Design: general population cohort with biochemical verification in a subset and replication in a second cohort
Population: 19,499 adults with both cardiac troponin I and T measured, median age 49 years, 58.3% women
Primary outcome: myocardial infarction, ischaemic stroke or cardiovascular death
Effect: myocardial injury adjusted HR 2.90 (95% CI 1.99–4.22) for troponin I; suspected macrotroponin HR 1.59 (0.94–2.70)

Cardiac troponin is increasingly used for risk prediction rather than only for diagnosing infarction, and this study asks how often a raised value in the general population is an artefact. Among 19,499 adults with both troponin I and T measured, troponin I was raised in 1% and troponin T in 6%. Of those raised values, 48% and 57% respectively were discordant — the two assays differing more than threefold. Biochemical testing with immunoglobulin depletion and ultracentrifugation attributed 92.5% of discordant troponin I results to macrotroponin complexes, but only 15.4% of discordant troponin T results.

The risk data are what make this actionable. Participants with true myocardial injury had a clearly raised risk of infarction, ischaemic stroke or cardiovascular death — adjusted hazard ratio 2.90 (95% CI 1.99–4.22) for troponin I and 1.99 (1.63–2.42) for troponin T. Those with suspected macrotroponin did not: 1.59 (0.94–2.70) and 1.27 (1.00–1.61). The findings replicated in a second general-population cohort.

The practical consequence is in the outpatient clinic rather than the emergency department. A patient sent to you because a screening or incidental troponin came back raised, with no symptoms and no ischaemic story, may be carrying an antibody complex rather than subclinical disease. Where both assays are available, measuring the other one is cheap and settles the question far faster than serial imaging.

In an asymptomatic patient with an isolated raised troponin, ask which assay was used
Where the laboratory offers both, request the alternative troponin before escalating investigation
Treat a greater than threefold discordance between I and T as a flag for interference, particularly for troponin I
Do not apply this reasoning to a patient with an acute presentation and a rising trend
Tell the patient why the number is being repeated — an unexplained raised troponin causes real anxiety

Why it matters

A raised troponin is being read as subclinical disease in people whose risk is no higher than anyone else's.

The statistics, in plain English

The hazard ratios for the suspected macrotroponin groups have confidence intervals that touch or cross 1.0 — 0.94 to 2.70 for troponin I and 1.00 to 1.61 for troponin T. That means the data are compatible with no excess risk at all, in contrast to the myocardial injury group where the whole interval sits well above 1.0. The gap between 92.5% and 15.4% also matters: discordance is a much more reliable sign of interference for troponin I than for troponin T, so the same rule of thumb does not carry equal weight for both assays.