Background:

Upper reference limits of highly sensitive cardiac troponin T (hs-cTnT) are typically determined in a healthy, population-based cohort, and are frequently exceeded in hospitalized patients, even in cases without the diagnosis or suspicion of an underlying cardiac disease. We hypothesized that hs-cTnT values of hospitalized patients without known heart disease are higher compared to a population-based cohort.

Methods:

Retrospective analyses were performed in two independent cohorts. We included 5,652 participants of the population-based LIFE-Adult cohort as well as 10,427 hospitalized patients admitted to the university hospital in Leipzig between 2014 and 2021. In both cohorts, subjects with diagnosed or suspected cardiac diseases or an impaired kidney function (eGFR < 60/ml/min/1.73 m²) were excluded. Additionally, inpatient participants with a principal diagnosis that may be associated with a release of hs-cTnT (e.g. stroke, sepsis, pulmonary embolism), were excluded. We used Spearman’s rank correlation for correlation analyses of hs-cTnT serum concentrations and age. Sex- and age-adjusted 99^th percentiles for hs-cTnT were obtained in both cohorts.

Results:

In both cohorts, hs-cTnT serum concentrations positively correlated with age (p < 0.001, ρ = 0.49/ 0.44 in population-based females/ males, ρ = 0.62/ 0.52 in hospitalized females/ males). Male sex was associated with higher hs-cTnT serum concentrations. Hospitalized individuals showed significantly higher hs-cTnT concentrations in nearly all age groups (Figure 1). In the population-based cohort only 99^th percentile hs-cTnT results of females aged above 70 and males aged above 60 years exceeded the assay’s upper reference limit. In contrast, the 99^th percentiles of all hospitalized age and sex groups lay above this threshold (Figure 2).

Conclusion:

Serum concentrations of hs-cTnT were higher in hospitalized patients compared to the general population. Consequently, besides age and sex, hospitalization appears to influence serum concentration of hs-cTnT. Applying current reference values may overdiagnose acute coronary syndrome. Further prospective studies are needed to approve this effect and to study underlying mechanisms. Our results raise the question, whether reference ranges from non-hospitalized cohorts can be uniformly applied to inpatients.

Figure 1: Boxplots of age-specific hs-cTnT values in the population-based (orange) and hospitalized (blue) cohorts in females and males. In both cohorts, hs-cTnT serum concentrations raise with increasing age. The hospitalized cohort is characterized by a wider interquartile range of hs-cTnT. The commonly used upper reference limit of the applied Elecsys® hs-cTnT assay (14 ng/l) is marked in black. Hs-cTnT: highly sensitive cardiac troponin T.

Figure 2: Hs-cTnT 99^th percentile values in relation to age in the population-based (orange curve) and hospitalized cohort (blue curve) divided by sex. The commonly used upper reference limit of the applied Elecsys® hs-cTnT assay (14 ng/l) is marked in black. Hs-cTnT: highly sensitive troponin T.