Coronary artery disease
Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
►► Additional material is
published online only. To view
please visit the journal online
(http://d​ x.​doi.o​ rg/​10.​1136/​
heartjnl-2​ 019-​315844).
For numbered affiliations see
end of article.
Correspondence to
Professor Sanjay Rajagopalan,
Division of Cardiovascular
Medicine, Case Western Reserve
Medical School, Cleveland, OH
44106, USA;
​Sanjay.​Rajagopalan@​
UHhospitals.​org
Received 14 August 2019
Revised 13 December 2019
Accepted 15 December 2019
Published Online First
26 February 2020
►► http://​dx.​doi.​org/​10.​1136/​
heartjnl-2​ 019-​316283
© Author(s) (or their
employer(s)) 2020. No
commercial re-­use. See rights
and permissions. Published
by BMJ.
To cite: Sundaram V,
Rothnie K, Bloom C, et al.
Heart 2020;106:677–685.
Original research
Impact of comorbidities on peak troponin levels and
mortality in acute myocardial infarction
Varun Sundaram  ‍ ‍,1,2 Kieran Rothnie,1 Chloe Bloom,1 Rosita Zakeri  ‍ ‍,1,3
Jayakumar Sahadevan,4 Ajay Singh,5 Toshiyuki Nagai,6 James Potts,1
Jadwiga Wedzicha,1 Liam Smeeth,7 Daniel Simon,2 Adam Timmis  ‍ ‍,8
Sanjay Rajagopalan,2 Jennifer Kathleen Quint1
Abstract
Objectives  To characterise peak cardiac troponin
levels, in patients presenting with acute myocardial
infarction (AMI), according to their comorbid condition
and determine the influence of peak cardiac troponin
(cTn) levels on mortality.
Methods  We included patients with the first admission
for AMI in the UK. We used linear regression to estimate
the association between eight common comorbidities
(diabetes mellitus, previous angina, peripheral arterial
disease, previous myocardial infarction (MI), chronic
kidney disease (CKD), cerebrovascular disease, chronic
heart failure (CHF) and chronic obstructive pulmonary
disease (COPD)) and peak cTn. Peak cTn levels were
adjusted for age, sex, smoking status and comorbidities.
Logistic regression and restricted cubic spline models
were employed to investigate the association between
peak cTn and 180-­day mortality for each comorbidity.
Results  330 367 patients with ST elevation myocardial
infarction and non-­ST elevation myocardial infarction
were identified. Adjusted peak cTn levels were
significantly higher in patients with CKD (adjusted %
difference in peak cTnT for CKD=42%, 95% CI 13.1 to
78.4) and significantly lower for patients with COPD,
previous angina, previous MI and CHF when compared
with patients without the respective comorbidities
(reference group) (cTnI; COPD=−21.7%, 95% CI −29.1
to −13.4; previous angina=−24.2%, 95% CI −29.6 to
−8.3; previous MI=−13.5%, 95% CI −20.6 to −5.9;
CHF=−28%, 95% CI −37.2 to −17.6). Risk of 180-­day
mortality in most of the comorbidities did not change
substantially after adjusting for peak cTn. In general, cTnI
had a stronger association with mortality than cTnT.
Conclusions  In this nationwide analysis of patients
presenting with AMI, comorbidities substantially influenced
systemic concentrations of peak cTn. Comorbid illness is a
significant predictor of mortality regardless of peak cTn levels
and should be taken into consideration while interpreting
cTn both as a diagnostic and prognostic biomarker.
Introduction
The use of cardiac troponin (cTn) as a protein marker
of acute myocardial infarction (AMI) was first recom-
mended two decades ago and is now the cornerstone
for establishing a diagnosis of AMI.1 Following AMI,
systemic concentrations of cTn rise and their peak
levels strongly correlate with the infarct size of the
myocardium and thus has excellent prognostic value.2
Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844
Patients presenting with an AMI, particu-
larly non-­ ST elevation myocardial infarction
(NSTEMI), frequently have multiple comorbidities
which may affect systemic concentrations of both
cardiac troponin T (cTnT) and cardiac troponin
I (cTnI).3 4 Previous studies have suggested that
cTn levels are higher in patients with AMI and
chronic kidney disease (CKD) compared with those
without, though the magnitude of cTn increase
and the impact of severity of CKD on peak cTn
levels have not been systematically evaluated.5–7
Conversely, patients with prior coronary artery
disease (CAD), congestive heart failure (CHF) or
chronic obstructive pulmonary disease (COPD)
presenting with AMI reportedly have lower peak
cTn levels compared with patients without the
respective comorbidities, although these studies
did not adjust for age, smoking status, medications
and other relevant comorbidities; moreover, the
relevance of lower peak troponin levels in these
patients have not been studied.8–10 Indeed, the
influence of common comorbidities on peak cTn
levels in patients presenting with AMI has not been
systematically evaluated and the quantum of change
in cTn caused by such comorbidities needs further
investigation. This has important clinical implica-
tions, particularly in the context of NSTEMI, the
most common type of AMI, where implementation
of guideline-­recommended therapies in real world
rely heavily on the degree of cTn elevation,11 and
in older patients with multiple comorbidities where
the interpretation of cTn could be marred due to a
complex interplay of age, sex, comorbidities, type
of cTn assay and AMI (ie ST elevation myocardial
infarction (STEMI) vs NSTEMI).12 13
Previous studies have demonstrated a linear and
graded relationship of peak cTn and mortality in
patients presenting with AMI.14 However, this linear
relationship between peak cTn and mortality could
vary based on the baseline comorbidities and the type
of cTn assay (cTnT vs cTnI). To characterise peak cTn
levels in common comorbid conditions and determine
the impact of these varying cTn levels on mortality
in AMI, we carried out this large population-­based
study across the UK.
Methods
Study population
We included all patients with hospitalisation for
acute coronary syndrome (ACS) between January
  677
 Coronary artery disease
2003 and June 2013 from the Myocardial Infarction National
Audit (MINAP). MINAP was linked to the Office of National
Statistics, providing mortality data at 180 days post-­discharge.
Patients missing data on peak troponin and those with prior
coronary artery bypass graft surgery were excluded from the
analysis, as coronary bypass grafts can significantly influence
peak levels of cTn and infarct size.15
Identification of comorbidities
Comorbid conditions were pre-­ defined by MINAP (online
supplementary appendix table S1). According to common prac-
tice when analysing MINAP data, missing clinical variables were
assumed to be absent.
Troponin values
The peak cTn levels in MINAP were entered by the physician as
part of the audit. We primarily used cTnT and cTnI for this analysis.
Studies using the MINAP registry have elucidated the prognostic
significance of varying troponin levels in AMI using the troponin
assays (T and I), indicating acceptable standardisation of troponin
values.16 The European Society of Cardiology (ESC) Study Group
on Biomarkers in Cardiology recommended the routine use of
high sensitivity troponin as a diagnostic biomarker for AMI in
2012; hence in the UK, high sensitivity troponin was seldom
measured prior to 2012.17 Despite the paucity of data, we were
able to perform a subgroup analysis of patients with high sensitivity
troponin T (hsTnT) (grouping together STEMI and NSTEMI) to
investigate the trends in peak hsTnT stratified by comorbidities.
Statistical analysis
For the primary analysis (difference in peak cTn), we log-­
transformed peak cTn levels and used linear regression to
compare peak cTn values between patients with and without
specific comorbidities of interest, stratified by the type of MI
(included STEMI and NSTEMI). We exponentiated linear regres-
sion coefficients to obtain a ratio of geometric means, adjusting
for age, sex, smoking status and comorbidities. We performed
additional analysis stratified by severity of CKD, to ascertain the
impact of the degree of CKD on cTn elevation.
To evaluate the relationship between comorbidity, peak
cTn and mortality, we restricted the analysis to patients with
NSTEMI, as peak cTn levels have greater therapeutic impli-
cations in the management of this group as compared with a
STEMI where primary percutaneous coronary intervention
is the standard of care regardless of cTn value. Three sets of
analysis were performed for the outcome of 180-­day all-­cause
mortality: (1) To investigate whether any differences in peak cTn
associated with a specific comorbidity had an impact on the risk
of death, we used logistic regression models to investigate the
risk of death associated with the specific comorbidity—before
and after adjustment for peak cTn level. We used likelihood ratio
tests to test for statistical evidence (p<0.05) that peak cTn modi-
fied the association between each comorbidity and mortality at
180 days. (2) Peak cTn was further categorised into quartiles
for each specific comorbidity (ie, 180-­day mortality in patients
with CKD with <25th percentile of troponin elevation among
all patients with CKD, etc) and adjusted mortality was estimated
for each category. (3) We used restricted cubic splines with three
knots (0.8, 3.82 and 22.57 for cTnI and 0.09, 0.38 and 2.15 for
cTnT for the overall analysis in NSTEMI (online supplemen-
tary appendix figure S1)) to evaluate the non-­linear association
between peak cTn level and risk of death at 180 days for each of
the comorbidities in our study. The knots differed when spline
678
Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
analysis was performed for each comorbidity (online supplemen-
tary appendix table S2–S3). Troponin was used as a continuous
variable for the spline analyses.
Data for 180-­day all-­cause mortality were presented as adjusted
OR (adjusting for age, sex, smoking status, comorbidities: periph-
eral arterial disease (PAD), cerebrovascular disease (CVD), COPD,
CKD, CHF, hypertension (HTN), previous myocardial infarction
(MI), diabetes mellitus (DM), previous angina; location of infarct
(anterior vs non-­anterior), medications (antiplatelet therapy, ACE
inhibitor, angiotensin receptor blocker, beta-­blockers and statins)
and revascularisation and maximally adjusted OR (adjusting for
peak cTn for each comorbidity). We did not perform multiple
imputation as previous studies have shown that imputation anal-
yses on the MINAP registry did not significantly alter effect sizes,
suggesting that missingness in MINAP is at random.18 Addition-
ally, the level of missingness of data in MINAP did not affect the
mortality ratios.19
Patient and public involvement
Patients were not involved in the design and conduct of the study
Ethics approval
Ethical approval for this study was obtained from the MINAP
Academic Group (Reference number: 13-­ MNP-07) and from
the Observational/Interventions Research Ethics Committee of
London School of Hygiene and Tropical Medicine (LSHTM
reference number: 6468). The National Institute for Cardiovas-
cular Outcomes Research, including the MINAP database, has
support under section 251 of the NHS Act 2006 to use patient
information for medical research without consent (NIGB:
ECC1-06(d)/2011). The study was conducted in compliance
with the Declaration of Helsinki.
Results
In all, 657 376 patients with ACS were identified in MINAP
database between January 2003 and June 2013. Among the
patients admitted, 32% (128,960) of the patients presented with
STEMI, 49.9% (201,407) with non-­STEMI and 18.1% (72,664)
had unstable angina (figure 1 and table 1).
Peak troponin levels stratified by comorbidities following
NSTEMI
In patients with NSTEMI, those with COPD (−21.7%, 95% CI
−29.1 to −13.4%), previous angina (−24.2%, 95% CI −29.6 to
−18.3 %), previous MI (−13.5%, 95% CI −20.6 to −5.9%), CVD
(−16%, 95% CI −24.2 to −7.0%) and CHF (−28.0%, 95% CI
−37.2 to −17.6%) had significantly lower adjusted difference in
peak cTnI (figure 2A and table 2). There were no significant differ-
ences in adjusted peak cTnI levels in patients with DM, PAD and
CKD compared with those without the respective comorbidity.
Similarly, in patients whose cTnT was measured, a similar asso-
ciation was seen for COPD (−14.3%, CI −18.7 to −9.7%) and
CHF (−9.1%, 95% CI −30.4 to −18.7%), although patients with
CKD had a substantially higher adjusted difference in peak cTnT
(+42%, CI +13.1 to+78.4%) (table 2) (figure 2A).
Peak troponin levels stratified by comorbidities following
STEMI
In patients presenting with STEMI, after adjustment for age, sex,
smoking status, MI location (anterior or non-­anterior) and other
comorbidities, a similar pattern was observed. Peak cTnI was
lower for patients with previous angina (−20.4%, 95% CI −29.7
to −17.1%), previous MI (−20.4%, 95% CI −27.2 to −13.0 %)
Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844
 Coronary artery disease

Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
Figure 1  Numbers of Patients in the base cohort and study cohort. ACS, acute coronary syndrome; CABG, coronary artery bypass grafting surgery;
cTnI, cardiac troponin I; cTnT, cardiac troponin T; MINAP, myocardial infarction national audit project; NSTEMI, non-­ST elevation myocardial infarction;
STEMI, ST elevation myocardial infarction; UA, unstable angina; hsTnT, high sensitivity troponin T.

and CHF (−26.7%, 95% CI −40.0 to −10.6%) compared with
patients with STEMI without the respective comorbidity (figure 2B
and table 3). In the case of cTnT, peak levels were found to be lower
for previous angina (−13.2%, 95% CI −21.5 to −4.1 %), previous
MI (−18.6%, 95% CI −26.9 to −8.0 %), CHF (−24.8%, 95% CI
−40.9 to −4.2%) and with COPD (−14.7%, 95 CI −23.8 to −4.6
%). There were no significant differences in adjusted peak cTnI
and cTnT levels in patients with DM, PAD and CVD (table 3).
There was a trend towards lower peak cTnI in patients with COPD
and higher cTnT in patients with CKD but this was not statistically
significant, possibly due to small sample sizes within these groups
(figure 2B and table 3).
Peak hsTnT following AMI
The trends in adjusted peak hsTnT values in patients with COPD,
CKD, CHF, previous MI and previous angina were quite similar
to the groups with cTnT (figure 3). However, the trends in peak
values did not reach statistical significance in certain groups (CKD
and CHF), possibly due to small sample size (table 4).

Table 1  Baseline characteristics of patients presenting with acute

myocardial infarction in the UK Risk of 180-day all-cause mortality with each comorbidity
NSTEMI STEMI before and after adjustment of peak troponin
201 407 128 960 In patients with NSTEMI, almost all comorbidities were asso-
Characteristic N (%) N (%) ciated with increased risk of death at 180 days after adjustment
Age group for age, sex, smoking status and other comorbidities. After
 60 and younger 45 063 (22.37) 45 752 (35.47) adjusting for peak cTn level, ORs for risk of death for each
 61–70 41 431 (20.57) 31 195 (24.19) comorbidity at 180 days did not change substantially for all
 71–80 54 516 (27.07) 28 041 (21.74) the comorbidities (table 2). In patients with STEMI, a similar
 81 and older 60 011 (29.80) 20 742 (16.08) pattern was observed (table 3). Additional analysis performed
Mean age in years 71.8 66.2 by categorising peak cTn into quartiles for each comorbidity
Sex did not reveal a clear relationship between increasing quar-
 Male 122 744 (60.94) 87 440 (67.80) tiles of peak cTn and mortality in most of the comorbidities
 Female 78 277 (38.86) 38 290 (29.69) (in NSTEMI and STEMI), with the exception of DM (online
Smoking status supplementary appendix tables S5–S8). The joint association
 Never smoker 65 221 (32.38) 33 905 (26.29) between increased cTn and comorbidity on the risk of 180 day
 Ex-­smoker 70 055 (34.78) 34 009 (26.37) mortality following NSTEMI and STEMI are presented in
 Current smoker 44 032 (21.86) 45 387 (35.19) online supplementary appendix table S12–S13.
 Non-­smoker: previous history unknown 21 713 (10.78) 12 429 (9.60)
Diabetes mellitus 40 693 (20.20) 16 128 (12.51) Relationship between peak troponin and mortality stratified
Hypertension 101 052 (50.17) 50 835 (39.41) by comorbidities in patients with NSTEMI; results of spline
Hyperlipidaemia 61 669 (30.09) 33 512 (25.99) analysis
Previous angina 56 016 (27.81) 15 247 (11.82) In patients with NSTEMI, the results of spline regression indicated
Previous myocardial infarction 38 473 (19.10) 11 549 (8.95) that mortality increased with rising peak cTnI up to at least 50 ng/
COPD 21 678 (10.76) 9883 (7.66) mL (online supplementary appendix figure S1). However, there
Chronic kidney disease 12 095 (6.01) 2890 (2.24) appeared to be a ceiling effect on the association between peak
Chronic heart failure 12 871 (6.39) 2171 (1.68) cTnT and mortality following NSTEMI after around 1.5 ng/mL,
Peripheral arterial disease 9321 (4.63) 3287 (2.55) after which rising peak cTn was not associated with increased
Cerebrovascular disease 19 313 (9.59) 6776 (5.25) mortality (online supplementary appendix figure S2). Splitting this
COPD, chronic obstructive pulmonary disease; NSTEMI, non-­ST elevation myocardial analysis by comorbidity produced associations with broadly similar
infarction; STEMI, ST elevation myocardial infarction; UA, unstable angina. shapes. There was a clear positive association between peak cTnI
Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844 679
 Coronary artery disease

Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
Figure 2  Adjusted percentage difference in peak troponin I and T in ACS (NSTEMI and STEMI) patients with specific comorbidities compared to
those without the respective comorbidities. The model was adjusted for adjusted for age, sex, smoking status and comorbidities for NSTEMI. The
model was adjusted for adjusted for age, sex, smoking status and co-­morbidities and location of infarct for STEMI. *Statistically significant (defined as
CIs not including zero). ACS, acute coronary syndrome; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; CHF, chronic heart
failure; h/o angina, history of angina; NSTEMI, non-­ST elevation myocardial infarction; Prev MI, previous myocardial infarction; STEMI, ST elevation
myocardial infarction.

and 180-­day mortality for most of the comorbidities (figure 4). No
such relationship was observed for cTnT (figure 5).
Subgroup analysis in CKD
We included 277 071 patients with serum creatinine values
for this analysis (online supplementary appendix figure S3).
glomerular filtration rate (GFR) was calculated using CKD-­
epi cohort equation. In patients with AMI (both STEMI and
NSTEMI), adjusted percentage difference in peak troponin
was significantly higher in CKD stages 3b, 4 and 5 (Stage
3b: 8.6%, 95% CI: 0.8% to 16.9%; Stage 4: 27.2%, 95% CI:
15.4% to 40.2%; Stage 5: 30.8%, 95% CI: 9% to 53.2%) when
compared with patients with CKD stage 1. The adjusted OR for
mortality increased with increasing severity of CKD (figure 6A
and B). There was no evidence of interaction between CKD

Table 2  Influence of comorbidities on peak troponin and 180-­day mortality (minimally and maximally adjusted—adjusting for peak troponins) in
patients with NSTEMI
Peak troponin with Peak troponin without Adjusted % difference N (%) died 180 days OR (95% CI) mortality at OR (95% CI) mortality at
the comorbidity, ng/ the comorbidity, ng/mL in peak troponin in the comorbidity 180 days before troponin 180 days adjusted for
mL (median, IQR) (median, IQR) (95% CI)* group adjustment† peak troponin level‡
Non-­STEMI troponin I
 COPD 1.92 (0.41–7.77) 2.30 (0.48–9.46) −21.7 (−29.1 to −13.4) 3124 (18.3) 1.48 (1.25 to 1.77) 1.51 (1.27 to 1.80)
 Peripheral vascular disease 2.68 (0.59–10.11) 2.25 (0.47–9.21) 5.8 (−8.5 to 22.2) 1674 (23.1) 1.31 (1.03 to 1.65) 1.29 (1.02 to 1.63)
 Cerebrovascular disease 2.20 (0.45–9.33) 2.27 (0.48–9.26) −16 (−24.2 to −7.0) 3850 (25.4) 1.50 (1.29 to 1.75) 1.54 (1.32 to 1.80)
 Previous MI 1.99 (0.39–8.20) 2.32 (0.50–9.55) −13.5 (−20.6 to −5.9) 6043 (18.8) 1.11 (0.96 to 1.28) 1.11 (0.96 to 1.28)
 Previous angina 1.86 (0.37–7.99) 2.42 (0.51–9.78) −24.2 (−29.6 to −18.3) 7898 (17.3) 1.19 (1.08 to 1.39) 1.22 (1.08 to 1.39)
 Chronic kidney disease 2.4 (0.51–10.07) 2.25 (0.47–9.21) 4.4 (−8.0 to 18.6) 3117 (32.1) 1.48 (1.24 to 1.78) 1.48 (1.24 to 1.78)
 Chronic heart failure 1.95 (0.38–8.50) 2.29 (0.48–9.32) −28.0 (−37.2 to −17.6) 3450 (35.0) 1.83 (1.52 to 2.19) 1.87 (1.56 to 2.25)
 Hypertension 2.14 (0.44–8.81) 2.40 (0.50–9.75) −8.0 (−13.3 to −2.35) 9516 (15.9) 0.90 (0.81 to 1.01) 0.89 (0.79 to 0.99)
 Diabetes mellitus 2.20 (0.45–9.05) 2.29 (0.48–9.32) −0.3 (−7.7 to 7.6) 6028 (17.5) 1.22 (1.07 to 1.40) 1.23 (1.08 to 1.40)
Non−STEMI Troponin T
 COPD 0.34 (0.13–1.00) 0.41 (0.15–1.28) −22.7 (−35.8 to −7.0) 2237 (19.6) 1.34 (1.02 to 1.77) 1.43 (1.08 to 1.90)
 Peripheral vascular disease 0.40 (0.15–1.31) 0.40 (0.15–1.23) −16.7 (−8.9 to 58.3) 1190 (24.3) 1.07 (0.74 to 1.56) 1.06 (0.73 to 1.55)
 Cerebrovascular disease 0.38 (0.14–1.22) 0.40 (0.15–1.24) 15.1 (5.5 to 39.8) 2671 (27.0) 1.06 (0.83 to 1.37) 1.07 (0.84 to 1.39)
 Previous MI 0.36 (0.14–1.15) 0.41 (0.15–1.26) 0.4 (−14 to 17.2) 4207 (20.8) 1.12 (0.90 to 1.39) 1.13 (0.91 to 1.40)
 Previous angina 0.35 (0.13–1.13) 0.42 (0.16–1.29) −0.9 (−13.5 to 13.5) 5809 (18.7) 1.09 (0.90 to 1.32) 1.10 (0.91 to 1.34)
 Chronic kidney disease 0.43 (0.15–1.65) 0.40 (0.15–1.21) 42 (13.1 to 78.4) 2013 (33.5) 1.48 (1.13 to 1.95) 1.44 (1.09 to 1.90)
 Chronic heart failure 0.33 (0.12–1.26) 0.41 (0.12–1.37) −9.1 (−30.4 to −18.7) 2369 (35.6) 1.70 (1.26 to 2.29) 1.72 (1.27 to 2.32)
 Hypertension 0.39 (0.14–1.24) 0.42 (0.16–1.24) 3.1 (−7.6 to 15.0) 7017 (17.7) 0.94 (0.80 to 1.13) 0.94 (0.79 to 1.12)
 Diabetes mellitus 0.38 (0.14–1.25) 0.41 (0.15–1.24) 6.2 (−7.7. to 21.5) 4028 (19.1) 1.23 (1.00 to 1.51) 1.22 (0.99 to 1.51)
*Adjusted for age, sex, smoking status, comorbidities and location of infarct.
†Adjusted for age, sex, smoking status, comorbidities, discharge medications (antiplatelet therapy, statins, beta-­blockers, ACE inhibitors or angiotensin receptor blockers) and revascularisation.
‡Adjusted for age, sex, smoking status, comorbidities, discharge medications (antiplatelet therapy, statins, beta-­blockers, ACE inhibitors or angiotensin receptor blockers), revascularisation and peak troponin
level.
COPD, chronic obstructive pulmonary disease; MI, myocardial infarction; NSTEMI, non-­ST elevation myocardial infarction.

680 Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844

 Coronary artery disease

Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
Table 3  Influence of comorbidities on peak troponin and 180-­day mortality (minimally and maximally adjusted—adjusting for peak troponins) in
patients with STEMI
Peak troponin with the Peak troponin without Adjusted % difference N (%) died 180 days OR (95% CI) mortality OR (95% CI) mortality at
comorbidity, ng/mL the comorbidity, ng/mL in peak troponin in the comorbidity at 180 days before 180 days adjusted for
(median, IQR) (median, IQR) (95% CI)* group troponin adjustment† peak troponin level‡
STEMI troponin I
 COPD 22.53 (4.72–50.00) 27.80 (5.73–50.00) −8.2 (−15.9 to 0.27) 921 (15.4) 1.49 (1.27 to 1.74) 1.52 (1.29 to 1.79)
 Peripheral vascular disease 26.00 (4.75–50.00) 27.30 (5.67–50.00) −9.1 (−21.5 to 5.3) 433 (21.3) 1.26 (1.00 to 1.60) 1.23 (0.95 to 1.58)
 Cerebrovascular disease 24.49 (4.70–50.00) 27.42 (5.70–50.00) −4.1 (−13.8 to 6.6) 1139 (28.3) 1.62 (1.39 to 1.88) 1.60 (1.37 to 1.87)
 Previous MI 18.00 (2.93–50.00) 28.40 (6.03–50.00) −20.4 (−27.2 to −13.0) 1339 (18.7) 1.24 (1.09 to 1.41) 1.22 (1.05 to 1.43)
 Previous angina (CAD) 19.97 (3.64–50.00) 28.36 (6.00–50.00) −20.4 (−29.7 to −17.1) 1699 (18.7) 1.08 (0.95 to 1.24) 1.03 (0.95 to 1.58)
 Chronic kidney disease 22.00 (3.50–50.00) 27.40 (5.70–50.00) 0.7 (−14.1 to 18) 631 (35.6) 1.85 (1.50 to 2.28) 1.79 (1.43 to 2.23)
 Chronic heart failure 15.95 (2.22–50.00) 27.50 (5.71–50.00) −26.7 (−40.0 to −10.6) 530 (41.4) 2.45 (1.92 to 3.14) 2.42 (1.87 to 3.13)
 Hypertension 25.96 (5.18–50.00) 28.38 (6.00–50.00) 0.1 (−4.7 to 5.1) 4294 (13.6) 0.92 (0.82 to 1.03) 0.92 (0.82 to 1.04)
 Diabetes mellitus 23.61 (4.55–50.00) 27.85 (5.82–50.00) 10.6 (3.1 to 18.7) 1734 (17.4) 1.49 (1.32 to 1.69) 1.46 (1.29 to 1.67)
STEMI troponin T
 COPD 2.07 (0.56–6.15) 2.46 (0.74–6.95) −14.7 (−23.8 to −4.6) 635 (14.5) 1.24 (1.02 to 1.52) 1.28 (1.05 to 1.58)
 Peripheral vascular disease 2.23 (0.58–6.75) 2.44 (0.73–6.90) 4.5 (−20.5 to 16.3) 321 (22.4) 1.17 (0.87 to 1.59) 1.20 (0.88 to 1.64)
 Cerebrovascular disease 2.50 (0.72–7.50) 2.42 (0.72–6.87) 15.2 (0.9 to 31.7) 940 (30.6) 1.66 (1.39 to 1.99) 1.67 (1.39 to 2.01)
 Previous MI 1.84 (0.48–5.87) 2.50 (0.76–7.00) −18.6 (−26.9 to −8.0) 1055 (20.5) 1.42 (1.18 to 1.70) 1.46 (1.21 to 1.77)
 Previous angina 1.95 (0.50–6.21) 2.50 (0.76–6.98) −13.2 (−21.5 to −4.1) 1465 (20.4) 1.12 (0.95 to 1.31) 1.13 (0.96 to 1.33)
 Chronic kidney disease 2.50 (0.65–9.09) 2.42 (0.72–6.85) 10 (−9.8 to 34.2) 451 (36.0) 1.66 (1.27 to 2.15) 1.67 (1.27 to 2.18)
 Chronic heart failure 2.11 (0.48–6.43) 2.43 (0.73–6.90) −24.8 (−40.9 to −4.2) 404 (38.9) 1.45 (1.07 to 1.95) 1.32 (0.97 to 1.81)
 Hypertension 2.39 (0.69–7.12) 2.46 (0.74–6.75) 8.2 (1.6 to 15.1) 3300 (14.9)
 Diabetes mellitus 2.29 (0.63–7.16) 2.45 (0.73–6.85) 5.7 (3.7 to 15.9) 1284 (18.3) 1.42 (1.22 to 1.66) 1.43 (1.22 to 1.67)
*Adjusted for age, sex, smoking status, comorbidities and location of infarct.
†Adjusted for age, sex, smoking status, comorbidities, location of infarct, discharge medications (antiplatelet therapy, statins, beta-­blockers, ACE inhibitors or angiotensin receptor blockers) and
revascularisation.
‡Adjusted for age, sex, smoking status, comorbidities, location of infarct, discharge medications (antiplatelet therapy, statins, beta-­blockers, ACE inhibitors or angiotensin receptor blockers), revascularisation
and peak troponin level
CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; MI, myocardial Infarction; STEMI, ST elevation myocardial Infarction.

severity and peak troponin (p>0.05) for the outcome of 180-­
day mortality.
Discussion
This nationwide study of patients presenting with AMI revealed
that comorbidities substantially influenced systemic concentra-
tions of cTn; however, the presence of comorbidity alone, irre-
spective of peak cTn, was the major determinant of mortality.
Major findings
The major findings of our study could be summarised as follows:
(1) In patients presenting with AMI and COPD, previous angina,
previous MI or CHF, peak cTn levels were significantly lower
compared with patients without the respective comorbidity, with
similar trends regardless of cTn assays (ie, peak cTnT and cTnI)
or AMI presentation (STEMI and NSTEMI). Results for hsTnT,
although limited by the small sample size, revealed similar

Figure 3  Adjusted percentage difference in peak troponin (I, T and high sensitivity troponin T*) in NSTEMI patients with specific comorbidities
compared to those without the respective comorbidities. For hsTnT, we included patients with both STEMI and NSTEMI due to a smaller sample size.
The model was adjusted for adjusted for age, sex, smoking status and comorbidities for NSTEMI. The model was adjusted for adjusted for age, sex,
smoking status and co-­and location of infarct for STEMI. CHF, chronic heart failure; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary
disease; h/o angina, history of angina; Prev MI, previous myocardial infarction; NSTEMI, non-­ST elevation myocardial infarction; STEMI, ST elevation
myocardial infarction.
Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844 681
 Coronary artery disease

Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
Table 4  Influence of comorbidities on peak high sensitivity troponin T in patients presenting with ACS (STEMI and NSTEMI together)
Peak troponin with the comorbidity, Peak troponin without the comorbidity, Adjusted % difference in peak troponin
Comorbidities ng/mL (median, IQR) ng/mL (median, IQR) (95% CI)*
COPD 244 (43–1021) 409 (69–1955) −26.5 (−45.5 to −0.9)
Peripheral arterial disease 295 (43–1418) 391 (66–1854) −13.8 (−27.7 to 86.2)
Cerebrovascular disease 247 (46–1360) 400 (68–1879) −1.2 (−30.4 to 40.3)
Previous angina 159 (32–847) 441 (75–2026) −35.3 (−50.0 to −16.3)
Previous MI 258 (66–1093) 396 (65–1865) −30.9 (−49.3 to −5.9)
Chronic kidney disease 147 (30–742) 502 (88–2172) 51.3 (−8.3 to 149.5)
Chronic heart failure 166 (53–681) 403 (66–1888) c15.3 (−55.9 to 62.6)
Hypertension 159 (32–847) 441 (75–2026) −65.1 (−24.4 to 98.6)
Diabetes mellitus 241 (37–1170) 438 (72–2027) −25.5 (−41.8 to −4.6)
Previous angina: patients with previous angina were categorised as stable coronary artery disease.
*Adjusted for age, sex, smoking status, comorbidities and location of infarct.
ACS, acute coronary syndrome; COPD, chronic obstructive pulmonary disease; MI, myocardial infarction; NSTEMI, non-­ST elevation myocardial infarction; STEMI, ST elevation
myocardial infarction.;

trends. (2) Among the comorbidities analysed, the quantum of
change in cTnT was highest for patients with CKD (42% higher)
as compared with those without CKD. (3) There was a signif-
icant impact of CKD severity (GFR <45 mL/min/1.73 m2) on
the degree of cTn elevation. (4) All comorbidities were associ-
ated with an increased risk of mortality in AMI, which was not
altered after adjusting for peak cTn with the exception of DM.
(5) CTnI had a stronger association with mortality than cTnT for
most of the comorbidities analysed.
Chronic kidney disease
The interpretation of cTn in the setting of CKD can be chal-
lenging, as patients with CKD can have elevated cTn levels even
in the absence of ischaemia.5 The AMI cut-­off based on a healthy
general population has been shown to be significantly lower
than the receiver operating curve optimal cut-­off for patients
with CKD.20 Twerenbold et al showed that while the ESC 0/1-­
hour algorithm (using hsTnT and hsTnI) had high sensitivity in
patients with CKD, the specificity of rule-­in and the overall effi-
cacy of the ESC algorithm was substantially reduced.6
Multiple mechanisms have been postulated for the elevation
of cTn and contrary to common belief, decreased renal clear-
ance seems to be an implausible explanation, as cTns are high
molecular weight compounds (37 kDa), unlikely to be cleared
by the kidneys.6 20 21 Some of the other possible explanations
are the presence of, increased left ventricular (LV) mass, LV
systolic dysfunction or subclinical CAD, resulting in increased
membrane permeability and cTn leaks.5 6 To our knowledge,
our study is the first to report the quantum of cTn elevation
in patients with CKD, impact of severity of CKD on cTn
elevation and its prognostic relevance after accounting for
important confounders including HF and type of MI.
Chronic heart failure
Our analysis demonstrated that while patients with HF had
substantially lower adjusted peak cTn levels than those without
HF, the presence of HF was associated with a higher mortality.
This was corroborated with the spline analysis, which revealed
a ceiling effect in prognostic value for both cTnT and cTnI in
patients with CHF (figures 4C and 5C). The lower peak cTn

Figure 4  (A –H) Association between peak cardiac troponin I and 180-­day mortality in NSTEMI stratified by comorbidities (Spline Analysis*). *We
used restricted cubic splines with three knots. The model was adjusted for adjusted for age, sex, smoking status, comorbidities, antiplatelet therapy,
beta-­blockers, statins, ACE inhibitors and revascularisation.NSTEMI, non-­ST elevation myocardial infarction.
682 Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844

Figure 5  (A –H) Association between peak cardiac troponin T and 180-­day mortality in NSTEMI stratified by comorbidities (Spline Analysis*). *We
used restricted cubic splines with three knots. The model was adjusted for adjusted for age, sex, smoking status, comorbidities, antiplatelet therapy,
beta-­blockers, statins, ACE inhibitors and revascularisation.NSTEMI, non-­ST elevation myocardial infarction.
and high mortality in this group could be due to a multitude of
reasons including the possibility that patient’s with ischaemic
cardiomyopathy may have reduced viable myocardium. Poor
myocardial reserve in this group of patients might therefore
help explain high mortality with smaller infarct size based on
lower peak cTn. Endogenous protective mechanisms following
ischaemia/reperfusion injury could be attenuated in patients
with ventricular remodelling or clinical HF.22 Morphological
and biological alterations in HF can impact the signal trans-
duction cascade of pre-­ conditioning and post-­conditioning
cardio protection, resulting in increased mortality despite
smaller initial infarct size.22
COPD, previous MI and previous angina
In patients with COPD and heightened cardiovascular risk,
plasma cTnI concentrations have been shown to be a major
predictor of future cardiovascular events and cardiovascular
death23; however, the diagnostic and prognostic accuracy of
cTn in patients with COPD with AMI have not been estab-
lished. We observed lower peak cTn levels in patients with
COPD, angina and previous MI after adjusting for important
confounders including smoking, age, CHF, HTN, DM, area of
infarct and renal impairment. The lower peak cTn level among
these patients may have arisen due to myriad reasons. Patients
with prior CAD are more likely have been prescribed drugs for
secondary prevention which could have influenced the infarct
size. The phenomenon of ischaemic pre-­ conditioning may
be an innate protective physiological mechanism for lower
cTn peak among COPD, previous MI and angina comorbidi-
ties.24–26 The chronic hypoxic state in patients with COPD may
pre-­condition the myocardium to become more resistant to
future infarcts.27 28 This may explain why patients with COPD
do not have an equivalent high peak cTn level as compared
with patients without COPD. Ischaemic pre-­conditioning still
remains a topic for considerable future research on its role in
MI among patients with COPD and its potential use in clinical
practice.
Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844
Coronary artery disease
Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
Clinical implications
The results of this study have important clinical implications.
It is well established that in patients with NSTEMI, imple-
mentation of guideline-­recommended treatment in real world
depends on the extent of cTn elevation that is, patients with
intermediate to major elevations are more likely to receive
guideline directed therapy as compared with those with minor
cTn elevations.11 In contrast to generally held assumptions,
our findings suggest that, patients with AMI with specific
comorbidities and lower peak cTn levels may still have a poor
prognosis. Clinicians should refrain from being reassured
by lower peak cTn levels in patients with AMI (particularly
NSTEMI) and concomitant HF, COPD, angina or a previous
MI. Conversely, a higher peak cTnT level may be less infor-
mative in the setting of CKD (Stage 3b and higher) where up
to 40% of the elevation in cTnT may relate to the presence
of CKD alone, irrespective of AMI type or other relevant
confounders.
Limitations
The diagnosis of ACS in the MINAP registry was made by the
physician treating the patient using standard investigations,
including clinical history, ECG and troponin. In 1998, the UK
launched the MINAP registry and required all hospital trusts
to report MIs. In a study performed by Herrett et al, only
50% of MIs reported in hospital episode statistics (nationwide
in-­hospital records in the UK) and Clinical Practice Research
Registry (CPRD; primary care records in the UK) were found
to be included in the MINAP registry. Moreover, patients with
AMI recorded in CPRD had about half the hazard of mortality
(at 30 days) of patients with AMI recorded in the MINAP, indi-
cating differences in case ascertainment between the registry,
nationwide hospital admissions and primary care databases.
Some patients with the AMI during the study period would be
possibly reclassified as acute myocardial injury alone since the
fourth universal definition of MI. We were not able to deter-
mine the impact of different combinations of comorbidities
(multi-­morbidity) on peak cTn using the data available, and
683
 Coronary artery disease
Figure 6  (A) Relationship between CKD severity and 180-­day
mortality; (B) adjusted % difference in troponin stratified by stages
of CKD. *A: statistically significant (defined as 95% CIs not including
one). *B: statistically significant (defined as 95% CIs not including
zero). (A) Patients with stage 1 CKD (GFR>90 mL/min/1.73 m2) was
used as the reference group; the model was adjusted for adjusted for
age, sex, smoking status, comorbidities, location of infarct, antiplatelet
therapy, beta-­blockers, statins, ACE inhibitors, revascularisation and
peak troponin. Interaction testing was performed between CKD stage
(categorical variable) and peak troponin (continuous variable) for the
outcome of 180-­day mortality and was negative (p>0.05). (B) Patients
with stage 1 CKD (GFR>90 mL/min/1.73 m2) was used as the reference
group. The model was adjusted for adjusted for age, sex, smoking status,
comorbidities and location of infarct. CKD, chronic kidney disease; GFR,
glomerular filtration rate.
further studies on comorbidity-­to-­comorbidity interactions are
needed. The MINAP database did not have all the informa-
tion we would have ideally included in the analysis such as
results of structural cardiac imaging and timing of revascu-
larisation. We were not able to conduct a complete analysis
with the hsTnT assay alone due to small sample size. However,
our analysis of hsTnT revealed that the magnitude and direc-
tionality of change in important comorbidities such as CKD,
COPD and CHF were similar to those of less sensitive cTnT
(eg, 40% to 50% increase in adjusted peak values of hsTnT
and regular cTnT in CKD, ~25% decrease in both hsTnT
and cTnT in COPD, ~10%–15% decrease in CHF). This is
because the ratio of peak troponin for each comorbidity to
those without the comorbidity is likely to be the same for both
highly sensitive and less sensitive troponins.
Conclusion
This large nationwide analysis of patients presenting with AMI
in the UK suggests that comorbidities significantly affect peak
cTn. Comorbidities should be taken in to consideration while
684
Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
Key questions
What is already known on this topic?
►► Previous studies have suggested that peak cardiac troponin
levels are higher in patients with acute myocardial infarction
(AMI) and chronic kidney disease (CKD) compared to those
without CKD. However, the magnitude of troponin increase
has not been evaluated.
►► Patients with prior coronary artery disease (CAD), congestive
heart failure or chronic obstructive pulmonary disease
(COPD) presenting with AMI reportedly have lower peak
troponin levels compared to patients without the respective
comorbidities; however, these studies had not adjusted for
age, sex and other relevant comorbidities.
What this study adds?
►► This nationwide study of patients presenting with AMI
revealed that comorbidities substantially influenced systemic
concentrations of troponin; however, the presence of
comorbidity alone, irrespective of peak troponin, was the
major determinant of mortality.
How might this impact on clinical practice?
►► Clinicians should refrain from being reassured by lower
peak troponin levels in patients with non-­ST elevation
myocardial infarction and concomitant HF, COPD, or prior
CAD. Conversely, a higher peak troponin level may be less
informative in the setting of CKD where up to 40% of the
elevation may relate to the presence of CKD alone.
interpreting cTn in the setting of AMI, as the prognostication
varies based on patient’s pre-­existing comorbid illness.
Author affiliations
1
National Heart and Lung Institute, Imperial College London, London, UK
2
Cardiovascular Medicine, University Hospitals Harrington Heart and Vascular
Institute,Case Western Reserve University, Cleveland, United States
3
Department of Cardiology, Royal Brompton and Harefield NHS Foundation Trust,
London, UK
4
Louis Stokes Veteran Affairs Medical Center, Cleveland, United States
5
Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts,
USA
6
National Heart and Lung Institute, Imperial College London, London, UK
7
Department of Epidemiology and Population Health, London School of Hygiene and
Tropical Medicine, London, UK
8
NIHR Cardiovascular Biomedical Research Unit, Bart’s Heart Centre, london, UK
Twitter Rosita Zakeri @RositaZakeri
Contributors  VS: study design, analyses, conducting and manuscript preparation.
KR: study design, analyses, CB: analyses and manuscript preparation. RZ: manuscript
preparation. JS: manuscript preparation. AS: manuscript preparation. TN: planning.
JP: analyses. JW: manuscript preparation. LS: design and manuscript preparation. DS:
design and manuscript preparation. AT: design and manuscript preparation. SR: study
design, conducting and manuscript preparation. JKQ: study design, conducting and
manuscript preparation.
Funding  This work was funded by an MRC Population Health Scientist Fellowship
held by JKQ[G0902135].
Competing interests  None declared.
Patient consent for publication  Not required.
Provenance and peer review  Not commissioned; externally peer reviewed.
Data availability statement  Data are available upon reasonable request.
We have used a deidentified participant data which is available on request after
obtaining specific permissions from the MINAP committee. Additional information
on statistical analyses plan and codes (statistical codes and diagnoses codes) are
available on request (email address: ​v.​sundaram@​imperial.​ac.​uk)
Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844

ORCID iDs
Varun Sundaram http://​orcid.​org/0​ 000-​0001-​6543-​8482
Rosita Zakeri http://​orcid.​org/​0000-​0002-4​ 225-​3693
Adam Timmis http://o​ rcid.​org/0​ 000-​0003-1​ 419-​112X
References
1 Braunwald E, Morrow DA. Unstable angina: is it time for a requiem? Circulation
2013;127:2452–7.
2 Hassan AKM, Bergheanu SC, Hasan-­Ali H, et al. Usefulness of peak troponin-­T to
predict infarct size and long-­term outcome in patients with first acute myocardial
infarction after primary percutaneous coronary intervention. Am J Cardiol
2009;103:779–84.
3 Chang W-­C, Boersma E, Granger CB, et al. Dynamic prognostication in non-­ST-­
elevation acute coronary syndromes: insights from GUSTO-­IIB and pursuit. Am Heart J
2004;148:62–71.
4 Hall M, Dondo TB, Yan AT, et al. Association of clinical factors and therapeutic
strategies with improvements in survival following non-­ST-­elevation myocardial
infarction, 2003-2013. JAMA 2016;316:1073–82.
5 deFilippi CR, Herzog CA. Interpreting cardiac biomarkers in the setting of chronic
kidney disease. Clin Chem 2017;63:59–65.
6 Twerenbold R, Badertscher P, Boeddinghaus J, et al. 0/1-­Hour triage algorithm
for myocardial infarction in patients with renal dysfunction. Circulation
2018;137:436–51.
7 deFilippi C, Wasserman S, Rosanio S, et al. Cardiac troponin T and C-­reactive protein
for predicting prognosis, coronary atherosclerosis, and cardiomyopathy in patients
undergoing long-­term hemodialysis. JAMA 2003;290:353–9.
8 Rothnie KJ, Quint JK. Chronic obstructive pulmonary disease and acute myocardial
infarction: effects on presentation, management, and outcomes. Eur Heart J Qual Care
Clin Outcomes 2016;2:81–90.
9 Enriquez JR, de Lemos JA, Parikh SV, et al. Association of chronic lung disease with
treatments and outcomes patients with acute myocardial infarction. Am Heart J
2013;165:43–9.
10 Bhatt HA, Sanghani DR, Lee D, et al. Predictors of peak troponin level in
acute coronary syndromes: prior aspirin use and SYNTAX score. Int J Angiol
2016;25:54–63.
11 Roe MT, Peterson ED, Li Y, et al. Relationship between risk stratification by cardiac
troponin level and adherence to guidelines for non-­ST-­segment elevation acute
coronary syndromes. Arch Intern Med 2005;165:1870–6.
12 Boeddinghaus J, Nestelberger T, Twerenbold R, et al. Impact of age on the
performance of the ESC 0/1h-­algorithms for early diagnosis of myocardial infarction.
Eur Heart J 2018;39:3780–94.
Sundaram V, et al. Heart 2020;106:677–685. doi:10.1136/heartjnl-2019-315844
Coronary artery disease
Heart: first published as 10.1136/heartjnl-2019-315844 on 26 February 2020. Downloaded from http://heart.bmj.com/ on April 17, 2020 at Bham/Bius/Pharma. Protected by copyright.
13 Shah ASV, Griffiths M, Lee KK, et al. High sensitivity cardiac troponin and the
under-­diagnosis of myocardial infarction in women: prospective cohort study. BMJ
2015;350:g7873.
14 Goldstein SA, Newby LK, Cyr DD, et al. Relationship between peak troponin values
and long-­term ischemic events among medically managed patients with acute
coronary syndromes. J Am Heart Assoc 2017;6. doi:10.1161/JAHA.116.005334.
[Epub ahead of print: 11 Apr 2017].
15 Sergeant PT, Blackstone EH, Meyns BP. Does arterial revascularization decrease
the risk of infarction after coronary artery bypass grafting? Ann Thorac Surg
1998;66:1–11. discussion 10-11.
16 Myint PK, Kwok CS, Bachmann MO, et al. Prognostic value of troponins in acute
coronary syndrome depends upon patient age. Heart 2014;100:1583–90.
17 Thygesen K, Mair J, Giannitsis E, et al. How to use high-­sensitivity cardiac troponins in
acute cardiac care. Eur Heart J 2012;33:2252–7.
18 Zaman MJ, Stirling S, Shepstone L, et al. The association between older age and
receipt of care and outcomes in patients with acute coronary syndromes: a cohort
study of the myocardial ischaemia national audit project (MINAP). Eur Heart J
2014;35:1551–8.
19 Zaman MJS, Philipson P, Chen R, et al. South Asians and coronary disease: is there
discordance between effects on incidence and prognosis? Heart 2013;99:729–36.
20 Twerenbold R, Wildi K, Jaeger C, et al. Optimal cutoff levels of more sensitive cardiac
troponin assays for the early diagnosis of myocardial infarction in patients with renal
dysfunction. Circulation 2015;131:2041–50.
21 Freda BJ, Tang WHW, Van Lente F, et al. Cardiac troponins in renal insufficiency: review
and clinical implications. J Am Coll Cardiol 2002;40:2065–71.
22 Seeger JPH, Benda NMM, Riksen NP, et al. Heart failure is associated with
exaggerated endothelial ischaemia-­reperfusion injury and attenuated effect of
ischaemic preconditioning. Eur J Prev Cardiol 2016;23:33–40.
23 Adamson PD, Anderson JA, Brook RD, et al. Cardiac troponin i and cardiovascular
risk in patients with chronic obstructive pulmonary disease. J Am Coll Cardiol
2018;72:1126–37.
24 Ferdinandy P, Hausenloy DJ, Heusch G, et al. Interaction of risk factors, comorbidities,
and comedications with ischemia/reperfusion injury and cardioprotection by
preconditioning, postconditioning, and remote conditioning. Pharmacol Rev
2014;66:1142–74.
25 Wiseman A, Waters DD, Walling A, et al. Long-­term prognosis after myocardial
infarction in patients with previous coronary artery bypass surgery. J Am Coll Cardiol
1988;12:873–80.
26 Tomai F, Crea F, Chiariello L, et al. Ischemic preconditioning in humans: models,
mediators, and clinical relevance. Circulation 1999;100:559–63.
27 Barnes PJ. Chronic obstructive pulmonary disease. N Engl J Med 2000;343:269–80.
28 Stone IS, Petersen SE, Barnes NC. Raised troponin in COPD: clinical implications and
possible mechanisms. Heart 2013;99:71–2.
685