The Duke treadmill score with bicycle ergometer: Exercise capacity is the most important predictor of cardiovascular mortality

Full esea ch pape The Duke eadmill sco e wi h bicycle e gome e : Exe cise capaci y is he mos impo an p edic o o ca dio ascula mo ali y Esko Saloka i 1 , Ja i A Laukkanen 2,3,4 , Te ho Leh imaki 5 , Sudhi Ku l 2 , Se o Kunu so 6 , F ancesco Zacca di 7 , Ja i Viik 8 , Rami Leh inen 9 , Kjell Nikus 10 , Tii Ko ¨o ¨bi 11 ,Va ¨ino ¨Tu janmaa 12 , Mika Ka ¨ho ¨nen 11 and Tuomo Nieminen 13,14 Abs ac Backg ound: The Duke eadmill sco e, a widely used eadmill es ing ool, is a weigh ed index combining exe cise ime o capaci y, maximum ST-segmen de ia ion and exe cise-induced angina. No p e ious s udies ha e in es iga ed whe he he Duke eadmill sco e and i s indi idual componen s based on bicycle exe cise es ing p edic ca dio ascula dea h. Design: Two popula ions wi h a s anda d bicycle es ing we e used: 3936 pa ien s e e ed o exe cise es ing (2371 men, age 56 13 yea s) om he Finnish Ca dio ascula S udy (FINCAVAS) and a popula ion-based sample o 2683 men (age 53 5.1 yea s) om he Kuopio Ischaemic Hea Disease s udy (KIHD). Me hods: Cox eg ession was applied o isk p edic ion wi h ca dio ascula mo ali y as he p ima y endpoin . Resul s: In FINCAVAS, du ing a median 6.3-yea (in e qua ile ange (IQR) 4.5–8.2) ollow-up pe iod, 180 pa ien s (4.6%) expe ienced ca dio ascula mo ali y. In KIHD, 562 pa ien s (21.0%) died om ca dio ascula causes du ing he median ollow-up o 24.1 (IQR 18.0–26.2) yea s. The Duke eadmill sco e was associa ed wi h ca dio ascula mo ali y in bo h popula ions (FINCAVAS, adjus ed haza d a io (HR) 3.15 o highes s. lowes Duke eadmill sco e e ile, 95% con idence in e al (CI) 1.83–5.42, P<0.001; KIHD, adjus ed HR 1.71, 95% CI 1.34–2.18, P<0.001). Howe e , a e p og essi e adjus men o he Duke eadmill sco e componen s, he sco e was no associa ed wi h ca dio ascula mo ali y in ei he s udy popula ion, as exe cise capaci y in me abolic equi alen s o ask was he dominan ha binge o poo p ognosis. Conclusions: The Duke eadmill sco e is associa ed wi h ca dio ascula mo ali y among pa ien s who ha e unde gone bicycle exe cise es ing, bu me abolic equi alen s o ask, a componen o he Duke eadmill sco e, p o ed o be a supe io p edic o . Keywo ds Ca dio ascula mo ali y, p ognos ic ac o s, s ess es Recei ed 23 Decembe 2017; accep ed 11 Sep embe 2018 1 Depa men o In e nal Medicine, Uni e si y o Helsinki, Finland 2 Ins i u e o Public Heal h and Clinical Nu i ion, Uni e si y o Eas e n Finland, Finland 3 Depa men o Medicine, Cen al Finland Heal h Ca e Dis ic , Finland 4 Facul y o Spo and Heal h Sciences, Uni e si y o Jy a ¨skyla ¨, Finland 5 Depa men o Clinical Chemis y, Uni e si y o Tampe e, Finland 6 School o Clinical Sciences, Uni e si y o B is ol, Sou hmead Hospi al, B is ol, UK 7 Diabe es Resea ch Cen e, Leices e Gene al Hospi al, Leices e , UK 8 Depa men o Biomedical Enginee ing, Tampe e Uni e si y o Technology, Finland 9 Depa men o ICT and Heal h Technology, Tampe e Uni e si y o Applied Sciences, Finland 10 Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e and Hea Cen e , Finland 11 Depa men o Clinical Physiology, Tampe e Uni e si y Hospi al and Uni e si y o Tampe e, Finland 12 Science Cen e, Pi kanmaa Hospi al Dis ic , Finland 13 Depa men o In e nal Medicine, Uni e si y o Helsinki, Finland 14 Sou h Ka elia Cen al Hospi al, Lappeen an a, Finland Co esponding au ho : Tuomo Nieminen, Uni e si y o Helsinki, In e nal Medicine PL 22 (Haa maninka u 4), 00014 Helsingin Yliopis o, Finland. Email: [email p o ec ed] Eu opean Jou nal o P e en i e Ca diology 2019, Vol. 26(2) 199–207 !The Eu opean Socie y o Ca diology 2018 A icle euse guidelines: sagepub.com/jou nals-pe missions DOI: 10.1177/2047487318804618 jou nals.sagepub.com/home/ejpc In oduc ion Al hough o he modali ies ha e pa ially eplaced clin- ical exe cise es as a diagnos ic ool o assessmen o co ona y a e y disease (CAD), he es has become mo e i al in e alua ing p ognosis and ocusing a ail- able ea men s o high- isk pa ien s. 1,2 One o he mos widely s udied and used p ognos ic a iables in exe cise es ing is he Duke eadmill sco e (DTS), which is a weigh ed index combining eadmill exe cise ime wi h s anda d B uce p o ocol, maximum ST-seg- men de ia ion on elec oca diog am (ECG) and exe - cise-induced angina. 3 The sco e was ini ially de eloped o add p ognos ic in o ma ion o ha p o ided by clin- ical da a, co ona y ana omy and le en icula ejec- ion ac ion in pa ien s wi h suspec ed CAD e e ed o ca he e isa ion. 3 The DTS was la e also alida ed o pa ien s wi h suspec ed CAD, 4 and i is s ill in clin- ical use o isk s a ifica ion o hese pa ien s. 5,6 In e es ingly, p io s udies ha e no p ope ly es ed whe he he DTS is a supe io p ognos ic ma ke in compa ison wi h i s componen s. As exe cise ime o he B uce p o ocol can be eplaced wi h me abolic equi alen s o asks (METs), he DTS is also sui able o o he exe cise es p o o- cols. 7 In many Eu opean coun ies, a bicycle e gome e a he han a eadmill is he p e e ed es ing modali y. Howe e , he p ognos ic alue o he DTS using he s anda d bicycle exe cise es ing has no ye been in es- iga ed. Thus he pu pose o his s udy was o assess he p ognos ic alue o he DTS in bicycle exe cise es ing. Combined analysis o se e al pa ame e s is commonly used o simpli y isk assessmen ; howe e , he combin- a ion needs o be supe io o an independen assess- men o each a iable. The e o e, we also compa ed he p ognos ic capabili y o he DTS o ha o i s com- ponen s. We hypo hesise ha he DTS is a use ul p og- nos ic ma ke o ca dio ascula dea h, supe io o i s componen s, in he exe cise es pe o med wi h a bicycle e gome e . Me hods S udy coho s We used wo independen da a coho s. In he Finnish Ca dio ascula S udy (FINCAVAS), all consecu i e pa ien s e e ed o an exe cise es a Tampe e Uni e si y Hospi al and willing o pa icipa e we e en olled be ween Oc obe 2001 and Decembe 2008. A de ailed desc ip ion o he s udy has been published. 8 A s udy popula ion o 4178 pa ien s (2537 men and 1641 women) was ec ui ed. Da a om a o al o 3758 pa ien s (2249 men and 1509 women) wi h ech- nically success ul exe cise es s – including exe cise capaci y and angina da a as well as con inuous digi al ECG wi h ST-segmen de ia ion and slope da a – we e analysed in he p esen s udy. The main indica ions o exe cise es ing we e a sus- picion o CAD (47%), e alua ion o wo k capaci y (27%), e alua ion o a hy hmias (26%), adequacy o CAD ea men (13%), as well as ob aining an exe cise es p ofile p io o an in asi e p ocedu e (9%) o a e myoca dial in a c ion (MI) (7%); some pa ien s had mo e han one indica ion. The e hics commi ee o Tampe e Uni e si y Hospi al dis ic app o ed he s udy p o ocol, and all pa ien s ga e in o med consen as s ipula ed in he Decla a ion o Helsinki. The o he da ase was om he Kuopio Ischaemic Hea Disease (KIHD) isk ac o s udy. This was a lon- gi udinal popula ion-based s udy designed o in es iga e isk ac o s o ca dio ascula diseases, a he oscle osis and ela ed ou comes. The baseline da a consis o an- domly selec ed 2683 men be ween 42 and 61 yea s o age li ing in he ci y o Kuopio and su ounding u al com- muni ies in eas e n Finland, o which 2392 men wi h success ul exe cise es da a we e included in ou s udy. The baseline examina ions we e conduc ed be ween Ma ch 1984 and Decembe 1989. The KIHD s udy has p e iously been desc ibed in de ail. 9 S udy low FINCAVAS. In FINCAVAS, w i en in o med consen , medical his o y, es ing and exe cise ECGs and exe cise sys olic and dias olic blood p essu e eco dings we e ob ained, as p e iously desc ibed. 8 In b ie , p io o he exe cise es using a bicycle e gome e , subjec s lay down in he supine posi ion o 10 minu es, and he es ing ECG was digi ally eco ded. The Mason– Lika 12-lead sys em was used. The exe cise es s we e in ended o be maximal es s excep o 7% o he es s, which e alua ed he s a us a e an MI and applied a hea a e limi o 120–130 bea s/min; e en hose we e included in he analysis. The ini ial wo kload o 20–30 W was inc eased s ep- wise by 10–30 W e e y minu e based on pa ien cha ac- e is ics. ECGs we e digi ally eco ded a 500 Hz wi h he Ca dioSo exe cise sys em ( e sion 4.14; GE Heal hca e, F eibu g, Ge many). The easons o e mina ion o he es we e ches pain (5.7%), dyspnoea (12.8%), a igue (61.6%), ECG changes (1.4%), blood p essu e changes (1.5%) o o he easons (11.5%). The objec i e exe cise le els as defined by he supe ising physician we e max- imal in 76.4%, almos maximal in 18.7%, easonable in 3.1% and poo in 0.6% o he es s. KIHD. In he KIHD s udy, he pa ien s unde wen a maximal symp om-limi ed exe cise ole ance es using an elec ically b aked cycle e gome e wi h inc ease in 200 Eu opean Jou nal o P e en i e Ca diology 26(2) he wo kload o 20 W/minu e. Maximal exe cise cap- aci y was calcula ed om VO 2 max, which was defined as he highes alue o he pla eau o di ec ly measu ed oxygen consump ion using a espi a o y gas analyse (Mijnha d and Medical G aphics, Minneapolis, MN, USA). ECG was eco ded con inuously wi h he Kone 620 elec oca diog aph (Kone, Tu ku, Finland). ECG was p in ed a 30 second in e als du ing exe cise and a leas 5 minu es o eco e y while he subjec was si ing on he bicycle. 9,10 The Duke eadmill sco e FINCAVAS. The DTS was calcula ed as ollows: exe cise ime – (5 max ST-segmen dep ession) – (4 Angina index). The exe cise ime in minu es was app oxima ed om he maximal METs using a linea app oach de i ed om he scien ific s a emen o he Ame ican Hea Associa ion ega ding exe cise s anda ds o es - ing and aining. 7 The ollowing equa ion was used: exe cise ime ¼(METs – 1)/1.044. METs we e es i- ma ed on he s anda dised basis o maximum wo kload and pa ien weigh , wi h 1 MET being equi alen o 3.5 ml oxygen up ake/kg/min. The o iginal DTS used maximal ST-segmen de i- a ion – whe he dep ession o ele a ion – compa ed o he es ing phase om he lead wi h he mos de i- a ion. 3 We ook in o accoun only he maximal ne ST-segmen dep ession (STD) because non-diagnos ic ST-segmen ele a ion is a he common in heal hy people due o ea ly epola isa ion. ST-segmen s we e analysed wi h modified CASE/Ca dioSo so wa e (V.1.84; GE Heal hca e, F eibu g, Ge many). The so - wa e also calcula es he ST slope on he basis o he cu en median complex, measu ed o e an in e al om he J poin o he 1/8 o he a e age R-R in e al (a maximum o 80 ms). This defini ion o slope akes in o accoun he co ela ion be ween hea a e and epola - isa ion ime, a oiding he effec s o he T wa e. 11 The STD and slope we e calcula ed o he end o he es ing phase and o each minu e o he exe cise and eco e y. Possible ST-segmen de ia ion a he es ing phase was ex ac ed om he alues du ing he exe cise and eco - e y o assess he diffe ence caused by he exe cise. I he simul aneous ST-segmen slope was a leas 0.5 mV/s, he dep ession was se o ze o, as up-sloping STD, whe he apid o slow, migh be associa ed wi h an inc eased isk o u u e co ona y e en s, bu is no p e- dic i e o he p esence o myoca dial ischaemia in he gene al popula ion. 12 The maximal STD in all he leads excep aVR was used in he analysis. I he maximal STD was less han 1 mm, i was se o ze o, as i is no conside ed pa hological. Values g ea e han 5 mm (n¼34) we e excluded om he da a because hey we e mos likely due o measu emen e o . The angina index has alues o 0 (no exe cise induced ischaemic ches pain), 1 (non-exe cise-limi ing angina) and 2 (exe cise-limi ing angina). To c ea e he angina index a iable, we had o me ge o iginal FINCAVAS a iables. I he pa ien had no pain o a ypical pain du ing he exe cise, he angina index has he alue 0. I he pa ien did no epo possible angina, bu he clinician es ima ed he pos - es p obabili y o CAD o be low, we also se he angina index alue o 0. I he pa ien had expe ienced ypical angina du ing he exe cise, he angina index was 1 and i angina was he eason o ending he es , he index was 2. KIHD. Exe cise ime was calcula ed om he METs as abo e. In he KIHD s udy, maximal ST-segmen de ia ion om he baseline du ing exe cise was eco ded om he leads V5 and aVF and he maximal alue o ST- segmen de ia ion was used in he analyses. ST-segmen ele a ions we e excluded, as well as dep essions less l han 1 mm and g ea e han 5 mm. The ECGs we e also coded manually by an expe ienced ca diologis , gi ing a a iable o ischaemia alues om 0 o 3, 0 meaning no ischaemia and 1–3 indica ing he se e i y o ischaemia. We ac o ed in his a iable so ha ou a i- able o he maximal STD o he DTS would be as close o ha we used in he FINCAVAS popula ion. I he ischaemia a iable had a alue o 0, he maximal STD a iable was se o 0, and i he ischaemia a iable had alues 1–3, he maximal STD a iable was cons i u ed as abo e. This way we app oxima ed ha he final ST- a iable akes in o accoun he slope o he ST-segmen . I he pa ien had no angina, he DTS angina index has a alue 0, and i a pa ien had exe cise-limi ing angina, he angina index has a alue o 2. Non-limi ing angina was no aken in o accoun in he o iginal s udy se ing. The DTS was cons i u ed as abo e om FINCAVAS. O iginally, i he o e all DTS is a leas 5, he pa ien is conside ed as ‘low isk’, a sco e be ween 11 and 5 deno es ‘mode a e isk’, and a sco e o 11 o less indi- ca es ‘high isk’. 3 Because only a limi ed numbe o pa ien s we e classified as belonging o he high- isk DTS ca ego y, we also di ided he DTS in o equal e - iles aking accoun he whole popula ion om wo coho s. De ini ion o como bidi ies The p esence o CAD (no/yes) was aken om he med- ical eco ds and ques ionnai e (p e ious MI, angina pec o is, e ascula isa ion, and/o use o ni oglyce in). Diabe es (no/yes) was defined as ha ing a clinical diag- nosis and egula ea men wi h die o medica ions. Saloka i e al. 201 In he KIHD s udy, as ing plasma glucose o 7.0 mmol/l o g ea e was an addi ional c i e ium. Smoking (no/yes) was defined as daily smoking (FINCAVAS) o smoking wi hin he pas 30 days (KIHD). The use o b-blocke s was defined as egula use (no/yes). Some o he egula b-blocke use s wi hin FINCAVAS we e ad ised o pause he d ug 3 days be o e he es ; hose pa ien s we e classified in he ‘no’ ca ego y. Asce ainmen o ollow-up e en s Dea h ce ifica es based on he 9 h o 10 h e ision o he In e na ional Classifica ion o Diseases (ICD-10) we e ecei ed om he causes o dea h egis e main- ained by S a is ics Finland; 8,13 he da a sou ce has been alida ed. 14 The e iewe s o dea h ce ifica es we e blinded o e en s and es esul s. The dea hs we e classified as ca dio ascula when he ICD-10 code was I00–I79. Ca dio ascula mo ali y was he p ima y endpoin and all-cause mo ali y he seconda y endpoin . In FINCAVAS, he au opsy a e was 46% o all dea hs and 57% o pa ien s who succumbed o ca dio ascula dea h. In he KIHD s udy he au opsy a e was 80% o all dea hs. S a is ical analysis Diffe ences be ween pa ien and exe cise es cha ac e - is ics acco ding o he DTS in e iles we e compa ed using one-way analysis o a iance, K uskal–Wallis es o chi-squa e es as app op ia e. The use ulness o he DTS in assessing he p ognosis was es ed wi h a Cox p opo ional haza ds model. Co a ia es included age, sex, body mass index (BMI), a diagnosed CAD, smoking s a us, b-blocke usage and diabe es. The componen s o he DTS (STD, angina index and pe o mance in METs) we e added o an addi ional model wi h o he co a ia es. The DTS was en e ed as he o iginal e sion ca ego ised in o h ee pa s (low isk, mode a e isk and high isk) and as a h ee-ca ego y e sion wi h e iles. S a is ical analyses we e pe o med using he SPSS elease 23.0 o Mac (SPSS Inc., Chicago, IL, USA). All s a is ical es s we e wo- ailed and used an alpha le el o 0.05. Resul s Pa ien cha ac e is ics a e gi en in Table 1. Wi hin he o iginal DTS ca ego isa ion, 3.5% o pa ien s o he FINCAVAS popula ion we e classified as being in high, 52% in mode a e and 40% in low isk. In he KIHD s udy popula ion, only 1.6% we e classified as high, 31.4% as mode a e and 67.0% as low- isk pa ien s. In ou new ca ego isa ion a DTS o e 6.7 belongs o he lowes isk e ile, sco es om 3.0 o 6.7 o he middle e ile and sco es below 3.0 o he highes isk e ile. FINCAVAS The age o he popula ion was 56 13 yea s. Dea h om any cause o om ca dio ascula causes was ec- o ded in 404 (10.3%) and 180 (4.6%) pa ien s, espec - i ely, o e he median ollow-up pe iod o 6.3 yea s (in e qua ile ange (IQR) 4.5–8.2). As a h ee-ca ego y a iable wi h he cu poin s om he o iginal eadmill s udies, he DTS (highes s. lowes g oup) was highly p edic i e o bo h ca dio as- cula dea h (HR 2.26, 95% CI 1.48–3.45, P<0.001) and all-cause mo ali y (HR 1.70, 95% CI 1.31–2.22, P<0.001) when adjus ing o o he co a ia es. Howe e , he p ognos ic signal o he DTS disappea ed en i ely when he model was also adjus ed wi h he h ee componen s o he DTS (METs, angina index and maximal STD): HR o ca dio ascula dea h was 1.10 (95% CI 0.63–1.92, P¼0.740). Impo an ly, exe - cise capaci y in METs was he only DTS componen wi h a s a is ically significan HR (0.77, 95% CI 0.70– 0.85, P<0.001) in his ully adjus ed model. The esul s we e simila o he DTS di ided in o equal e iles: he DTS was p edic i e o ca dio ascula dea h (HR 3.15, 95% CI 1.83–5.43, P<0.001) and all-cause dea h (HR 2.08, 95% CI 1.49–2.90, P<0.001). Howe e , when adjus ed wi h he h ee indi- idual componen s, he DTS was no p edic i e o ca dio ascula dea h and, again, METs was he sole componen linked wi h p ognosis, as shown in Table 2 and Figu e 1. KIHD The age o he KIHD s udy popula ion was 53 5.1 yea s. A o al o 1200 dea hs (44.7%) occu ed om any cause and 562 (21.0%) om ca dio ascula causes o e he median ollow-up pe iod o 24.1 yea s (IQR 18.0–26.2 yea s). As a adi ional h ee-ca ego y a iable and adjus ed wi h co a ia es, he DTS was p edic i e o ca dio as- cula dea h and all-cause mo ali y when compa ing he lowes wi h he highes g oup (HR 1.52, 95% CI 1.25–1.85, P<0.001 and 1.37, 95% CI 1.20–1.56, P<0.001, espec i ely). When also adjus ed wi h all h ee componen s o he DTS, he h ee-ca ego ical DTS was no associa ed wi h ca dio ascula dea h (HR 1.11, 95% CI 0.83–1.49, P¼0.48). O he h ee indi idual componen s, only exe cise capaci y in METs was ela ed o he isk o ca dio ascula dea h (HR 0.85, 95% CI 0.79–0.90, P<0.001). 202 Eu opean Jou nal o P e en i e Ca diology 26(2) Table 1. Pa ien and exe cise es cha ac e is ics o bo h coho s acco ding o he DTS isk e iles. FINCAVAS KIHD DTS DTS Lowes isk e ile (n¼1209) Middle isk e ile (n¼1024) Highes isk e ile (n¼1525) Lowes isk e ile (n¼1158) Middle isk e ile (n¼704) Highes isk e ile (n¼530) Mean SD Mean SD Mean SD P alue Mean SD Mean SD Mean SD P alue Age (yea s) 50.3 12.6 57.9 11.7 59.1 12.7 <0.001 51.6 5.3 54.0 4.5 54.1 4.6 <0.001 BMI (kg/m 2 ) 26.5 4.1 28.4 4.7 27.7 4.7 <0.001 26.2 3.0 28.1 3.9 26.5 3.1 <0.001 METs 10.0 2.2 6.8 2.2 6.3 2.7 <0.001 10.0 1.6 7.0 1.7 7.9 2.3 <0.001 Median IQR Median IQR Median IQR Median IQR Median IQR Median IQR STD (mm) 0.0 0.0–0.0 0.0 0.0–0.0 1.3 0.0–1.9 <0.001 0.0 0.0–0.0 0.0 0.0–0.0 1.6 1.3–2.1 <0.001 n(%) n(%) n(%) n(%) n(%) n(%) Angina index 3 (0.2%) 23 (2.2%) 122 (8.0%) <0.001 0 (0.0%) 3 (0.3%) 216 (14.2%) 0 (0.0%) 0 (0.0%) 86 (16.2%) <0.001 Women 448 (37.1%) 446 (43.6%) 615 (40.3%) 0.008 CAD 219 (18.1%) 129 (12.8%) 564 (37.0%) <0.001 151 (13.0%) 229 (32.5%) 194 (36.6%) <0.001 b-Blocke 471 (39.0%) 628 (61.3%) 1011 (66.3%) <0.001 88 (7.6%) 183 (26.0%) 140 (26.4%) <0.001 Diabe es 92 (7.8%) 129 (12.8%) 205 (13.8%) <0.001 35 (3.0%) 46 (6.5%) 42 (7.9%) <0.001 Smoking 278 (23.0%) 251 (24.5%) 319 (20.9%) 0.095 340 (29.4%) 258 (36.6%) 163 (30.9%) 0.004 P io MI 152 (12.6%) 216 (21.1%) 344 (22.6%) <0.001 SD: s anda d de ia ion; DTS: Duke eadmill sco e; METs: me abolic equi alen s o asks; STD: ¼maximal ischaemic ST-segmen dep ession du ing exe cise; CAD: diagnosed co ona y a e y disease; MI: myoca dial in a c ion. Saloka i e al. 203 When di ided in o equal e iles, he DTS was p e- dic i e o ca dio ascula dea h (adjus ed HR 1.43, 95% CI 1.13–1.80, P¼0.003) and all-cause dea h (HR 1.43, 95% CI 1.22–1.66, P<0.000). Howe e , when all com- ponen s o he DTS we e included in he mul i a iable model, only METs and maximal STD we e s a is ically significan , as shown in Table 2 and Figu e 1. Discussion Ou findings demons a e ha he DTS based on he bicycle e gome e es is consis en ly p edic i e o a al ca dio ascula and all-cause mo ali y. The DTS was p edic i e o ca dio ascula dea h in bo h popula ions as an o iginal h ee-ca ego y a iable. Howe e , exe - cise capaci y alone p o ed o be a supe io p edic o o ou comes in compa ison o he DTS. One uni inc emen in exe cise capaci y (METs) was ela ed o a 17–20% lowe isk o ca dio ascula mo ali y. F om he o he wo DTS componen s, he angina index was no p edic i e o ca dio ascula dea h in ei he popu- la ion; howe e , maximal STD was s a is ically signifi- can in he KIHD s udy popula ion. The esul s a e unequi ocal in demons a ing ha he DTS does no p o ide any p ognos ic in o ma ion independen o i s componen s. This is he fi s s udy o assess he p ognos ic alue o he DTS wi h he s anda d bicycle e gome e . A p e- ious s udy wi h a limi ed numbe o pa icipan s (n¼211) was conduc ed o compa e he bicycle e gom- e e in he supine posi ion and he eadmill es when assessing he alue o he DTS. 15 The use o he supine bicycle e gome e ins ead o he eadmill led o a lowe maximal wo kload and METs le el. 15 This was a eason ha pa ien s go lowe DTS esul s and hus hey we e mos ly classified in o highe isk ca ego ies. The e a e se e al possible explana ions as o why BMI, smoking s a us and diabe es we e associa ed wi h ou come in he KIHD s udy bu no in FINCAVAS pa ien s. Fi s , he s udy popula ions we e qui e diffe en . In he FINCAVAS coho , pa ien s we e e e ed o an exe cise es due o clinical easons, whe eas in he KIHD s udy, pa icipan s we e a andomly selec ed popula ion-based sample o men. The KIHD s udy included middle-aged men based on a gene ically and e hnically homogeneous popula ion om eas e n Finland, an a ea p e iously known o i s high p e alence and incidence o ca diome abolic diseases. In addi ion, i is possible ha he assessmen me hods o baseline cha ac e is ics and isk ac o s influenced he obse ed findings. Finally, he KIHD s udy da a had an essen ially longe ollow-up han he o he da ase , g an ing he isk ac o s mo e ime o cause clinical e en s and ca dio ascula mo ali y. Gene al diffe ences be ween eadmill and s anda d bicycle e gome e exe cise es ing ha e been s udied ea lie , and pa ien s seem o ge 10–20% lowe maximal METs i hey a e no accus omed o cycling, 16 possibly because muscle a igue p e en s hem om eaching hei VO 2 max. When he DTS was o iginally de eloped o ead- mill exe cise es ing, ST-de ia ion was he mos Table 2. Cox eg ession su i al analysis o he DTS as e iles, i s componen s and co a ia es, ca dio ascula -mo ali y as an endpoin . FINCAVAS KIHD 95% CI 95% CI HR Lowe Uppe P alue HR Lowe Uppe P alue DTS (3 d s. 1s e ile) 1.149 0.604 2.185 0.672 0.900 0.677 1.195 0.466 METs 0.799 0.726 0.88 <0.001 0.821 0.767 0.879 <0.001 STD 1.042 0.869 1.251 0.656 1.209 1.012 1.444 0.036 Angina 0.757 0.571 1.003 0.053 1.057 0.863 1.295 0.592 Age 1.034 1.018 1.05 <0.001 1.068 1.044 1.094 <0.001 Sex 2.929 1.985 4.323 <0.001 BMI 0.999 0.964 1.035 0.965 1.058 1.029 1.087 <0.001 CAD 0.952 0.69 1.313 0.763 1.403 1.138 1.731 0.002 b-Blocke 1.387 0.935 2.057 0.104 1.379 1.109 1.715 0.004 Diabe es 1.216 0.823 1.796 0.327 2.018 1.509 2.700 <0.001 Smoking 1.183 0.832 1.681 0.349 2.199 1.823 2.652 <0.001 HR: haza d a io; DTS: Duke eadmill sco e; METs: me abolic equi alen s o asks; STD: maximal ischaemic ST-segmen dep ession du ing exe cise; Angina: non-limi ing (FINCAVAS) and limi ing (FINCAVAS and KIHD) angina du ing exe cise; BMI: body mass index; CAD: diagnosed co ona y a e y disease. 204 Eu opean Jou nal o P e en i e Ca diology 26(2) powe ul p ognos ic a iable, and adding angina and exe cise ime inc eased p ognos ic powe . 3 Exe cise capaci y (METs) has been epea edly alida ed o be he single mos powe ul p ognos ic ma ke p edic ing ca dio ascula mo ali y 17–20 and majo ca dio ascula e en s. 21 The Ame ican Hea Associa ion ecen ly made a s a emen ha ca dio espi a o y fi ness should be assessed ou inely in clinical p ac ice. 22 In a p e ious s udy, classic ca dio ascula isk ac o s including dia- be es, hype ension, obesi y and smoking, we e linked o dec eased exe cise capaci y. 23 In a coho o middle- aged men wi h mo e han 40 yea s o ollow-up, low exe cise capaci y led o inc eased o e all mo ali y a es, independen o smoking, blood p essu e and se um choles e ol. 24 A la ge longi udinal s udy o Swedish adolescen men ound ha poo exe cise cap- aci y was also associa ed wi h a g ea e isk o inciden hea ailu e. 25 These examples ou line he independen p ognos ic impo ance o exe cise capaci y. In con as , he p ognos ic significance o ST-de ia ion and angina is s ill unclea , despi e he ac ha he diagnos ic powe o bo h ST-segmen changes and angina is well es ablished. 26–29 The ela i e impo ance o each o he DTS compo- nen s p obably depends la gely on he popula ion. The s udy popula ion om which he DTS was ini ially de i ed had a high p obabili y o CAD, confi med by consequen co ona y ca he e isa ion, whe eas mos 1.0 Low isk Mode a e isk High isk Low isk DTS DTS DTSDTS FINCAVAS(a) (b) (c) (d) KIHD Mode a e isk High isk Low isk Mode a e isk High isk Low isk Mode a e is k High isk 0.9 0.8 Su i al Su i al Su i al Su i al 0.7 HR 3.15, p< 0.001 HR 1.71, p< 0.001 HR 1.15, p= 0.672 HR 0.90, p=0.466 0.6 1.0 0.9 0.8 0.7 0.6 1.0 0.9 0.8 0.7 0.6 1.0 0.9 0.8 0.7 0.6 0 20 40 60 80 100 120 0 20 40 60 80 100 120 0 100 200 300 4000100 Time o dea h o o he end o he ollow-up ( mon hs ) Time o dea h o o he end o he ollow-up ( mon hs ) Time o dea h o o he end o he ollow-up (mon hs) Time o dea h o o he end o he ollow-up (mon hs) 200 300 400 Figu e 1. Adjus ed Cox eg ession su i al analysis acco ding o Duke eadmill sco e isk e iles in bo h coho s, ca dio ascula mo ali y as an endpoin ; (a) and (c) wi hou he componen s o he DTS; (b) and (d) wi h he componen s. Saloka i e al. 205 o he la e s udies ha e consis ed o pa ien s e e ed o exe cise es ing because o a milde suspicion o CAD. This could pa ially explain he shi o he bes p ognos ic ma ke om ST-de ia ion o exe cise capaci y. Whe he combining he h ee componen s in o he DTS yields addi ional p ognos ic in o ma ion in com- pa ison o using he componen s sepa a ely has no been well explo ed. P e iously, using mul i a iable ana- lyses in assessing he p ognosis o pa ien s has enhanced he p ognos ic powe compa ed o analysing independen a iables o he han he DTS compo- nen s. 19,30,31 In hese s udies, he analysed a iables we e hea a e eco e y, T-wa e al e nans and METs. This supplemen a y p ognos ic in o ma ion may ha e been due o he diffe en pa hophysiological bases assessed by diffe en a iables. In he DTS, he e a e wo a iables indica ing myoca dial ischaemia which may no be help ul. The s eng h o his s udy is he usage o wo inde- penden la ge p ospec i e da ase s and he assessmen o a comp ehensi e ange o po en ial co a ia es. The ollow-up o he KIHD s udy is pa icula ly long; how- e e , he s udy is limi ed by he absence o emale pa - icipan s. Da a on angina ha did no limi exe cise was no conside ed in he o iginal KIHD s udy se ing. The s eng hs o FINCAVAS a e a la ge popula ion and elec onically sa ed bea - o-bea ECG da a. Measu ing exe cise capaci y as METs may ha e i s weaknesses ega ding pa ien s’ age and a - ee mass. 32–34 Nei he s udy akes in o accoun he changes in pa ien cha ac e is ics and ea men du ing he ollow-up pe iod. The e can be esidual con ounde s affec ing ca dio ascula mo ali y in addi ion o co a i- a es, which we accoun ed o . Na u ally, we canno d aw causal in e ences om obse a ional s udies. In conclusion, he DTS seemed o be p edic i e o ca dio ascula dea h o pa ien s who unde wen bicycle exe cise es ing in wo diffe en popula ions. Howe e , when adjus ed wi h i s componen s, he p e- dic i e powe o he DTS disappea ed, as exe cise cap- aci y p o ed o be a supe io p edic o o a al ca dio ascula ou comes. Mo e s udies a e needed o confi m ou findings using bicycle exe cise es ing. Measu ing and ca ego ising exe cise capaci ies should be u he cul i a ed. Fu u e esea ch should also con- cen a e on de eloping mo e accu a e mul i a iable p ognos ic sco es which ake in o accoun mo e e sa- ile unde lying pa hophysiological bases. Au ho con ibu ion ES, TN and JAL con ibu ed o he concep ion and design o he wo k. All au ho s con ibu ed o he acquisi ion, analysis, o in e p e a ion o da a o he wo k. ES d a ed he manu- sc ip . All au ho s c i ically e ised he manusc ip . All au ho s ga e final app o al and ag ee o be accoun able o all aspec s o he wo k ensu ing in eg i y and accu acy. Decla a ion o con lic ing in e es s The au ho (s) decla ed no po en ial conflic s o in e es wi h espec o he esea ch, au ho ship, and/o publica ion o his a icle. Funding The au ho (s) disclosed eceip o he ollowing financial sup- po o he esea ch, au ho ship, and/o publica ion o his a icle: he Finnish Ca dio ascula S udy (FINCAVAS) has been financially suppo ed by he compe i i e esea ch und- ing o Tampe e Uni e si y Hospi al (G an s X51001), he Finnish Cul u al Founda ion, he Finnish Founda ion o Ca dio ascula Resea ch (TL), he Emil Aal onen Founda ion (TL), Finland, and he Tampe e Tube culosis Founda ion and by he Academy o Finland (g an 286284 o TL). Re e ences 1. Gula i M, Black HR, Shaw LJ, e al. The p ognos ic alue o a nomog am o exe cise capaci y in women. N Engl J Med 2005; 353: 468–475. 2. Klig ield P and Laue MS. Exe cise elec oca diog am es ing beyond he ST segmen . Ci cula ion 2006; 114: 2070–2082. 3. Ma k DB, Hla ky MA, Ha ell FE, e al. Exe cise ead- mill sco e o p edic ing p ognosis in co ona y a e y disease. Ann In e n Med 1987; 106: 793. 4. Ma k DB, Shaw L, Ha ell FE, e al. P ognos ic alue o a eadmill exe cise sco e in ou pa ien s wi h suspec ed co ona y a e y disease. N Engl J Med 1991; 325: 849–853. 5. Cassa A, Holmes DR, Rihal CS, e al. Ch onic co ona y a e y disease: diagnosis and managemen . Mayo Clin P oc 2009; 84: 1130–1146. 6. Gibbons RJ, Balady GJ, B icke JT, e al. ACC/AHA 2002 Guideline upda e o exe cise es ing: summa y a - icle. A epo o he Ame ican College o Ca diology/ Ame ican Hea Associa ion Task Fo ce on P ac ice Guidelines (Commi ee o Upda e he 1997 Exe cise Tes ing Guidelines). J Am Coll Ca diol 2002; 40: 1531–1540. 7. Fle che GF, Ades PA, Klig ield P, e al. Exe cise s and- a ds o es ing and aining a scien i ic s a emen om he Ame ican Hea Associa ion. Ci cula ion 2013; 128: 873–934. 8. Nieminen T, Leh inen R, Viik J, e al. The Finnish Ca dio ascula S udy (FINCAVAS): cha ac e ising pa ien s wi h high isk o ca dio ascula mo bidi y and mo ali y. BMC Ca dio asc Diso d 2006; 6: 9. 9. Salonen JT. Is he e a con inuing need o longi udinal epidemiologic esea ch? – The Kuopio Ischaemic Hea Disease Risk Fac o S udy. Ann Clin Res 1988; 20: 46–50. 10. Laukkanen JA, Ma ¨kikallio TH, Rau amaa R, e al. Ca dio espi a o y i ness is ela ed o he isk o sudden 206 Eu opean Jou nal o P e en i e Ca diology 26(2) ca diac dea h: a popula ion-based ollow-up s udy. JAm Coll Ca diol 2010; 56: 1476–1483. 11. GE Heal hca e. Use ’s Guide, CASEWS, e sion 1.84. F eibu g, Ge many: GE Heal hca e, 2012. 12. Desai MY, C ugnale S, Mondeau J, e al. Slow upsloping ST-segmen dep ession du ing exe cise: does i eally sig- ni y a posi i e s ess es ? Am Hea J 2002; 143: 482–487. 13. Laukkanen JA, Ku l S, Lakka TA, e al. Exe cise- induced silen myoca dial ischemia and co ona y mo bid- i y and mo ali y in middle-aged men. J Am Coll Ca diol 2001; 38: 72–79. 14. Pajunen P, Koukkunen H, Ke onen M, e al. The alidi y o he Finnish Hospi al Discha ge Regis e and Causes o Dea h Regis e da a on co ona y hea disease. Eu J Ca dio asc P e Rehabil 2005; 12: 132–137. 15. Rahimi K, Thomas A, Adam M, e al. Implica ions o exe cise es modali y on mode n p ognos ic ma ke s in pa ien s wi h known o suspec ed co ona y a e y disease: eadmill e sus bicycle. Eu J Ca dio asc P e Rehabil 2006; 13: 45–50. 16. Mye s J, Buchanan N, Walsh D, e al. Compa ison o he amp e sus s anda d exe cise p o ocols. J Am Coll Ca diol 1991; 17: 1334–1342. 17. Gup a S, Roha gi A, Aye s CR, e al. Ca dio espi a o y i ness and classi ica ion o isk o ca dio ascula disease mo ali y. Ci cula ion 2011; 123: 1377–1383. 18. Kodama S, Sai o K, Tanaka S, e al. Ca dio espi a o y i ness as a quan i a i e p edic o o all-cause mo ali y and ca dio ascula e en s in heal hy men and women: A me a-analysis. JAMA 2009; 301: 2024–2035. 19. Minkkinen M, Nieminen T, Ve ie RL, e al. P ognos ic capaci y o a clinically indica ed exe cise es o ca dio- ascula mo ali y is enhanced by combined analysis o exe cise capaci y, hea a e eco e y and T-wa e al e - nans. Eu J P e Ca diol 2015; 22: 1162–1170. 20. Mye s J, P akash M, F oeliche V, e al. Exe cise cap- aci y and mo ali y among men e e ed o exe cise es - ing. N Engl J Med 2002; 346: 793–801. 21. Kokkinos PF, Faselis C, Mye s J, e al. Ca dio espi a o y i ness and incidence o majo ad e se ca dio ascula e en s in US e e ans: a coho s udy. Mayo Clin P oc 2017; 92: 39–48. 22. Ross R, Blai SN, A ena R, e al. Impo ance o assessing ca dio espi a o y i ness in clinical p ac ice: A case o i ness as a clinical i al sign: a scien i ic s a emen om he Ame ican Hea Associa ion. Ci cula ion 2016; 134: e653–e699. 23. Ch ysohoou C, Skoumas J, Geo giopoulos G, e al. Exe cise capaci y and haemodynamic esponse among 12,327 indi iduals wi h ca dio-me abolic isk ac o s unde going eadmill exe cise. Eu J P e Ca diol 2017; 24: 1627–1636. 24. Laden all P, Pe sson CU, Mandalenakis Z, e al. Low ae obic capaci y in middle-aged men associa ed wi h inc eased mo ali y a es du ing 45 yea s o ollow-up. Eu J P e Ca diol 2016; 23: 1557–1564. 25. Lindg en M, A ˚be g M, Schau elbe ge M, e al. Ca dio espi a o y i ness and muscle s eng h in la e ado- lescence and long- e m isk o ea ly hea ailu e in Swedish men. Eu J P e Ca diol 2017; 24: 876–884. 26. She le K, Ma cus R, F oeliche VF, e al. Hea a e eco e y: alida ion and me hodologic issues. J Am Coll Ca diol 2001; 38: 1980–1987. 27. Gian ossi R, De ano R, Mul ihill D, e al. Exe cise- induced ST dep ession in he diagnosis o co ona y a e y disease. A me a-analysis. Ci cula ion 1989; 80: 87–98. 28. Goldschlage N, Selze A and Cohn K. T eadmill s ess es s as indica o s o p esence and se e i y o co ona y a e y disease. Ann In e n Med 1976; 85: 277–286. 29. Weine DA, McCabe C, Hue e DC, e al. The p edic i e alue o anginal ches pain as an indica o o co ona y disease du ing exe cise es ing. Am Hea J 1978; 96: 458–462. 30. Leino J, Minkkinen M, Nieminen T, e al. Combined assessmen o hea a e eco e y and T-wa e al e nans du ing ou ine exe cise es ing imp o es p edic ion o o al and ca dio ascula mo ali y: he Finnish Ca dio ascula S udy. Hea Rhy hm 2009; 6: 1765–1771. 31. Mo a S, Redbe g RF, Sha e AR, e al. Enhanced isk assessmen in asymp oma ic indi iduals wi h exe cise es ing and amingham isk sco es. Ci cula ion 2005; 112: 1566–1572. 32. Tompu i TT. Me abolic equi alen s o ask a e con- ounded by adiposi y, which dis u bs objec i e measu e- men o physical ac i i y. F on Physiol [In e ne ]. 2015 [ci ed 8 Decembe 2017]; 6. h ps://www. on ie sin.o g/ a icles/10.3389/ phys.2015.00226/ ull (accessed 18 Sep embe 2018). 33. Kokkinos P, Mye s J, F anklin B, e al. Ca dio espi a o y i ness and heal h ou comes: a call o s anda dize i ness ca ego ies. Mayo Clin P oc 2017; 93: 333–336. 34. K achle B, Sa onen K, Komulainen P, e al. Ca diopulmona y i ness is a unc ion o lean mass, no o al body weigh : he DR’s EXTRA s udy. Eu J P e Ca diol 2015; 22: 1171–1179. Saloka i e al. 207