Reference interval and upper decission limit for serum uric acid – an evidence-based approach on Romanian population using an a posteriori method

Introduction. There is accumulating evidence that high normal serum uric acid (SUA) levels of 6-7 mg/dL are associated with cardiovascular morbidity and metabolic syndrome (MetS), hence the need to redefine its upper limit of normal (ULN). We aimed to derive ULN based on statistics and evidence in a representative sample of the population and to observe its relation to MetS components.

Methods. All SUA measurements form a university rheumatology hospital were extracted between January 5th 2010 and March 21st 2018. SUA levels were measured by a single biochemist a unique type of commercially available kit. Follow-up measurements, patients with diagnoses influencing SUA levels and outlying measurements were excluded. ULNs were studied using least square analysis.

Results. Of the 22503 SUA measurements in the database, only 3318 came from normal individuals: 33.3% men (n=1105), 66.7% women (n = 2213). Least square analysis revealed the following SUA reference intervals (RI): 3.43-6.19 mg/dL for the combined sample; 4.44-7.01 mg/dL for men, 3.28-5.56 mg/dL for women. The values corresponding to the 66th percentile of each group presented lower ULNs: 5.36 mg/dL for all, 6.10 mg/dL for men, 4.90 mg/dL for women. The prevalence of hyperuricemia increased from 13.8% (manufacturer’s gender-specific ULN) to 19.9% (derived ULN). Mean SUA levels significantly increased with the number of MetS components.

Conclusion. We recommend that hyperuricemia should be defined using a statistical approach of ULN selection corresponding to the genderand population-specific 66th percentile of data range.