Clinical use of conventional reference intervals in the frail elderly.

RATIONALE, AIMS AND OBJECTIVES Reference intervals provided by the laboratory are commonly established by measuring samples from apparently healthy subjects in the ages 18-65 years, excluding elderly individuals with chronic diseases and medication. The aim of our study was to establish whether current reference intervals for immune parameters and chemical biomarkers are valid for older individuals including those with chronic diseases, so-called frail elderly. METHODS Data from our cohort of 138 non-infected nursing home residents (NHR), mean age 86.8 years, range 80-98, were compared with raw data, as basis for the development of reference intervals, obtained from reference populations, like blood donors (IgA, IgG, IgM, C3 and C4) and from the Nordic Reference Interval Project (NORIP) (alanine aminotransferase, albumin, aspartate aminotransferase, creatinine, gamma-glutamyl transferase, lactate dehydrogenase, phosphate, sodium and urea). Immune parameters were measured by nephelometry and in NORIP the measurements were performed by means of different routine methods, in more than 100 laboratories. RESULTS Only nine individuals (7%) of NHR were found to be free from chronic disease. C3, C4 (P < 0.001) and IgG levels (P < 0.05) were higher, while IgM levels (P < 0.001) were lower in NHR compared with reference blood donors. Levels of alanine aminotransferase, phosphate (P < 0.001), albumin (P < 0.05) and sodium (P < 0.01) were lower while creatinine and urea levels were higher (P < 0.001) in NHR compared with NORIP subjects. CONCLUSION Comparing laboratory results from elderly people with conventional reference intervals can be misleading or even dangerous, as normal conditions may appear pathological, or vice versa and thus lead to unnecessary or even harmful treatment.