Association between electronic cigarette use and respiratory outcomes among people with no established smoking history: a comprehensive review and critical appraisal

Vol.:(0123456789)
In e nal and Eme gency Medicine (2025) 20:667–680
h ps://doi.o g/10.1007/s11739-025-03894-7
IM - REVIEW
Associa ion be weenelec onic ciga e e use and espi a o y
ou comes amongpeople wi hnoes ablished smoking his o y:
acomp ehensi e e iew andc i ical app aisal
A ielleSelya1,2· GiusyRi aMa iaLaRosa3· LuciaSpicuzza2,3,4· JayminB.Mo ja ia5· G aziaCaci6·
Ricca doPolosa2,3,7,8
Recei ed: 6 No embe 2024 / Accep ed: 4 Feb ua y 2025 / Published online: 24 Feb ua y 2025
© The Au ho (s) 2025
Abs ac
Nico ine consump ion in many coun ies is shi ing away om combus ible ciga e es and owa d elec onic ciga e es (ECs).
Unde s anding he o e all popula ion-le el impac equi es weighing hei possible bene i s (e.g., o smoking cessa ion/
swi ching) s ha ms (e.g., long- e m heal h isks). Howe e , cu en e idence on heal h isks is limi ed by he absence o
long- e m da a and con ounding by p io ciga e e smoking his o y. Focusing on sho - o medium- e m espi a o y ou comes
associa ed wi h EC use among people who ne e smoked (PWNS) is in o ma i e. We pe o m a na a i e e iew and c i ical
app aisal o s udies examining he p ospec i e associa ion be ween exclusi e EC use and espi a o y ou comes among PWNS
(ei he  ue ne e -smoking o ne e -es ablished smoking). We included 12 s udies wi h p ospec i e designs ha examine
a ange o espi a o y ou comes subsequen o EC use among PWNS. Eigh s udies did no ind s a is ically signi ican di -
e ences in espi a o y isk associa ed wi h baseline EC use. The emaining i e s udies epo ed a signi ican associa ion
in a leas one analysis, bu in ou o hese s udies, associa ions we e no obus ac oss models. Limi a ions included o e -
eliance on da a om he U.S. Popula ion Assessmen o Tobacco and Heal h, unce ain di ec ionali y (i.e., p e-exis ing
espi a o y condi ions we e no always uled ou ), con ounding by o he combus ible obacco use, and small sample sizes.
All bu one s udy lacked clea and s a is ically signi ican e idence o sel - epo ed espi a o y diagnoses associa ed wi h
EC use among PWNS, o showed a enuous associa ion wi h mild espi a o y symp oms. This has a o able implica ions o
ECs’ popula ion heal h impac ; howe e , small sample sizes and s a is ical biases limi his e idence. A o mal sys ema ic
e iew on his opic is o hcoming.
Keywo ds As hma· Ch onic obs uc i e pulmona y disease· Elec onic ciga e es· Ne e -smoking· Respi a o y illness·
Respi a o y symp oms
Abb e ia ions
AHR Adjus ed haza d a e
ASH Ac ion on Smoking and Heal h
AOR Adjus ed odds a io
ARR Adjus ed isk a io
CI Con idence in e al
* Ricca do Polosa
[email p o ec ed]
A ielle Selya
aselya@pinneyassocia es.com
1 Pinney Associa es, Inc., 201 N. C aig S . Sui e 320,
Pi sbu gh, PA15213, USA
2 Cen e o Excellence o  heAccele a ion o HA m
Reduc ion (CoEHAR), Uni e si y o Ca ania, Via S. So ia,
78-Ed. 4, P. 2, 95123Ca ania, I aly
3 Depa men o Clinical andExpe imen al Medicine,
Uni e si y o Ca ania, Ca ania, I aly
4 Respi a o y Uni , Uni e si y Teaching Hospi al
“Policlinico-S.Ma co”, Uni e si y o Ca ania, Ca ania, I aly
5 Depa men o Respi a o y Medicine, Ha e ield Hospi al,
Guy’s & S Thomas’ NHS Founda ion T us , Ha e ield, UK
6 UOC MCAU, Uni e si y Teaching Hospi al
“Policlinico-S. Ma co”, Uni e si y o Ca ania, Ca ania, I aly
7 Ins i u e o In e nal Medicine, Uni e si y Teaching Hospi al
“Policlinico-S.Ma co”, Uni e si y o Ca ania, Ca ania, I aly
8 Depa men o Medicine andSu ge y, “Ko e” Uni e si y
o Enna, Enna, I aly
668 In e nal and Eme gency Medicine (2025) 20:667–680
COPD Ch onic obs uc i e pulmona y disease
EC Elec onic ciga e e
eCO Expi ed ca bon monoxide
eNO Expi ed ni ic oxide
EPA En i onmen al P o ec ion Agency
HRCT High- esolu ion compu ed omog aphy
NHIS Na ional Heal h In e iew Su ey
PATH Popula ion Assessmen o Tobacco and
Heal h
PICO Popula ion, in e en ion, compa a o ,
ou comes
PRISMA P e e ed Repo ed I ems o Sys em-
a ic Re iews and Me a-Analyses
RCT Randomized con olled ial
VERITAS S udy The Vaping E ec s: Real-Wo ld In e -
na ional Su eillance S udy
WHO Wo ld Heal h O ganiza ion
In oduc ion
Combus ible ciga e e smoking emains he p ima y cause
o p e en able p ema u e dea h in many coun ies wo ld-
wide. Elec onic ciga e es (ECs) a e a non-combus ible
nico ine p oduc and as such a e subs an ially less ha m ul
han ciga e es [1–3]. Gi en ha ECs a e a he lowe end o
he isk con inuum [3–6] and a e e ec i e smoking cessa ion
aids [7] among people who a e ac i ely ying o qui smok-
ing in he nea u u e—o al e na i ely, an appealing and
lowe - isk al e na i e consume p oduc o hose who a e
no immedia ely planning o qui [8]—ECs ha e he po en-
ial o subs an ially educe smoking-a ibu able mo ali y in
he popula ion i hey displace ciga e e smoking, acco ding
o simula ion modeling s udies [9–13].
Howe e , he ul ima e popula ion impac s o ECs also
depend on hei possible de imen al e ec s, especially
om long- e m and cumula i e use. While ECs mos likely
pose subs an ially lowe heal h isks ela i e o ciga e es,
he e may be some absolu e le el o isk om EC use alone
(Fig.1). Such absolu e isks could a enua e some o he
p ojec ed bene i s o swi ching comple ely om ciga e es o
ECs, as well as inc ease isks o EC use among people who
smoke (i.e., dual use) and among people who ne e smoked
(PWNS). We ocus he e on espi a o y heal h ou comes, as
e-ciga e es ha e been a ound o su icien ly long o heo-
e ically impac espi a o y heal h.
Cu en e idence on whe he EC use uniquely poses
measu able espi a o y heal h isks has majo limi a ions
due o he na u e o obse a ional s udies, which con ain
selec ion bias and con ounding. In pa icula , since mos
people who use ECs ha e a his o y o smoking combus i-
ble ciga e es (ei he cu en ly o o me ly), any appa en
associa ion be ween EC use and a heal h ou come is likely
con ounded by smoking his o y. While many s udies adjus
o smoking s a us, his is o en insu icien : adjus ing o
mo e de ailed smoking his o y (e.g., using pack-yea s) is
essen ial o accoun o he deg ee o smoking exposu e [14,
15], ye mos s udies me ely adjus o smoking s a us (cu -
en s. o me s. ne e smoking; o e en simply cu en s.
non-cu en smoking). The esul ing esidual con ounding
sugges s ha he appa en associa ion be ween EC use and
heal h ou comes could in eali y be pa ly, o pe haps ully,
explained by cumula i e smoking his o y. S udies ocus-
ing on PWNS can a oid his con ounding bias, and p o ide
s onge e idence examining ECs’ possible di ec heal h
isks in humans; howe e , li le is known abou espi a o y
isks o EC use in his g oup, as EC use among PWNS is
a e [16–18].
Fig. 1 Possible mechanisms by
which e-ciga e es could a ec
espi a o y heal h
669In e nal and Eme gency Medicine (2025) 20:667–680
He e, we conduc a na a i e e iew o exis ing li e a u e
on EC use by PWNS and espi a o y ou comes, de ined as
b oadly as possible. We ocused on s udies wi h p ospec i e
designs o a oid a common bias in c oss-sec ional s udies
on his opic (i.e., due o cases wi h he e e se empo al
sequence whe e he espi a o y ou come p eceded EC use)
[19]. We also c i ically app aise each s udy wi h espec o
s eng hs and weaknesses and make ecommenda ions o
u u e esea ch.
Ma e ials andme hods
We pe o med a li e a u e e iew o iden i y pee - e iewed
s udies ha examined EC use and espi a o y symp oms
among PWNS and used a p ospec i e s udy design. PubMed
and Scopus da abases we e sea ched in Ap il 2024 using he
sea ch e ms: (“ne e -smoke s” OR “ne e smoke s” OR
“ne e smok*” OR “näi e” OR “heal hy”) AND (“e-cig*”
OR “e-ciga e e” OR ENDS OR “elec onic ciga e e*”)
AND (“ espi *” OR “lung”) AND (“coho ” o “obse a-
ional” o “ ollow-up” OR “ andomized con olled ial” OR
“RCT”). Addi ional manual sea ching was done in e e -
ence lis s o included a icles and in ele an pee - e iewed
jou nals on ei he espi a o y disease esea ch o obacco
esea ch.
All ypes o p ospec i e s udy we e included [e.g., clinical
obse a ional s udies, andomized clinical ials (RCTs), and
popula ion su eys]. C oss-sec ional s udies we e excluded
due o unce ain y abou he empo al sequence o exposu e
and ou come [19], as we e labo a o y s udies, e iews, s udy
p o ocols, case epo s, con e ence abs ac s, and a icles no
w i en in English.
Due o he small numbe o eligible s udies, we e ained
s udies ega dless o hei de ini ion o ne e -smoking,
i.e.,ei he “ ue” ne e -smoking (i.e., ne e smoked e en a
pu in one’s li e ime) o ne e -es ablished smoking (< 100
ciga e es/li e ime); see Limi a ions. Addi ionally, some
s udies we e e ained which analyzed people who cu en ly
o o me ly smoke bu which analyzed da a in a way ha
allowed an app oxima e es ima ion o he e ec among
PWNS.
Fi s , i les and abs ac s o all sea ch esul s we e
sc eened independen ly by wo e iewe s (GRMLR and GC),
and eligible o po en ially eligible a icles we e e iewed in
ull by h ee e iewe s (AS, GRMLR, and GC) o de e mine
inal eligibili y. Disag eemen s we e esol ed h ough dis-
cussion, b inging in a ou h e iewe (AS, GRMLR, GC,
and RP). Some a icles ha lacked one o he c i e ia we e
e ained o discussion on a case-by-case basis i all e iew-
e s ag eed hey we e ne e heless in o ma i e.
C i ical app aisal was conduc ed h ough de ailed eading
o he ull ex and ocused on he ollowing conside a ions:
(1) possible iola ions o he co ec empo al sequence, e.g.,
i espi a o y symp oms o condi ions could ha e been p e-
sen om baseline and he e o e co-occu ed o p eceded
EC use; (2) whe he all ele an con ounding ac o s we e
included in he model (e.g., o he non-ciga e e combus i-
ble obacco use); (3) possible sample size limi a ions; (4)
p esence o sensi i i y/supplemen a y analyses, and obus -
ness o indings ac oss di e en analyses; (5) plausibili y
o esul s in he model (e.g., whe he a ailable esul s align
wi h a dose– esponse e ec and a e la ge o cu en EC
use han o me -EC use); (6) how PWNS we e handled in
he model (i.e., subg oup analysis o adjus men o smok-
ing s a us).
Resul s
A e applying inclusion and exclusion c i e ia, a o al o 12
s udies we e included (see Table1; see Table2 o excluded
s udies). In his sec ion, we p o ide only basic cha ac e is ics
and main indings o each s udy, and in Discussion p o ide
a mo e ho ough summa y o each pape along wi h an in e-
g a ed c i ical app aisal o each s udy.
Respi a o y ou comes a ied ac oss s udies: h ee s ud-
ies analyzed sel - epo ed as hma (incidence/onse o sel -
epo ed as hma in Pa el e al. [20], age o sel - epo ed new
onse in Pe ez e al. [21], and sel - epo ed p e alence and
pas -yea as hma a acks in To e al. [22]); wo s udies ana-
lyzed wheezing symp oms (sel - epo ed pas -yea wheezing
in Sanchez-Rome o e al. [23] and onse o sel - epo ed
wheezing symp oms in Xie e al. [24]); ou s udies analyzed
an index o sel - epo ed espi a o y symp oms, some which
used cu -o alues deno ing unc ionally impo an symp-
oms (Ka ey e al. [25]; Reddy e al. [26]; S e ens e al. [27];
Sa gen e al. [14]); one s udy analyzed any sel - epo ed
espi a o y diagnoses (COPD, ch onic b onchi is, emphy-
sema, o as hma; Kenkel e al. [28]); one s udy analyzed lung
unc ion (using expi ed bioma ke s, spi ome y es s, and
high- esolu ion compu ed omog aphy; Polosa e al. [29]);
and one s udy was a e iew and commen a y on you h EC
use (Polosa e al. [30]). Se en s udies ocused on he gene al
adul popula ion, wo ocused on young adul s (ages 18–24),
and i e (some o which also examined adul s) ocused on
you h. A o al o 5 s udies we e iden i ied o adul s and 2
o young people. Table3 p esen s he agg ega e indings o
each o he abo e ou come ca ego ies, along wi h no able
limi a ions (discussed in de ail in Discussion).
Se en s udies epo ed no signi ican associa ion be ween
baseline EC use and subsequen espi a o y ou comes among
PWNS. Fi e epo ed a leas one signi ican associa ion, bu in
ou o hese s udies, his esul was no obus ac oss di e en
models p esen ed (see Table1). The majo i y o s udies (n = 9)
analyzed da a om he Popula ion Assessmen o Tobacco and
670 In e nal and Eme gency Medicine (2025) 20:667–680
Table 1 Included s udies and hei cha ac e is ics
A icle Age g oup Da a sou ce No. o PWNS Exposu e/con ol g oup Ou come Signi ican associa ion among
PWNS?
Ka ey e al. [25]Adul s (18 +) PATH,
Wa es 4–5
65 Cu en s. o me -EC use s.
ne e -EC use
Respi a o y symp oms ( unc-
ionally impo an : cu o o 2
on 0–9 scale; 7 symp oms)
No; “Among ne e combus ible
obacco smoke s, no signi i-
can associa ion was de ec ed
be ween e-ciga e e use and
impo an espi a o y symp-
oms”
Reddy e al. [26] You h (12–17),
Adul s (18 +)
PATH,
Wa es 3–4
Cu en (some days o e e yday)
EC use s. non-cu en use
Incidence/onse o sel - epo ed
espi a o y symp oms in pas
yea
No; non-signi ican due o con i-
dence in e al con aining 1.00:
“exclusi e [EC] use s (adjus ed
odds a io [AOR] s. noncu en
use s, 1.17; 95% CI, 0.79–1.74)”
Sa gen e al. [14] Young adul s (18–24) PATH,
Wa es 1–3
Cu en (some days o e e yday)
EC use s. non-cu en use
Respi a o y symp oms ( unc-
ionally impo an : cu o o 3
on 0–9 scale; 7 symp oms)
Some imes; p e alence ou come:
“Compa ed o ne e use s, he
isk o unc ionally impo an
espi a o y symp oms we e no
signi ican ly di e en o exclu-
si e use s o e-ciga e e”
wo sening ou come:
“… indings o exclusi e
e-ciga e e use we e sensi i e
o symp om se e i y, show-
ing a signi ican associa ion
wi h wo sening symp oms a a
h eshold o ≥ 2… bu no a a
symp om h eshold o ≥ 3”
imp o emen ou come:
“…e-ciga e e use s a a h eshold
o ≥ 3… we e also mo e likely
[ o] show symp om imp o e-
men compa ed o ne e use s”
[No e: hese compa isons we e
no solelyamong PWNS and
may be biased; see Discussion]
Kenkel e al. [28]Adul s (18 +) PATH,
Wa es 1–3
12 Cu en (some days o e e yday)
EC use and used ECs ai ly
egula ly ( s. non-cu en o
ne e egula )
Sel - epo ed diagnoses o
espi a o y disease (COPD,
ch onic b onchi is, emphy-
sema, o as hma)
No; “…among esponden s who
had ne e smoked combus ible
obacco, we ind no e idence
ha cu en o o me e-ciga e e
use is associa ed wi h espi a-
o y disease”
671In e nal and Eme gency Medicine (2025) 20:667–680
Table 1 (con inued)
A icle Age g oup Da a sou ce No. o PWNS Exposu e/con ol g oup Ou come Signi ican associa ion among
PWNS?
S e ens e al. [27] You h (12–17) PATH,
Wa es 3–4
2998 Cu en s. o me -EC use s.
ne e -EC use
Respi a o y symp oms ( unc-
ionally impo an : cu o o 2
on 0–9 scale; 7 symp oms)
No; “Baseline e-ciga e e use
did no inc ease he odds o
ha ing unc ionally impo an
espi a o y symp oms a ollow-
up ega dless o combus ible
obacco use s a us”
Polosa e al. [29]Adul s (18 +) Bespoke sample
o I alian adul s,
3.5-yea ollow-
up
9 Daily EC use o 3 + mon hs s.
ne e -EC use
Lung unc ion, espi a o y
symp oms, eNO, eCO, and
HRCT o lungs
No; “ his s udy did no demon-
s a e any heal h conce ns asso-
cia ed wi h long- e m use o EC
in ela i ely young use s who
did no also smoke obacco”
Sanchez-Rome o e al. [23]Adul s (18 +) PATH,
Wa es 1–5
51 Cu en (some days o e e yday)
EC use s. non-cu en use
Sel - epo ed wheezing in pas
12mon hs
No; “Associa ions we e small
and no s a is ically signi ican
o he odds o sel - epo ed
wheezing among ne e ciga e e
and cu en ENDS use when
compa ed wi h ne e ciga e e
and noncu en [EC] use”
Polosa e al. [30] You h (US middle
and high school
s uden s)
Re iew and commen a y o US you h su ey esul s and espi a o y isks Some imes; “Al hough aping has
been linked o espi a o y symp-
oms, hey end o be ansien
and o unce ain signi icance”
Pa el e al. [20] You h (12–17) PATH,
Wa es 1–5
142 P30D EC use s. no P30D use Incidence/onse o sel - epo ed
as hma diagnosis
No; “…adolescen s using [EC]
exclusi ely… did no [ha e a
s a is ically signi ican highe
isk o inciden diagnosed
as hma a ollow-up]”

672 In e nal and Eme gency Medicine (2025) 20:667–680
Table 1 (con inued)
A icle Age g oup Da a sou ce No. o PWNS Exposu e/con ol g oup Ou come Signi ican associa ion among
PWNS?
Pe ez e al. [21] You h (12–17),
Adul s (18 +)
PATH,
Wa es 1–6
96 you h
160 adul s
P30D EC use s. no P30D use Age o as hma onse Some imes; main analysis:
“adul s who epo ed ne e
using ciga e es… and P30D
ENDS use a he i s wa e o
pa icipa ion had inc eased isk
o as hma incidence a ea lie
ages in compa ison o adul s
who epo ed no P30D [EC]
use”
Supplemen a y analysis:
Non-signi ican due o con i-
dence in e al con aining 1.00:
(ne e use o combus ible TP
and P30D [EC] use ( s. ne e
use o combus ible TP and no
P30D [EC] use), AHR = 0.93
(0.70–1.22))
To e al. [22] Adul s (15–60) CCHS wi h link-
age o admin-
is a i e heal h
eco ds
75 Cu en s. non-cu en EC use As hma a ack in pas
12mon hs, among hose wi h
as hma
Some imes; as hma p e alence
ou come: “EC use s had
inc eased odds o p e alen
as hma compa ed wi h nonus-
e s… bu he associa ion was
no s a is ically signi ican ”
As hma a acks ou come:
“EC use… showed signi ican
in e ac ion wi h sex… Female
EC use s and nonuse s had a
signi ican wo old inc ease in
odds o as hma a acks com-
pa ed wi h male nonuse s”
[No e: hese compa isons we e
no solely among PWNS and
may be biased; see Discussion]
Xie e al. [24] Young adul s (18–24) PATH, Wa es 1–5 312 Cu en s. o me -EC use s.
ne e -EC use
Onse o espi a o y symp oms
(wheezing)
Yes; “Cu en e-ciga e e use was
associa ed wi h highe odds
o any espi a o y symp om…
and wheezing in he ches …
Associa ions pe sis ed among
pa icipan s who ne e smoked
combus ible ciga e es”
CCHS Canadian Communi y Heal h Su ey, COPD ch onic obs uc i e pulmona y disease, EC elec onic ciga e e, eCO expi ed ca bon monoxide, eNO expi ed ni ic oxide, HRCT high-
esolu ion compu ed omog aphy, P30D pas 30-day, PATH Popula ion Assessmen o Tobacco and Heal h, PWNS people who ne e smoked (i.e., < 100 ciga e es/li e ime, including “no e en a
pu ”), TP obacco p oduc s. Bold ex : o e all conclusion abou he p esence o a signi ican associa ion. Unde lined ex : subheadings o speci ic analyses, i applicable
673In e nal and Eme gency Medicine (2025) 20:667–680
Heal h (PATH), a na ionally ep esen a i e longi udinal US
su ey o you h and adul s. Wi h one excep ion (Polosa e al.
[29]; see Discussion), none o he s udies p o ided de ailed
EC de ice cha ac e is ics (e.g., nico ine concen a ion, la o ,
o de ice ype).
Discussion
We iden i ied 12 ele an s udies examining he associa-
ion be ween EC use and subsequen espi a o y ou comes
among PWNS. He e, we discuss each pape ’s indings in
de ail and c i ically app aise he s eng hs and weaknesses o
each s udy, o ganized i s by main inding (whe he o no a
signi ican associa ion be ween EC use and espi a o y ou -
come was ound in a leas one s udy) and nex by age g oup.
S udies epo ing noassocia ion: summa y
andc i ical app aisal
Adul s
In adul s, a 2-yea s udy by Ka ey e al. [25] and a simi-
la ime ame s udy by Reddy e al. [26] bo h ound no
signi ican associa ion be ween baseline EC use and he
de elopmen o unc ionally impo an espi a o y symp-
oms (using a 7-i em index o symp oms, wi h a cu o o
a mean sco e o 2+ on a 0–9 scale) in adul s who ne e
smoked. Bo h s udies analyzed consecu i e wa es o he
longi udinal Popula ion Assessmen on Tobacco and Heal h
(PATH) su ey in adul s.
A limi a ion o exis ing (p edominan ly c oss-sec-
ional) s udies o e-ciga e e use and heal h ou comes is
bias due o he p esence o pa icipan s whose espi a-
o y symp oms p eceded EC use [19]. To o e come his
limi a ion, bo h Ka ey e al. [25] and Reddy e al. [26]
excluded pa icipan s who al eady had a diagnosis o es-
pi a o y disease [25] o espi a o y symp oms [26] a he
baseline wa e. This is a no able s eng h o bo h s udies
as i ensu es he co ec empo al sequence o examine
whe he e-ciga e es ha e a causal e ec on he de elop-
men o espi a o y symp oms. An addi ional s eng h o
Ka ey e al. [25] was o s a i y by ciga e e s a us (cu en ,
o me , o ne e ) which allowed an examina ion o espi a-
o y symp oms uniquely associa ed wi h e-ciga e e use
(i.e., among ne e -smoke s, whose espi a o y symp oms
canno be a ibu ed o smoking his o y). Reddy e al. [26]
Table 2 Excluded s udies and
eason o exclusion
a S udy only examined cu en s. non-cu en smoking and hus combines o me and ne e -smoke s
b While some o he analyses we e uly longi udinal, hey we e in he e e se di ec ion o he esea ch
ques ion (i.e., as hma p edic ing la e ciga e e and e-ciga e e use)
A icle T uly longi udinal? In es iga es ne e -
smoking indi idu-
als?
Za ala-A ciniega e al. 2024, Res Squa e
h ps:// doi. o g/ 10. 21203/ s.3. s- 37931 49/ 1
YES NOa
Delmas e al. 2024, Respi Med
h ps:// doi. o g/ 10. 1016/j. med. 2023. 107496
NO YES
Muke jee e al. 2024, Am J P e Med
h ps:// doi. o g/ 10. 1016/j. amep e. 2023. 12. 005
YES NOa
Cheney e al. 2023, P e Med Rep
h ps:// doi. o g/ 10. 1016/j. pmed . 2023. 102473
NObNOa
Tacke e al. 2024, Tho ax
h ps:// doi. o g/ 10. 1136/ ho ax- 2022- 218670
YES NOa
Ma ingly e al. 2023, P e Med
10.1016/j.ypmed.2023.107512
YES NOa
Chai on e al. 2024, Tob Induc Dis
h ps:// doi. o g/ 10. 18332/ id/ 156839
NO YES
Be lowi z e al. 2023, Am J P e Med
h ps:// doi. o g/ 10. 1016/j. amep e. 2022. 10. 006
YES NOa
Co do a e al. 2022, P e Med Rep
h ps:// doi. o g/ 10. 1016/j. pmed . 2022. 102016
YES NOa
Dai e al. 2020, NTR
h ps:// doi. o g/ 10. 1093/ n / n aa1 80
YES NOa
Tacke e al. 2020, JAMA Ne w Open
h ps:// doi. o g/ 10. 1001/ jaman e wo kopen. 2020. 20671
YES NOa
Bha a and Glan z 2020, Am J P e Med
h ps:// doi. o g/ 10. 1016/j. amep e. 2019. 07. 028
YES NOa
674 In e nal and Eme gency Medicine (2025) 20:667–680
Table 3 O e iew o agg ega e esul s by ype o ou come and no able limi a ions
EC elec onic ciga e e, eCO expi ed ca bon monoxide, eNO expi ed ni ic oxide, HRCT high- esolu ion compu ed omog aphy, P30D pas 30-day, PATH Popula ion Assessmen o Tobacco
and Heal h, PWNS people who ne e smoked (i.e., < 100 ciga e es/li e ime, including “no e en a pu ”), TP obacco p oduc s
Type o ou come S udies Agg ega e indings (see Table1 o de ailed indings) Limi a ions (see Discussion o de ails)
Sel - epo ed as hma ou comes Pa el e al. [20]
Pe ez e al. [21]
To e al. [22]
No e idence o associa ion wi h as hma incidence/
onse [20]; enuous associa ion wi h age o as hma
onse [21] (see Limi a ions);
No e idence o associa ion wi h as hma p e alenceo
pas -yea as hma a acks o e all Annals ATS [39],
bu possible in e ac ion wi h sex o pas -yea as hma
a acks [22] (see Limi a ions)
In Pe ez e al. [21], enuous associa ion wi h age o
as hma onse may be due o esidual con ounding by
o he combus ible obacco use; N.S. a e accoun ing
o his
In To e al. [22], in e ac ion wi h sex may be en i ely due
o sex; no clea e idence o a unique associa ion wi h
EC use
Wheezing symp oms Sanchez-Rome o e al. [23]
Xie e al. [24]
No e idence associa ion wi h sel - epo ed pas -yea
wheezing [23]; associa ion wi h onse o wheezing
[24]
Xie e al. [24] ha e possible con ounding by o he com-
bus ible obacco use, which was no accoun ed o
Respi a o y symp om index Ka ey e al. [25]
Reddy e al. [26]
S e ens e al. [27]
Sa gen e al. [14]
No e idence o associa ion wi h p e alence o unc-
ionally impo an symp oms [14, 25, 26]; enuous
associa ion wi h bo h wo sening and imp o emen
o unc ionally impo an symp oms depending on
cu -o alue [14]
Reddy e al. [26] did no accoun o p io smoking
his o y and associa ions could be due o cumula i e
smoking his o y
In Ka ey e al. [25], s onge associa ion among o me
( s cu en EC use) is no plausible, sugges ing
unaccoun ed- o con ounding
Sa gen e al. [14] adjus ed o smoking s a us a he han
analyzing PWNS sepa a ely, which may in oduce bias
In all s udies he e, small numbe s o PWNS who used
ECs may limi s a is ical powe
Any sel - epo ed espi a o y diagnosis Kenkel e al. [28] No e idence o associa ion wi h p e alence o any
sel - epo ed espi a o y diagnosis
Small sample size o PWNS who used ECs (n = 12)
p e en ed eliable es ima es
Lung unc ion, eNO, eCO, and HRCT o lungs Polosa e al. [29] No e idence o e ec s on lung unc ion Small sample size o PWNS who used ECs (n = 9) p e-
en ed s a is ical analysis
Re iew and commen a y on you h EC use Polosa e al. [30] Some e idence o ansien espi a o y symp oms o
unce ain clinical signi icance
Re lec s limi a ions o unde lying s udies, e.g., unac-
coun ed- o con ounding
675In e nal and Eme gency Medicine (2025) 20:667–680
on he o he hand, only examined cu en e-ciga e e use,
wi hou accoun ing o p io smoking his o y: had he e
been a signi ican associa ion be ween e-ciga e e use and
onse o espi a o y symp oms, i could be due o p io
ciga e e smoking, in which case adjus ing o pack-yea s
would be necessa y ( hough no necessa ily su icien ) o
accoun o his [14].
Limi a ions s ill apply o bo h Ka ey e al. [25] and Reddy
e al. [26] due o he na u e o obse a ional da a. Speci i-
cally, he e may be spu ious associa ions due o o he ac-
o s ha we e unadjus ed o : o example, Ka ey e al. [25]
showed ha among adul s who o me ly smoked, o me EC,
bu no cu en EC use, was associa ed wi h highe odds o
de eloping espi a o y symp oms. Since i is no biologi-
cally plausible ha o me -EC use could ha e a causal e ec ,
while cu en use does no , his likely indica es he in luence
o addi ional con ounding ac o s. Addi ionally, since he e
we e ew ne e -smoking adul s who used ECs, he abili y o
de ec signi ican associa ions wi h espi a o y symp oms
may ha e been limi ed, calling o u u e esea ch o seek
ou la ge samples o adul s who use ECs bu ne e -smoked
ciga e es. O e all, ne e heless, hese indings lack clea
e idence o mode a e o se e e espi a o y isks o EC use
o e he medium e m among ne e -smoking adul s.
Simila ly, Sa gen e al. [14] analyzed adul PATH pa -
icipan s wi hou a diagnosis o COPD a he baseline wa e
and ound no signi ican di e ences in he p e alence o
espi a o y symp oms a he ollow-up wa e be ween adul s
who exclusi ely used ECs s. did no use ECs a baseline.
(This same s udy did, howe e , ind signi ican associa ions
be ween EC use and changes in espi a o y symp oms—
bo h wo sening and imp o emen ; see sec ion below.)
Simila ly o he s udies abo e, a no able s eng h o his
s udy is including only pa icipan s wi hou baseline diag-
noses o espi a o y disease, ensu ing he co ec empo al
sequence o examining he possible e ec s o e-ciga e e
use. A limi a ion o his analysis [14] is ha Sa gen e al.
did no un a dedica ed analysis con aining solely ne e -
smoking adul s, bu a he s a is ically adjus ed o smoking
s a us (i.e., had smoking s a us as a con ol a iable, using
ne e -smoking as he e e ence g oup). In he con ex o a
mul i a ia e eg ession, each es ima e is in e p e ed as he
change in ha a iable wi h all o he a iables in he model
held cons an : i.e., he odds a io o EC use should e lec
he e ec o exclusi e EC use ega dless o whe he a pa -
icipan ne e , o me ly, o cu en ly smoked (p o ided ha
all model assump ions a e me ). Howe e , he e s ill may be
some bias ( owa d a posi i e associa ion) om including all
smoking s a us g oups in he same model, especially i he e
is an unaccoun ed- o in e ac i e e ec be ween EC use and
smoking his o y (such ha hose wi h longe smoking his-
o ies ha e highe heal h isks) which could alsely appea
o ca y o e o PWNS.
A s udy by Kenkel e al. [28] conduc ed a eplica ion
and ex ension o a p io c oss-sec ional s udy by Bha a and
Glan z [31], he la e o which epo ed ha EC use was
associa ed wi h signi ican lyhighe odds o de eloping es-
pi a o y symp oms among adul s in PATH, despi e exclud-
ing hose who al eady had espi a o y disease a baseline.
Howe e , Kenkel e al. [28] no ed ha in Bha a and Glan z’s
analysis [31], he majo i y o EC use s cu en ly o o me ly
smoked: only 12 pa icipan s ou o 17,601 used ECs bu had
ne e smoked ciga e es, and none o hese 12 de eloped es-
pi a o y symp oms in he PATH wa es examined. Since his
g oup was oo small o a o mal s a is ical analysis, Kenkel
e al. [28] ins ead eplica ed Bha a and Glan z’s analysis
[31] bu examined ca ego ies o e-ciga e e and ciga e e
use. Resul s showed no e idence o espi a o y disease o e
a 3-yea pe iod in ne e -smoking adul s who used ECs.
Fu he , among adul s who o me ly and cu en ly smoked,
he e was no ma ginal independen associa ion be ween EC
use and espi a o y ou comes o e and abo e smoking s a-
us, which is consis en wi h Sa gen e al.’s conclusion [14]
ha ciga e e smoking la gely explains onse o espi a o y
symp oms, wi h no addi ional isk in oduced by ECs. Limi-
a ions also apply o Kenkel e al.’s analysis, howe e , due
o he obse a ional na u e o he da a and he small sample
size o adul s who used ECs bu ne e smoked (N = 12) [28].
These s udies collec i ely indica e ha while some indi idu-
als who use ECs may expe ience mild espi a o y symp oms,
e idence is lacking o an o e all medium- e m impac o
ECs on lung heal h in he absence o an es ablished smok-
ing his o y.
S e ens e al. [27] examined e-ciga e e use and espi a-
o y symp oms among you h in PATH Wa es 3–4. They
ocused on you h wi hou as hma a baseline, which is a
s eng h o he s udy as i excludes you h who may ha e
p e-exis ing espi a o y symp oms due o as hma. S e ens
e al. [27] ound ha you h who exclusi ely used ECs a
baseline did no ha e signi ican ly highe odds o epo ing
unc ionally impo an (cu o o 2 on a 0–9 scale) espi a-
o y symp oms (7-i em scale ela ed o wheezing). This was
ue ega dless o combus ible obacco his o y (i.e., he e
was no signi ican associa ion speci ically among ne e -
smoking you h).
Simila ly, a 3.5-yea p ospec i e obse a ional s udy o
daily ECs’ use s wi hou a his o y o smoking by Polosa
e al. [29] ound no signi ican al e a ions in lung unc ion,
espi a o y symp oms, o exhaled b ea h ni ic oxide (eNO),
and no s uc u al abno mali ies in high- esolu ion compu ed
omog aphy (HRCT) scans. Howe e , he s udy aced limi-
a ions, such as a small sample size, which diminished i s
powe o de ec abno mali ies o signi ican changes o e
ime, he po en ial o selec ion bias (as he s udy may ha e
disp opo iona ely included heal hie e-ciga e e use s), and
a medium- e m ollow-up du a ion. This was he only s udy