High-Sensitive Cardiac Troponin T in Elderly Men
High-Sensitive Cardiac Troponin T in Elderly Men
The Uppsala Longitudinal Study of Adult Men (ULSAM) was initiated in 1970. All 50-year-old men born between 1920 and 1924 and living in Uppsala, Sweden were invited to participate in a health survey that focused on identifying CV risk factors (described in detail at www.pubcare.uu.se/ULSAM). The present analysis is based on the third examination cycle of the ULSAM cohort that was conducted between 1991 and 1995 when the participants were approximately 71 years old. Of the 1,221 participants, 940 had remaining plasma samples available for measurements of cTnT and all covariates examined.
We assessed both subgroups with (n = 379) and without prevalent CV disease (n = 561) at baseline. Cardiovascular disease was defined as follows: coronary artery disease (previous myocardial infarction [MI] or angina pectoris as noted in the medical history); Q or QS waves or left bundle-branch block (Minnesota codes 1.1 to 1.3 and 7.1) on the baseline electrocardiogram; current treatment with nitroglycerine or glycosides; or a history of any CV disease, as noted in the Swedish Hospital Discharge Register (International Classification of Diseases, 10th Revision [ICD-10], codes I00 to I99).
Information on clinical history and anthropomorphic measures was obtained at the baseline visit. Plasma glucose (fasting and 120 minutes after an oral glucose load), serum total cholesterol, low-density lipoprotein (LDL) cholesterol, and high-density lipoprotein (HDL) cholesterol levels were measured by routine laboratory analysis.
Venous blood samples were drawn at baseline and stored at −70°C for a mean (±SD) of 11 ± 2 years until analysis at the Department of Clinical Chemistry, Uppsala University. Cardiac troponin T was measured on an Elecsys 2010/cobas e411 instrument (Roche Diagnostics, Mannheim, Germany) using a high-sensitive assay (reagents from lot 153401) with a measurement range of 3.0 to 10,000 ng/L, and 14.0 ng/L as the 99th percentile among healthy subjects. The lowest concentration measurable with a coefficient of variation <10% has been reported at 13.0 ng/L. Cardiac troponin T levels <3.0 ng/L were assigned a value of 3.0 ng/L. N-terminal pro-B-type natriuretic peptide (NT-proBNP) was analyzed using a sandwich immunoassay (Roche Diagnostics) on an Elecsys 2010 instrument. C-reactive protein (CRP) and cystatin C were measured with the use of latex-enhanced reagents on a Behring BN ProSpec analyzer (Siemens Healthcare Diagnostics, Deerfield, IL). The analytical imprecisions of the assays for NT-proBNP, CRP, and cystatin C were not separately assessed within this study but are not allowed to exceed 3%, 2%, and 2%, respectively, to fulfill the local requirements for quality assurance.
Complete mortality and morbidity data from all subjects were collected from the Swedish Cause of Death and Hospital Admission registries held by the National Board of Health and Welfare in Sweden. The events were classified as total and CV mortality, fatal and nonfatal coronary events (ICD-10 codes I00 to I99), fatal and nonfatal stroke (ICD-10 codes I60 to I69), and the composite of these outcomes. The median follow-up was 10.0 (range 0.9–12.4) years.
Continuous variables are described as means (with SD) or medians (with interquartile ranges), as appropriate. For continuous variables, the Shapiro-Wilk's statistic W was calculated where the region W ≥0.95 led to the use of a parametric method while otherwise a nonparametric method was used. Dichotomous variables are expressed as numbers and percentages with differences being analyzed with the χ test.
In cross-sectional analyses, we assessed the associations of cTnT to the continuous variables body mass index (BMI), waist circumference, systolic and diastolic blood pressure, plasma glucose, total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, NT-proBNP, CRP, and cystatin C. Results are given as parametric Pearson correlation coefficients or nonparametric Spearman rank correlation coefficients. We also tested the associations of cTnT to the dichotomous variables current smoking, obesity (defined as BMI ≥30 kg/m), diabetes (defined as fasting glucose >7.0 mmol/L or the use of oral hypoglycemic agents or insulin), previous or prevalent atrial fibrillation (defined as any previous diagnosis of atrial fibrillation [ICD-10 code I49]), prevalent coronary artery disease, previous stroke, prevalent CV disease, and previous or prevalent cancer (defined as any diagnosis of malignancy [ICD-10 codes C00 to D48]). These results are given as odds ratios (ORs) with 95% CIs of a 1-SD increase in ln-transformed cTnT levels. All cross-sectional analyses were adjusted for current smoking, hypertension, diabetes, and (where applicable) LDL cholesterol and HDL cholesterol.
The prognostic value of cTnT levels was investigated using Cox proportional hazards regression models. Results are presented as estimated hazard ratios (HRs) with 95% CI of a 1-SD increase of cTnT (ln). Adjustment was made in model 1 for established risk indicators: age, current smoking, BMI, systolic blood pressure, antihypertensive treatment, total cholesterol, HDL cholesterol, lipid-lowering treatment, diabetes mellitus, and previous or prevalent cancer. Model 2 was additionally adjusted for levels of NT-proBNP (ln), CRP (ln), and cystatin C (ln). The proportional hazard assumptions of the Cox regression models were confirmed using the Schoenfeld's test. The linearity of the relation between cTnT and outcome was examined visually in GAM plots. To assess the calibration of the Cox models, we used the Grønnesby and Borgan test that compares the number of events that are observed with those that are expected based on the estimation from the models.
The discriminative ability of cTnT was estimated as C-statistics for Cox regression models. The increased discriminative ability of one regression model versus another was assessed calculating the absolute and relative integrated discrimination improvement (IDI). Kaplan-Meier curves were constructed to illustrate the timing of events. In all tests, a 2-sided P < .05 was considered significant without adjustments for multiplicity. The software package SAS 9 (SAS Institute, Inc, Cary, NC) was used for the statistical analyses.
All participants gave written informed consent. The study complies with the Declaration of Helsinki and had been approved by the ethics committee at the Faculty of Medicine of Uppsala University. No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the manuscript, and its final contents.
Methods
Study Population
The Uppsala Longitudinal Study of Adult Men (ULSAM) was initiated in 1970. All 50-year-old men born between 1920 and 1924 and living in Uppsala, Sweden were invited to participate in a health survey that focused on identifying CV risk factors (described in detail at www.pubcare.uu.se/ULSAM). The present analysis is based on the third examination cycle of the ULSAM cohort that was conducted between 1991 and 1995 when the participants were approximately 71 years old. Of the 1,221 participants, 940 had remaining plasma samples available for measurements of cTnT and all covariates examined.
We assessed both subgroups with (n = 379) and without prevalent CV disease (n = 561) at baseline. Cardiovascular disease was defined as follows: coronary artery disease (previous myocardial infarction [MI] or angina pectoris as noted in the medical history); Q or QS waves or left bundle-branch block (Minnesota codes 1.1 to 1.3 and 7.1) on the baseline electrocardiogram; current treatment with nitroglycerine or glycosides; or a history of any CV disease, as noted in the Swedish Hospital Discharge Register (International Classification of Diseases, 10th Revision [ICD-10], codes I00 to I99).
Baseline Measurements
Information on clinical history and anthropomorphic measures was obtained at the baseline visit. Plasma glucose (fasting and 120 minutes after an oral glucose load), serum total cholesterol, low-density lipoprotein (LDL) cholesterol, and high-density lipoprotein (HDL) cholesterol levels were measured by routine laboratory analysis.
Biochemical Analysis
Venous blood samples were drawn at baseline and stored at −70°C for a mean (±SD) of 11 ± 2 years until analysis at the Department of Clinical Chemistry, Uppsala University. Cardiac troponin T was measured on an Elecsys 2010/cobas e411 instrument (Roche Diagnostics, Mannheim, Germany) using a high-sensitive assay (reagents from lot 153401) with a measurement range of 3.0 to 10,000 ng/L, and 14.0 ng/L as the 99th percentile among healthy subjects. The lowest concentration measurable with a coefficient of variation <10% has been reported at 13.0 ng/L. Cardiac troponin T levels <3.0 ng/L were assigned a value of 3.0 ng/L. N-terminal pro-B-type natriuretic peptide (NT-proBNP) was analyzed using a sandwich immunoassay (Roche Diagnostics) on an Elecsys 2010 instrument. C-reactive protein (CRP) and cystatin C were measured with the use of latex-enhanced reagents on a Behring BN ProSpec analyzer (Siemens Healthcare Diagnostics, Deerfield, IL). The analytical imprecisions of the assays for NT-proBNP, CRP, and cystatin C were not separately assessed within this study but are not allowed to exceed 3%, 2%, and 2%, respectively, to fulfill the local requirements for quality assurance.
Prognostic Evaluation
Complete mortality and morbidity data from all subjects were collected from the Swedish Cause of Death and Hospital Admission registries held by the National Board of Health and Welfare in Sweden. The events were classified as total and CV mortality, fatal and nonfatal coronary events (ICD-10 codes I00 to I99), fatal and nonfatal stroke (ICD-10 codes I60 to I69), and the composite of these outcomes. The median follow-up was 10.0 (range 0.9–12.4) years.
Statistical Analysis
Continuous variables are described as means (with SD) or medians (with interquartile ranges), as appropriate. For continuous variables, the Shapiro-Wilk's statistic W was calculated where the region W ≥0.95 led to the use of a parametric method while otherwise a nonparametric method was used. Dichotomous variables are expressed as numbers and percentages with differences being analyzed with the χ test.
In cross-sectional analyses, we assessed the associations of cTnT to the continuous variables body mass index (BMI), waist circumference, systolic and diastolic blood pressure, plasma glucose, total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, NT-proBNP, CRP, and cystatin C. Results are given as parametric Pearson correlation coefficients or nonparametric Spearman rank correlation coefficients. We also tested the associations of cTnT to the dichotomous variables current smoking, obesity (defined as BMI ≥30 kg/m), diabetes (defined as fasting glucose >7.0 mmol/L or the use of oral hypoglycemic agents or insulin), previous or prevalent atrial fibrillation (defined as any previous diagnosis of atrial fibrillation [ICD-10 code I49]), prevalent coronary artery disease, previous stroke, prevalent CV disease, and previous or prevalent cancer (defined as any diagnosis of malignancy [ICD-10 codes C00 to D48]). These results are given as odds ratios (ORs) with 95% CIs of a 1-SD increase in ln-transformed cTnT levels. All cross-sectional analyses were adjusted for current smoking, hypertension, diabetes, and (where applicable) LDL cholesterol and HDL cholesterol.
The prognostic value of cTnT levels was investigated using Cox proportional hazards regression models. Results are presented as estimated hazard ratios (HRs) with 95% CI of a 1-SD increase of cTnT (ln). Adjustment was made in model 1 for established risk indicators: age, current smoking, BMI, systolic blood pressure, antihypertensive treatment, total cholesterol, HDL cholesterol, lipid-lowering treatment, diabetes mellitus, and previous or prevalent cancer. Model 2 was additionally adjusted for levels of NT-proBNP (ln), CRP (ln), and cystatin C (ln). The proportional hazard assumptions of the Cox regression models were confirmed using the Schoenfeld's test. The linearity of the relation between cTnT and outcome was examined visually in GAM plots. To assess the calibration of the Cox models, we used the Grønnesby and Borgan test that compares the number of events that are observed with those that are expected based on the estimation from the models.
The discriminative ability of cTnT was estimated as C-statistics for Cox regression models. The increased discriminative ability of one regression model versus another was assessed calculating the absolute and relative integrated discrimination improvement (IDI). Kaplan-Meier curves were constructed to illustrate the timing of events. In all tests, a 2-sided P < .05 was considered significant without adjustments for multiplicity. The software package SAS 9 (SAS Institute, Inc, Cary, NC) was used for the statistical analyses.
All participants gave written informed consent. The study complies with the Declaration of Helsinki and had been approved by the ethics committee at the Faculty of Medicine of Uppsala University. No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the manuscript, and its final contents.
Source...