In e na ional Jou nal o
En i onmen al Resea ch
and Public Heal h
A icle
Impac o Ex eme Tempe a u es on Ambulance
Dispa ches Due o Ca dio ascula Causes in
No h-Wes Spain
San iago Ges al Romani 1,2,* , Dominic Royé2,3,4 , Luis Sánchez San os 5and
Adol o Figuei as 2,3,4
1Hospi al o Mon ecelo Pon e ed a, 36071 Pon e ed a, Spain
2Depa men o P e en i e Medicine and Public Heal h, Uni e si y o San iago de Compos ela,
15782 San iago de Compos ela, Spain; [email p o ec ed] (D.R.); [email p o ec ed] (A.F.)
3Conso ium o Biomedical Resea ch in Epidemiology and Public Heal h Ne wo k, (CIBERESP),
15782 San iago de Compos ela, Spain
4Heal h Resea ch Ins i u e o San iago de Compos ela (IDIS), 15782 San iago de Compos ela, Spain
5
Galician Public Heal h Eme gencies Founda ion-061, 36680 A Es ada, Spain; Luis.Sanchez.San os@se gas.es
*Co espondence: san iago.ges al. omani@se gas.es
Recei ed: 31 Oc obe 2020; Accep ed: 1 Decembe 2020; Published: 3 Decembe 2020
Abs ac : In oduc ion and objec i es
. The inc ease in mo ali y and hospi al admissions associa ed
wi h high and low empe a u es is well es ablished. Howe e , less is known abou he in luence o
ex eme ambien empe a u e condi ions on ca dio ascula ambulance dispa ches. This s udy seeks
o e alua e he e ec s o minimum and maximum daily empe a u es on ca dio ascula mo bidi y
in he ci ies o Vigo and A Co uña in No h-Wes Spain, using eme gency medical calls du ing
he pe iod 2005–2017.
Me hods.
Fo he pu poses o analysis, we employed a quasi-Poisson ime
se ies eg ession model, wi hin a dis ibu ed non-linea lag model by exposu e a iable and ci y.
The ela i e isks o cold- and hea - ela ed calls we e es ima ed o each ci y and empe a u e model.
Resul s.
A o al o 70,537 calls we e e alua ed, mos o which we e associa ed wi h low maximum
and minimum empe a u es on cold days in bo h ci ies. A maximum empe a u es, signi ican
cold- ela ed e ec s we e obse ed a lags o 3–6 days in Vigo and 5–11 days in A Co uña. A minimum
empe a u es, cold- ela ed e ec s egis e ed a simila pa e n in bo h ci ies, wi h signi ican ela i e
isks a lags o 4 o 12 days in A Co uña. Hea - ela ed e ec s did no display a clea ly signi ican
pa e n.
Conclusions.
An inc ease in ca dio ascula mo bidi y is obse ed wi h mode a ely low
empe a u es wi hou ex emes being equi ed o es ablish an e ec . Public heal h p e en ion plans
and wa ning sys ems should conside including mode a e empe a u e ange in he p e en ion o
ca dio ascula mo bidi y.
Keywo ds:
ambulance dispa ches; ex eme empe a u e; Galicia; ca dio ascula diseases;
quasi-Poisson eg ession model
1. In oduc ion
The impo ance cu en ly a ached o global clima e change has a oused g ea in e es in s udying
he ela ionships be ween clima e and heal h. One o he clima ic ac o s which has caused g ea es
conce n is empe a u e. The ela ionship be ween empe a u e and mo ali y is well es ablished,
wi h a J, V o U-shaped ela ionship, s a ing om a minimum mo ali y empe a u e (lowe isk) and
inc easing as empe a u es ise o all [
1
,
2
]. I has also been obse ed ha he minimum mo ali y
empe a u e and he unc ional exposu e- esponse cu e a y om one place o ano he due o clima ic
condi ions and he esilience o he popula ion, among o he ac o s [3].
In . J. En i on. Res. Public Heal h 2020,17, 9001; doi:10.3390/ije ph17239001 www.mdpi.com/jou nal/ije ph
In . J. En i on. Res. Public Heal h 2020,17, 9001 2 o 10
Ca dio ascula diseases a e he leading cause o dea h in de eloped coun ies, wi h hei incidence
being associa ed wi h ambien empe a u e [
3
]. On days wi h ex eme empe a u e, an inc ease
has been obse ed in he isk o acu e myoca dial in a c ions [
4
] and ce eb o ascula acciden s [
5
].
Likewise, he incidence o episodes o pa oxysmal a ial ib illa ion, which can cause a decompensa ion
and inc ease ca dio ascula mo bidi y, has also been associa ed wi h empe a u e [
6
]. Unde ce ain
ho and cold en i onmen al condi ions, physiological body empe a u e egula ion mechanisms may
become incapable o main aining he mal equilib ium, leading o al e a ions ha place conside able
s ess on he ca dio ascula sys em [7].
Many s udies published o da e ha e employed ou comes such as mo ali y, hospi al admissions
and eme gency depa men da a [
8
–
10
]. Mo eo e , a con as ing pa e n o empe a u e- ela ed
mo ali y and hospi alisa ion can be obse ed [11].
These ypes o s udy only eco d he mos se e e cases and, hus, only co e a limi ed p opo ion
o pa ien s wi h ca dio ascula p oblems. These a iables may no be su icien ly sensi i e o mode a e
en i onmen al changes [
12
,
13
]. Calls o eme gency medical se ices due o ca dio ascula causes may
be e e lec he in luence o ambien empe a u e on he incidence o ca dio ascula mo bidi y as hey
a e capable o egis e ing small decompensa ions no equi ing hospi al ca e. To da e, e y ew s udies
ha e examined he use o eme gency medical calls due o ca dio ascula causes and hei ela ionship
wi h empe a u e.
We sough o s udy he ela ionship be ween ex eme daily (minimum and maximum) empe a u es
and eme gency medical se ice calls due o ca dio ascula causes in wo medium-sized ci ies in
No h-Wes Spain (Vigo and A Co uña) wi h a p edominan ly oceanic clima e.
2. Me hods
2.1. Se ings
This s udy was ca ied ou in wo ci ies (A Co uña and Vigo) loca ed on he A lan ic coas o
Galicia (No h-Wes Spain). A Co uña (la i ude 43.3
◦
, longi ude
−
8.4
◦
) has a popula ion o 245,711
inhabi an s, 24.5% o whom a e o e 65 yea s o age, while Vigo (la i ude 42.2
◦
, longi ude
−
8.7
◦
) has
a popula ion o 295,364, wi h 22.4% o e 65 yea s o age. A Co uña has a sou he n oceanic clima e and
empe a u es emain mild h oughou he yea , wi h a low annual empe a u e ange. Vigo’s clima e is
one o ansi ion be ween he Medi e anean oceanic clima e and he oceanic clima e, wi h high ain all.
The Galicia-061 Public Heal h Eme gency Founda ion—Fully In eg a ed Eme gency Ca e Cen e
(Fundaci
ó
n P
ú
blica U xencias Sani a ias de Galicia-061—Cen o In eg ado de A enci
ó
n
á
s Eme xencias)
is a public and uni e sal heal h ca e sys em which manages he demand o u gen heal hca e,
o e s medical ad ice ia elephone and mobilises he necessa y heal h esou ces (ambulances) o be
sen o he scene o medical inciden s.
2.2. Heal hca e Da a
Eme gency medical calls we e ob ained om he Galicia-061 Public Heal h Eme gency
Founda ion—Fully In eg a ed Eme gency Ca e Cen e o he pe iod 1 Janua y 2005 o 14 No embe
2017 in he ci ies o Vigo and A Co uña. The medical ale da e was used as a ime e e ence.
The da abase was il e ed by ca dio ascula cause (In e na ional Classi ica ion o Diseases, 9 h and
10 h e isions; ICD-9: 390–459, ICD-10: I00-I99), based on h ee a iables wi h a diagnos ic eco d o
he ale . The h ee diagnos ic eco ds co esponded o: he suspec ed diagnosis by cen al eme gency
ca e s a ; he diagnosis by medicalised-ambulance s a ; and he diagnosis by p ima y ca e s a .
We included all ale s egis e ed wi h a leas one diagnosis o ca dio ascula causes in one o he
h ee a iables. No dis inc ion was d awn conce ning he na u e o e mina ion o he ale (dea h,
ans e o hospi al, e c.).
In . J. En i on. Res. Public Heal h 2020,17, 9001 3 o 10
2.3. Tempe a u e Da a
Da a on minimum (Tmin) and maximum daily empe a u es (Tmax) eco ded by he me eo ological
s a ions loca ed a Vigo ai po and in he ci y cen e o A Co uña we e ob ained om he S a e
Me eo ological Agency (Agencia Es a al de Me eo olog
í
a), o he s udy pe iod (1 Janua y 2005 o
14 No embe 2017).
2.4. S a is ical Analysis
The associa ions be ween eme gency medical calls due o ca dio ascula causes and maximum
and minimum empe a u e we e modelled using a Quasi-Poisson ime se ies eg ession model
wi hin a non-linea lag model dis ibu ed acco ding o each exposu e a iable and ci y [
14
,
15
].
A dis ibu ed lag non-linea model (dlnm) simul aneously i s complex non-linea and lagged e ec s o
an en i onmen al a iable on a esponse a iable in gene alised linea models o gene alised addi i e
models. The s a is ical de ails o he model a e desc ibed by Bhaska an e al. [
16
] and Gaspa ini e
al. [17].
To check o long- e m end and seasonali y, we included a na u al cubic spline wi h 7 deg ees
o eedom (d ) pe yea [
17
]. In addi ion, adjus men o he possible con ounding e ec s o he day
o he week was made wi h an indica o . The lag esponse cu e o he a iable o exposu e was
modelled using a na u al cubic B-spline and h ee in e io kno s (equally-spaced log alues o lags).
Fo he ela ionship be ween empe a u e and eme gency call, we used a c oss-basis ma ix consis ing
o a quad a ic B-spline wi h h ee in e io kno s a speci ic empe a u e pe cen iles o each ci y (10%,
75% and 90%). The lag pe iod was ex ended o 14 days o cap u e he lag in cold- ela ed e ec s [
17
].
The h eshold empe a u e ( he empe a u e a which he ela i e isk o calls is minimal), used as
a e e ence alue was es ima ed using he me hod applied by Tobías e al. [18].
Fo each ci y and empe a u e model, we es ima ed he o e all cumula i e ela i e isks o calls
associa ed wi h cold and hea , espec i ely. This was de ined as he isk o an inc ease in he 5%
and 95% pe cen iles o each empe a u e dis ibu ion in ela ion o minimum mo bidi y empe a u e
(MMT).
The isk o mo ali y a ibu able o a empe a u e x
o a gi en day in he ime se ies is de ined
as he numbe AN
x,
and ac ion AF
x,
o calls egis e ed in he nex Ldays, wi h Lbeing he maximum
lag pe iod, de ined by:
AFx, =1−exp
−
L
X
l=0
βx ,l
and
NAx, =AFx, ·
L
X
l=0
n +1
L+1
whe e
Pβx ,l
is he o e all log-cumula i e isk o empe a u e x
on day , and n
is he numbe o calls
on day . This me hod is desc ibed in de ail by Gaspa ini and Leone [19].
In he sensi i i y analysis, di e en pa ame e isa ions we e es ed, checking o a ious deg ees
o eedom o he a iables (3 o 6 d ) o seasonali y and end (6 o 10 d pe yea ). All s a is ical
analyses we e pe o med wi h he R so wa e sys em ( e sion 3.6.2), using he mgc ( e sion 1.8-15)
and dlnm packages ( e sion 2.2.6) [20].
3. Resul s
Th oughou he s udy pe iod (1 Janua y 2005 o 14 No embe 2017), 70,537 calls we e made o
he Galician eme gency heal hca e sys em due o ca dio ascula causes (37,278 in Vigo and 33,259
in A Co uña). Figu e 1shows he end in he daily se ies o hese calls om 2005 o 2017, showing
a simila seasonal a iabili y o bo h ci ies.
In . J. En i on. Res. Public Heal h 2020,17, 9001 4 o 10
In . J. En i on. Res. Public Heal h 2020, 17, x FOR PEER REVIEW 4 o 10
Figu e 1. O e all cumula i e exposu e- esponse associa ions be ween maximum (A) and minimum
empe a u e (B), and eme gency medical calls in he ci ies o Vigo and A Co uña.
The dis ibu ion acco ding o sex in he wo ci ies displayed he same pa e n, wi h 59% o
eme gency call pa ien s being women and only 40.8% men; 76.2% o all pa ien s we e o e 64 yea s
o age. The mos equen ly diagnosed diseases we e s oke, ca diac ailu e, and ches pain o an
ischaemic na u e.
The Tmax and Tmin alues in Vigo and A Co uña we e simila , wi h o al calls and maximum
alues being highe in Vigo (Table 1, Supplemen a y Ma e ials Figu e S2).
Table 1. Dis ibu ion o daily ex eme empe a u e alues and eme gency calls (2005–2017).
Maximum Tempe a u es (°C) Minimum Tempe a u es (°C) Calls (Numbe )
Ci y Mean Minimum Maximum Mean Minimum Maximum To al Mean Maximum
Vigo 18.9 4.2 40.8 10.1 −2.8 22.5 37278 7.9 23
A Co uña 19.2 4.5 39.5 10.5 −2.0 23.6 33259 7.1 18
The MMT, a which he e ec o empe a u e on calls is es ima ed o be minimal, wi h a ela i e
isk (RR) o 1.0, was obse ed a maximum and minimum empe a u es o 28.3 °C (95% con idence
in e al (CI), 4.5–39.5 °C) and 16.5 °C (95% CI, 15.0–22.5 °C) in A Co uña and 28 °C (95% CI, 14.1–40.8
°C) and 15.4 °C (95% CI, −2.7–16.7 °C) in Vigo (Figu e 1). The impac s o ex eme empe a u es on he
Figu e 1.
O e all cumula i e exposu e- esponse associa ions be ween maximum (
A
) and minimum
empe a u e (B), and eme gency medical calls in he ci ies o Vigo and A Co uña.
The dis ibu ion acco ding o sex in he wo ci ies displayed he same pa e n, wi h 59% o
eme gency call pa ien s being women and only 40.8% men; 76.2% o all pa ien s we e o e 64 yea s
o age. The mos equen ly diagnosed diseases we e s oke, ca diac ailu e, and ches pain o
an ischaemic na u e.
The Tmax and Tmin alues in Vigo and A Co uña we e simila , wi h o al calls and maximum
alues being highe in Vigo (Table 1, Supplemen a y Ma e ials Figu e S2).
Table 1. Dis ibu ion o daily ex eme empe a u e alues and eme gency calls (2005–2017).
Maximum Tempe a u es (◦C) Minimum Tempe a u es (◦C) Calls (Numbe )
Ci y Mean Minimum Maximum Mean Minimum Maximum To al Mean Maximum
Vigo 18.9 4.2 40.8 10.1 −2.8 22.5 37278 7.9 23
A Co uña 19.2 4.5 39.5 10.5 −2.0 23.6 33259 7.1 18
The MMT, a which he e ec o empe a u e on calls is es ima ed o be minimal, wi h a ela i e isk
(RR) o 1.0, was obse ed a maximum and minimum empe a u es o 28.3
◦
C (95% con idence in e al
(CI), 4.5–39.5
◦
C) and 16.5
◦
C (95% CI, 15.0–22.5
◦
C) in A Co uña and 28
◦
C (95% CI,
14.1–40.8 ◦C
)
and 15.4
◦
C (95% CI,
−
2.7–16.7
◦
C) in Vigo (Figu e 1). The impac s o ex eme empe a u es on he
RR o mo bidi y a e summa ised in Figu e 1. The co ela ion be ween eme gency calls and ex eme
In . J. En i on. Res. Public Heal h 2020,17, 9001 5 o 10
empe a u es was posi i e, wi h a non-linea U- and J-shaped inc ease in e ec s obse ed a high
empe a u es in he wo ci ies s udied. Cold- ela ed e ec s we e mo e clea ly obse able compa ed o
hose es ima ed o hea .
Figu e 2shows he lags o he o e all cumula i e hea - and cold- ela ed e ec s o Tmax and Tmin
on eme gency medical calls due o ca dio ascula diseases in A Co uña and Vigo. In he case o he
e ec s o Tmax, he e was a 2-day lag in he appea ance o cold- ela ed e ec s in Vigo, wi h a signi ican
RR a e lags o 3 o 6 days and 5 o 11 days in A Co uña. The highes RRs we e obse ed as ollows:
1.06 (95% CI, 1.03–1.09) a a lag o 3 o 4 days in Vigo, and 1.03 (95% CI, 1.01–1.05) a a lag o 5 o 6 days
in A Co uña due o he e ec s o cold. Es ima es o cold-induced e ec s we e highe o Vigo han o
A Co uña. Rega ding he possible e ec s a ibu able o hea in Tmax, no signi ican RRs we e ound.
In . J. En i on. Res. Public Heal h 2020, 17, x FOR PEER REVIEW 5 o 10
RR o mo bidi y a e summa ised in Figu e 1. The co ela ion be ween eme gency calls and ex eme
empe a u es was posi i e, wi h a non-linea U- and J-shaped inc ease in e ec s obse ed a high
empe a u es in he wo ci ies s udied. Cold- ela ed e ec s we e mo e clea ly obse able compa ed
o hose es ima ed o hea .
Figu e 2 shows he lags o he o e all cumula i e hea - and cold- ela ed e ec s o Tmax and
Tmin on eme gency medical calls due o ca dio ascula diseases in A Co uña and Vigo. In he case
o he e ec s o Tmax, he e was a 2-day lag in he appea ance o cold- ela ed e ec s in Vigo, wi h a
signi ican RR a e lags o 3 o 6 days and 5 o 11 days in A Co uña. The highes RRs we e obse ed
as ollows: 1.06 (95% CI, 1.03–1.09) a a lag o 3 o 4 days in Vigo, and 1.03 (95% CI, 1.01–1.05) a a lag
o 5 o 6 days in A Co uña due o he e ec s o cold. Es ima es o cold-induced e ec s we e highe
o Vigo han o A Co uña. Rega ding he possible e ec s a ibu able o hea in Tmax, no signi ican
RRs we e ound.
A Tmin, cold- ela ed e ec s displayed a simila pa e n, wi h signi ican RRs a lags o 4 o 12
days in A Co uña, ising o a maximum o 1.03 (95% CI, 1.02–1.05) a a lag o 6 days. Cold- ela ed
e ec s in Vigo emained simila , al hough non-signi ican , in all he lags, wi h an RR o 1.01 being
obse ed (95% CI, 0.99–1.03) a a lag o 6 days.
Figu e 2. S uc u e o lags o o e all cumula i e hea -and cold- ela ed e ec s o maximum (A) and
minimum empe a u e (B) on eme gency medical calls in he ci ies o Vigo and A Co uña.
Figu e 2.
S uc u e o lags o o e all cumula i e hea -and cold- ela ed e ec s o maximum (
A
) and
minimum empe a u e (B) on eme gency medical calls in he ci ies o Vigo and A Co uña.
A Tmin, cold- ela ed e ec s displayed a simila pa e n, wi h signi ican RRs a lags o 4 o 12 days
in A Co uña, ising o a maximum o 1.03 (95% CI, 1.02–1.05) a a lag o 6 days. Cold- ela ed e ec s in
Vigo emained simila , al hough non-signi ican , in all he lags, wi h an RR o 1.01 being obse ed
(95% CI, 0.99–1.03) a a lag o 6 days.
Possible hea - ela ed e ec s showed no clea ly signi ican pa e n. Howe e , men ion should be
made o he hea -induced e ec in he ci y o Vigo, wi h an RR o 1.01 (95% CI, 0.99–1.02) a a lag o 1
In . J. En i on. Res. Public Heal h 2020,17, 9001 6 o 10
day. This is ex emely clea in Figu e 2B a a high minimum empe a u e. In con as , no hea - ela ed
e ec s we e de ec ed in A Co uña.
Table 2shows excess eme gency calls a ibu able o hea and cold, acco ding o he exposu e
a iable and ci y. Cold- ela ed ca dio ascula mo bidi y a maximum empe a u es was es ima ed
a 14% (95% CI, 5.4–21.6%) and 7.4% (95% CI,
−
1.8–15.3%) o A Co uña and Vigo, espec i ely.
This ac ion ansla ed as 4650 excess annual calls (95% CI, 1782–7199) o A Co uña and 2751 (95% CI,
−
659–5710) o Vigo. Compa ison be ween he ci ies showed ha excess calls due o cold empe a u es
we e much highe in he case o A Co uña. The obse ed e ec s showed a clea associa ion be ween
medical calls and low empe a u es o maximum and minimum exposu e. In con as , excesses due o
hea we e mani es ly lowe in magni ude, wi h 0.3% (95% CI,
−
0.3–0.8%) and 0.4% (95% CI,
−
0.2–0.9%)
o exposu e o Tmax in A Co uña and Vigo, espec i ely.
Table 2. Eme gency medical calls a ibu able o hea and cold by exposu e a iable and ci y.
Ci y Exposu e E ec * AF% (95 CI%) ‡AN (95 CI%)
A Co uña Maximum cold 14 (5.4. 21.6) 4650 (1782, 7199)
A Co uña Maximum hea 0.3 (−0.3. 0.8) 103 (−96, 270)
Vigo Maximum cold 7.4 (−1.8. 15.3) 2751 (−659, 5710)
Vigo Maximum hea 0.4 (−0.2. 0.9) 142 (−59, 319)
A Co uña Minimum cold 15.7 (7.3. 23.5) 5224 (2443, 7823)
A Co uña Minimum hea 0.3 (−0.2. 0.7) 89 (−81, 231)
Vigo Minimum cold 3.9 (−5.6. 12.5) 1448 (−2083, 4666)
Vigo Minimum hea 0.5 (−0.1. 0.9) 169 (−32, 339)
* AF: a ibu able ac ion, ‡AN: a ibu able numbe .
4. Discussion
This s udy examines he e ec s o ex eme empe a u es on eme gency medical calls due o
ca dio ascula issues in wo medium-sized ci ies (Vigo and A Co uña) in No h-Wes Spain wi h
an oceanic clima e. The esul s sugges ha he numbe o daily eme gency calls a e associa ed wi h
lowe minimum and maximum empe a u es. These indings could indica e ha low empe a u es
a e associa ed wi h mo e cases o eme gency calls han high empe a u es, a leas in ci ies wi h
an oceanic clima e.
The e ec ound o mode a e empe a u es on ambulance dispa ches due o ca dio ascula causes
is consis en wi h ha ound in a na ionwide s udy in Japan [
21
]. In ha s udy, i was obse ed ha
ex eme empe a u es ha e less impac [a ibu able ac ion (AF) =0.1%] on he numbe o calls han
mode a e empe a u es, which a e esponsible o 18% o calls o ca dio ascula causes. A simila AF
was ound o A Co uña (15.7%). The esul s ob ained om he eme gency medical sys em in Hambu g
(Ge many) [
22
] also demons a ed his g ea e impac o mode a e empe a u es. The mos plausible
explana ion is ha he equency o days wi h mode a ely low o high empe a u es is highe han
ha o days wi h ex eme empe a u es. As a as sub ypes o ca dio ascula e en s a e conce ned,
g ea e e ec s a e shown o be induced by low a he han by high empe a u es [23–30].
Ou esul s con as wi h hose o o he s udies ha ha e also e alua ed he impac o empe a u e
on ambulance callou s, speci ically in London [
31
], and in Luoyang (cen al China) [
32
], which ound
a ela ionship be ween ambulance calls o hea and cold. Howe e , he exposu e- esponse cu e was
clea ly dependen on he dispa ch ca ego y [31].
No all s udies use he same measu e o empe a u e o e alua e e ec s on heal h. In ou case,
we op ed o s udy he e ec s o ex eme daily empe a u e (Tmax and Tmin) due o he ac ha his,
as compa ed o mean empe a u e, is a measu e wi h which less exposu e da a a e los , gi en ha
he mean empe a u es a e no capable o e lec ing daily ex emes [
33
]. O he s udies [
34
,
35
] ha e
In . J. En i on. Res. Public Heal h 2020,17, 9001 7 o 10
e alua ed he e ec ia he appa en empe a u e o a ia ion in daily empe a u e, which a e o he
ypes o independen e ec s on empe a u e- ela ed mo bidi y.
The obse ed MMT was es ima ed a 15.4
◦
C in Vigo and 16.5
◦
C in A Co uña o he minimum
and a 28
◦
C in Vigo and 28.3
◦
C in A Co uña o he maximum empe a u e. The small di e ence in he
MMT be ween he wo ci ies can be accoun ed o by he di e ence in he loca ion o he me eo ological
s a ions. I may also be a e lec ion o he mild clima e in bo h ci ies in e ms o low empe a u es.
Simila esul s we e epo ed in Adelaide [36] and B isbane [37].
Ou s udy ound a U- o J-shaped ela ionship be ween empe a u e and eme gency calls due
o ca dio ascula causes, simila o ha obse ed in p e ious s udies [
34
–
38
]. The e ec s a e no
e y p onounced o mode a e empe a u es, al hough hey g ow as hey mo e owa ds he ex emes
(Figu e 1) [2,3].
In e ms o lag s uc u e, he appea ance o he e ec a low empe a u es is ound in lags o
3–6 days in Vigo and 4–11 days in A Co uña, and a high empe a u es in lags o 1–2 days in Vigo.
These esul s a e compa able o hose o s udies in which empe a u e and mo bidi y a e co ela ed
wi h hospi al admissions o eme gency ca e episodes [
8
,
9
], whe e he e ec s due o high empe a u es
a e obse ed on he same day o he ollowing day. These s udies show ha he e ec s due o low
empe a u es end o appea wo days a e exposu e and a e hen seen o pe sis o 2 o 3 days.
In mo ali y s udies, by con as , he e ec s induced by low empe a u es a e obse ed o e a longe
pe iod [39].
The pa hophysiological mechanism in ol ed in he inc ease in ca dio ascula e en s on
days wi h low empe a u es is no clea . I would appea ha in he case o suscep ible
indi iduals, changes induced a he di e en le els desc ibed below could p ecipi a e ca dio ascula
decompensa ion [
40
]. A a cu aneous le el, exposu e o cold s imula es he he mo- ecep o s, p oducing
ca echolamine sec e ion and asocons ic ion [
41
], and inc eases hea a e, blood p essu e and oxygen
demand. A a enal le el, diu esis is inc eased and, wi h plasma olumes deple ed, blood iscosi y
inc eases [
40
]. A a pulmona y le el, mucocilia y ac i i y dec eases, wi h he possibili y o causing
b onchocons ic ion, he elease o biochemical media o s by lung cells, and a endency o de elop
in ec ions. In e ms o blood, he e is an inc ease in acu e phase eac an s, pla ele agg egabili y and
choles e ol alues, all o which a e ca dio ascula isk ac o s [41].
A hea - ela ed e ec wi h a high Tmin was obse ed in he ci y o Vigo, which was close o
s a is ical signi icance, RR =1.01 (95% CI, 0.99–1.02) a lag 1 (Figu e 2B. This is an e ec which has
no been seen in many s udies and may be ela ed o he highe sensi i i y o he measu e used.
Ele a ed nigh empe a u es can p ecipi a e ca dio ascula decompensa ions in people wi h impai ed
he mo egula ion mechanisms and igge dis u bances in adequa e nigh es (ci cadian hy hm) [
41
].
Ou s udy has se e al ad an ages; he i s is he use o ambulance calls as an indica o o
ca dio ascula mo bidi y. We belie e ha his is a mo e sensi i e measu e o mo bidi y since: (1)
i co e s a la ge popula ion and is mo e easily accessible han eme gencies in heal hca e cen es,
as no means o anspo is equi ed. This can be impo an in he case o olde people who li e
alone o people wi h low income; (2) some pa hologies can be ea ed in si u (hype ensi e c ises,
mild hea ailu e symp oms, pa oxysmal a hy hmias, dizziness in ela ion o hypo ensi e condi ions,
among o he s) and, he e o e, a e no egis e ed as hospi al eme gencies. The second ad an age o ou
s udy is ha he es ima ion o he a ibu able ac ion has allowed us o assess which empe a u e
anges ha e a g ea e impac on public heal h, he eby making i possible o (1) design in e en ions
o mi iga e exposu e o hese empe a u es; and (2) adap and op imise heal h se ices acco ding
o empe a u es.
As a as possible limi a ions o his s udy a e conce ned, he ollowing should be no ed. Fi s ly,
al hough well-es ablished me hods we e used o con ol o end and seasonali y, i canno be uled ou
ha he esul s ob ained may ha e been a ec ed by con ounding empo al pa e ns due o unknown
a iables. Secondly, he diagnos ic da a used we e hose supplied by he Fully In eg a ed Eme gency
Ca e Cen e and, al hough such da a may a y a e hospi al ca e, he eby possibly de ac ing om
In . J. En i on. Res. Public Heal h 2020,17, 9001 8 o 10
he accu acy o he s udy, hey none heless ha e he ad an age o g ea e sensi i i y when i comes
o de ec ing less se e e p ocesses. Thi dly, he exposu e assigned o he popula ion was he ex e io
empe a u e, a ixed exposu e which is applied equally o bo h ci ies, wi hou aking in o accoun
in a-u ban di e ences, indi idual acclima isa ion, beha iou al and socioeconomic ac o s o housing
cha ac e is ics, all o which could modi y exposu e [
39
]. Fou hly, he ac o ha ing a single moni o ing
s a ion a ailable in each ci y means ha exposu e da a migh no be su icien ly ep esen a i e o he
whole popula ion. Fu he mo e, he e ec s ob ained may be unde es ima ed as he s a ions chosen
migh no include he in luence o he u ban hea island. Tha said, howe e , he loca ion o he
me eo ological s a ion a Vigo ai po , a an al i ude o app oxima ely 200 me es, means ha he
popula ion exposu e da a could be somewha highe han he alues ob ained. While his migh
change he minimum mo bidi y alue sligh ly, i would no change he o e all e ec s due o he mal
a ia ions. Fi hly, possible con ounding ac o s such as ai pollu ion, which has been associa ed wi h
episodes o ca dio ascula mo bidi y [
41
], we e no aken in o accoun . Howe e , he e a e s udies
which sugges ha ailu e o adjus o en i onmen al pollu ion does no al e esul s [42].
5. Conclusions
This is he i s s udy conduc ed in No h-Wes Spain o e alua e he e ec s o ambien empe a u e
on ca dio ascula mo bidi y using calls o he medical eme gency sys em. E idence was ound o
show ha daily ca dio ascula disease mo bidi y is associa ed wi h mode a ely low maximum and
minimum empe a u es.
Fu he esea ch will be necessa y o s udy deepe eme gency call subg oups and o he
geog aphical a eas as well as popula ion subg oups. The esul s ob ained should be included in
mo bidi y- ela ed p e en ion plans o wa ning sys ems and in ambulance o ecas models. They should
no only ake in o accoun ex eme alues bu also mode a ely low o high empe a u es, and, in addi ion,
should be implemen ed a bo h a egional and a local le el. This could p o e highly ele an in he
cu en con ex o popula ion ageing and global wa ming.
Supplemen a y Ma e ials:
Supplemen a y ma e ials can be ound a h p://www.mdpi.com/1660-4601/17/23/
9001/s1.
Au ho Con ibu ions:
Concep ualiza ion, S.G.R., D.R., A.F. and L.S.S.; me hodology, D.R. and A.F.; o mal
analysis, S.G.R. and D.R.; w i ing-o iginal d a p epa a ion S.G.R.; w i ing- e iew and edi ing, S.G.R., D.R., A.F.
All au ho s ha e ead and ag eed o he published e sion o he manusc ip .
Funding: D.R. was suppo ed by a pos doc o al esea ch ollowship o he Xun a de Galicia (Spain).
Con lic s o In e es : The au ho s decla e no con lic o in e es .
Abb e ia ions
AN/AF A ibu able numbe / ac ion
RR Rela i e isk
Tmin Minimum empe a u e
Tmax Maximum empe a u e
MMT Minimum mo bidi y empe a u e
95% CI 95% con idence in e al
Re e ences
1.
A ms ong, B.G.; Chalabi, Z.; Fenn, B.; Haja , S.; Ko a s, S.; Miloje ic, A.; Wilkinson, P. Associa ion o
mo ali y wi h high empe a u es in a empe a e clima e: England and Wales. J. Epidemiol. Communi y Heal h
2011,65, 340–345. [C ossRe ] [PubMed]
2.
Baccini, M.; Bigge i, A.; Acce a, G.; Kosa sky, T.; Ka souyanni, K.; Anali is, A.; Ande son, H.R.; Bisan i, L.;
D’Ippoli i, D.; Dano a, J.; e al. Hea e ec s on mo ali y in 15 Eu opean ci ies. Epidemiology
2008
,19, 711–719.
[C ossRe ] [PubMed]
In . J. En i on. Res. Public Heal h 2020,17, 9001 9 o 10
3.
B aga, A.L.; Zanobe i, A.; Schwa z, J. The e ec o wea he on espi a o y and ca dio ascula dea hs in 12
U.S. ci ies. En i on. Heal h Pe spec . 2002,110, 859–863. [C ossRe ] [PubMed]
4.
Bhaska an, K.; Haja , S.; Haines, A.; He e , E.; Wilkinson, P.; Smee h, L. E ec s o ambien empe a u e on
he incidence o myoca dial in a c ion. Hea 2009,95, 1760–1769. [C ossRe ]
5.
Roy
é
, D.; Za abei ia, M.T.; Riancho, J.; San u
ú
n, A. A ime se ies analysis o he ela ionship be ween
appa en empe a u e, ai pollu an s and ischemic s oke in Mad id, Spain. En i on. Res.
2019
,173, 349–358.
6.
Comelli, I.; Fe o, J.; Lippi, G.; Comelli, D.; Sa o i, E.; Ce ellin, G. Incidence o acu e-onse a ial ib illa ion
co ela es wi h ai empe a u e. Resul s o a nine-yea su ey. J. Epidemiol. Glob. Heal h
2014
,4, 151–157.
[C ossRe ]
7.
Bła˙zejczyk, A.; Bła˙zejczyk, K.; Ba anowski, J.; Kuchcik, M. Hea s ess mo ali y and desi ed adap a ion
esponses o heal hca e sys em in Poland. In . J. Biome eo ol. 2018,62, 307–318. [C ossRe ]
8.
Zhang, Y.; Yan, C.; Kan, H.; Cao, J.; Peng, L.; Xu, J.; Wang, W. E ec o ambien empe a u e on eme gency
depa men isi s in Shanghai, China: A ime se ies s udy. En i on. Heal h 2014,13, 100. [C ossRe ]
9.
Schwa z, J.; Same , J.M.; Pa z, J.A. Hospi al admissions o hea disease: The e ec s o empe a u e and
humidi y. Epidemiology 2004,15, 755–761. [C ossRe ]
10.
Liu, Y.; Hoppe, B.O.; Con e ino, M. Th eshold E alua ion o Eme gency Risk Communica ion o Heal h
Risks Rela ed o Haza dous Ambien Tempe a u e. Risk Anal. 2018,38, 2208–2221. [C ossRe ]
11.
Iñiguez, C.; Roy
é
, D.; Tob
í
as, A. Con as ing pa e ns o empe a u e ela ed mo ali y and hospi alisa ion by
ca dio ascula and espi a o y diseases in 52 Spanish ci ies. En i on. Res. 2020,192, 110191.
12.
Roy
é
, D.; Figuei as, A.; Ta acido, M. Sho - e m e ec s o hea and cold on espi a o y d ug use. A ime-se ies
epidemiological s udy in A Co uña, Spain. Pha macoepidemiol. D ug Sa . 2018,27, 638–644.
13.
Ca acedo-Ma inez, E.; Sanchez, C.; Ta acido, M.; Saez, M.; Ja o, V.; Figuei as, A. E ec o sho - e m
exposu e o ai pollu ion and pollen on medical eme gency calls: A case-c osso e s udy in Spain. Alle gy
2008,63, 347–353. [C ossRe ] [PubMed]
14.
Gaspa ini, A.; A ms ong, B. Reducing and me a-analysing es ima es om dis ibu ed lag non-linea models.
BMC Med. Res. Me hodol. 2013,13, 1. [C ossRe ]
15.
Gaspa ini, A.; A ms ong, B.; Kenwa d, M.G. Mul i a ia e me a-analysis o non-linea and o he
mul i-pa ame e associa ions. S a . Med. 2012,31, 3821–3839. [C ossRe ]
16.
Bhaska an, K.; Gaspa ini, A.; Haja , S.; Smee h, L.; A ms ong, B. Time se ies eg ession s udies in
en i onmen al epidemiology. In . J. Epidemiol. 2013,42, 1187–1195. [C ossRe ]
17.
Gaspa ini, A.; Guo, Y.; Hashizume, M.; La igne, E.; Zanobe i, A.; Schwa z, J.; Tobias, A.; Tong, S.; Rocklö , J.;
Fo sbe g, B.; e al. Mo ali y isk a ibu able o high and low ambien empe a u e: A mul icoun y
obse a ional s udy. Lance 2015,386, 369–375. [C ossRe ]
18.
Tob
í
as, A.; A ms ong, B.; Gaspa ini, A. B ie Repo : In es iga ing Unce ain y in he Minimum Mo ali y
Tempe a u e: Me hods and Applica ion o 52 Spanish Ci ies. Epidemiology 2017,28, 72–76. [C ossRe ]
19.
Gaspa ini, A.; Leone, M. A ibu able isk om dis ibu ed lag models. BMC Med. Res. Me hodol.
2014
,
14, 55. [C ossRe ]
20.
Co e Team, R. A Language and En i onmen o S a is ical Compu ing; R Founda ion o S a is ical Compu ing:
Vienna, Aus ia, 2019; A ailable online: h ps://www.R-p ojec .o g/(accessed on 3 Ma ch 2020).
21.
Onozuka, D.; Hagiha a, A. All-Cause and Cause-Speci ic Risk o Eme gency T anspo A ibu able o
Tempe a u e: A Na ionwide S udy. Medicine 2016,94, e2259. [C ossRe ]
22.
Hensel, M.; S uh , M.; Geppe , D.; Ke s en, J.F.; Lo enz, J.; Ke ne , T. Rela ionship be ween ambien
empe a u e and equency and se e i y o ca dio ascula eme gencies: A p ospec i e obse a ional s udy
based on ou -o -hospi al ca e da a. In . J. Ca diol. 2017,228, 553–557. [C ossRe ] [PubMed]
23.
McGuinn, L.; Haja , S.; Wilkinson, P.; A ms ong, B.; Ande son, H.R.; Monk, V.; Ha ison, R. Ambien
empe a u e and ac i a ion o implan able ca dio e e de ib illa o s. In . J. Biome eo ol.
2013
,57, 655–662.
[C ossRe ] [PubMed]
24.
Venclo iene, J.; Baba skiene, R.M.; Dobozinskas, P.; Dedele, A.; Lopa iene, K.; Ragaisy e, N. The sho - e m
associa ions o wea he and ai pollu ion wi h eme gency ambulance calls o pa oxysmal a ial ib illa ion.
En i on. Sci. Pollu . Res. In . 2017,24, 15031–15043. [C ossRe ] [PubMed]
25.
Venclo iene, J.; Baba skiene, R.; Dobozinskas, P.; Siu kai e, V. E ec s o wea he condi ions on eme gency
ambulance calls o acu e co ona y synd omes. In . J. Biome eo ol.
2015
,59, 1083–1093. [C ossRe ] [PubMed]
