The routine co-ordering of the erythrocyte sedimentation rate and C-reactive protein is an entrenched diagnostic habit, accounting for nearly half of inflammatory biomarker orders in health insurance registries. Despite widespread reliance on this paired workup, clinicians routinely encounter discordant results where one marker indicates active inflammation while the other remains quiescent. A comprehensive systematic review led by Xing Pan and colleagues, encompassing seventy published investigations, has evaluated the clinical and analytical factors underlying this divide across three operational layers: raw laboratory measurements, composite disease-activity scores, and diagnostic ratios. Rather than representing an inexplicable anomaly, the divergent behavior of these two markers reflects predictable physiological differences, shifting analytical definitions, and unstandardized laboratory methods that directly affect diagnostic categorization and patient management across rheumatic diseases.

The fundamental divergence between the erythrocyte sedimentation rate and C-reactive protein originates within their vastly distinct biological substrates and clearance kinetics. C-reactive protein is an acute-phase protein synthesized rapidly by hepatocytes under interleukin-6 stimulation, possessing a brief plasma half-life of approximately nineteen hours that enables it to mirror real-time fluctuations in systemic inflammatory activity. In contrast, the sedimentation rate quantifies the settling velocity of erythrocytes in plasma, an indirect physical phenomenon determined primarily by fibrinogen and immunoglobulin concentrations alongside red cell morphology and plasma viscosity. Because immunoglobulins and fibrinogen turn over across days to weeks, the sedimentation rate exhibits substantial physiological inertia, rising days after C-reactive protein during an acute insult and lingering long after active inflammation has resolved.

Across the published medical literature, reported frequencies of laboratory discordance span from four percent to thirty-four percent, while dual-normal findings encompass four percent to seventy-five percent of cohorts. The review establishes that these published rates cannot be pooled into a single statistical summary because every frequency depends on the analytical cut-offs selected and the specific patient population. Investigators have relied on incomparable operational frameworks, contrasting studies that anchor abnormality to standard physiological reference ranges against those defining discordance through internal sample distributions such as quartiles or tertiles. Percentile-based criteria inevitably generate mathematical discordance even within completely healthy cohorts devoid of inflammatory pathology, whereas fixed clinical thresholds identify instances where a single marker exceeds a diagnostic limit, demonstrating that reported rates reflect arbitrary boundary lines rather than immutable biological constants.

Empirical evidence directly demonstrates how sensitive discordance rates are to shifting mathematical criteria applied to identical clinical datasets. When researchers applied three distinct definitions to a single Romanian cohort comprising 44,427 same-day test pairs, the measured discordance frequency shifted from 6.4 percent under a tertile model to 17.6 percent using quartile boundaries, reaching 27.3 percent when evaluated against age-adjusted reference limits. Similarly, when investigators reanalyzed 748 rheumatoid arthritis observations after doubling standard laboratory thresholds, dual-normal results climbed from 37.3 percent to 64.8 percent. Furthermore, this threshold manipulation inverted the relative proportions of discordant classifications, transforming near-equivalent rates of isolated marker elevation into a greater than two-fold predominance of isolated C-reactive protein elevation, underscoring that altering numerical cut-offs fundamentally reshapes the apparent clinical picture.

At the composite index layer, the review reveals that substituting the erythrocyte sedimentation rate for C-reactive protein within disease assessment formulas introduces systematic evaluation errors. Across eighteen studies evaluating the 28-joint Disease Activity Score in rheumatoid arthritis, every investigation reporting directional trends observed that the sedimentation formulation generated consistently higher numeric activity scores than the protein-based variant. Categorical agreement statistics between the two index versions varied widely, spanning kappa values from 0.28 to 0.80, with this variability driven predominantly by whether clinicians applied sedimentation-based thresholds to protein calculations. When investigators employed score cut-offs calibrated specifically for the protein formulation, categorical concordance improved substantially, proving that much of the perceived disagreement between composite scores represents an artifact of misaligned threshold boundaries rather than true clinical incongruity between patient disease states.

The systematic elevation of sedimentation-based activity scores produces significant consequences for therapeutic eligibility in rheumatoid arthritis management. Because the protein-based score yields lower values, applying standard thresholds underestimates high disease activity while inappropriately inflating apparent remission, as demonstrated in Brazilian data showing 37.7 percent remission under the protein version versus 26.4 percent under the sedimentation index. Large registry analyses encompassing over 31,000 paired assessments show this numerical divergence is roughly twice as pronounced in female patients and individuals aged over fifty, reflecting the progressive physiological elevation of baseline sedimentation rates that occurs with advancing age. Consequently, unadjusted score substitutions risk excluding eligible patients from advanced biological therapies or falsely classifying active synovitis as remission, particularly among older women whose sedimentation rates reflect age-related shifts rather than joint inflammation.

Importantly, the directional discrepancies observed in rheumatoid arthritis indices do not generalize across other rheumatic disease assessment tools. In axial spondyloarthritis, evaluating the Ankylosing Spondylitis Disease Activity Score revealed that the sedimentation and protein variants disagreed in 22.9 percent to 27.2 percent of evaluations, yet demonstrated no consistent directional offset, yielding a negligible mean difference of negative 0.001. Conversely, when researchers assessed juvenile idiopathic arthritis using the 27-joint Juvenile Arthritis Disease Activity Score, the two index formulations achieved 91.1 percent categorical agreement and a kappa of 0.867, prompting authors to deem them clinically interchangeable. A similarly tight intraclass correlation of 0.99 appeared between index versions in polymyalgia rheumatica, confirming that index interchangeability cannot be deduced from biomarkers alone but must be determined independently for every composite scoring formula.

A crucial finding of the synthesis is that close agreement at the composite index level can obscure profound discordance between raw laboratory tests. In juvenile arthritis cohorts where composite scores achieved greater than ninety percent concordance, the correlation between raw sedimentation rates and protein concentrations remained poor at an r value of 0.335. Composite activity scores successfully buffer laboratory discordance because the biomarker represents only one of four scoring components, with joint counts and global assessments absorbing disparities across active cohorts. However, this protective buffering mechanism deteriorates as patients approach minimal disease activity, where index correlation dropped from 0.925 down to 0.496, demonstrating that index formulations diverge precisely when therapeutic decisions surrounding remission and drug withdrawal become most sensitive to minor numerical differences.

The most clinically urgent insight highlighted by the review concerns the phenomenon of dual-normal results occurring in patients with confirmed, active inflammatory disease. Across every investigated rheumatic condition, obtaining normal values across both laboratory tests fails to exclude active pathological processes. In a registry analysis of 9,135 rheumatoid arthritis patients exhibiting clinically active disease, 58.0 percent presented with normal levels for both markers, while primary presentation cohorts observed dual-normal readings in 33 percent to 42 percent of patients. Similarly, both markers were normal in 10.2 percent of biopsy-proven giant cell arteritis cases, dropping to 4.0 percent among patients unexposed to glucocorticoids, and in 14.8 percent of imaging-confirmed polymyalgia rheumatica cases, demonstrating that absent systemic acute-phase markers cannot be interpreted as diagnostic proof of quiescent disease.

This failure of systemic markers reflects a fundamental biological limitation rather than an analytical detection deficiency within modern laboratory instruments. True inflammatory tissue pathology can remain localized within synovium, arterial walls, or bursal sacs without generating an acute-phase response sufficient to elevate circulating markers above normal limits. Arthroscopic investigations confirmed active histological inflammation in 49.4 percent of rheumatoid arthritis patients whose C-reactive protein remained below five milligrams per liter, and in 71.4 percent of those classified in remission by composite scoring. Furthermore, in polymyalgia rheumatica presentations exhibiting normal conventional markers, serum amyloid A was universally elevated alongside sonographically confirmed bursitis, confirming that localized inflammation frequently proceeds without triggering systemic cascades, meaning that lowering reference thresholds would merely produce false positives rather than resolve this physiological limitation.

When true laboratory discordance occurs, the direction of marker separation provides valuable diagnostic clues regarding underlying patient pathology and co-existing clinical conditions. An isolated elevation of the erythrocyte sedimentation rate frequently indicates hypergammaglobulinemia associated with multiple myeloma, Waldenström macroglobulinaemia, or immunoglobulin G4-related disease, as well as chronic occult bone or prosthetic joint infections, which carried a fourteen-fold increased risk. Multivariable modeling has disproven the clinical assumption that anaemia independently causes sedimentation elevation, showing no significant association once adjusted for inflammatory status. Conversely, an isolated rise in C-reactive protein characteristically points toward hyperacute tissue injury, early bacterial infections of mucosal, pulmonary, or urinary origin, elevated body mass, or non-alcoholic fatty liver disease, which significantly suppresses sedimentation rates while leaving protein concentrations unaffected.

Technical and reporting deficits within the published literature further complicate clinical interpretation, as laboratory methodologies remain largely invisible to treating physicians. Forty of sixty-four evaluated studies failed to report their specific sedimentation method, despite quality surveys showing automated techniques deviate from the reference Westergren technique by up to 142 percent. Pre-analytical delays between blood collection and testing were documented in only a single investigation, and thirty-eight studies omitted protein assay specifications. Furthermore, therapeutic agents targeting the interleukin-6 receptor directly suppress hepatic protein synthesis and dampen sedimentation rates, rendering conventional biomarker readings uninformative regarding disease activity or superimposed infections. Because longitudinal data show that two-thirds of discordant results are transient and resolve on subsequent testing, clinicians should resist hasty management changes, verify testing parameters, and repeat ambiguous laboratory pairs in stable patients.

Subject of Research: Clinical interpretation and discordance patterns between erythrocyte sedimentation rate and C-reactive protein measurements in rheumatic diseases

Article Title: Discordance between the erythrocyte sedimentation rate and C-reactive protein in rheumatic diseases: reported frequencies, patterns, and clinical interpretation

Article References: Pan, X., Zhang, W., Liu, Y., Luo, T., Huang, S., & Li, X. (2026). Discordance between the erythrocyte sedimentation rate and C-reactive protein in rheumatic diseases: reported frequencies, patterns, and clinical interpretation. Frontiers in Medicine, 13, Article 2005901. https://doi.org/10.3389/fmed.2026.2005901

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.2005901

Keywords: erythrocyte sedimentation rate, C-reactive protein, rheumatic diseases, biomarker discordance, rheumatoid arthritis, giant cell arteritis, polymyalgia rheumatica, systemic lupus erythematosus, axial spondyloarthritis, juvenile idiopathic arthritis, DAS28-ESR, DAS28-CRP, ASDAS, JADAS27, Westergren technique, systemic inflammation, acute-phase response, interleukin-6 receptor inhibitors