Adult CSF total protein upper reference limits should be age-partitioned and significantly higher than 0.45 g/L: a systematic review

Background: Cerebrospinal fluid total protein (CSF-TP) is measured in the diagnosis of a range of immune or infectious disorders of the nervous system. Most laboratories and the medical literature use an antiquated, age-independent upper limit of 0.45 g/L. Therefore, we performed a systematic review of reference studies in the medical literature, with the primary objective of determining the CSF total protein upper reference limit (URL). Secondary objectives were to assess the effects of age, gender, laboratory methods, and methodological quality.

Methods: A pre-planned and peer-reviewed electronic search strategy was used to search Ovid Medline and EMBASE for 1960-2017. All records underwent title/abstract review, and potentially relevant records underwent independent full-text review by two researchers. The remaining studies underwent quality assessment using a modification of the QUADAS2 revised tool. CSF-TP upper reference limits extracted from these studies were used to compute weighted means.

Results: Twenty-two articles were retained for qualitative analysis and 20 for quantitative analysis. The weighted average of CSF-TP URL was 0.55 g/L, in studies with high methodological quality. Studies that examined the effect of age reported consistent correlations with advancing age, and CSF-TP URL values incrementally exceeded 0.60 g/L after age 50. There were no meaningful differences according to gender, laboratory method, or quality assessment score.

Conclusions: There is concordance in available literature to recommend increasing CSF total protein upper reference limits, and to consider implementing age-adjusted values above 0.60 g/L starting at age 50. This information merits worldwide dissemination, to reduce the risk of over-diagnosis.