Background: Test descriptions from major diagnostic manufacturers do not include ferritin reference intervals (RIs) for individuals aged 60 and older. The absence of older adults-specific RIs contrasts with the widespread use of serum ferritin testing in older adults. We aimed to establish and verify RIs using two common analytical methods. Methods: For this study, 1467 older adults were prospectively enrolled and monitored for morbidity and mortality, and exclusion criteria were applied. Ferritin was measured using chemiluminescent microparticle immunoassay (CMIA) and transferred to an electrochemiluminescence immunoassay (ECLIA) using method comparison. RIs were evaluated using a direct method with a prospective observational study based on healthy individuals according to the Clinical and Laboratory Standards Institute (CLSI) 28-A3c guideline and compared with RIs obtained using an indirect approach based on data obtained in clinical routine outpatients, where normal and abnormal values are supposed to be statistically differentiated to determine RIs. When applied within a countrywide population-based setting in Liechtenstein, the impact of novel RIs on the frequency of abnormal values was analyzed. Results: A total of 386 men and 532 women were included in the direct RI determination. Women (W) had significantly lower ferritin levels than men (M), while age over the age of 60 years had no significant association with ferritin in men and women. RIs were 23–241 ng/mL (W) and 19–396 ng/mL (M) for CMIA and 27–293 ng/mL (W) and 23–480 ng/mL (M) for ECLIA. These RIs are higher than those mentioned in the test descriptions in both tests. In comparison, the indirect method for both assays showed comparably lower reference limits, whereas upper reference limits were only approximately similar. The prevalence of high abnormal ferritin levels was considerably lower with this study’s RIs compared with manufacturer RIs. Conclusions: Employing older adults-specific RIs in clinical routine seems to be advisable. This reduces the frequency of abnormal high values in comparison with the widely applied practice of extrapolating RIs obtained from younger age groups to older adults and therefore leads to fewer follow-up investigations.
Ludin et al. (Mon,) studied this question.