CSF Aβ42/Aβ40 Ratio Calculator
CSF Aβ42/Aβ40 Ratio Calculator
Divide CSF Aβ42 by Aβ40 and read it against the FDA-authorised Lumipulse cut-offs — the Core 1 amyloid biomarker, and far less platform-dependent than either peptide alone.
CSF Aβ42/Aβ40 ratio
Aβ42 ÷ Aβ40CSF Aβ42 500 pg/mL and Aβ40 7,000 pg/mL, both on Lumipulse
The ratio, and the two ways of taking it
Positive at or below 0.058 · likely positive 0.059 to 0.072 · negative at or above 0.073 (Lumipulse G)
molar ratio = mass ratio × 0.95919 — 4.08% lower
- a mass ratio
- both peptides are reported in pg/mL and divided directly, so every published threshold is a ratio of mass concentrations
- 0.95919
- 4,329.87 ÷ 4,514.1. Aβ42 is the heavier peptide, so the same mass is fewer moles and the molar ratio is lower. The published cut-offs do not apply to it
- same sample, same run
- a ratio across two assays or two draws reintroduces exactly the variation it exists to remove
Worked example
CSF Aβ42 500 pg/mL and Aβ40 7,000 pg/mL, both on Lumipulse
500 ÷ 7,000 = 0.0714
Between 0.059 and 0.072, so the authorised result is likely positive rather than positive
In molar terms: 0.0714 × 0.95919 = 0.0685 — 4.08% lower, and the published thresholds do not apply to it
The Aβ42 of 500 pg/mL is itself below Lumipulse's 526 ng/L cut-off and well below Elecsys's 834 pg/mL — which is why the ratio, not the peptide, is the transferable number
The authorised result has three tiers, not a threshold
| Ratio (Lumipulse G, mass) | Reported as | Agreement with amyloid PET |
|---|---|---|
| at or below 0.058 | Positive | 97% PET-positive; PPV 96.6% (95% CI 92.8–98.4%) |
| 0.059 to 0.072 | Likely positive | the band where the assay does not decide |
| at or above 0.073 | Negative | 84% PET-negative |
| 0.110 | a research cut-off | derived in a preclinical cohort on an earlier Lumipulse calibration |
Where this ratio sits in the 2024 criteria
| Biomarker | Tier | Platform-dependence |
|---|---|---|
| CSF Aβ42/Aβ40 ratio | Core 1 | moderate — the ratio cancels much of it |
| CSF p-tau181/Aβ42 ratio | Core 1 | moderate; Mayo’s Elecsys cut-off is 0.028 |
| CSF t-tau/Aβ42 ratio | Core 1 | moderate |
| Amyloid PET | Core 1 | the comparator the ratios are validated against |
| CSF Aβ42 alone | not listed | high — Elecsys 834 pg/mL against Lumipulse 526 ng/L |
Why a ratio, and why this one
The CSF amyloid measurement has a long-standing problem: the absolute Aβ42 concentration is not comparable between assays. Mayo’s Elecsys-based evaluation puts the amyloid-negative boundary above 834 pg/mL and the published Lumipulse validation puts it above 526 ng/L — the same unit, a 58% lower number — and within Lumipulse alone the December 2018 recalibration to certified reference material moved reported values by about 31.5%. On top of assay bias sits biological variation: how much amyloid precursor protein an individual processes varies severalfold, so a low Aβ42 can reflect low production rather than plaque, and a mid-range Aβ42 can conceal substantial deposition in a high producer.
Dividing by Aβ40 addresses all three at once. Aβ40 comes from the same precursor through the same secretases, so it carries the production term; it is measured on the same platform in the same run, so it carries most of the calibration bias; and it adsorbs to tube walls much as Aβ42 does, so it carries most of the pre-analytical loss. What survives the division reflects selective sequestration of Aβ42 into plaque. That is why the 2024 Alzheimer’s Association criteria list the ratio among the Core 1 biomarkers and do not list absolute Aβ42 at all, and why the first FDA-authorised in-vitro diagnostic for Alzheimer’s disease reports a ratio rather than a concentration.
Two cautions on the output. It has three tiers and not two — the band between 0.059 and 0.072 is the assay declining to decide — and its agreement with PET is asymmetric, 97% for positives against 84% for negatives, so a negative ratio is the weaker statement.
The third caution is arithmetical and is why the two converter pages exist beside this one. Everything above is a ratio of mass concentrations. Convert both peptides to pmol/L and re-divide and the answer falls by 4.08%, because Aβ42 is the heavier molecule — enough to carry a negative result into the indeterminate band. The molar figure is shown beside the result so the difference is visible rather than discovered.
Frequently asked questions
What is a normal CSF Aβ42/Aβ40 ratio?
On the FDA-authorised Lumipulse G assay, at or above 0.073 is reported as negative for amyloid pathology, 0.059 to 0.072 as likely positive, and at or below 0.058 as positive. Those cut-offs belong to that assay: a ratio from another platform must be read against that platform’s own threshold, and a research cut-off of 0.110 has been published from a preclinical cohort on an earlier calibration.
Why is the ratio better than Aβ42 alone?
Because it cancels three sources of noise at once — assay calibration bias, individual amyloid production and pre-analytical adsorption loss — all of which enter Aβ42 directly. The first alone is a 58% difference between the Elecsys and Lumipulse Aβ42 cut-offs. The 2024 criteria list the ratio as Core 1 and absolute Aβ42 not at all.
Should I use mass units or molar units for the ratio?
Mass units, in pg/mL, because that is what every published threshold rests on. The molar ratio is a legitimate quantity and is 4.08% lower, since Aβ42 is 4,514.1 Da against Aβ40’s 4,329.87 Da. Applying a mass-ratio cut-off to a molar ratio shifts the verdict in the amyloid-positive direction, and 4% is enough to cross the 0.073 boundary.
Does a positive ratio mean the patient has Alzheimer’s dementia?
No. It is evidence of amyloid pathology, which is present in cognitively normal people and becomes commoner with age. The 2024 criteria treat the biological and clinical axes separately: biomarkers establish the pathology, symptoms and cognitive testing establish the stage.
Can I take Aβ42 and Aβ40 from different reports?
No. The ratio works because the two peptides share a sample, a tube, a run and a calibration, and those shared factors are exactly what it cancels. Two platforms also add a bias of their own: one published comparison found Lumipulse reads about twice the MSD V-PLEX Aβ40 value for the same sample.
Related calculators
References
- Fujirebio Diagnostics. Lumipulse G beta-Amyloid Ratio (1-42/1-40) — cut-offs and clinical performance. Product information sheet, 2022.
- Jack CR Jr, Andrews JS, Beach TG, et al. Revised criteria for diagnosis and staging of Alzheimer’s disease: Alzheimer’s Association Workgroup. Alzheimers Dement. 2024. doi:10.1002/alz.13859.
- Lau K, Lin KW. Lumipulse G beta-amyloid ratio test for the diagnosis of Alzheimer disease. Am Fam Physician. 2023;107(5):550-551.
- Validation of the Lumipulse automated immunoassay for the measurement of core AD biomarkers in cerebrospinal fluid. Clin Chem Lab Med. 2021; doi:10.1515/cclm-2021-0651.
- Keshavan A, Wellington H, Chen Z, et al. Concordance of CSF measures of Alzheimer’s pathology with amyloid PET status in a preclinical cohort. Alzheimers Dement (Amst). 2020;12:e12097.
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/