Resea ch A icle
Po J Public Heal h 2022;40:43–51
Op imis ic You h: Young Adul s P edic ed a Fas e
Dec ease in Risk du ing COVID-19 Eme gency S a e
in Po ugal
Sa a Scale i a, b Inês Dua e a, c Ca a ina Sen a a Jo ge Almeida a, c
An ónio Jo ge Fe ei a d Jon Walb in a, c A u Pilacinski a, c
aP oac ion Lab, Facul y o Psychology and Educa ional Sciences, Uni e si y o Coimb a, Coimb a, Po ugal;
bFacul y o Psychology, Uni e si y o Pado a, Pado a, I aly; cCINEICC, Facul y o Psychology and Educa ional Sciences,
Uni e si y o Coimb a, Coimb a, Po ugal; dFacul y o Medicine, Uni e si y o Coimb a, Coimb a, Po ugal
Recei ed: Ap il 12, 2021
Accep ed: Feb ua y 23, 2022
Published online: Ap il 20, 2022
Co espondence o:
Sa a Scale i, sa a.scale i @ s uden i.unipd.i
A u Pilacinski, a .pilacinski @ gmail.com
© 2022 The Au ho (s). Published by S. Ka ge AG, Basel
on behal o NOVA Na ional School o Public Heal h
Ka ge @ka ge .com
www.ka ge .com/pjp
DOI: 10.1159/000524076
Keywo ds
Risk pe cep ion · Public heal h · COVID-19 · Age di e ences
Abs ac
Pe cep ion o isk is known o change h oughou he li e-
span. P e ious s udies showed ha younge adul s a e
mo e p one o isk beha iou s han olde adul s. Do hese
age- ela ed di e ences in luence isk pe cep ion du ing a
pandemic c isis? He e, we in es iga ed how age in luenced
p edic ed isk du ing he COVID-19 eme gency s a e in Po -
ugal. We show ha ime-p ojec ed es ima ions (e.g., ap-
p aisals based on ‘now’ s. ‘in wo weeks’ ime’, o ‘in ou
weeks’ ime’) o bo h isk beha iou and impo ance o
ansmission p e en ion dec ease o e ime. Impo an ly,
p ojec ed isk dec eased mo e s eeply o younge han
olde adul s. Ou indings sugges ha younge adul s
ha e a di e en pe cep ion o epidemic- ela ed isk han
olde adul s. This seems o suppo he iew ha public
heal h policy making du ing epidemics should di e en ial-
ly a ge younge adul s.
© 2022 The Au ho (s). Published by S. Ka ge AG, Basel
on behal o NOVA Na ional School o Public Heal h
Ju en ude o imis a: Os jo ens adul os p e i am uma
diminuição mais ápida do isco du an e o es ado de
eme gência ela i o à COVID-19 em Po ugal
Pala as Cha e
Pe ceção do isco · Saúde pública · COVID-19 · Di e enças
de idade
Resumo
A pe ceção do isco muda ao longo da ida. Es udos an e i-
o es mos a am que os jo ens adul os são mais p opensos
a compo amen os de isco do que os adul os mais elhos.
Se á que es as di e enças elacionadas com a idade in luen-
ciam a pe ceção de isco du an e uma c ise pandémica?
Aqui, in es igámos como a idade in luenciou o isco p e-
is o du an e o es ado de eme gência da COVID-19 em Po -
ugal. Demons amos que as es ima i as de empo p oje a-
das (e.g., a aliações baseadas em ago a s. "daqui a duas
semanas", ou "daqui a qua o semanas") an o do compo -
amen o de isco como da impo ância da p e enção da
ansmissão diminuem ao longo do empo. É impo an e
no a que o isco p oje ado diminuiu mais acen uadamen e
pa a os jo ens adul os do que pa a os adul os mais elhos.
Os nossos esul ados suge em que os jo ens adul os êm
Sa a Scale i, Inês Dua e and Ca a ina Sen a con ibu ed equally o
he manusc ip .
This is an Open Access a icle licensed unde he C ea i e Commons
A ibu ion-NonComme cial-4.0 In e na ional License (CC BY-NC)
(h p://www.ka ge .com/Se ices/OpenAccessLicense), applicable o
he online e sion o he a icle only. Usage and dis ibu ion o com-
me cial pu poses equi es w i en pe mission.
Scale i/Dua e/Sen a/Almeida/Fe ei a/
Walb in/Pilacinski
Po J Public Heal h 2022;40:43–51
44
DOI: 10.1159/000524076
uma pe ceção do isco elacionado com epidemias di e-
en e do que os adul os mais elhos e pa ecem co obo a a
pe spe i a de que a elabo ação de polí icas de saúde públi-
ca du an e epidemias de e isa de o ma di e en e os jo-
ens adul os.
© 2022 The Au ho (s). Published by S. Ka ge AG, Basel
on behal o NOVA Na ional School o Public Heal h
In oduc ion
Risk pe cep ion e e s o an indi idual’s subjec i e
judgemen s abou he likelihood o nega i e occu ences,
such as inju y, illness, disease, and dea h. Risk pe cep ion
is impo an in heal h and isk communica ion because i
de e mines which haza ds people ca e abou and how
hey deal wi h hem [1]. A common assump ion is ha
knowledge and ce ain y abou isk de e mine how i is
pe cei ed. Based on a ional decision-making, his as-
sump ion is p edominan ly asc ibed o expe s, who ely
on scien i ic in o ma ion and objec i e assessmen [2]. By
con as , laypeople commonly e alua e isks by using
heu is ics. Few s udies ha e examined he age- ela ed di -
e ences in isk pe cep ion using hypo he ical eal-li e si -
ua ions [3, 4]. Recen wo k has ocused on isk in he con-
ex o he COVID-19 pandemic [5–8]. These s udies e-
ealed mixed esul s, i.e., pe cei ed ulne abili y o
COVID-19 ended o dec ease wi h olde age, ye olde
adul s pe cei ed highe isk o dying because o CO-
VID-19 [5]. Rosi e al. [9] also ound ha olde adul s
(70+ yea s) pe cei ed highe isk se e i y o COVID-19
compa ed o he o he age g oups, and, con e sely,
younge adul s (18–29 yea s) pe cei ed lowe isk se e -
i y o COVID-19 compa ed o all o he age g oups.
Risk pe cep ion and isk a i ude a e o en ea ed as
dis inc ye in e dependen concep s by many au ho s
(howe e , hese e ms a e some imes used in e change-
ably in he li e a u e [10]). Risk a i ude e e s o a b oad-
e concep – beyond me e isk- aking (i.e., he likelihood
o engaging in isky beha iou s) – ha in ol es addi ion-
al ac o s, such as isk pe cep ion and he pe cei ed ben-
e i s o engaging in isky beha iou [11]. Al hough isk
a i ude is conside ed a ela i ely s able psychological
ai , he e is e idence ha i may change o e ime; peo-
ple end o become mo e isk-a e se as hey age [12], al-
hough he e can be a ia ions depending on he domain.
Fo example, younge adul s we e ound o be mo e like-
ly o ake isks in he domain o heal h and sa e y, com-
pa ed o olde adul s [13]. Addi ionally, p e ious wo k
shows ha isk- aking declines wi h age ac oss se e al
di e en domains, including heal h [14]. These age- ela -
ed di e ences in isk a i ude may be impo an when
conside ing public heal h guidelines in si ua ions such as
he ecen COVID-19 pandemic.
Mos heo e ical models on isk pe cep ion ecognize
ha i has wo main dimensions [1]: he cogni i e dimen-
sion, which ela es o how much people know and unde -
s and abou he isk, and he emo ional dimension, which
ela es o how hey eel abou isk. Indi iduals o en e al-
ua e isk acco ding o subjec i e app aisals, in ui i e
judgemen s, and in e ences d awn om he media [15].
Fu he mo e, ce ain cogni i e biases may also po en ial-
ly in luence isk pe cep ion (e.g., he op imism bias – he
endency o belie e ha isks pose a less se ious h ea o
onesel han hey do o o he people [16]).
The ampli ica ion o pe cei ed isk du ing a pandem-
ic is pa ially accoun ed o by a model p oposed by Po-
le i e al. [17] ha desc ibes he dynamics o isk pe cep-
ion du ing he H1N1 in luenza pandemic in I aly, in
2009. These au ho s p oposed he suscep ible-in ec ed-
eco e ed model [17], which emphasizes ha suscep ible
indi iduals may pe o m “no mal” o “al e ed” beha -
iou s, based on he pe cei ed isk o in ec ion. In pa icu-
la , he au ho s p opose ha , based on he pe cei ed isk
o in ec ion, some indi iduals adop “al e ed” beha iou s
o educe i al exposu e and in ec ion. These beha iou s
include a educ ion in physical con ac wi h o he s, e-
quen hand washing, s ong adhe ence o espi a o y e i-
que e (e.g., minimizing sneezing o coughing in public
places), a oidance o c owded en i onmen s, o limi ing
a el. The au ho s desc ibe an ini ial o e es ima ion o
isk, accompanied by “al e ed” beha iou s (such as pu -
chasing la ge amoun s o cold/ lu medicine) aimed a
educing he sp ead o he in ec ion. Subsequen ly, he e
was a no ed dec ease in pe cei ed isk, due o he slow
inc ease in he numbe o cases, ha was accompanied by
a educ ion in “al e ed” beha iou s which, in u n, con-
ibu ed o a e-accele a ion o in ec ion a e.
Li le is known abou whe he he dynamics o isk
pe cep ion (and isk a i ude), as desc ibed in he model
o Pole i e al. [17], di e ac oss di e en age g oups.
P e ious esea ch has shown ha younge adul s ake
mo e isks, especially heal h-associa ed isks (e.g., using
d ugs o alcohol o ha ing unp o ec ed sex) [18–24]
when compa ed o o he adul s [21, 25]. This endency
o isk- aking migh also in luence u u e p edic ions o
isk among young adul s du ing he pandemic.
Adhe ence o he go e nmen -manda ed p e en i e
measu es is shown o be c i ical o cu b he sp ead o he
in ec ion, bu he e a e indi idual di e ences in he ex-
Op imis ic You h
45
Po J Public Heal h 2022;40:43–51
DOI: 10.1159/000524076
en people apply hese p e en i e measu es. Fo exam-
ple, a su ey by Uni e si y College London including da a
om o e 10 weeks showed ha guideline adhe ence de-
c eased subs an ially, om 70% a he s a o he su ey,
o 50% a he end [26]. Howe e , his decline di e ed be-
ween age g oups: while mo e han 60% o adul s epo -
ed ollowing he guidelines en i ely, app oxima ely only
40% o younge adul s epo ed doing he same. O he
s udies du ing p io epidemics ha e shown simila e-
sul s; du ing he SARS epidemic in Canada in 2003,
younge adul s epo ed ollowing he guidelines less, and
also pe cei ed he si ua ion o be less isky, compa ed
wi h olde adul s [27]. P io esea ch also shows ha old-
e adul s pe cei e highe isk and beha e mo e cau ious-
ly han younge adul s conce ning heal h- ela ed ac i i-
ies [28] and ha sel - epo ed isk pe cep ion in social,
inancial and ec ea ional domains inc eases wi h age
[29]. Fu he mo e, olde adul s epo pe cei ing mo e
isk o mo ali y i in ec ed wi h COVID-19 [5]. Toge he ,
hese indings indica e a clea end o lowe isk pe cep-
ion in younge adul s, ela i e o olde adul s.
The go e nmen -manda ed p e en i e measu es and
he e ec o he pandemic mo e gene ally caused d ama -
ic changes in he li es o all ci izens [30]. These measu es
ha e been pa icula ly ha d o accep o he younge peo-
ple, namely, hose who could no a end classes, mee
iends, o engage in spo [30, 31]. Olde people we e
likewise ins uc ed o emain a home and sel -isola e
since hey cons i u e a high- isk g oup o COVID-19
[31, 32]. These social es ic ions may in u n play an im-
po an ole in how he popula ion pe cei es he changes
o he pandemic si ua ion and he isk o in ec ion o e
ime.
In his wo k, we in es iga ed age di e ences ega ding
he pe cep ion o isk ela ed o e e yday beha iou s du -
ing he ea ly epidemic eme gency s a e in Po ugal (Po -
ugal was in a s a e o eme gency om Ma ch 18 o May
2, 2020). Speci ically, we ocus on indi idual judgemen s
o isk and o he ac o s du ing he COVID-19 pandem-
ic: (1) beha iou s ca ying isk in ec ion ( isk beha iou s,
e.g. “ isi ing iends”, “using an ele a o ”); (2) beha iou s
aimed a p e en ing disease sp eading and p epa edness
o c ises (p e en ion beha iou s, e.g. “using a ace mask”,
“ha ing enough medical supplies”); and (3) ole ance o
dep i a ion o social needs (social needs, e.g. “social dis-
ancing om amily”, “no going o wo k”). We in es i-
ga ed indi iduals’ subjec i e pe cep ion ac oss 4 ime-
p ojec ed in e als: (1) 2 weeks ago; (2) now; (3) in 2
weeks; and (4) in 4 weeks. Speci ically, we compa e di -
e ences in he pe cep ion o p ojec ed isk p e en ion
beha iou s and social needs dep i a ion among wo age
g oups: eme ging adul s (younge adul s in he 18–25
yea s age ange) and adul s (adul s aged 26 yea s and old-
e ). We hypo hesize ha : (1) he p ojec ed pe cep ion o
isk will dec ease o e u u e-p ojec ed ime poin s [17],
and ha his educ ion will be enhanced (i.e., dec ease
mo e sha ply) o younge adul s; (2) he impo ance gi -
en o p e en ion beha iou s will dec ease ac oss u u e-
p ojec ed ime poin s, and ha his dec ease will be en-
hanced o younge adul s; (3) he dep i a ion o social
needs will be pe cei ed o be less ole able o e successi e
p ojec ed ime poin s, and ha his will be mo e p o-
nounced (i.e., ole ance educes mo e sha ply) among
younge adul s.
Me hods
Pa icipan s
E hical app o al was g an ed om he Facul y o Psychology
and Educa ional Sciences o he Uni e si y o Coimb a, Po ugal.
Ini ially, a o al o 199 online pa icipan s we e ec ui ed. Due o
he quickly e ol ing na u e o he pandemic si ua ion, pa icipan s
we e ec ui ed h ough di e en channels (e.g., email, social me-
dia) and he esponse a e was no measu ed. Subjec s ga e consen
and we e in o med o hei igh o wi hd aw hei pa icipa ion a
any ime and ha hei da a would be anonymous. Subjec s we e
deb ie ed abou he aims o he s udy upon comple ion o he ques-
ionnai e o no sys ema ically in luence hei answe s. 26 subjec s
we e excluded o no esiding in Po ugal. Addi ionally, ano he
16 we e excluded o missing demog aphic in o ma ion. The inal
sample consis ed o 157 pa icipan s esiding in Po ugal, ac oss
a ious egions o he coun y. Mos pa icipan s esided in owns
be ween 5,000 and 150,000 (n = 101) inhabi an s, abou a i h in
owns wi h 150,000 o mo e inhabi an s (n = 33), and he emain-
ing pa icipan s esided in owns wi h 5,000 o less (n = 23). Pa -
icipan s wi hin he age ange o 18–25 (n = 85) we e designa ed as
“younge adul s” [33], while pa icipan s aged 26+ (n = 72) we e
designa ed as “adul s”. The o al sample comp ised 33 male and
124 emale pa icipan s.
P ocedu e
Th ee ques ionnai es we e cons uc ed o ou su ey. Fo each
o hese ques ionnai es, we pooled se e al ep esen a i e i ems o
di e en isk beha iou s, p e en ion beha iou s and social needs.
I ems o he ques ionnai es we e based on public heal h in o ma-
ion a ailable a he ime om heal h au ho i ies and media (e.g.,
washing hands, main aining physical dis ance om o he s). Those
i ems we e hen ansla ed in o Po uguese. A e he de ini ion o
he pool o ques ions o be included in he su ey, he English and
Po uguese e sions we e se up in he pla o m PsyToolKi [34,
35].
The ec ui men p ocess consis ed o wo pa allel ph ases. Uni-
e si y s uden s we e ec ui ed ia email. Addi ionally, he links
o he su ey we e sha ed on social media o each a b oade sam-
ple (in e ms o age ange and geog aphical loca ion). Da a collec-
ion ook place be ween Ap il 13 and May 15, 2020, app oxima e-
Scale i/Dua e/Sen a/Almeida/Fe ei a/
Walb in/Pilacinski
Po J Public Heal h 2022;40:43–51
46
DOI: 10.1159/000524076
ly 4–9 weeks a e he onse o he na ional s a e o eme gency (i.e.,
manda o y sel -qua an ine). We e i ied ha he da e o da a col-
lec ion did no bias he ob ained a ings (see online suppl. ma e-
ial 2, o all online suppl. ma e ial, see www.ka ge .com/
doi/10.1159/000524076). We adop ed a c oss-sec ional design; all
da a om a gi en indi idual we e ob ained in a single session,
whe e hey p oduced p ojec ed es ima es ac oss he ollowing 4
ime poin s: (1) 2 weeks ago, (2) now, (3) in 2 weeks, (4) in 4 weeks.
As such, we in es iga ed changes in he p ojec ed isk pe cep ion
ac oss ime.
Ques ionnai e
Pa icipan s we e asked o comple e a ques ionnai e ha ook
app oxima ely 15 min. Fi s , hey answe ed se e al demog aphic
ques ions (see below, also see online suppl. Appendix 2), ollowed
by he h ee main ques ionnai es. We de ised i ems based on el-
e an public heal h ad ice (e.g., minimizing he isk o in ec ion by
washing hands, social dis ancing, e c.). Because hese i ems we e
no p e iously alida ed, and because we had no p io knowledge
o which speci ic con ex s we e mos ele an o es ima ing each
measu e (e.g., hand washing o isk minimiza ion), we sough o
cap u e each measu e in a ela i ely b oad ashion. To his end, we
gene a ed a ela i ely la ge se o i ems o each measu e, o e ain
i ems ha collec i ely yielded high C onbach’s alpha. Due o high
in e -i em co ela ions o each ques ionnai e, we e ained all
i ems wi hin and gene a ed a single sco e pe pa icipan (pe each
ques ionnai e and ime poin ).
Demog aphic Da a
We collec ed subjec s’ basic demog aphic in o ma ion, such as
age, gende , p o ession, eligion, and own size (popula ion) and
whe he hey had a known medical condi ion ha migh place
hem a highe isk o COVID-19. The impac o he media upon
hei pe cep ion o ques ionnai e ou come measu es was e alua ed
wi h dicho omous yes/no ques ions ega ding he use o a ious
sou ces o in o ma ion (TV/ adio, in e ne , social media o o he s,
u he asking which o he sou ces hey used) and wi h a 5-poin
Like scale o he pe cei ed c edibili y o na ional news in gen-
e al.
Risk Pe cep ion
Risk pe cep ion was measu ed wi h a 10-poin scale, whe e 1
was “low pe cei ed isk” and 10 was “high pe cei ed isk”. The
ques ionnai e consis ed o 29 i ems – co esponding o 29 isk be-
ha iou s. Fo each i em, esponses we e equi ed o sepa a e 4
ime poin s (2 weeks ago, now, in 2 weeks, and in 4 weeks). See
online supplemen a y Table 1 o mos and leas a ed i ems o all
ime poin s (Fig.1).
In ec ion P e en ion Beha iou s
The impo ance a ibu ed o 9 in ec ion p e en ion beha -
iou s was measu ed in a simila way o he p e ious ques ionnai e
(e.g., ac oss all 4 ime poin s), whe e 1 was “low impo ance” and
10 was “high impo ance”. See online supplemen a y Table 2 o
mos and leas highly a ed i ems o all ime poin s (Fig.2).
Social Needs
The measu emen o social needs (i.e., ole ance o es ic ed
social con ac ) was assessed wi h 6 i ems whe e 1 was “low ole -
ance” and 10 was “high ole ance”. Again, a ings we e ob ained
o all 4 ime poin s. See online supplemen a y Table 3 o mos
and leas highly a ed i ems o all ime poin s.
S a is ical Analysis
S a is ical analyses we e pe o med using IBM SPSS S a is ics
( e sion 25). Missing alues we e eplaced using he ‘se ies means’
8.0
7.5
Risk beha iou s a ing
7.0
6.5
2 weeks ago In 2 weeks In 4 weeks
Time
Now
No e. A e age isk pe cep ion a ings ac oss 4 ime poin s,
o each age g oup. E o ba s deno e he s anda d e o o he
mean (SEM).
Age ca ego y
18–25
>26 6.6
6.4
P e en ion beha iou s a ing
6.2
Age ca ego y
18–25
>26
2 weeks ago In 2 weeks In 4 weeks
Time
Now
No e. A e age a ings o p e en ion beha iou s ac oss
4 ime poin s, o each age g oup. E o ba s deno e
he s anda d e o o he mean (SEM).
Fig. 1. Risk pe cep ion ac oss p ojec ed ime poin s. Fig. 2. Pe cei ed impo ance o p e en ion beha iou s ac oss p o-
jec ed ime poin s.
Colo e sion a ailable online
Colo e sion a ailable online
Op imis ic You h
47
Po J Public Heal h 2022;40:43–51
DOI: 10.1159/000524076
me hod (i.e., eplacing missing alues wi h column means). Fo
each ques ionnai e, he in e i em co ela ions (C onbach’s alpha)
we e conduc ed ( o each ime poin sepa a ely) o e alua e he
in e nal consis ency ac oss i ems ( he equencies o he highes -
sco ed and lowes -sco ed i ems a e p esen ed in he Resul s sec-
ion). Because in e i em co ela ions we e consis en ly high o
each ques ionnai e (and ime poin ), we calcula ed i em mean al-
ues pe ques ionnai e, ime poin , and subjec , and used hese da a
in he main analyses.
Two-way mixed ANOVAs (age g oup × ime poin ) we e pe -
o med on subjec s’ mean ques ionnai e i em alues (pe ime
poin and ques ionnai e). Fo conciseness, we epo only he
highes signi ican ANOVA e ms (e.g., in e ac ions bu no main
e ec s when in e ac ions a e signi ican ; and when no , bo h in e -
ac ions and main e ec s a e epo ed ins ead). Follow-up analyses
we e conduc ed o es i linea ends (i.e., a linea educ ion in
p ojec ed a ing ac oss he 4 ime poin s) was signi ican ; speci i-
cally, o each subjec , a linea eg ession slope (i.e., be a coe i-
cien ) was gene a ed o he 4 ime poin s (i.e., dec easing linea
ends yield nega i e alues), and g oup le el e ec s we e es ed ia
es s on subjec s’ eg ession slope alues. Las ly, explo a o y anal-
yses we e conduc ed o see i he obse ed age di e ences in isk
pe cep ion and p e en ion beha iou s in e ac ed wi h o he de-
mog aphic a iables (see online supplemen 1) and also, whe he
hese measu es in e ac ed wi h he da e o subjec s’ da a collec ion
(online supplemen 2).
Resul s
Risk Pe cep ion
Fi s , we analysed he isk pe cep ion (C onbach’s al-
pha alues anged om 0.952 o 0.967 o he 4 sepa a e
ime poin s, indica ing e y high in e -i em co ela ions,
hus jus i ying mean a e aging o i ems in o indi idual
measu es [pe ime poin ]). A wo-way mixed ANOVA
e ealed a signi ican in e ac ion be ween age g oup and
ime poin (F(1.36, 210.65) = 7.39, p = 0.003). A ollow-up
independen samples es on linea slopes o each age
g oup demons a ed a g ea e dec ease in p ojec ed isk
a ings o younge adul s han o adul s ( (155) = –2.957,
p = 0.002).
In ec ion P e en ion Beha iou s
Nex , we analysed in ec ion p e en ion beha iou s
(high C onbach’s alpha alues o each ime poin we e
obse ed, anging be ween 0.782 and 0.814). A wo-way
mixed ANOVA e ealed a signi ican in e ac ion e ec
be ween age g oup and ime poin (F(1.56, 242.36) = 4.96,
p = 0.013). A ollow-up independen es on subjec s’
linea slopes yielded a signi ican g oup di e ence, indi-
ca ing a sha pe educ ion in he pe cei ed impo ance o
p e en ion beha iou s in younge adul s ( (155) = –2.438,
p = 0.008).
Social Needs
Finally, we analysed ole ance o dep i a ion o social
needs esul ing om manda o y qua an ine and o he
p e en i e policies ( ime poin -speci ic C onbach’s al-
pha alues anged be ween 0.827 and 0.855). A wo-way
mixed ANOVA e ealed ha he in e ac ion be ween
age g oup and ime poin was no signi ican (F(1.263,
195.742) = 1.03, p = 0.329). The end owa ds a main
e ec o age g oup was also no signi ican (F(1, 155) =
2.17, p = 0.143), while he main e ec o ime poin was
signi ican (F(1.263, 195.74) = 152.15, p < 0.001). A ol-
low-up one-sample es on all subjec s’ slope alues e-
ealed a linea dec ease in ole ance owa ds dep i ed
social needs o e ime ( (156) = –13.451, p < 0.001;
Fig.3).
Discussion
The cen al aim o he p esen s udy was o in es iga e
po en ial age- ela ed di e ences in he pe cep ion o isk,
along wi h p e en a i e beha iou , and dep i a ion o so-
cial needs ela ed o e e yday beha iou s du ing he CO-
VID-19 pandemic. Ou esul s p o ide e idence o age-
ela ed di e ences in how isk pe cep ion is p ojec ed
ac oss di e en ime poin s. Bo h he adul s and younge
adul s g oups showed a simila le el o pe cei ed p esen
7.5
7.0
6.5
Social needs a ing
6.0
5.5
5.0
2 weeks ago In 2 weeks In 4 weeks
Time
No e. A e age a ings o ole ance o es ic ed social
needs ac oss 4 ime poin s, o each age g oup.
E o ba s deno e he s anda d e o o he
mean (SEM).
Now
Age ca ego y
18–25
>26
Fig. 3. Tole ance o dep i a ion o social needs ac oss ime poin s.
Colo e sion a ailable online

Scale i/Dua e/Sen a/Almeida/Fe ei a/
Walb in/Pilacinski
Po J Public Heal h 2022;40:43–51
48
DOI: 10.1159/000524076
isk (i.e., a he ime he su ey was illed in). Impo an -
ly, howe e , he p ojec ed isk pe cep ion dec eased o e
ime and his dec ease was signi ican ly mo e p onounced
o he younge adul s. P e en ion beha iou s also
showed he same e ec , sugges ing ha eme gen adul s,
as compa ed o olde adul s, end o pe cei e he ime-
p ojec ed impo ance o p e en ion beha iou s o de-
c ease mo e sha ply o e ime. Finally, unlike he 2 p e-
ceding measu es, when conside ing dep i a ion o social
needs, bo h g oups showed a simila dec ease in in ole -
ance ac oss p ojec ed ime poin s.
Rega ding isk pe cep ion, he obse ed dec eases
o e ime pa allel changes in eal isk beha iou de-
sc ibed by Pole i e al. [17] du ing he H1N1 in luenza
pandemic in I aly, in 2009, demons a ing ha o bo h
eal and p ojec ed da a, pe cei ed isk ends o be highe
a he beginning o pandemic s a es, be o e g adually e-
ducing. Mo e speci ically, he obse ed age di e ence in
he p ojec ed isk pe cep ion is in ag eemen wi h p e i-
ous wo k showing ha olde adul s end o be mo e cau-
ious han younge adul s in heal h- ela ed con ex s [28],
and ha olde adul s end o show a highe le el o sel -
epo ed isk [29] and a e mo e isk-a e se in gene al
[12].
This sha pe educ ion in p ojec ed isk in younge
adul s is po en ially ela ed o expe ien ial di e ences be-
ween he wo age g oups du ing he COVID-19 pan-
demic. Bi di e al. [36] ound ha younge adul s e-
po ed highe le els o pandemic- ela ed s ess, mo e li e
changes, and a g ea e sense o social isola ion, and ha
con ibu ed o o e all poo e psychological well-being.
Po en ially, hese indings indica e ha he COVID-19
pandemic was o e all less ole able o younge adul s
han o olde adul s, which migh con ibu e o a sha p-
e dec ease o he p ojec ed isk pe cep ion. I is impo -
an o no e, howe e , ha while Bi di e al. [36] mea-
su ed he ac ual well-being o hei subjec s, in he p es-
en s udy we used ime-p ojec ed measu es, showing
people’s pe cep ions abou he u u e; he e, we did no
ind a subs an ial age di e ence in p ojec ed ole ance o
dep i ed social needs ac oss ime (Fig.3). We p opose an
al e na i e explana ion o ou esul , conside ing pe -
sonali y di e ences be ween age g oups, po en ially a -
ec ing hei es ima ions abou he u u e. Se e al s udies
ha e ound ha olde adul s a e less op imis ic abou he
u u e in gene al [37–39]. Acco ding o hese indings,
di e ences in ai le els o disposi ional op imism migh
be a be e explana ion o he p esen ly obse ed sha p-
e dec ease in p ojec ed isk pe cep ion among he
younge adul s.
Se e al o he ac o s may also explain age di e ences
in isk pe cep ion. I has been sugges ed ha in ec ion-
ela ed isk is modula ed by he pe cei ed se iousness o
a pa icula disease, such ha he pe cei ed likelihood o
de eloping a disease may pa ly depend on how se ious
i s consequences a e conside ed o be [30]. This could
happen o many easons, such as a gene al endency o
unde es ima e ha m ul exposu e in younge adul s [20],
o a ailu e o ecognize pe sonal ulne abili y o nega i e
ou comes associa ed wi h he isk beha iou s – so-called
“un ealis ic op imism” [16, 25, 40]. Simila ly, Commo-
da i and La Rosa [30] and Commoda i e al. [41] showed
ha younge adul s unde es ima e he isk o in ec ion
because o he widesp ead e idence ha younge people
a e gene ally less ulne able o he ad e se consequences
o COVID-19.
In addi ion o disposi ional op imism, di e ences in
isk pe cep ion a e po en ially in luenced by socio-demo-
g aphic ac o s, such as gende , educa ion and employ-
men . Indeed, ecen s udies conduc ed du ing he CO-
VID-19 pandemic ound ha hese ac o s p edic ed bo h
isk pe cep ion [5, 42] and he adop ion o p e en i e/
p o ec i e beha iou s [43, 44]. Ano he a iable ha is
likely o in luence isk pe cep ion (and subsequen deci-
sion-making) is an indi idual’s emo ional s a e [45]. In-
deed, indi iduals in a nega i e emo ional s a e end o
e alua e e en s mo e nega i ely han indi iduals in a pos-
i i e emo ional s a e [46]. These indings a e ele an o
he cu en esul s because changes in he emo ional ex-
pe ience a e obse ed wi h age: olde adul s a ibu e
g ea e impo ance o emo ions han younge adul s [47].
In pa icula , indi iduals wi h high le els o anxie y end
o pe cei e nega i e ou comes as mo e likely and se e e
[48]. We, he e o e, specula e ha olde adul s exhibi ed
a less s eep decline in he p ojec ed es ima es o isk pe -
cep ion based on hese socio-emo ional ac o s. Finally,
we specula e ha isk pe cep ion (in he p esen s udy)
also depends on knowledge and p e ious expe ience: The
“a ailabili y heu is ic” sugges s ha people wi h a ecen
pe sonal expe ience o a speci ic e en ha e a highe isk
pe cep ion o ha e en [49]. In pa icula , p e ious wo k
shows ha indi iduals who knew someone in ec ed wi h
COVID-19 may show an inc eased pe cep ion o isk [7,
50].
I is wo h no ing ha subjec i e isk pe cep ion is a
key ac o in modula ing heal h beha iou s and migh
in luence he adop ion o p e en i e measu es [30, 41,
51, 52], such as social dis ancing. As such, i is o i al
impo ance o add ess subjec i e aspec s o isk pe cep-
ion ac oss popula ions ha migh pe cei e i di e en ly.
Op imis ic You h
49
Po J Public Heal h 2022;40:43–51
DOI: 10.1159/000524076
In pa allel o he p ojec ed dec ease in isk, we obse ed
a p ojec ed dec ease in he impo ance o in ec ion p e-
en ion beha iou s, especially in younge adul s. This
could be a by-p oduc o he dec ease in pe cei ed isk,
in u n, leading o a educ ion in he need o adop ing
p e en i e measu es. Con a y o wha one migh in ui-
i ely expec , and wha he da a om Bi di e al. [36]
would sugges , we did no ind any age- ela ed di e -
ences in he ole ance o social dep i a ion (al hough a
weak end was e iden o he main e ec o age; see
Fig.3). We sugges ha his may ela e o he gene al use
o digi al communica ion (e.g., ideo calls o ins an
messaging) by he wo age g oups ha was widely used
o educe eelings o social isola ion du ing he ea lie
s ages o he COVID-19 pandemic [53]. We u he spec-
ula e ha he lack o g oup di e ences may a ise om
di e ences in he equency o use be ween ages g oups;
ha is, younge adul s, who end o use digi al commu-
nica ion mo e equen ly o suppo ela i ely wide so-
cial ne wo ks, may ely mo e on digi al communica ion
o egula e eelings o social isola ion, compa ed o o he
adul s ha end o use i less, and may ha e mo e es ic-
i e social ne wo ks.
Limi a ions
We acknowledge se e al limi a ions o ou indings.
Fi s , we used ime-p ojec ed measu es o assess how pa -
icipan s p edic ed ha isk would change h oughou
ime and he ex en o which hese esponses e lec eal-
ime pe cep ion (i.e., app aisals ob ained a di e en ime
in e als). Second, we deployed an in e ne -based sel - e-
po ques ionnai e ha may pa ially be subjec o a deg ee
o sel - epo bias. Thi d, he echnical challenge o con-
duc ing a su ey and collec ing da a quickly du ing he
s a e o eme gency esul ed in a ela i ely low sample size,
hus limi ing he gene alizabili y o ou esul s o he o e -
all popula ion; u he , because o hese cons ain s, we did
no pe o m ex ensi e alida ion o he ques ionnai e
i ems (al hough we did demons a e high in e -i em eli-
abili y wi hin each ques ionnai e). Las ly, since pa ici-
pan s we e ec ui ed p edominan ly h ough uni e si y-
ela ed channels, ou indings a e la gely con ined o a
sample o uni e si y-a ilia ed indi iduals (s uden s and
s a ).
Fu u e Di ec ions
The obse ed sha pe educ ion o p ojec ed isk pe -
cep ion in younge adul s is pa icula ly impo an in he
con ex o accina ion campaigns. Indeed, al hough
many people ecei ed a COVID-19 accine, a high a e o
non-complian people could unde mine accine e o s.
Case o i e al. [54] demons a e ha a pe cei ed highe
isk o he i us p edic ed accine up ake and ha as
eme gency measu es we e li ed (e.g., lockdown), indi-
iduals we e less likely o op o accina ion. We, he e-
o e, highligh he impo ance o isk pe cep ion in a pol-
icy-making con ex . These esul s a e pa icula ly impo -
an gi en ha adhe ence o ecommended sa e y
p ac ices depends on indi iduals’ isk pe cep ion, ac-
co ding o se e al heal h beha iou models, such as he
Heal h Belie Model [55], he Theo y o Reasoned Ac ion
[56], he Theo y o Planned Beha iou [57], and he Sub-
jec i e Expec ed U ili y Theo y [58], which, oge he ,
demons a e ha isk pe cep ion is a c ucial ac o in
shaping isk beha iou .
Concluding Rema ks
The p esen esul s ha e se e al social, psychological,
and poli ical implica ions. In pa icula , hese indings
may se e o in luence policymake s and o he social o -
ganiza ions o design campaigns ha conside age- ela -
ed di e ences in he pe cep ion o isk [20] hus enhanc-
ing he e ec i eness o heal h campaigns a ge ing young
people [59, 60].
Acknowledgmen s
We would like o hank he membe s o P oac ion Lab and es-
pecially Daniel Ribei o and S ephanie K is ensen o hei sugges-
ions while c ea ing he i em lis o his s udy.
S a emen o E hics
S udy app o al s a emen : E hical app o al was g an ed by he
Facul y o Psychology and Educa ional Sciences o he Uni e si y
o Coimb a and he s udy was conduc ed e hically in acco dance
wi h he Wo ld Medical Associa ion Decla a ion o Helsinki.
Consen o pa icipa e s a emen : Subjec s ga e consen and
we e in o med o hei igh o wi hd aw hei pa icipa ion a any
ime and ha hei da a would be anonymised. Subjec s we e de-
b ie ed abou he aims o he s udy upon comple ion o he ques-
ionnai e in o de o no sys ema ically in luence hei answe s.
Con lic o In e es S a emen
The au ho s ha e no con lic s o in e es o decla e.
Scale i/Dua e/Sen a/Almeida/Fe ei a/
Walb in/Pilacinski
Po J Public Heal h 2022;40:43–51
50
DOI: 10.1159/000524076
Re e ences
1 Slo ic P. Pe cep ion o isk. In: Slo ic P, edi-
o . The pe cep ion o isk. S e ling, VA:
Ea hscan; 2000. p. 220–31. O iginal wo k
published in 1987.
2 Kahneman D, Slo ic P, T e sky A. Judgmen
unde unce ain y: heu is ics and biases. Cam-
b idge, UK: Camb idge Uni e si y P ess; 1982.
3 Hanoch Y, Rolison JJ, F eund AM. Does med-
ical isk pe cep ion and isk aking change
wi h age? Risk Anal. 2018; 38(5): 917–28.
4 Sun Y, Sun J. Pe cep ion, p epa edness, and e-
sponse o sunami isks in an aging socie y: e -
idence om Japan. Sa Sci. 2019; 118: 466–74.
5 B uine de B uin W. Age di e ences in CO-
VID-19 isk pe cep ions and men al heal h:
e idence om a na ional U.S. su ey conduc -
ed in Ma ch 2020. J Ge on ol. 2020; 76: e24–9.
6 Pasion R, Pai a TO, Fe nandes C, Ba bosa F.
The AGE e ec on p o ec i e beha io s du -
ing he COVID-19 ou b eak: sociodemo-
g aphic, pe cep ions and psychological ac-
coun s. F on Psychol. 2020; 11: 561785.
7 Guas a ie o E, Toppo C, Magnani FG, Ro-
mano R, Facchini C, Campioni R, e al. Olde
adul s’ isk pe cep ion du ing he COVID-19
pandemic in Lomba dy egion o I aly: a
c oss-sec ional su ey. J Ge on ol Soc Wo k.
2021; 64(6): 585–98.
8 Ki i M, Hansson I, Bjälkeb ing P. Up and
abou : olde adul s’ well-being du ing he
COVID-19 pandemic in a Swedish longi udi-
nal s udy. J Ge on ol B Psychol Sci Soc Sci.
2021; 76(2): e4–9.
9 Rosi A, an Vug FT, Lecce S, Cecca o I, Val-
la ino M, Rapisa da F, e al. Risk pe cep ion
in a eal-wo ld si ua ion (COVID-19): how i
changes om 18 o 87 yea s old. F on Psy-
chol. 2021; 12: e646558.
10 an Winsen F, Wau e s E, Lauwe s LH, de May
Y, Van Passel S, Vancau e en M. Combining
isk pe cep ion and isk a i ude: a comp ehen-
si e indi idual isk beha io model. In: In e -
na ional Cong ess o he Eu opean Associa ion
o Ag icul u al Economis s (EAAE), Augus
30-Sep embe 2, 2011, Zu ich, Swi ze land.
Change and unce ain y challenges o ag icul-
u e, ood and na u al esou ces. Wageningen,
The Ne he lands: Eu opean Associa ion o Ag-
icul u al Economis s; 2011. p. 1–12.
11 Webe EU, Blais AR, Be z NE. A domain-spe-
ci ic isk-a i ude scale: measu ing isk pe -
cep ions and isk beha io s. J Beha Decis
Making. 2002; 15(4): 263–90.
12 Dohmen T, Falk A, Gols eyn BHH, Hu man
D, Sunde U. Risk a i udes ac oss he li e
cou se. Econ J. 2017; 127(605): F95–116.
13 Rolison JJ, Hanoch Y, Wood S, Liu PJ. Risk-
aking di e ences ac oss he adul li e span: a
ques ion o age and domain. J Ge on ol B Psy-
chol Sci Soc Sci. 2014; 69(6): 870–80.
14 Jose AK, Rich e D, Samanez-La kin GR,
Wagne GG, He wig R, Ma a R. S abili y and
change in isk- aking p opensi y ac oss he
adul li e span. J Pe s Soc Psychol. 2016;
111(3): 430–50.
15 Skjong R, Wen wo h BH. Expe judgmen
and isk pe cep ion. In: 11 h In e na ional
O sho e and Pola Enginee ing Con e ence,
S a ange , No way, June 17-22, 2001. P o-
ceedings. Moun ain View, CA: In e na ional
Socie y o O sho e and Pola Enginee s;
2001. p. 537–44.
16 Weins ein ND. Un ealis ic op imism abou
u u e li e e en s. J Pe s Soc Psychol. 1980;
39(5): 806–20.
17 Pole i P, Ajelli M, Me le S. The e ec o isk
pe cep ion on he 2009 H1N1 pandemic in-
luenza dynamics. PLoS One. 2011; 6(2):
e16460.
18 Donohew L, Lo ch EP, Palmg een P. Sensa-
ion seeking and a ge ing o ele ised an i-
d ug PSAs. In: Donohew L, Syphe HE, Bu-
koski WJ, edi o s. Pe suasi e communica ion
and d ug abuse p e en ion. Hillsdale, NJ:
Law ence E lbaum Associa es; 1991. p. 209–
26.
19 Donohew L, Lo ch EP, Palmg een P. Applica-
ions o a heo e ic model o in o ma ion ex-
posu e o heal h in e en ions. Hum Com-
mun Res. 1998; 24(3): 454–68.
20 Johnson RJ, McCaul KD, Klein WM. Risk in-
ol emen and isk pe cep ion among adoles-
cen s and young adul s. J Beha Med. 2002;
25(1): 67–82.
21 Mille CH, Quick BL. Sensa ion seeking and
psychological eac ance as heal h isk p edic-
o s o an eme ging adul popula ion. Heal h
Commun. 2010; 25(3): 266–75.
22 Shee VC, Cline RW. Indi idual di e ences
in sensa ion seeking and sexual beha io : im-
plica ions o communica ion in e en ion
o HIV/AIDS p e en ion among college s u-
den s. Heal h Commun. 1995; 7(3): 205–23.
23 Ska a S, Sussman S, Den CW. P edic ing eg-
ula ciga e e use among con inua ion high
school s uden s. Am J Heal h Beha . 2001;
25(2): 147–56.
24 S ephenson M, Palmg een P. Sensa ion seek-
ing, pe cei ed message sensa ion alue, pe -
sonal in ol emen , and p ocessing o an i-
ma ijuana PSAs. Commun Monog . 2001;
68(1): 49–71.
25 Cohn LD, MacFa lane S, Yanez C, Imai WK.
Risk-pe cep ion: di e ences be ween adoles-
cen s and adul s. Heal h Psychol. 1995; 14(3):
217–22.
26 Fancou D, S ep oe A. COVID-19 social
s udy [In e ne ]. London: Uni e si y College
London; 2020 [ci ed 2021 Dec 10]. A ailable
om: h ps: //www.co idsocials udy.o g/.
27 Blendon RJ, Benson JM, DesRoches CM, Ra-
leigh E, Taylo -Cla k K. The public’s esponse
o se e e acu e espi a o y synd ome in To-
on o and he Uni ed S a es. Clin In ec Dis.
2004; 38(7): 925–31.
28 Bonem EM, Ellswo h PC, Gonzalez R. Age
di e ences in isk: pe cep ions, in en ions
and domains: age di e ences in isk aking. J
Beha Decis Mak. 2015; 28(4): 317–30.
29 Rolison JJ, Hanoch Y, F eund AM. Pe cep ion
o isk o olde adul s: di e ences in e alua-
ions o sel e sus o he s and ac oss isk do-
mains. Ge on ology. 2019; 65(5): 547–59.
30 Commoda i E, La Rosa VL. Adolescen s in
qua an ine du ing COVID-19 pandemic in
I aly: pe cei ed heal h isk, belie s, psycholog-
ical expe iences and expec a ions o he u-
u e. F on Psychol. 2020; 11: 559951.
31 Ped osa AL, Bi encou L, F óes ACF, Ca-
zumbá MLB, Campos RGB, B i o SBC, e al.
Emo ional, beha io al, and psychological im-
pac o he COVID-19 pandemic. F on Psy-
chol. 2020; 11: e566212.
32 A mi age R, Nellums LB. COVID-19 and he
consequences o isola ing he elde ly. Lance
Public Heal h. 2020; 5(5): e256.
Funding Sou ces
This wo k was suppo ed by an ERC s a ing g an “Con en -
MAP” (#802553) o J.A. J.W. was suppo ed by an ERC s a ing
g an “Con en MAP” (#802553). A.P. was suppo ed by a Founda-
ion o Science and Technology o Po ugal and P og ama COM-
PETE g an (PTDC/PSI-GER/30745/2017).
Au ho Con ibu ions
S.S. analyzed da a, co-w o e and ewo ked he manusc ip . I.D.
and C.S. collec ed da a, co-analyzed da a, d a ed he manusc ip .
J.A. and A.J.F. con ibu ed o manusc ip w i ing. J.W. and A.P.
designed he s udy, co-analyzed da a, co-w o e and ewo ked he
manusc ip .
Da a A ailabili y S a emen
The da a ha suppo he indings o his s udy a e a ailable on
eques om he co esponding au ho .
Op imis ic You h
51
Po J Public Heal h 2022;40:43–51
DOI: 10.1159/000524076
33 A ne JJ. Socializa ion in eme ging adul -
hood: om he amily o he wide wo ld,
om socializa ion o sel -socializa ion. In:
G usec JE, Has ings PD, edi o s. Handbook o
socializa ion: heo y and esea ch. New Yo k,
NY: The Guil o d P ess; 2007. p. 208–31.
34 S oe G. PsyToolki : a so wa e package o
p og amming psychological expe imen s us-
ing Linux. Beha Res Me hods. 2010; 42(4):
1096–104.
35 S oe G. PsyToolki : a no el web-based me h-
od o unning online ques ionnai es and e-
ac ion- ime expe imen s. Teach Psychol.
2017; 44(1): 24–31.
36 Bi di KS, Tu kelson A, Finge man KL, Pole-
nick CA, Oya A. Age di e ences in s ess, li e
changes and social ies du ing he COVID-19
pandemic: implica ions o psychological
well-being. Ge on ologis . 2021; 61(2): 2015–
216.
37 Chessell ZJ, Ra hbone CJ, Souchay C, Cha les-
wo h L, Moulin CJA. Au obiog aphical
memo y, pas and u u e e en s, and sel -im-
ages in younge and olde adul s. Sel Iden i y.
2014; 13(4): 380–97.
38 Ko e -G ühn D, Smi h J. When ime is un-
ning ou : changes in posi i e u u e pe cep-
ion and hei ela ionships o changes in
well-being in old age. Psychol Aging. 2011;
26(2): 381–7.
39 Ry CD. Possible sel es in adul hood and old
age: a ale o shi ing ho izons. Psychol Aging.
1991; 6(2): 286–95.
40 Weins ein ND, Klein WM. Un ealis ic op i-
mism: p esen and u u e. J Soc Clin Psychol.
1996; 15(1): 1–8.
41 Commoda i E, La Rosa VL, Coniglio MA.
Heal h isk pe cep ions in he e a o he new
co ona i us: a e he I alian people eady o a
no el i us? A c oss sec ional s udy on pe -
cei ed pe sonal and compa a i e suscep ibil-
i y o in ec ious diseases. Public Heal h.
2020; 187: 8–14.
42 D yhu s S, Schneide CR, Ke J, F eeman
AL, Recchia G, Van De Bles AM, e al. Risk
pe cep ions o COVID-19 a ound he wo ld.
J Risk Res. 2020; 23(7–8): 994–1006.
43 Ca lucci L, D’Amb osio I, Balsamo M. Demo-
g aphic and a i udinal ac o s o adhe ence
o qua an ine guidelines du ing COVID-19:
he I alian model. F on Psychol. 2020; 11:
559288.
44 Li S, Feng B, Liao W, Pan W. In e ne use, isk
awa eness, and demog aphic cha ac e is ics
associa ed wi h engagemen in p e en i e be-
ha io s and es ing: c oss-sec ional su ey on
COVID-19 in he Uni ed S a es. J Med In e -
ne Res. 2020; 22(6): e19782.
45 Loewens ein GF, Webe EU, Hsee CK, Welch
N. Risk as eelings. Psychol Bull. 2001; 127(2):
267–86.
46 Slo ic P, Pe e s E. Risk pe cep ion and a ec .
Cu Di Psychol Sci. 2006; 15(6): 322–5.
47 Ca s ensen LL, Sha i YZ, Ba nes JT. Age ad-
an ages in emo ional expe ience pe sis e en
unde h ea om he COVID-19 pandemic.
Psychol Sci. 2020; 31(11): 1374–85.
48 S öbe J. T ai anxie y and pessimis ic ap-
p aisal o isk and chance. Pe s Indi id Di .
1997; 22(4): 465–76.
49 Slo ic P, Finucane ML, Pe e s E, MacG ego
DG. Risk as analysis and isk as eelings: some
hough s abou a ec , eason, isk, and a io-
nali y. Risk Anal. 2004; 24(2): 311–22.
50 Liu M, Zhang H, Huang H. Media exposu e
o COVID-19 in o ma ion, isk pe cep ion,
social and geog aphical p oximi y, and sel -
a ed anxie y in China. BMC Public Heal h.
2020; 20: 1649.
51 B ewe NT, Weins ein ND, Cui e CL, He -
ing on JE. Risk pe cep ions and hei ela ion
o isk beha io . Ann Beha Med. 2004; 27(2):
125–30.
52 Ibuka Y, Chapman GB, Meye s LA, Li M, Gal-
ani AP. The dynamics o isk pe cep ions
and p ecau iona y beha io in esponse o
2009 (H1N1) pandemic in luenza. BMC In-
ec Dis. 2010; 10: 296.
53 Sha ma RA, Maheshwa i S, B ons he R. CO-
VID-19 in he e a o loneliness. Cu Psychi-
a . 2020; 19(5): 31–3.
54 Case o i M, Gi a di P, Rubal elli E, Tasso A,
Lo o L, Ga a uzzi T. Associa ions o CO-
VID-19 isk pe cep ion wi h accine hesi an-
cy o e ime o I alian esiden s. Soc Sci Med.
2021; 272: 113688.
55 Rosens ock IM. His o ical o igins o he
heal h belie model. Heal h Educ Monog .
1974; 2(4): 328–35.
56 Fishbein M, Ajzen I. Belie , a i ude, in en-
ion, and beha io : an in oduc ion o heo y
and esea ch. Reading, MA: Addison-Wesley;
1975.
57 Ajzen I. The heo y o planned beha io . O -
gan Beha Hum Decis P ocess. 1991; 50(2):
179–211.
58 Edwa ds W. The heo y o decision making.
Psychol Bull. 1954; 51(4): 380–417.
59 Maibach EW, Ro hschild ML, No elli WD.
Social ma ke ing. In: Glanz K, Rime BK,
Lewis FM, edi o s. Heal h beha io and
heal h educa ion: heo y, esea ch, and p ac-
ice. San F ancisco, CA: Jossey-Bass; 2002. p.
437–61.
60 Rima BK, K eu e MW. Ad ancing ailo ed
heal h communica ion: a pe suasion and
message e ec s pe spec i e. J Commun. 2006;
56(Suppl 1): S184–201.