Skip main navigation

Troponin I as a Predictor of Coronary Heart Disease and Mortality in 70-Year-Old Men

A Community-Based Cohort Study
Originally publishedhttps://doi.org/10.1161/CIRCULATIONAHA.105.570762Circulation. 2006;113:1071–1078

Abstract

Background— Cardiac troponin I (cTnI), a standard for detection of myocardial damage, has recently been reported to predict acute myocardial infarction or death in patients with unstable coronary heart disease (CHD). Cardiac TnI concentrations increase with age in subjects free from clinical signs of CHD, suggesting silent myocardial damage. We investigated the association between cTnI and future CHD and mortality in a community-based cohort of men.

Methods and Results— A community-based study was conducted from August 1991 to May 1995 among 1203 men in Uppsala, Sweden, aged 70 years at baseline with a follow-up of up to 10.4 years with the use of registry data (National Board of Health and Welfare, Sweden). CHD was defined with the use of data taken from the Cause of Death Registry or from first-time hospitalization for CHD as recorded in the Hospital Discharge Registry. Cardiac TnI concentrations were measured blinded for outcome, in frozen baseline plasma samples, with the use of the AccuTnI from Beckman Coulter, Inc. Hazard ratios (HRs) from Cox proportional hazards are presented with 95% confidence intervals (CIs) for a 1-SD increase. In men free from cardiovascular disease (CVD), cTnI predicted death (HR, 1.26; 95% CI, 1.08 to 1.46; P=0.003) or first CHD event (HR, 1.31; 95% CI, 1.11 to 1.54; P=0.001) after adjustment for conventional risk factors: total and HDL cholesterol, plasma glucose, body mass index, smoking, and systolic blood pressure.

Conclusions— In this first longitudinal report, cTnI was shown to predict death and first CHD event in men free from CVD at baseline, indicating the importance of silent cardiac damage in the development of CHD and mortality.

Cardiac troponins are considered the “gold standard” measurements for diagnosing myocardial damage in patients with chest pain. Minor myocardial damage is observed in patients with unstable coronary artery disease, in which troponins are of prognostic importance and are instrumental for the clinical management of these patients.

Clinical Perspective p 1078

In subjects free from clinical signs of coronary heart disease (CHD), we previously reported the finding of an age-dependent increase in plasma cardiac troponin I (cTnI) levels, with the 99th percentile upper reference level (URL) in subjects aged >60 years being significantly higher than in those aged ≤60 years.1 It was also obvious from that study that only a minority of the older subjects had elevated levels compared with the younger subjects. We therefore speculated that these slightly elevated levels might predict a forthcoming CHD, in analogy with findings in patients with CHD in whom increased plasma concentrations are strong predictors of future events such as myocardial infarction or CHD death.2–9 Our previous findings also raised the question of whether the 99th percentile URL of healthy subjects is a relevant cutoff limit. Should the group of subjects with slightly elevated concentrations be considered healthy or in a subclinical phase of CHD? Such a decision can only be made on the basis of results that show that such a group has an ongoing disease process in their myocardium. In previous studies we showed that the second-generation cTnI assay from Beckman Coulter exhibited superior clinical performance because it identified &10% more patients with unstable CHD and a poor prognosis than some other first-generation sensitive cTnI assays and the cTnT assay.1 We therefore undertook a prospective study that sought to investigate the relationship between cTnI and first CHD event and death in a community-based cohort10 of 70-year-old men with a follow-up of up to 10 years (median 7.9 years). We determined baseline cTnI concentrations by the AccuTnI method (Beckman Coulter, Inc) and adjusted associations observed for conventional risk factors for CHD.

Methods

Study Population

In 1970, all men born between 1920 and 1924 and residing in Uppsala were invited to a health survey, the Uppsala Longitudinal Study of Adult Men (ULSAM), in which 82% (n=2322) participated.11 After 20 years, at the age of 70 years, they were invited for reinvestigation, which formed the baseline of the present study, comprising 1221 men of 1681 still alive (73%).12

CHD mortality and morbidity data were collected from the official Swedish registries held by the Center for Epidemiology, National Board of Health and Welfare in Sweden, where all deaths and hospitalizations in Sweden are registered with International Classification of Diseases (ICD) codes and dates. Information on medical history and current pharmacological treatment was obtained with the use of the original protocol questionnaire.11

To select subjects free of cardiovascular disease (CVD) at baseline (n=835), as atherosclerosis in CVD other than CHD and CHD covariates, for the analyses of first CHD event as the outcome, the following exclusion criteria were used: presence of prior myocardial infarction or angina pectoris in medical history; Q or QS complexes or left bundle-branch block (Minnesota codes 1.1 to 1.3 or 7.1, respectively) in baseline ECG registration; previous or incident CVD (ICD, 9th Revision [ICD-9] codes 390 to 459, equivalent to International Classification of Diseases, 10th Revision [ICD-10] codes I00 to I99) including congestive heart failure or current treatment with nitroglycerin or cardiac glycosides.

The study was designed, initiated, and managed by the authors, who were responsible for data collection, data analysis, preparation of the report, and the decision to submit the results for publication.

The Ethics Committee of the Faculty of Medicine at Uppsala University approved the study. Written informed consent was obtained from all subjects.

Follow-Up Data

CHD morbidity, defined by combining data from the Cause of Death Registry (CDR) and the Hospital Discharge Registry (HDR), is an efficient, validated alternative to revised hospital discharge notes and death certificates.13,14

CHD was defined with the use of the registry data as death, as recorded in the CDR, or first-time hospitalization for CHD (ICD-9 codes 410 to 414, equivalent to ICD-10 codes I20 to I25), as recorded in the HDR (censor date December 31, 2001).

All-cause mortality was obtained from the CDR. No subject was lost to follow-up because of missing registry data.

cTnI Determinations and Baseline Characteristics

Venous blood samples were drawn in the morning after an overnight fast. Plasma was prepared and anticoagulated with EDTA, freshly frozen, and stored frozen since baseline (−70°C). In August 2004, with the use of 1 manufacturer lot, cTnI in plasma was measured by the AccuTnI assay,15,16 ie, no interassay drift would be expected. The plasma samples used had been stored for 11±2 years and had been thawed a maximum of 1 time after being aliquoted. The stability of cTnI has previously been confirmed after 5 freeze-thaw cycles.16 According to the manufacturer, the minimum detectable concentration was <0.01 μg/L. Total imprecision for the AccuTnI was 4.1% to 8% (range, 0.05 to 11 μg/L). The 99th percentile URL was given as 0.04 μg/L. A multicenter trial showed the imprecision of 10% coefficient of variation (CV) at 0.06 μg/L and the 20% CV at 0.03 μg/L, whereas our own single-center evaluation of the assay gave an imprecision of 10% CV at a concentration of 0.03 μg/L and a 20% CV at 0.0085 μg/L.16 Determinations of cTnI, available from 1203 subjects, were performed blinded for outcome with reagents supplied by the company and performed according to their instructions on the instrument supplied by the manufacturer at the Department of Clinical Chemistry, Uppsala University Hospital.

Fasting concentrations of plasma glucose and serum cholesterol were measured by routine laboratory analysis at the Department of Clinical Chemistry, Uppsala University Hospital.12,17 Weight, height, body mass index (BMI), ECG, and supine systolic (SBP) and diastolic (DBP) blood pressures were measured under standardized conditions.12,17 Hypertension was defined as use of antihypertensive drugs or SBP >160 mm Hg or DBP >95 mm Hg at a single visit to comply as much as possible with guidelines current at that time (the diagnosis is usually based on several BP measurements over a given time period). In addition, current cutoffs of SBP >140 mm Hg or DBP >90 mm Hg were used in an additional set of analyses. Diabetes was defined as a fasting plasma glucose ≥7.0 mmol/L or the use of oral hypoglycemic agents or insulin. Smoking status, classified as current smoking versus never or previous smoking combined, was obtained from the questionnaire.

Statistical Analysis

Analyses were defined a priori. The statistical software package STATA 8.0 for PC (STATA Corporation) was used. All tests were 2-tailed, and a probability value <0.05 was considered significant. Skewed variables (cTnI and glucose) were log transformed to achieve normal distribution. Normally distributed variables were used in all statistical analyses. Group differences were tested with ANOVA and the Student t test. In the prospective analyses, Cox proportional hazards regression models were used. Hazard ratios (HRs) with 95% confidence intervals (CIs) and probability values were estimated for a 1-SD increase in a continuous variable and for a 1-step increase in the dichotomous variable smoking to determine the magnitude of the relationship to and the statistical significance of the predictors of the defined outcome, which was first CHD event, CHD death, or censor date of follow-up period, whichever came first. No violation to the proportional assumption was found with the use of Schoenfeld’s residuals in a linear correlation test for each Cox model presented.

Models defined a priori were also performed with the 2 predefined cutoff levels for cTnI obtained from the calculations of the 99th percentile URL in the previously examined cohort of subjects free from clinical signs of CHD.1 The cutoff level of 0.040 μg/L was the 99th percentile of the whole cohort, and the cutoff level of 0.021 μg/L was the 99th percentile of subjects below the age of 60 years. The lower cutoff level of 0.021 μg/L from that study1 also coincided with the cutoff level of the highest quartile in this study. Kaplan-Meier plots were performed for both cutoff levels and are shown in the figures.

In multivariable models, adjustments were made for the risk factors serum total and HDL cholesterol, plasma glucose, smoking, SBP, and BMI in men free from CVD, with CHD and mortality as the outcome, and further adjustments were made for previous CVD for mortality as the outcome in the total sample. In addition, log likelihood ratio tests were performed for multivariable models including and excluding the cTnI variable.

In additional multivariable models, the adjustments performed were the same with the exception of the continuous variables SBP and glucose, which were exchanged for the dichotomized variables hypertension and diabetes.

Results

Table 1 shows baseline clinical characteristics at age 71 years with known prevalent CVD included and excluded and probability values for Pearson product moment correlations with cTnI and t tests with cTnI in subjects free from CVD. The subjects had a median follow-up time of 7.9 years (up to 10.4 years), with a total of 9389 person-years at risk (PYAR) for all-cause mortality as the outcome and a total of 6170 PYAR for CHD as the outcome. During follow-up, 257 of 1203 subjects (rate 2.73/100 PYAR) died, and 116 of 835 subjects (rate 1.88/100 PYAR) had a first CHD event, of whom 51 died (44%).

TABLE 1. Clinical Characteristics at Baseline With Known Previous CVD Included or Excluded and Pearson Product Moment Correlation Coefficients or Student t Test for Associations With cTnI

CVD Included (n=1203)CVD Excluded (n=835)Correlation With cTnIP for CorrelationsP for t Test With cTnI
Values are arithmetic mean±SD. CVD refers to ICD-9 codes 390 to 459 and ICD-10 codes I00 to I99.
Age, y71.0±0.670.9±0.6
cTnI, μg/L0.017±0.060.016±0.061.00
Serum cholesterol, mmol/L5.8±1.05.8±1.00.0430.216
HDL cholesterol, mmol/L1.28±0.341.30±0.35−0.0510.142
Triglycerides, mmol/L1.45±0.771.39±0.760.0750.030
Smoking, %20.922.00.165
SBP, mm Hg147±18147±180.0570.101
DBP, mm Hg84±984±90.0430.215
Hypertension (160/95), %46.040.60.036
Hypertension (140/90), %72.069.90.319
Plasma glucose, mmol/L5.8±1.45.7±1.40.0680.048
Diabetes, %10.76.30.024
BMI, kg/m226.3±3.426.0±3.20.0590.088
Prevalent CVD, % (n)30.6 (368)0 (0)

Crude HRs relative to a 1-SD difference in variables for a first CHD event and for all-cause mortality with known prevalent CVD included and excluded during follow-up are presented in Table 2. In univariate analysis, cTnI showed a strong association with all-cause mortality and first CHD event (Table 2). Serum total and HDL cholesterol, smoking, SBP, BMI, and plasma glucose were all associated with a first CHD event in the univariate analyses, whereas triglycerides were not (Table 2). Prevalent CVD at baseline was significantly associated with all-cause mortality (Table 2).

TABLE 2. Crude HRs for Mortality With Known CVD at Baseline Included or Excluded and for CHD for up to 10 Years of Follow-Up (Median 7.9 Years)

Mortality Baseline CVD Included (n=257/1203)Mortality Baseline CVD Excluded (n=155/835)CHD Baseline CVD Excluded (n=116/835)
HR (95% CI)PHR (95% CI)PHR (95% CI)P
Values are arithmetic mean±SD. HRs with 95% CIs were estimated with Cox proportional hazards regression models, applied to variables standardized to 1 SD (except smoking: 0/1). CVD refers to ICD-9 codes 390 to 459 and ICD-10 codes I00 to I99; CHD refers to ICD-9 codes 410 to 414 and ICD-10 codes I20 to I25.
Age1.11 (0.91–1.36)0.2861.12 (0.84–1.47)0.4201.02 (0.78–1.41)0.920
cTnI1.44 (1.30–1.59)<0.0011.42 (1.28–1.57)<0.0011.45 (1.24–1.69)<0.001
Serum cholesterol0.94 (0.83–1.07)0.3471.20 (1.01–1.43)0.0361.20 (1.01–1.43)0.036
HDL cholesterol0.81 (0.71–0.93)0.0020.78 (0.65–0.93)0.0050.66 (0.53–0.81)<0.001
Triglycerides1.04 (0.92–1.18)0.5161.04 (0.85–1.23)0.5591.33 (1.12–1.57)0.001
Smoking1.68 (1.28–2.21)<0.0011.79 (1.22–2.63)0.0031.79 (1.22–2.63)0.003
SBP1.19 (1.06–1.34)0.0041.30 (1.10–1.54)0.0021.30 (1.10–1.54)0.002
DBP1.11 (0.97–1.25)0.1061.13 (0.94–1.35)0.1831.13 (0.94–1.35)0.183
Hypertension (160/95)1.64 (1.28–2.11)0.0012.29 (1.60–3.27)0.0012.29 (1.60–3.27)0.001
Hypertension (140/90)1.68 (1.22–2.31)0.0011.65 (1.12–2.42)0.0122.39 (1.48–3.84)<0.001
Plasma glucose1.17 (1.04–1.30)0.0051.25 (1.09–1.44)0.0021.25 (1.09–1.44)0.002
Diabetes1.87 (1.39–2.50)0.0012.38 (1.59–3.56)0.0012.38 (1.59–3.56)0.001
BMI1.17 (1.04–1.32)0.0081.24 (1.04–1.49)0.0171.24 (1.04–1.49)0.017
Prevalent CVD1.34 (1.04–1.73)0.023

Figure 1 shows the risk of death in relation to cTnI levels in men free from CVD at baseline (n=835) and in subjects with prevalent CVD (n=368) at baseline. In both groups we found an increased risk of all-cause mortality in relation to increasing cTnI levels (P<0.001), which was more prominent in subjects with prevalent CVD. In subjects with cTnI <0.021 μg/L, all-cause mortality did not differ between subjects with or without prevalent CVD (P=0.075).

Figure 1. All-cause mortality in relation to cTnI levels in the group of men free from CVD at baseline and in the group with prevalent CVD at baseline. The limits for cTnI concentrations were as follows: ≥0.040 μg/L; <0.040 to ≥0.021 μg/L; and <0.021 μg/L. There was a significant trend in mortality risk related to cTnI levels in both groups (P<0.001). No statistical differences in mortality was found between the 2 groups with levels <0.021 μg/L.

Figure 2 presents unadjusted Kaplan-Meier survival curves for men free from CVD at baseline, which showed that the mortality risk was higher for men with cTnI ≥0.021 μg/L (n=210/835) than for men with cTnI <0.021 μg/L (P<0.001) (Figure 2a) and was higher for men with a cTnI ≥0.040 μg/L (22/835) than for men with cTnI <0.040 μg/L (P<0.001) (Figure 2b).

Figure 2. Unadjusted Kaplan-Meier survival curves for all-cause mortality in men free from CVD at baseline during up to 10 years of follow-up (median 7.9 years) for men with cTnI ≥0.021 μg/L compared with men with cTnI <0.021 μg/L (P<0.001) (a) and for men with a cTnI concentration ≥0.040 μg/L compared with men with cTnI <0.040 μg/L (P<0.001) (b).

Similarly, the risk of first CHD event was higher for men with cTnI ≥0.021 μg/L (n=210/835) than for men with cTnI <0.021 μg/L (P=0.024) (Figure 3a) and was higher for men with cTnI ≥0.040 μg/L (n=22/835) than for men with cTnI <0.040 μg/L (P<0.001) (Figure 3b).

Figure 3. Unadjusted Kaplan-Meier survival curves for first CHD events in men free from CVD at baseline during up to 10 years of follow-up (median 7.9 years) for men with cTnI ≥0.021 μg/L compared with men with cTnI <0.021 μg/L (P=0.024) (a) and for men with a cTnI concentration ≥0.040 μg/L compared with men with cTnI <0.040 μg/L (P<0.001) (b).

In the multivariable models, in men free from CVD, after adjustments for the conventional risk factors, the association between cTnI and first CHD event and mortality remained significant (Table 3), as well as after further adjustment for prevalent CVD at baseline for all-cause mortality as the outcome in the total sample. Log likelihood ratio tests were all significant (P=0.022 to 0.001). In the additional analysis with adjustment for hypertension (with the use of the 160/95 mm Hg or treatment definition) and diabetes instead of SBP and glucose, which is presented in Table 3, results were similar for the association between cTnI and mortality (HR, 1.36; 95% CI, 1.22 to 1.51; P<0.001, equivalent to model 1), mortality excluding prevalent CVD at baseline (HR, 1.26; 95% CI, 1.09 to 1.46; P=0.002, equivalent to model 2), and CHD (HR, 1.33; 95% CI, 1.14 to 1.58; P<0.001, equivalent to model 3). When the alternative definition of hypertension with the use of the 140/90 mm Hg or treatment definition, corresponding cTnI results were as follows: model 1: HR, 1.36; 95% CI, 1.22 to 1.52; P<0.001; model 2: HR, 1.26; 95% CI, 1.09 to 1.47; P=0.002; and model 3: HR, 1.31; 95% CI, 1.12 to 1.55; P=0.001.

TABLE 3. HRs From 3 Multivariable Models for Mortality With Known CVD at Baseline Included or Excluded and for CHD as the Outcome for up to 10 Years of Follow-Up (Median 7.9 Years)

Mortality Model 1 Baseline CVD Included (n=257/1203)Mortality Model 2 Baseline CVD Excluded (n=155/835)CHD Model 3 Baseline CVD Excluded (n=116/835)
HR (95% CI)PHR (95% CI)PHR (95% CI)P
Data are HRs with 95% CIs. P values were estimated with Cox proportional hazards regression models, applied to variables standardized to 1 SD (except smoking: 0/1). CVD refers to ICD-9 codes 390 to 459 and ICD-10 codes I00 to I99; CHD refers to ICD-9 codes 410 to 414 and ICD-10 codes I20 to I25.
cTnI1.35 (1.21–1.50)<0.0011.26 (1.08–1.46)0.0031.31 (1.11–1.54)0.001
Total cholesterol0.96 (0.84–1.09)0.5300.91 (0.76–1.09)0.3081.28 (1.06–1.55)0.010
HDL cholesterol0.87 (0.75–1.01)0.0590.84 (0.69–1.01)0.0670.66 (0.52–01.83)0.001
Smoking (0/1)1.71 (1.30–2.25)<0.0012.09 (1.50–2.92)<0.0011.74 (1.18–2.56)0.005
Systolic blood pressure1.16 (1.03–1.31)0.0131.22 (1.04–1.43)0.0161.29 (1.08–1.55)0.004
BMI1.07 (0.94–1.22)0.3010.97 (0.80–1.17)0.7361.01 (0.82–1.24)0.918
Plasma glucose1.06 (0.94–1.20)0.3111.09 (0.94–1.27)0.2321.11 (0.95–1.31)0.170
Previous CVD1.35 (1.04–1.76)0.024

Table 4 shows crude and adjusted HRs for the 2 predefined cutoffs. In the univariate analyses, cTnI ≥0.040 or ≥0.021 μg/L, respectively, was predictive of all-cause mortality with baseline prevalent CVD included and excluded, respectively, and for first CHD event. In the adjusted analyses, cTnI ≥0.040 or ≥0.021 μg/L was still predictive of all-cause mortality after adjustments were made for conventional risk factors for CHD and prevalent CVD at baseline. A cTnI concentration ≥0.040 μg/L was still predictive of first CHD event when further adjustments were made for conventional risk factors. In the additional analysis with adjustment for hypertension (with either definition) and diabetes instead of SBP and glucose, as presented in Table 3, results were similar for CHD and for mortality.

TABLE 4. HRs From Multivariable Models for Mortality With Known CVD at Baseline Included or Excluded and CHD as the Outcome for up to 10 Years of Follow-Up (Median 7.9 Years) by Different Cutoff Values for cTnI

Mortality Baseline CVD Included (n=257/1203)Mortality Baseline CVD Excluded (n=155/835)CHD Baseline CVD Excluded (n=116/835)
HR (95% CI)PHR (95% CI)PHR (95% CI)P
Data are HRs with 95% CIs estimated with Cox proportional hazards regression models, applied to cTnI variables dichotomized by the lower limit of quartile IV (cTnI ≥0.021 μg/L) and by the cutoff value used in the clinic for patients with chest pain (cTnI ≥0.040 μg/L). CVD refers to ICD-9 codes 390 to 459 and ICD-10 codes I00 to I99; CHD refers toICD-9 codes 41 to 414 and ICD-10 codes I20 to I25.
*Adjusted for smoking, total and HDL cholesterol, SBP, BMI, and fasting glucose and further adjusted for previous CVD for mortality as the outcome when baseline CVD was included.
cTnI ≥0.021 μg/L
    Crude1.75 (1.36–2.25)<0.0011.67 (1.21–2.34)0.0021.53 (1.06–2.23)0.184
    Adjusted*1.63 (1.26–2.12)<0.0011.53 (1.09–2.15)0.0141.29 (0.88–1.90)0.193
cTnI ≥0.040 μg/L
    Crude3.20 (2.11–4.85)<0.0012.92 (1.54–5.54)0.0013.75 (1.96–7.15)<0.001
    Adjusted*2.76 (1.77–4.30)<0.0012.12 (1.06–4.22)0.0332.38 (1.18–4.81)0.016

Table 5 presents predictive capacities of the 2 a priori specified cutoff levels for cTnI and outcomes.

TABLE 5. Predictive Capacities of 2 Predefined Cutoff Values for cTnI and Mortality With Known Previous CVD Included or Excluded and CHD as the Outcome for up to 10 Years of Follow-Up (median 7.9 Years)

Mortality Baseline CVD Included (n=257/1203)Mortality Baseline CVD Excluded (n=155/835)CHD Baseline CVD Excluded (n=116/835)
PPV indicates positive predictive value; NPV, negative predictive value; and PCC, percentage correctly classified. CVD refers to ICD-9 codes 390 to 459 and ICD-10 codes I00 to I99; CHD refers to ICD-9 codes 410 to 414 and ICD-10 codes I20 to I25.
cTnI ≥0.021 μg/L
    Sensitivity, %40.034.432.5
    Specificity, %74.677.076.2
    PPV, %30.025.719.5
    NPV, %82.183.586.4
    PCC, %67.369.069.6
cTnI ≥0.040 μg/L
    Sensitivity, %9.76.48.0
    Specificity, %98.098.298.3
    PPV, %56.845.545.5
    NPV, %80.081.985.7
    PCC, %79.179.984.7

Discussion

An elevated concentration of cTnI in 70-year-old men free from clinical signs of CVD was shown to be associated with an increased risk of first CHD event over a follow-up period of up to 10 years. Furthermore, cTnI was predictive of all-cause mortality in this cohort independent of baseline prevalent CVD. These findings are novel and suggest that subclinical silent myocardial damage makes a major contribution to CHD morbidity and mortality. Most interesting, the observed associations between cTnI concentrations and increased risk of first CHD event and mortality were independent of major conventional CHD risk factors.

The release of cTnI from myocardial cells is probably caused by many different mechanisms. One such mechanism is the release through the irreversibly permeabilized cell membrane, as seen in cell death and necrosis in situations of acute severe ischemia in the acute coronary syndrome.18 Leakage of troponin molecules might also occur through reversibly permeabilized plasma membranes as a consequence of transient and less severe ischemia19 or as a consequence of the exposure of the myocardial cells to cytokines such as tumor necrosis factor-α.20 These mechanisms are probably operative in severe diseases with other signs of multiorgan failure, such as in septicemia,19 in which elevated troponin levels are often found without any indications of coronary artery occlusion. Still another mechanism might be impairment of the plasma membrane integrity due to aging of the cells and the apoptotic elimination of cells.18 The mechanisms contributing to the elevation of cTnI in a cohort of our 70-year-old men with poor prognosis cannot be discerned from the present study. Several mechanisms are likely to be involved, including asymptomatic thrombotic or embolic occlusions of small coronary arteries,21 ongoing activation of inflammation, and genetically determined or acquired acceleration of programmed cell death.

Smoking, cholesterol, high blood pressure, diabetes,22 and high glucose concentrations23 are considered major, well-established conventional risk factors for CHD morbidity and mortality, and high BMI24 is considered a major risk factor for mortality. The processes, influenced by the conventional risk factors that are ongoing in the myocardium, are long-term processes over decades followed by a subclinical period before the onset of clinical signs of CHD, manifest CHD, and finally death. Our findings indicate that cTnI is a marker of myocardial damage during such a subclinical period and that this period consists of several years before the onset of clinical signs of CHD and death. We adjusted the observed association between cTnI and CHD and mortality for smoking, cholesterol, blood pressure, hypertension, fasting glucose, diabetes, and BMI without major reduction of the observed associations. Thus, we consider cTnI to be a marker of subclinical myocardial damage, necrosis, or underlying atherosclerosis and not as a mediator of the effect of the conventional risk factors on the atherosclerotic process, which may explain the role of cTnI in the multivariable models as a predictor of first CHD event independent of conventional risk factors. However, cTnI could also be a marker of the effect of genetic factors, apoptosis, or aging itself. A mutation in the cTnI encoding gene and an association with impairment of troponin interactions and diminished myocardial contractility in cardiomyopathy has been reported recently.25 A possible effect of genetic factors or the effects of aging processes themselves on cTnI levels require further studies. Influences of lifestyle or lifestyle modification on cTnI levels, if any, are not known. Most importantly, because this is an observational study, mechanistic conclusions cannot be drawn but can only be matters of speculation.

In our previous study we reported increasing cTnI concentrations with age in the apparently healthy reference population.1 The age relationship raised the question of whether the URL should be age related or whether the lower reference limit for cTnI of 0.021 μg/L actually represents the “true” URL of healthy subjects. This question could not be answered in that study because no follow-up data on the outcome of these subjects were available. In the present study we took advantage of a population-based cohort of men in which baseline data were collected at the age of 70 years, which made baseline data standardized for age. Women were not included in the ULSAM study for historical reasons, and therefore generalizability to women is uncertain. We lack data for renal function, ie, glomerular filtration and C-reactive protein as a marker for inflammation; however, there is no report on the effect of renal function on cTnI levels.

Our findings of a close association between the elevated cTnI levels in the cohort of 70-year-old men free from CVD at baseline and the development of CHD indicates that the answer to the aforementioned question is that even slightly raised cTnI levels above the lower 99th percentile of 0.021 μg/L are signs of ongoing subclinical processes in the myocardium. Furthermore, because these slightly raised cTnI levels are associated with an increased mortality risk, our results may suggest that a cutoff <0.040 μg/L may be a clinically relevant limit, at least in elderly men, but this needs to be determined in future studies. We have used EDTA plasma samples, which are not ideal because of EDTA disruption of troponin. Thus, there may be a lowering of cTnI by &12% over the whole range of cTnI concentrations, which, however, will not affect the relative risk estimates. Because of this, we are reluctant to propose a cutoff value. Instead, we see this investigation as an exploratory study. Future studies should determine cTnI in fresh samples for the purpose of defining a possible cutoff. Efforts to improve the CV for the assay and to increase precision in the lower range of cTnI concentrations should also be encouraged. Our results further emphasize the notion that the development of CHD is a continuum and that the measurement of cTnI with highly sensitive assays allows us to detect this process at a very early stage. Thus far, the determination of cTnI with the highly sensitive second-generation AccuTnI assay is unique in detecting this cohort of subjects without known CVD but with this sign of myocardial damage because the determinations of cTnT in another large cohort of subjects free from clinical signs of CHD did not identify any measurable levels even among older subjects,5 which precluded any calculations of associations to outcome. Whether these differences between generations of cTnI and cTnT assays are related to the fact that they measure different molecules with different kinetics and patterns of release from injured myocardium26,27 or whether they are merely a matter of lack of sensitivity of the latter assay is unknown.

In the present study we present data on an association between cTnI and first CHD event and mortality, independent of conventional CHD risk factors, using registry data for defining the outcomes. There is always a possibility of misclassification bias in registry data. However, a quality control of the CDR by the Swedish centers of the World Health Organization MONICA (multinational monitoring of trends and determinants in cardiovascular disease) study have shown good agreement for registration of myocardial infarction.13,14 The possible limitation of using registry data including misclassification would have been an underestimation of true risk. However, we present significant risk estimates of important magnitude.

The implication of identification of subclinical myocardial damage in elderly men without clinical symptoms raises the question of the use of this information in the clinical setting. In patients with detectable cTnI and symptoms suggestive of unstable CHD treatment, guidelines have been well established28 because these patients benefit from early coronary intervention and medical treatment with low-molecular-weight heparin.29 Management strategies for silent ischemia, however, are less clearly defined. Findings from our study raise the question of which strategies are to be established for identifying subclinical myocardial damage and for subsequent actions. Furthermore, they clearly highlight the need for clinical research on this topic. Above all, our findings suggest that there is a need for measuring cTnI with highly sensitive cTnI assays.1

Conclusions

We conclude that cTnI in elderly men is a predictor of first CHD event and of all-cause mortality independent of conventional major CHD risk factors. Thus, cTnI should be regarded as a risk indicator for future CHD or death both in men free from clinical signs of CVD and, as previously shown, in patients presenting with clinical symptoms of unstable CHD.

This study was supported by research grants from the Swedish Medical Research Council No. 5446, Foundation for Geriatric Research, Uppsala Geriatric Fund, and Uppsala University. The reagents and the instrument for the AccuTnI assay were supplied by the company Beckman Coulter, Inc (Chaska, Minn). There was no other external sponsor of the study. The technical expertise of Martin Venge in the performance of troponin assays is greatly appreciated.

Disclosures

None.

Footnotes

Correspondence to Björn Zethelius, Md, PhD, Department of Public Health and Caring Sciences/Geriatrics, Uppsala Science Park, 75185 Uppsala, Sweden. E-mail

References

  • 1 Venge P, Lagerqvist B, Diderholm E, Lindahl B, Wallentin L. Clinical performance of three cardiac troponin assays in patients with unstable coronary artery disease (a FRISC II substudy). Am J Cardiol. 2002; 89: 1035–1041.CrossrefMedlineGoogle Scholar
  • 2 Diderholm E, Andren B, Frostfeldt G, Genberg M, Jernberg T, Lagerqvist B, Lindahl B, Venge P, Wallentin L. The prognostic and therapeutic implications of increased troponin T levels and ST depression in unstable coronary artery disease: the FRISC II invasive troponin T electrocardiogram substudy. Am Heart J. 2002; 143: 760–767.CrossrefMedlineGoogle Scholar
  • 3 Frostfeldt G, Gustafsson G, Lindahl B, Nygren A, Venge P, Wallentin L. Possible reasons for the prognostic value of troponin-T on admission in patients with ST-elevation myocardial infarction. Coron Artery Dis. 2001; 12: 227–237.CrossrefMedlineGoogle Scholar
  • 4 James S, Armstrong P, Califf R, Simoons ML, Venge P, Wallentin L, Lindahl B. Troponin T levels and risk of 30-day outcomes in patients with the acute coronary syndrome: prospective verification in the GUSTO-IV trial. Am J Med. 2003; 115: 178–184.CrossrefMedlineGoogle Scholar
  • 5 James SK, Armstrong P, Barnathan E, Califf R, Lindahl B, Siegbahn A, Simoons ML, Topol EJ, Venge P, Wallentin L. Troponin and C-reactive protein have different relations to subsequent mortality and myocardial infarction after acute coronary syndrome: a GUSTO-IV substudy. J Am Coll Cardiol. 2003; 41: 916–924.CrossrefMedlineGoogle Scholar
  • 6 Lindahl B, Venge P, Wallentin L, for the FRISC Study Group. Relation between troponin T and the risk of subsequent cardiac events in unstable coronary artery disease. Circulation. 1996; 93: 1651–1657.CrossrefMedlineGoogle Scholar
  • 7 Lindahl B, Andren B, Ohlsson J, Venge P, Wallentin L, for the FRISK Study Group. Risk stratification in unstable coronary artery disease: additive value of troponin T determinations and pre-discharge exercise tests. Eur Heart J. 1997; 18: 762–770.CrossrefMedlineGoogle Scholar
  • 8 Lindahl B, Toss H, Siegbahn A, Venge P, Wallentin L, for the FRISC Study Group: Fragmin During Instability in Coronary Artery Disease. Markers of myocardial damage and inflammation in relation to long-term mortality in unstable coronary artery disease. N Engl J Med. 2000; 343: 1139–1147.CrossrefMedlineGoogle Scholar
  • 9 Lindahl B, Diderholm E, Lagerqvist B, Venge P, Wallentin L. Mechanisms behind the prognostic value of troponin T in unstable coronary artery disease: a FRISC II substudy. J Am Coll Cardiol. 2001; 38: 979–986.CrossrefMedlineGoogle Scholar
  • 10 Zethelius B, Byberg L, Hales CN, Lithell H, Berne C. Proinsulin is an independent predictor of coronary heart disease: report from a 27-year follow-up study. Circulation. 2002; 105: 2153–2158.LinkGoogle Scholar
  • 11 Hedstrand H. A study of middle-aged men with particular reference to risk factors for cardiovascular disease. Upsala J Med Sci. 1975; 80 (suppl 19): 1–61.CrossrefMedlineGoogle Scholar
  • 12 Byberg L, McKeigue PM, Zethelius B, Lithell HO. Birth weight and the insulin resistance syndrome: association of low birth weight with truncal obesity and raised plasminogen activator inhibitor-1 but not with abdominal obesity or plasma lipid disturbances. Diabetologia. 2000; 43: 54–60.CrossrefMedlineGoogle Scholar
  • 13 Merlo J, Lindblad U, Pessah-Rasmussen H, Hedblad B, Rastam J, Isacsson SO, Janzon L, Rastam L. Comparison of different procedures to identify probable cases of myocardial infarction and stroke in two Swedish prospective cohort studies using local and national routine registers. Eur J Epidemiol. 2000; 16: 235–243.CrossrefMedlineGoogle Scholar
  • 14 Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas AM, Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project: registration procedures, event rates, and case-fatality rates in 38 populations from 21 countries in four continents. Circulation. 1994; 90: 583–612.CrossrefMedlineGoogle Scholar
  • 15 Uettwiller-Geiger D, Wu AH, Apple FS, Jevans AW, Venge P, Olson MD, Darte C, Woodrum DL, Roberts S, Chan S. Multicenter evaluation of an automated assay for troponin I. Clin Chem. 2002; 48 (pt 1): 869–876.CrossrefMedlineGoogle Scholar
  • 16 Venge P, Lindahl B, Wallentin L. New generation cardiac troponin I assay for the access immunoassay system. Clin Chem. 2001; 47: 959–961.CrossrefMedlineGoogle Scholar
  • 17 Zethelius B, Hales CN, Lithell HO, Berne C. Insulin resistance, impaired early insulin response, and insulin propeptides as predictors of the development of type 2 diabetes: a population-based, 7-year follow-up study in 70-year-old men. Diabetes Care. 2004; 27: 1433–1438.CrossrefMedlineGoogle Scholar
  • 18 Olivetti G, Abbi R, Quaini F, Kajstura J, Cheng W, Nitahara JA, Quaini E, De Loreto C, Beltrami CA, Krajewski S, Reed JC, Anversa P. Apoptosis in the failing human heart. N Engl J Med. 1997; 336: 1131–1141.CrossrefMedlineGoogle Scholar
  • 19 Wu AH. Increased troponin in patients with sepsis and septic shock: myocardial necrosis or reversible myocardial depression? Intensive Care Med. 2001; 27: 959–961.CrossrefMedlineGoogle Scholar
  • 20 Prabhu S. Cytokine-induced modulation of cardiac function. Circ Res. 2004; 95: 1140–1153.LinkGoogle Scholar
  • 21 Del Carlo CH, O’Connor CM. Cardiac troponins in congestive heart failure. Am Heart J. 1999; 138 (pt 1): 646–653.CrossrefMedlineGoogle Scholar
  • 22 Wannamethee SG, Shaper AG, Whincup PH, Walker M. Role of risk factors for major coronary heart disease events with increasing length of follow up. Heart. 1999; 81: 374–379.CrossrefMedlineGoogle Scholar
  • 23 Balkau B, Shipley M, Jarrett RJ, Pyorala K, Pyorala M, Forhan A, Eschwege E. High blood glucose concentration is a risk factor for mortality in middle-aged nondiabetic men: 20-year follow-up in the Whitehall Study, the Paris Prospective Study, and the Helsinki Policemen Study. Diabetes Care. 1998; 21: 360–367.CrossrefMedlineGoogle Scholar
  • 24 Fontbonne A, Tchobroutsky G, Eschwege E, Richards JL, Claude JR, Rosselin GE. Coronary heart disease mortality risk: plasma insulin level is a more sensitive marker than hypertension or abnormal glucose tolerance in overweight males: the Paris Prospective Study. Int J Obes. 1988; 12: 557–565.MedlineGoogle Scholar
  • 25 Murphy RT, Mogensen J, Shaw A, Kubo T, Hughes S, McKenna WJ. Novel mutation in cardiac troponin I in recessive idiopathic dilated cardiomyopathy. Lancet. 2004; 363: 371–372.CrossrefMedlineGoogle Scholar
  • 26 Collinson PO, Stubbs PJ. Are troponins confusing? Heart. 2003; 89: 1285–1287.CrossrefMedlineGoogle Scholar
  • 27 Sharma S, Jackson PG, Makan J. Cardiac troponins. J Clin Pathol. 2004; 57: 1025–1026.CrossrefMedlineGoogle Scholar
  • 28 Bertrand ME, Simoons ML, Fox KA, Wallentin LC, Hamm CW, MacFadden E, De Feyter PJ, Specchia G, Ruzyllo W. Management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J. 2002; 23: 1809–1840.CrossrefMedlineGoogle Scholar
  • 29 Wallentin L. Low-molecular-weight heparin as a bridge to timely revascularization in unstable coronary artery disease: an update of the Fragmin During Instability in Coronary Artery Disease II Trial. Haemostasis. 2000; 30 (suppl 2): 108–113; discussion 106–107.MedlineGoogle Scholar
circulationahaCirculationCirculationCirculation0009-73221524-4539Lippincott Williams & Wilkins
CLINICAL PERSPECTIVE28022006

Cardiac troponin I (cTnI), a standard for detection of myocardial damage, predicts acute myocardial infarction or death in patients with unstable coronary heart disease (CHD). cTnI concentrations increase with age in subjects free from clinical signs of CHD, suggesting silent myocardial damage. The question is whether slightly elevated cTnI has clinical importance in elderly subjects without clinical symptoms. Is it prognostic for future manifest CHD and death? In this cohort study, cTnI was predictive of mortality and of first CHD events in subjects free from cardiovascular disease diagnosis at baseline. The overall mortality risk, associated with elevations of cTnI, suggests that subclinical silent myocardial damage makes a major contribution to clinical manifestations of CHD and mortality. In patients with detectable cTnI and unstable CHD, treatment guidelines have been well established because these patients benefit from early coronary intervention and treatment with low-molecular-weight heparin. Management strategies for silent ischemia are, however, less clearly defined. Findings from the present study raise the questions of which strategies are to be established for identifying subclinical silent myocardial damage and, additionally, how this should be put into effect in clinical practice. Because this is an observational study, we cannot provide insights into the mechanisms of prognostic information, but our results may at least change the thinking about silent ischemia and clearly highlight a need for further interventional trials to answer questions about whether and how to change future clinical practice.

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.