A population model of age and gender specific cardiac troponin levels

The range of the high-sensitivity cardiac troponin (HSCTN) values in the gen-eral population increases progressively with age, being several times higher in older patients, beyond 70 years old, than in younger ones, below 40. More accurate normal limits for HSCTN, taking into account age and gender, are needed for the differential diagnosis and evaluation of the prognostic signif-icance of increases that do not reach the vendor-supplied upper reference limits (URL). We performed an analysis of the high sensitivity cardiac troponin (HSCTN) of 21743 individuals, representative for the general US population, that were studied in the NHANES survey performed by the Center for Disease Control of the USA. The vendor supplied URL values are typically several times higher than the actual upper limit in subjects under 40 and correspond to the URL at ages between 51 and 84 for specific vendors and genders. For each HSCTN test variant, we considered each one year age group between 1 and 85 and either gender. The distribution of the logarithmed HSCTN for a given test variant and gender, in each subsample i, is relatively close to Gaussian (N ({micro}i, {sigma}i)). Two quadratic models, for the 1-15 and the 16-85 age ranges, were found to fit well with the {micro}i, while the{sigma} i follow a linear model. These theoretical distributions can be used used to estimate any quantiles of the HSCTN distribution as functions of age and gender, including the upper reference limit (URL) for the HSCTN I and T in the general population. Consequently, it is possible to construct a calculated HSCTN indicator that represents the estimated age and gender specific centile in which a measurement with a specific test kit falls, or the percentual distance above the 99th centile. This indicator would be independent of age, gender and kit-vendor, and thus more accurate and easier to employ in clinical practice.