Handling and transportation of CSF samples

Cerebrospinal fluid (CSF) can be collected in the lumbar region by an experienced physician. Taking into account the necessary precautions (e.g. sterile environment) and other medical investigations (e.g. brain imaging), the lumbar puncture is generally accepted by the medical society as a safe procedure.

Adsorption to plastic polymers can alter the biomarker concentrations and the extent of this adsorption differs between tubes. Therefore, always use the same polypropylene reference tube (within one lab).

Recommended procedure

  1. The CSF sample must be sent to the local laboratory without delay.

    • A CSF cell count is usually performed
    • The CSF sample is centrifuged in the original tube
  2.  The CSF sample is aliquoted for basic and specific neurodegenerative disease biomarkers

    • CSF should be aliquoted in tubes made of polypropylene
    • The volume needed for CSF analyses may vary between laboratories
    • The volume needed for Alzheimer specific biomarkers is between 0.5 mL and 1.0 mL
  3. Transportation: CSF samples can be sent by ordinary mail, at room temperature if the shipping time is less than two days. If the CSF sample is taken on a Friday, it can be frozen and sent to the central laboratory on dry ice the next week.
  4. Handling of CSF samples before analyses of AD biomarkers

    • It is recommended to freeze all CSF samples if the analysis cannot be performed within 48 hours. Samples sent frozen should be kept frozen until analysis.
    • CSF samples sent at room temperature should be frozen, and CSF samples sent frozen should be kept frozen.

Sample analysis on fresh samples upon arrival may result in slightly different CSF biomarker values compared to frozen samples but is unlikely to result in a different CSF biomarker profile.


CSF biochemical pattern interpretation

Alzheimer’s disease and pre-clinical stages

Alzheimer’s disease is the most common neurodegenerative disease and demonstrates exponential increase in prevalence with advancing age beyond 60 years. There are three stages usually described: latent stage when the disease has started but is asymptomatic, prodromal stage when the disease has progressed and very mild clinical signs and symptoms are present, and clinical stage when the disease has advanced and the full clinical spectrum is expressed.

It is critical to realize that latent disease cannot be distinguished from absence of disease by clinical examination or neuropsychological testing, but rather requires some ensemble of laboratory-based methods to detect disease initiation in the absence of symptoms.

How fast the patient will progress in the disease depends on risk-enhancing factors, such as age, modifying genes, cognitive reserve, comorbidities, and so forth.

Pathophysiological biomarkers

Individuals can now be identified as being in the preclinical state by the in vivo evidence of Alzheimer pathology (AP), by a biological or molecular “signature” of AD.

CSF Aβ42 and amyloid PET are highly concordant when used to dichotomize individuals as amyloid positive or amyloid-negative, showing 80%–90% agreement across studies. The CSF Aβ42/40 ratio typically shows agreement with PET above 90%. 

As Tau PET ligand is not routinely available, CSF T-tau and P-tau are the easiest tools for evidence of tauopathy. Alternatively, topographical markers include volume changes in the brain (hippocampal atrophy, cortical thickness) assessed by MRI and hypometabolism of neocortical regions measured by fluorodeoxyglucose (FDG)-PET..

Optimal and reliable blood-based biomarkers are not yet ready for clinical application.

Only the association of both pathologic hallmarks defines AD even in the absence of cognitive symptoms.

Definition of abnormality threshold for CSF biomarkers

The threshold for “abnormality” for CSF biomarkers is difficult to assess, especially in clinically healthy elderly subjects. There are several approaches to define what is abnormal.

  1. abnormality may be defined based on comparison between cognitively normal (having a CSF collection for any other causes than NDD) and AD groups.
  2. abnormality can be defined based on the distribution of values within a cognitively normal population, where subjects with values exceeding, for example, 2 standard deviations below or above the mean can be considered “abnormal”.
  3. abnormality can be defined based on longitudinal observation of clinical progression in a group starting as healthy and declining to AD at follow-up evaluations.

In healthy control subjects, the cortical uptake of Aβ agents is low in comparison with patients suffering from prodromal AD of the hippocampal type/MCI-due-to-AD(*) or fully developed AD dementia. However, a significant proportion of cognitively healthy elderly show increased cortical Aβ binding and decreased CSF Aβ42. This finding is supported by postmortem histopathological data showing Aβ plaques upwards of 30% of the non-demented elderly population above 75 years of age, likely representing preclinical AD. (*) MCI-due-to-AD, Mild cognitive impairment due to Alzheimer etiology.

CSF T-tau levels increased with age and were higher in apolipoprotein E (APOE) carriers. The APOE polymorphism is the most widely accepted genetic factor increasing the risk for sporadic AD. APOE ε4 carriers might be predisposed to vascular diseases which in turn could contribute to age-related brain damage and therefore to elevated T-tau levels.

In conclusion, a substantial number of healthy subjects over age 60 (25-40%) has at least one CSF biomarker concentration in range that can be considered abnormal. To minimize the age-related risk factor, the “normality” may be defined using results from clinically and cognitively normal individuals below the age of 50.

Remark:  Commercial assays for measurement of CSF biomarkers labeled CE-mark for in-vitro diagnostic propose estimated range of normal values for specific populations.

Combination of CSF biomarkers to be more useful in prediction

CSF T-tau, P-tau, and Aß42 are valuable as biomarkers of AD. At present, their strength relies mostly in supporting neurodegenerative etiology criteria for MCI and AD, and their reasonable capacity to predict the conversion from MCI to AD. A combination of biomarkers seems to be more useful in prediction than a single analyte.

The below table published by Mattson N. et al (2012) summarizes the specificities and likelihood ratios at cutoffs for 85% sensitivity for AD dementia according age categories. The specificity of CSF biomarkers decreases with age, as an effect of the high AD prevalence in older ages, but the likelihood ratios are improved when CSF biomarkers are combined.


The CSF biomarkers in combination, eg. low CSF Aβ42 peptide with high total tau and phosphorylated tau, are sensitive and specific biomarkers highly predictive of progression to AD dementia in patients with mild cognitive impairment and of presence of AD etiology even in older populations.


CSF markers for AD risk stratification and predictive value

Model 1: AD risk adapted from Hansson O et al, Lancet Neurol 2006

Association between CSF biomarkers and incipient AD – Monocentric longitudinal study by Hansson O, et al (2006). The follow-up period was extended (5-10 years) published by Buchhave P, et al (2012).

95.5 % of patients with MCI and abnormal CSF converted to AD.


Model 2: AD risk adapted from Lewczuk P, et al. J Neural Transm. 2009; J Alzheimers Dis. 2015; Somers C, et al., J Alzheimers Dis. 2019

The Erlangen score validated using two cohorts of pre-dementia subjects, the German Dementia Competence Network (n = 190 subjects with MCI) and the US Alzheimer's Disease Neuroimaging Initiative 1 (n = 292 MCI or cognitively normal subjects). The Erlangen score uses a risk graph approach.

The CSF results of a given patient are scored between 0 and 4 points. A CSF result with all biomarkers entirely normal is scored 0 points; a pattern with only marginal alterations in one biomarkers group (either Aβ or Tau, but not both) results in the score of 1; a CSF result with the alterations in either Aβ metabolism (decreased Aβ42 concentration and/or decreased Aβ42/40 ratio) or Tau metabolism (increased concentrations of T-tau and/or P-Tau) but not both is scored 2 points; a result with clear alterations in one biomarkers’ group (either Aβ or Tau) accompanied by marginal alterations in the other group is scored 3 points; clear alterations in both Aβ and T-tau/P-Tau result in 4 points.


Model 3: AD risk adapted from Lehmann S, et al. Alzheimer's Research & Therapy 2014; Front Aging Neurosci. 2018

The scale’s overall predictive value for AD for the different categories (N= 1,273 patients included 646 AD and 627 non-AD) from six independent memory-clinic cohorts.

AD risk assessment may integrate the Aβ42/40 ratio (instead of Aβ42) which accounts for interindividual difference in amyloidogenic APP-processing.

These simple scales using the presence of two or three pathologic biomarkers as a criterion of AD can be used to facilitate the interpretation of CSF pattern in routine.

Remark: The two last illustrations of risk scoring are cutoff values independent, meaning each laboratory can easily supplement it with the cutoff values and normal/abnormal ranges according to analytical method used for biomarker measurement.


CSF biosignature: a dynamic of neuropathologic changes - Not a standalone diagnostic

The combination of CSF biomarkers permits a diagnosis of AD in earlier stages of the disease. Nevertheless, the clinical identification of cognitive impairment and the use of both structural (CT/MRI) and functional (SPECT/PET) brain imaging are necessary for an accurate differential diagnosis with other neurodegenerative diseases. Mixed pathology, especially in elderly subjects, is frequent.


  • Amyloid-β PET—Correlation with cerebrospinal fluid biomarkers and prediction of Alzheimer’s disease diagnosis in a memory clinic. Müller EG, et al. PLoS One. 2019; 14(8): e0221365. Observational Study.
  • Preclinical Alzheimer’s disease: Definition, natural history, and diagnostic criteria. Dubois B, et al. Proceedings of the Meeting of the International Working Group (IWG) and the American Alzheimer’s Association on “The Preclinical State of AD”; July 23, 2015; Washington DC, USA. Alzheimers Dement. 2016; 12(3): 292-323. Review.
  • Clinical utility of cerebrospinal fluid biomarkers in the diagnosis of early Alzheimer’s disease. Blennow K, et al. Alzheimers Dement. 2015; 11(1): 58-69. Review.
  • Rethinking on the concept of biomarkers in preclinical Alzheimer’s disease. Berti V, et al. Neurol Sci. 2016; 37(5): 663-672. Review.
  • Interpreting Biomarker Results in Individual Patients with Mild Cognitive Impairment in the Alzheimer’s Biomarkers in Daily Practice (ABIDE) Project. van Maurik IS, et al. Alzheimer’s Disease Neuroimaging Initiative. JAMA Neurol. 2017; 74(12): 1481-1491.
  • The effects of normal aging and ApoE genotype on the levels of CSF biomarkers for Alzheimer’s disease. Glodzik-Sobanska L, et al. Neurobiol Aging. 2009; 30(5): 672-681.
  • Cerebrospinal fluid biomarkers of Alzheimer’s disease in cognitively healthy elderly. Randall C, et al. Front Biosci (Landmark Ed). 2013; 18: 1150-1173. Review.
  • Increased cerebrospinal fluid F2-isoprostanes are associated with aging and latent Alzheimer’s disease as identified by biomarkers. Montine TJ, et al. Neuromolecular Med. 2011; 13(1): 37-43.
  • Cerebrospinal fluid markers for Alzheimer’s disease in a cognitively healthy cohort of young and old adults. Paternicò D, et al. Alzheimers Dement. 2012; 8(6): 520-527. Comparative
  • Age and diagnostic performance of Alzheimer disease CSF biomarkers. Mattsson N, et al. Neurology. 2012; 78(7): 468-476.
  • Prevalence of cerebral amyloid pathology in persons without dementia: a meta-analysis. Jansen WJ, et al. JAMA. 2015; 313(19): 1924-1938. Meta-Analysis.
  • Cerebrospinal fluid biomarkers and prediction of conversion in patients with mild cognitive impairment: 4-year follow-up in a routine clinical setting. Lanari A, et al. Scientific World Journal. 2009; 9: 961-966.
  • Association between CSF biomarkers and incipient Alzheimer’s disease in patients with mild cognitive impairment: a follow-up study. Hansson O, et al. Lancet Neurol. 2006; 5(3): 228-234. Comparative Study.
  • Cerebrospinal fluid levels of β-amyloid 1-42, but not of tau, are fully changed already 5 to 10 years before the onset of Alzheimer dementia. Buchhave P, et al. Arch Gen Psychiatry. 2012; 69(1): 98-106. Comparative Study.
  • A diagnostic scale for Alzheimer’s disease based on cerebrospinal fluid biomarker profiles. Lehmann S, et al. Alzheimers Res Ther. 2014; 6(3): 38.
  • Relevance of Aβ42/40 Ratio for Detection of Alzheimer Disease Pathology in Clinical Routine: The PLMR Scale. Lehmann S, et al. Front Aging Neurosci. 2018; 28(10):138
  • Neurochemical dementia diagnostics: a simple algorithm for interpretation of the CSF biomarkers. Lewczuk P, et al. J Neural Transm (Vienna). 2009; 116(9): 1163-1167.
  • Validation of the Erlangen Score Algorithm for the Prediction of the Development of Dementia due to Alzheimer’s Disease in Pre-Dementia Subjects. Lewczuk P, et al. J Alzheimers Dis. 2015; 48(2): 433-441.
  • Validation of the Erlangen Score Algorithm for Differential Dementia Diagnosis in Autopsy-Confirmed Subjects. Somers C, et al. J Alzheimers Dis. 2019; 68(3): 1151-1159.

Added value of Aβ42/40 ratio

Aβ peptide is produced from a transmembrane Aβ precursor protein, sequentially cleaved by β- and γ-secretase. Cleavage of APP by γ-secretase generates a number of Aβ isoforms. Aβ42, a 42 amino acid-long peptide, has the highest propensity for aggregation and appears to be the predominant species in neuritic plaques. Although the concentration of Aβ40, has been reported to be unaltered in AD, the Aβ42/40 ratio has been suggested to be superior to the concentration of Aβ42 alone in discriminating patients with AD.

In a population of normal subjects and AD patients, the distribution of total Aβ (40 and 42) follows a Gaussian distribution for both normal subjects and AD patients, with Aβ40 making up about 70% of total Aβ. Further although many cases fall into the middle of the distribution with the majority having normally total Aβ, outliers are still present. Some AD patients will have high total Aβ (enhanced amyloidogenic processing of APP, also called “high producers” and some cognitively normal subjects will have low total Aβ (reduced amyloidogenic) processing of APP, also called “low producers”.

This means that AD patients with a high total Aβ will show an incomplete CSF pattern and vice versa, normal subjects with low total Aβ will be classify as individual with sign of cerebral amyloidosis. In all three cases (normal, low and high total Aβ) the ratio can correctly classify some doubtful CSF pattern. The ratio led to a reduction by half of the number of indeterminate profi les without changing the conclusion when usual biomarkers (Aβ42 and P-tau) were concordant.

There is a consensus to recognize that the Aβ42/40 ratio is helpful to:

  • reflect interindividual difference in amyloidogenic-APP processing
  • solve undetermined core biomarker profiles of AD
  • decrease the impact of preanalytical and analytical sources of variability within and among centers

Lewczuk P, et al., J Alzheimers Dis, 2015:

Added value of Aβ40 alone

CAA diagnosis based on MRI findings alone is not clear. Amyloid imaging with amyloid-binding PET ligands can detect CAA, although they cannot discriminate vascular from parenchymal amyloid deposits. In addition, CSF markers may be useful, including levels of Aβ40 for CAA and anti-Aβ antibody for CAA-related inflammation (CAA-ri). Recent findings indicate that the presence of one or more biomarkers plus one or more risk factors may be suggestive of CAA:

  • Amyloid imaging with greater occipital uptake
  • A decrease in CSF Aβ40 levels

Risk factors:

  • General factors:

    • Old age
    • AD
  • Genetic factors:

    • CAA-related gene mutations in familial cases
    • APOE in sporadic cases: ε4 as a risk factor for CAA

Impact of carrying the APOE4 allele

The brain Aβ pathology is inarguably associated with APOE ε4 status. Carrying the ε4 allele of the APOE gene encoding apolipoprotein E (APOE ε4) markedly increases the risk for AD and CAA. APOE ε4-mediated amyloid pathology depends on its neuronal LDL receptor–related protein 1 (LRP1). APOE ε4 decreases Aβ clearance without affecting Aβ production. According to the current concept, Aβ that accumulate in the brain in AD is likely due to its faulty clearance from the brain. LRP1 is a major efflux transporter for Aβ at the blood-brain barrier (BBB). Binding of Aβ to LRP1 at the abluminal side of the BBB initiates a rapid Aβ clearance from brain to blood via transcytosis across the BBB.

In summary, cognitive impairment in the ageing brain is typically driven by overlapping neurodegenerative and cerebrovascular pathologies. The impaired perivascular clearance of Aβ and the deficient neuronal LRP1 exacerbate the brain accumulation of Aβ peptides and subsequent deposition – the most likely cause of CAA and AD.



  • APOE4-mediated amyloid-beta pathology depends on its neuronal receptor LRP1. Tachibana M, et al. J Clin Invest. 2019; 129(3): 1272-1277.
  • Core cerebrospinal fluid biomarker profile in cerebral amyloid angiopathy: A meta-analysis. Charidimou A, et al. Neurology. 2018; 90(9): e754-e762. Review.
  • Cerebral amyloid angiopathy and Alzheimer disease - one peptide, two pathways. Greenberg SM, et al. Nat Rev Neurol. 2020; 16(1): 30-42. Review.
  • Cerebral amyloid angiopathy: emerging concepts. Yamada M. J Stroke. 2015; 17(1): 17-30. Review.
  • Cerebrospinal Fluid Aβ42/40 Corresponds Better than Aβ42 to Amyloid PET in Alzheimer’s Disease. Lewczuk P, et al. J Alzheimers Dis. 2017; 55(2): 813-822.
  • Cerebrospinal fluid amyloid-β 42/40 ratio in clinical setting of memory centers: a multicentric study. Dumurgier J, et al. Alzheimers Res Ther. 2015; 7(1): 30.
  • Additional use of Aβ42/Aβ40 ratio with cerebrospinal fluid biomarkers P-tau and Aβ42 increases the level of evidence of Alzheimer’s disease pathophysiological process in routine practice. Sauvée M, et al. J Alzheimers Dis. 2014; 41(2): 377-386. Comparative Study.
  • Amyloid beta peptide ratio 42/40 but not A beta 42 correlates with phospho-Tau in patients with low- and high-CSF A beta 40 load. Wiltfang J, et al. J Neurochem. 2007; 101(4): 1053-1059. Comparative Study.