Age- and sex-dependent reference intervals for D-dimer: evidence for a marked increase by age

A low D-dimer is commonly used to exclude venous thromboembolism in low risk patients. However, the reference intervals are poorly defined and D-dimer has been shown to increase by patient age. We aimed to establish age- and sex-dependent D-dimer reference intervals and to test the consequence of different cut-off limits.

By means of the STA-Lia test D-dimer was measured in 1,352 subjects aged between 23 and 93years from the general population. A further 94 subjects were measured with the HemosIL test.

The reference intervals were age-dependent with the upper 95% limit increasing from 0.92 FEU mg/L in the age group <50years to 2.39 FEU mg/L in the group ≥70years. Minor differences in the reference intervals between sexes were found. The commonly used cut-off at 0.5 FEU mg/L resulted in a decrease of negatives from 91% in <50years to 56 % in ≥70years. The age-dependent cut-off value (patient's age*0.01 FEU mg/L in 50+ years) resulted in an increase of negatives by age. The linear increase in D-dimer has a slope at approximately 0.0077*age, meaning that a cut-off at (patient's age*0.0077) FEU mg/L, gives an almost constant number of positives across age groups. The two assays showed a poor correlation and a higher percentage of positives using the STA Lia test.

The age-dependent increase in D-dimer seen in patients was confirmed, which questions the clinical value of the present cut-off for venous thromboembolism.