CSF GFAP Rises Sharply With Age, New Reference Limits Show

Glial fibrillary acidic protein, or GFAP, has moved from a niche research marker to one of the most closely watched indicators of astrocytic activation and central nervous system injury. As laboratories increasingly measure GFAP in cerebrospinal fluid to investigate neurological and neurodegenerative disorders, a persistent problem remains: without reliable upper reference limits, it is difficult to judge whether a given concentration is normal or may represent a pathological elevation. A cross-sectional study from the Danish Pain Research Biobank, published in Fluids and Barriers of the CNS, now provides age-stratified reference values for CSF GFAP measured on the Single Molecule Array HD-X platform.

The researchers set out to establish biologically refined upper reference limits, or URLs, for CSF GFAP. They drew on CSF samples from the Danish Pain Research Biobank and applied strict inclusion criteria. Participants had no known neurological or neurodegenerative disease, and their samples showed no evidence of neuroinflammation or significant CSF abnormalities. This filtering step matters because reference limits derived from an unselected population can be inflated by subclinical disease, which would push the apparent upper limit upward and blur the boundary between health and pathology.

In total, 325 participants entered the reference population, with a median age of 50 years and 72.6 percent women. The cohort was divided into three adult age strata: 93 participants aged 18 to 39, 143 aged 40 to 59, and 89 aged 60 to 79. CSF GFAP concentrations were quantified with the Simoa HD-X platform. The researchers evaluated associations between GFAP concentrations and demographic and clinical variables using linear regression and restricted cubic spline analyses.

The results show a marked age gradient. Median CSF GFAP concentrations were 5,988 nanograms per liter in the youngest group, 8,406 nanograms per liter in the middle group, and 13,766 nanograms per liter in the oldest group, with interquartile ranges of 4,773 to 7,582, 6,867 to 11,083, and 10,877 to 18,127 nanograms per liter respectively. The trend across groups was statistically significant, with a p value below 0.001. In multivariable analyses, age was the only independent determinant among the demographic and clinical variables evaluated, again with p below 0.001.

Building on that age dependence, the investigators estimated upper reference limits non-parametrically as one-sided 95th percentiles, each with a 90 percent confidence interval. For adults aged 18 to 39, the URL was 10,580 nanograms per liter, with a confidence interval of 6,576 to 14,584. For those aged 40 to 59, the limit rose to 15,814 nanograms per liter, with a confidence interval of 12,734 to 18,893. For the 60 to 79 group, the URL reached 29,066 nanograms per liter, with a confidence interval of 24,748 to 33,385. The nearly threefold difference between the youngest and oldest strata underscores how much a single, age-blind cutoff would misclassify older adults.

The choice of a one-sided 95th percentile reflects standard practice in reference interval studies: it identifies the concentration below which 95 percent of the reference population falls, so values above it are candidates for consideration as pathological elevations. Reporting 90 percent confidence intervals around each estimate conveys the uncertainty inherent in percentile estimation from finite samples. The non-parametric estimation also avoids imposing a specific distributional shape on the data, which is useful when biomarker distributions are skewed.

The study’s conclusions are measured. The authors state that CSF GFAP concentrations increase markedly with age, supporting age-specific interpretation of measurements, and that the study provides age-specific URL estimates across three adult age strata, establishing a reference framework for future clinical and translational applications of CSF GFAP. They do not claim diagnostic performance for any specific disease, nor do they propose treatment thresholds. The reference limits are descriptive anchors for a screened population, not validated decision rules for diagnosing any condition.

Several limitations deserve emphasis when weighing these numbers. The reference population is cross-sectional, meaning each participant was sampled once, so the age gradient reflects differences between birth cohorts rather than tracked changes within individuals over time. The cohort is Danish, and demographic and clinical variables beyond those evaluated could influence GFAP in other populations. The oldest stratum tops out at 79 years, so the limits cannot simply be extrapolated to the very old. In addition, the participants were drawn from a pain research biobank, and although individuals with known neurological disease and CSF abnormalities were excluded, the population is not a random sample of the general population. The article is also being shared early as a peer-reviewed, accepted version that remains subject to further edits before the final Version of Record appears.

Even with those caveats, the practical implications are immediate. Any laboratory or research group interpreting Simoa-measured CSF GFAP now has a screened, age-stratified benchmark: roughly 10,580 nanograms per liter for adults under 40, 15,814 for those 40 to 59, and 29,066 for those 60 to 79, each with a quantified confidence interval. Applying a single cutoff across all adults would flag many healthy older individuals as abnormal and could mask true elevations in younger patients. As GFAP assays spread from specialized centers into broader clinical and translational use, age-specific reference limits of this kind are the groundwork that turns a raw concentration into an interpretable result.

Subject of Research: Age-specific upper reference limits for cerebrospinal fluid glial fibrillary acidic protein measured with the Simoa platform in a Danish biobank cohort

Article Title: Age-specific upper reference limits for Simoa-measured cerebrospinal fluid glial fibrillary acidic protein: a cross-sectional DANPAIN biobank study

Article References: Obel, L. M., Hunderup, M. W., Madsen, J. S., Petersen, E. R. B., Olsen, D. A., Wetche, J. L., Lambertsen, K. L., & Blichfeldt-Eckhardt, M. (2026). Age-specific upper reference limits for Simoa-measured cerebrospinal fluid glial fibrillary acidic protein: a cross-sectional DANPAIN biobank study. Fluids and Barriers of the CNS. https://doi.org/10.1186/s12987-026-00863-w

Image Credits: AI Generated

DOI: 10.1186/s12987-026-00863-w

Keywords: GFAP, cerebrospinal fluid, Simoa, upper reference limits, astrocytes, biomarker, aging, Danish Pain Research Biobank, neurodegeneration, reference intervals, cross-sectional study, central nervous system injury