e
F om each pee - e iewed a icle, all a ailable loca ion in o ma ion
(coo dina es) was ex ac ed and included in he da abase. F om he
publica ions, we compiled a geo-posi ioning occu ence da abase o
con i med AMR p esence, comp ising ei he poin s, in he case o ci ies,
owns o illages, o polygons in he case o coun ies o p o inces. Geo-
posi ioning was ex ac ed o he smalles possible le el (e.g., coun y,
p o ince, dis ic , ci y/ own o egion) and he coo dina es we e
ex ac ed using Google Maps (h ps://www.maps.google.p /). I he
pee - e iewed a icle clea ly s a ed a speci ic geo-posi ioned place (e.g.,
ci y, own o illage), i s cen al coo dina e was eco ded. Bu i he
s udy a ea could only be iden i ied a an adminis a i e a ea le el (e.g.,
p o ince o dis ic ), he cen oid o he polygon was eco ded. The
assigning o geo-posi ions o da a om he pee - e iew li e a u e was
done in a geog aphic in o ma ion sys em (A cGIS Desk op: Release 10.6.
Redlands, CA: En i onmen al Sys ems Resea ch Ins i u e).
3. Resul s and discussion
The sys ema ic analysis o pee - e iewed pape s epo ing an imi-
c obial esis ance in li es ock species o he las 60 yea s shows ha
esea ch in his opic has been exponen ially g owing in in e es by he
scien i ic communi y and ha he use o an ibio ics in ood p oducing
animals is pe asi e. In o al, we e ie ed 1341 alid publica ions
co e ing a pe iod om 1957 o 2018 (Fig. 1).
3.1. Tempo al and geog aphical e olu ion o publica ions
The numbe o publica ions on an imic obial esis ance in ood-
p oducing animals has s eadily inc eased h ough ime, e lec ing he
amoun o esea ch e o s wo ldwide (Fig. 2). The i s a icle dealing
wi h his opic appea ed in 1957 and he numbe o a icles p oduced
a e wa ds g ew o e all. Pe haps he ini ial in e es s in AMR esea ch in
ood animals a e linked o he “ala m call” made du ing he 1950’s by a
B i ish mic obiologis who d ew a en ion in o he human heal h im-
plica ions o he ca eless use o an ibio ics in animal a ming. E.S.
Fig. 1. Scoping e iew lowcha o he pee - e iewed da ase selec ion p ocess. All sea ch ca ego ies (li es ock and an imic obial esis ance) we e joined by he
Boolean exp ession “AND,” esul ing in he in e sec ion used o analysis. The PRISMA low diag am o he sea ch s a egy, s udy selec ion and da a manage-
men p ocedu e.
R.T. To es e al.
One Heal h 13 (2021) 100324
4
“Andy” Ande son desc ibed he inc ease o an ibio ic esis ance genes
ho izon ally om animals o humans [35]. Howe e , i was no un il
a e 1985 ha publica ions began o appea annually. Since he 90s’,
he annual a e age g ow h a e o scien i ic publica ions dealing wi h
an imic obial esis ance in ood-p oducing animals was 26.61% (Fig. 2).
A e he i s published s udy on AMR and ood-p oducing animals
in he U.S.A., subsequen epo s eme ged om Eu ope (Uni ed
Kingdom) in 1970 and Asia (India) in 1977. I was i s epo ed in
Oceania (Aus alia) in 1983 and, inally, in A ica (Zambia), in 1993
(Fig. 3). A e he 2000’s, AMR esea ch in ood-p oducing animals
showed a ema kable inc easing end (Fig. 2). In e es ingly, a e 2000,
mea p oduc ion has s abilized in mos de eloped coun ies and has,
con as ingly, inc eased in A ica (68%), Asia (64%) and Sou h Ame ica
(40%) [36]. As de eloping coun ies ha e become iche , he e was an
inc ease in he demand o animal p o ein die s. Such shi was
esponsible o a majo ans o ma ion o global ood animals p oduc-
ion, whe e an imic obials a e ou inely used o main ain heal h and
p oduc i i y [35; 36], he ea e he inc ease in publica ions can be a
consequence o his. Fo example, global mea p oduc ion has ipled in
de eloping coun ies be ween 1980 and 2002, om 45 o 134 million
ons [34]. This expansion was mos ly suppo ed by coun ies expe i-
encing apid economic g ow h, pa icula ly in Eas Asia, accompanied
by a ema kable inc ease o he published s udies in his egion (Fig. 3,
Fig. 1 Supplemen a y ma e ial).
Publica ions we e no dis ibu ed e enly on all con inen s: he ma-
jo i y o publica ions we e spa ialized in o Eu ope (38%), ollowed by
Asia (24%), no h Ame ica (23%), A ica (8%), sou h Ame ica (6%), and
Oceania (2%) (Fig. 3). I is clea ha in Asia, A ica and Sou h Ame ica
he inc ease o publica ions only occu ed in he las decade, which can
be a consequence o he inc easing impo ance o he AMR opic a a
global le el, bu also o go e nmen al in es men in science, pa icula ly
in o he One Heal h p ism and he ad en o high h oughpu echnol-
ogies. Se e al c i ical poin s a e c ucial o unde s anding ou esul s
ha a e deeply linked o he na ional poli ics o app o als, es ic ions
and bans o an ibio ics use in ood p oduc ion. An ibio ics use in ood
p oduc ion was al eady an es ablished p ac ice in he 1960’s. Fo
example, in 1938, P on osil ( he i s e ec i e d ug agains G am-
posi i e in ec ions) was ma ke ed in B i ain o use in animals [38],
ollowed by U.S.A.. Soon a e , non he apeu ic use o an ibio ic as
g ow h p omo e s (AGPs) also p o ed o be e ec i e o p o ec agains
bac e ial in ec ions and i s use g ew a e 1949 whe e i became licensed
o be used wi hou e e ina y p esc ip ion in 1951 in Wes Ge many.
The expansion o an ibio ic use was e en mo e d ama ic in China bu
ou ine use o an ibio ics in a ms was no be o e he 1980s. Du ing he
Fig. 2. Global publica ion ends o an imic obial esis ance in ood animals, n =1341 (a) and he cumula i e numbe o publica ions o he 1957–2018 pe iod (b)
(n =1341).
R.T. To es e al.
One Heal h 13 (2021) 100324
5
la e 1990s and ea ly 2000s, he Eu opean Union (EU) phased in mo e
s ingen egula ions on he use o an imic obial subs ances in animal
p oduc ion, pa icula ly hei use o g ow h p omo ion. Also in he
ea ly 2000s, he EU enac ed policies o on- a m su eillance. I was in
1996, ha U.S.A. began o sys ema ically moni o esis ance in ood-
bo ne pa hogens, wi h he implemen a ion o he Na ional An imic o-
bial Resis ance Moni o ing Sys em (NARMS).
When he publica ion ends a e analyzed by decade, he e is a
isible consis en inc ease in publica ions globally bu also signi ican
shi poin s in he geog aphical a ea o esea ch p oduc ion (Fig. 4),
which a e na u ally coinciden wi h impo an key poin s in he an i-
bio ic esis ance his o y [39]. E en hough he WHO o ganized some
wo king g oups and mee ings ocused on an ibio ic esis ance du ing he
la e 70’s (WHO 1976; WHO 1978), i was only when a clinician-scien is
om Tu s Uni e si y (U.S.A.), S ua Le y, in 1981, ga he ed 147 sci-
en is s om 27 coun ies (bu mos o hem om U.S.A.) on a con e ence
in San o Domingo (Dominican Republic) o sign he “S a emen
Rega ding Wo ldwide An ibio ic Misuse”, emphasizing ha AMR was a
“wo ldwide public heal h p oblem” – his con e ence was a on he basis
o he Alliance o he P uden Use o An ibio ics (APUA) [39].
Th oughou he decade, he e was an inc ease o he U.S.A. go e nmen
agency unding o AMR bu he headship on AMR esea ch shi ed o
Eu ope by he mid-1990s. Ala med by he inc ease in mul id ug-
esis an Salmonella in he 1960s, he “Swann epo ” (1969) banned
he he apeu ic use o impo an an ibio ics (e.g., penicillin and e a-
cyclines) o ag icul u al g ow h p omo ion, in G ea B i ain [38].
Despi e he wa nings, esponses o an ibio ic esis ance as a sha ed,
global p oblem we e mu led. Bu G ea B i ain, quaked by he bo ine
spongi o m encephali is (“mad cow disease”) ga he ed ac ion, which
peaked in a 1998 epo by he House o Lo ds, ha inally led in 2001 o
he launching o he Eu opean An imic obial Resis ance Su eillance
Sys em (EARS-Ne ) [39].
Bene i ing by he go e nmen in e es and awa eness o he an i-
bio ic esis ance p oblem, up un il he 90’s mos o he esea ch was
mos ly ocused on he U.S.A. and some Eu opean coun ies, pe ec ly
e lec ed by Fig. 4. Ne e heless, he low de elopmen aid o heal h,
and pa icula ly o AMR, in de eloping coun ies su ely led o a slow
implemen a ion o na ional su eillance amewo ks [40], which hen
ansla ed in o a small numbe o publica ions du ing his decade. By he
1990’s, he WHO was ully p omised wi h he global AMR p oblem and
o ganized se e al wo king g oups and mee ings on he “majo public
heal h p oblem in bo h de eloped and de eloping coun ies” (WHO,
1994; WHO, 1997; WHO, 1998), p obably mo i a ing u he esea ch
in e es on his p oblema ic in o he coun ies. The WHO epo s
s essing he impo ance o including AMR on a global esea ch agenda,
and he ele ance o ackling his opic on de eloping coun ies, was an
impo an u ning poin . This, in combina ion wi h he inc eased in-
as uc u es and quali ied esea ches in hese coun ies (e.g., India,
China), c ea ed new pe spec i es and oppo uni ies, which likely ed
in o he shi s o na ional agendas, ansla ed in o he consis en inc ease
in publica ions in hese coun ies. Fo example, China, one o he wo ld’s
la ges p oduce s and consume s o an ibio ics, began moni o ing bac-
e ial esis ance na ionwide since 2005 h ough he China An imic obial
Su eillance Ne wo k [41] (Fig. 5). Finally, in 2013, AMR was indica ed
as one o he majo global isks a he Wo ld Economic Fo um: “While
i uses may cap u e mo e headlines, a guably he g ea es isk o hub is o
human heal h comes in he o m o an ibio ic- esis an bac e ia” (WEF
2013).
The inc eased numbe o publica ions ocused o AMR on ood-
p oducing animals in de eloping coun ies is likely a e lec ion o
o e all shi ing esea ch p io i ies and unding a ailabili y [41] bu also
a e lec ion o he inc ease in ele ance o he use and misuse o an i-
mic obials in ood-p oducing animals and hei po en ial impac on
human heal h [32]. A ecen e iew jus highligh ed his obse a ion in
ha educing he le el o an ibio ic use in ood-p oducing animals is
likely o be bene icial o bo h animals and human beings [32]. I also
unde lines he One Heal h pa adigm and possibly he sensi iza ion o
esea che s in he e e ina y ield o conduc ing s udies on AMR-
ela ed opics. In ac , AMR pe ec ly i s he One Heal h concep ,
de ined as “… he collabo a i e e o o mul iple disciplines – wo king
locally, na ionally, and globally – o a ain op imal heal h o people,
animals and ou en i onmen …” [42,43]. The ecogni ion o he in e -
disciplina y na u e o AMR led he Food and Ag icul u e O ganiza ion o
he Uni ed Na ions (FAO) and Wo ld O ganiza ion o Animal Heal h
(OIE) o suppo he Global Ac ion Plan on An imic obial Resis ance
(GAP) de ined on he 68 h Wo ld Heal h Assembly in May 2015 (44).
Bu while he numbe o publica ions has s eadily inc eased h ough
ime, e lec ing he amoun o esea ch e o s wo ldwide, i also high-
ligh s poo in es iga ion, and he e o e su eillance, in de eloped
coun ies. The managemen s a egies o an imic obial esis ance in
Fig. 3. Tempo al and geog aphical (pe con inen ) dis ibu ion o publica ions o he 1957–2018 pe iod (n =1341).
R.T. To es e al.

One Heal h 13 (2021) 100324
6
hese esou ce-limi ed a eas a e specially challenging because s a egies
o igh and manage AMR canno mi o hose used in de eloping
coun ies o he sake o deple ing esou ces.
3.2. Tempo al and geog aphical e olu ion o publica ions pe bac e ial
genus
Publica ions end o cen e mo e on bac e ia ela ed o zoono ic,
oodbo ne casual ies, ep oducing policy and p io i ies wo ldwide and a
ocus on public heal h. Campylobac e spp., E. coli, En e ococcus spp.,
Salmonella spp. and S aphylococcus spp. a e he op i e bac e ia pub-
lished globally, wi h E. coli and Salmonella spp. ep esen ing he majo i y
o bac e ia unde ocus (Fig. 5). The op i e bac e ia in e ms o pub-
lica ion coun s we e all zoono ic, which likely illus a es he engaged-
men o he esea ch communi y in human heal h esea ch and/o he
way unding o in ec ious disease esea ch is d i en by human heal h.
The di e ences on he p opo ion o publica ion numbe s in he op-
i e bac e ia mi o s he eali y o each coun y conce ning ood-
p oducing animals and de elopmen s a us o he p oduc ion sys ems
(Fig. 2 Supplemen a y ma e ial). Fo example, in Eu ope and USA he
di e ence be ween he numbe o s udies ocused on Salmonella sp.
(>USA) and Campylobac e sp. (>Eu ope) is a consequence o he
oodbo ne disease eali y in each coun y. In Eu ope, he mos common
bac e ial oodbo ne disease is campylobac e iosis, ollowed by salmo-
nellosis, and non- yphoidal Salmonella spp. ollowed by Campylobac e
spp. cause he highes bu den (WHO, 2020). Con e sely, he main
oodbo ne disease (e.g., dia hea, nausea, abdominal pain, omi ing) in
USA is caused by Salmonella spp. [45]. Rep esen a ion o he numbe s
o A ica, Asia and Sou h Ame ica show Salmonella as he bac e ia
second mos epo ed, highligh ing he de eloping s a us o hese con-
inen s, whe e he su eillance and e ec i e measu es o ood sa e y
may no be as e icien when compa ing wi h de eloped coun ies [46].
E. coli was he mos epo ed in Asia, Eu ope, No h Ame ica and Oce-
ania, which was expec ed since E. coli is he mos abundan commensal
and indica o bac e ia o he in es inal ac o mammals, wi h some
ex a-in es inal pa hogenic s ains being a public heal h p oblem and i s
ubiqui y and adap i e beha io playing a key ole in he sp ead o AMR
h ough he human-animal-en i onmen in e ace [47]. Since E. coli a e
po en ial ese oi s o esis ance de e minan s, i has been egula ly
moni o ed in ood animals o assess he p e alence o an imic obial
esis ance (EFSA & ECDC, 2014).
3.3. Class o an ibio ics
An ibio ics ha e been ou inely used in ood-p oducing animals o
con ol diseases ( he apeu ic use), o p e en diseases in heal hy animals
wi h somehow isk o in ec ion (p ophylac ic use) and o imp o e eed
e iciency (g ow h p omo ion use) [44]. Wi h he isk o losing hei
e iciency, which would lead o a global public heal h p oblem, and o
p ese e he success o medically impo an an imic obials, pa icula ly
hose c i ically impo an o human medicine, he WHO de eloped
guidelines on use o an imic obials in ood p oducing animals.
Ou esul s show ha se e al classes o an ibio ics a e widely ound
in ood-p oducing animals and ha hei use has been inc easing
h ough ime, pa icula ly in he las 15 yea s, wi h highligh o e a-
cycline, quinolones and be a-lac ams (Fig. 6). E en hough e acyclines
a e no , cu en ly, a i s op ion an imic obial in he human clinic, hey
emain as one o hose mos used an ibio ic classes in ood-p oducing
animals, ep esen ing a classical example o AMR eme gence due o
he selec i e p essu e exe ed by ex ended use. Quinolones a e among
he “Highes P io i y C i ically Impo an An imic obials“ (WHO, 2019).
Quinolones and be a-lac ams a e i s -line choice an imic obials o ea
in ec ions in humans, due o hei b oad-spec um ac ion and low
oxici y, also commonly adminis e ed o ood-p oducing animals.
Resis ance o ex ended-spec um be a-lac ams (ESBLs), especially hi d-
and ou h-gene a ion cephalospo ins and ca bapenems, ha e wo ied
Fig. 4. Global spa ial dis ibu ion o AMR pee - e iewed epo s pe selec ed ime in e als (a) and since 1957 (b) (n =1341).
R.T. To es e al.
One Heal h 13 (2021) 100324
7
medical p ac ice and scien is s in he las decade. WHO ecen ly de ined
hi d- and ou h-gene a ion cephalospo ins as being “c i ically impo -
an ” o human medicine since inc easing esis ance o his an imic o-
bial could comp omise he ea men o bac e emia and bac e ial
meningi is [49]. Bu hi d-gene a ion (ce io u ) and ou h-gene a ion
cephalospo in (ce quinome) ha e been de eloped o e e ina y use
[49].
F om a pe spec i e o he numbe o publica ions by class o an i-
bio ics in he las h ee decades (1990–2018), he Fig. 6 shows an in-
c ease in he numbe o publica ions globally. Howe e , some classes o
an ibio ics call a en ion o hei pa icula i ies, such as polypep ides
and ni o u ans. Colis in is cu en ly one o he las he apeu ic op ions
o se e e in ec ions by MDR bac e ia in hospi als, i is an an ibio ic ha
has been clinically used since he 60’s and whose use has also been e-
po ed ood-p oducing animals. S ikingly, an inc ease in he numbe o
publica ions conce ning colis in is egis e ed only since he 2000’s,
p obably due o i s no o ie y as a he apeu ic op ion [50]. Ni o u an
an ibio ics, used o ea men o u ina y in ec ions, ha e no shown an
inc ease in he numbe o publica ions in he las ew yea s when
compa ed wi h he o he classes, which is p obably ela ed wi h he
p og essi e aboli ion o hei use in ood-p oducing animals bu also as
hese a e p e e ed an ibio ics used in poul y. Wo ks ela ed wi h
ni o u an esis ance a e mainly cen ed in he desc ip ion o c oss-
esis ance wi h o he an ibio ic classes [51]. The con inued use o he
abo emen ioned an imic obials ha could po en ially selec o esis-
an o ganisms is wo ying because o he impo ance o ood animals in
sp eading and main enance o esis ance genes in o he en i onmen . In
his line, WHO has been ecommended he educ ion o es ic ion o
use, ei he o g ow h p omo ion o p ophylac ical pu poses, o all
classes o medically impo an an imic obials in ood-p oducing animals
(WHO, 2017). The a ailable da a sugges s ha aboli ion o hese an i-
mic obials in ood-p oducing animals would ep esen minimal unde-
si ed consequences (e.g., less e icien animal heal h managemen o
inc eased p oduc ion cos ) compa ed o he ad an ages he eo [52].
3.4. Geog aphical dis ibu ion o publica ions by li es ock species
The mos common li es ock species being epo ed in he selec ed
publica ions includes poul y, ca le (bee and bu alo mea ), pig, sheep
and goa o a lesse ex en (Fig. 7). In e es ingly, hese igu es a e
coinciden wi h wo ldwide ood animal p oduc ion numbe s (FAO,
2018). Wi h he excep ion o no h Ame ica, poul y was he g oup o
animals wi h mo e AMR s udies (Fig. 7); his g ea e p ominence o
publica ions in he poul y sec o is p obably ela ed o he ac ha his
Fig. 5. Tempo al (pe decade) and geog aphical (pe con inen ) e olu ion o AMR publica ions ocusing on he op- i e bac e ial gene a (s udies be o e 1970 a e
included in he 1970–1980 decade) (n =1341).
R.T. To es e al.
One Heal h 13 (2021) 100324
8
is he mos p oduced mea wo ldwide (FAO, 2019) bu also he one
whose g owing in ensi ica ion was mo e p o ound in he las decades
[53].
Be ween 1961 and 2016, poul y mea p oduc ion has inc eased
globally om 9 o 120 million ones, and in 2016 his ep esen ed 36%
o global mea p oduc ion (FAO, 2020). Likewise, in he las h ee de-
cades, egg p oduc ion has inc eased by mo e han 150% wo ldwide,
om 15 o 81 million ones (FAO, 2020). Mos o his inc ease has
occu ed in Asia, whe e p oduc ion ose almos ou old [54]. In Uni ed
S a es o Ame ica, ca le was unde ocus o mos publica ions, p obably
because his coun y is he wo ld’s la ges bo ine (e.g., bee and bu alo)
p oduce , gene a ing 11–12 million ones in 2014 (USDA-FAS, 2019).
Be ween 1961 and 2014, ca le mea p oduc ion has mo e han doubled
globally, om 28 million ones yea ly o 68 million ones. O he majo
p oduce s a e B azil and China.
Globally, pig and poul y p oduc ion a e he as es g owing ood-
p oducing animals segmen s. Inc easingly, pigs and ca le a e also
aised unde simila condi ions o con inemen and high densi y like
poul y (FAO, 2018) which poses a pa icula ly signi ican h ea , no
only due o in ensi e animal p oduc ion, bu also exace ba ed by poo
sani a ion in as uc u e [55]. Se e al de eloping coun ies a e also
showing a shi owa ds indus ializa ion o ood animals p oduc ion by
eplacing adi ional sys ems o in ensi e uni s, wi h much o ha ise
occu ing in Asia, Sou h Ame ica and No h A ica [34]. Ou analyses
sugges s ha he e has been an inc eased in e es in ood-p oducing
animals esea ch issues in de eloped coun ies, which a e hose p e-
cisely whe e he e has been an inc emen o mea p oduc ion and whe e
human popula ion g ow h is expec ed o inc ease. In one sense, his
means ha an imic obial esis ance in ood-p oducing animals and hei
h ea o pass AMR de e minan s o human ia ood chain is now
acknowledged and ha e ec i e con ol and biosecu i y measu es
associa ed may ha e been applied in he meanwhile.
3.5. S udy limi a ions
We ecognize ha ou s udy has some limi a ions. Fi s ly, we only
ocused on a icles ha ha e been published in academic jou nals
indexed in Web o Science, excluding subs an ial p opo ion o knowl-
edge ha may ha e been published in o he o ma s (e.g., books, epo s,
and na ional jou nals) and in go e nmen and indus y epo s, sum-
ma izing signi ican su eillance e o s a ound he globe. Publica ions
ha did no include he used sea ch e ms in he i le migh ha e been
le ou o ou analysis, al hough we suspec ha ou esul s e lec he
gene al ends wi hin he hema ic esea ch landscape. Whe eas ou
esul s a e based only on pee - e iew li e a u e, hey likely e lec he
obse ed in e es by he scien i ic communi y, measu ed he e by he
Fig. 6. Tempo al dis ibu ion o publica ions pe class o an ibio ics.
R.T. To es e al.
One Heal h 13 (2021) 100324
9
numbe o publica ions, and he global dynamics wi hin his esea ch
ield i sel , and no he ac ual le els o an imic obial esis ance and/o
use. Howe e i is also impo an o no e ha esea ch in e es should
no be con used wi h incidence o an imic obial esis ance. In addi ion,
his analysis was es ic ed o in e na ional jou nals in English, he e o e
a linguis ic bias may exis . Bu he ca ea s discussed a e ou weighed by
he indings assembled. We ha e e iewed a la ge body o e idence
sugges ing a clea empo al and geog aphical bias owa ds he pe asi e
use o an ibio ics in ood-p oducing animals, o e ing a alid ep esen-
a ion wi hin his esea ch ield a a global le el.
3.6. Conclusions
The e was an inc ease in he numbe o publica ions on he AMR in
he las decade which is p obably a e lec ion o he p o ound changes
ha ood-animal p oduc ion sys ems ha e unde gone o e he pas de-
cades: unde he umb ella o apid popula ion u no e , la ge numbe o
animals a e aised in con inemen (56). In mode n ood-animal p o-
duc ion, an imic obial agen s a e egula ly used as heal h managemen
ools. Today, ood animals and animal-de i ed p oduc s a e aded
wo ldwide. Thus, he selec ion o an imic obial esis ance in one
coun y becomes a global p oblem. This has ale ed go e nmen s and
s akeholde s emphasizing he need o b oad ini ia i es, con ol and
moni o ing sys ems o con ainmen o an imic obial esis ance, which
is laid ou in he end o publica ions. Ou esul s show ha h oughou
he ime he e was a shi ing in esea ch p io i ies and his was mos
likely ela ed o egional di e ences in ood-p oducing animals p o-
duc ion and changing p ac ices in he ood p oduc ion indus y. An i-
mic obial esis ance esea ch in ood-p oducing animals in de eloped
coun ies is s ill g owing, e lec ing ei he changes in li es ock p o-
duc ion sys ems [57] bu also because hey ealized he nega i e e ec
o he esis ance eme gency globally.
Ou assessmen ou lines h ee global p io i ies o ac ion. Fi s , ou
maps show egions ha a e poo ly su eyed and whe e sampling
in ensi ica ion e o s could be mos aluable, namely no h Asia and
A ica. Second, ou indings clea ly indica e ha he highes publica ion
a e is cu en ly alloca ed o India and China. The inc ease in mea
p oduc ion and he shi in ood animals p oduc ion sys ems s ess he
impo ance o hese coun ies o implemen ac ions o p e en u he
agg a a ion o he AMR p oblem. Thi d, coun ies such as he U.S.A. and
some Eu opean coun ies, whe e na ional e o s o ha monized su -
eillance, uni o m epo ing and p oac i e policy ac ions occu ed
ea lie , and whe e an ibio ic ban and es ic ions o g ow h p omo ion
is a leas olun a y, should se e as encou aging examples o he
change o new s a egies in ood animal p oduc ion [17]. Since cu en
ends owa ds in ensi e li es ock p oduc ion a e unlikely o e e se,
he esea ch in ood-p oducing animals will be an ongoing a ea o in-
e es . Ou e iew shows whe e esea ch in his a ea has been ocused
o e he las decades and whe e i should mo e o wa d. We conclude
ha he global e o o AMR moni o ing and esea ch is poo ly allo-
ca ed, wi h mos o he scien i ic ools and esou ces ocused on egions
whe e mea p oduc ion has pla eaued and s ong na ional plans a e
ongoing, he e o e whe e he nex impo an p oblems a e leas likely o
p e ail [17,37]. Sys ema ic e o s should include an ibio ic su eillance
in egions o coun ies whe e no, o li le da a, is a ailable. Howe e , he
lack o hese s uc u es in de eloping coun ies, whe e he e is o en
inadequa e su eillance and minimal labo a o y capaci y and whe e
li es ock p oducing is swi ly inc easing, can be a se ious se back [16].
Al e na i e models on “sma su eillance” on middle and low-income
coun ies mus ocus on inexpensi e es ing, de eloped by ained
wo ke s which could eed a web-based sys em, na ional and in e na-
ionally connec ed. This will ob iously no eplace he necessi y o
labo a o ies and s anda dized p ac ical p ocedu es bu ins ead would
allow a sense o moni o ing and su eillance in a eas whe e no labo a-
o ies a e a ailable. Addi ionally, in e na ional collabo a ion is essen ial
as AMR is a global p oblem, wi h no bo de s, and we encou aged his
Fig. 7. Geog aphic dis ibu ion o he numbe o AMR publica ions pe speci ic li es ock species (n =1341).
R.T. To es e al.