Can Biomarker and Genetic Analysis Help Predict Alzheimer Disease?

migraine and dna concept
Researchers sought to evaluate the accuracy of biomarkers and to identify novel related genes to define Alzheimer disease.

A study published in Brain identified novel, plausible, genome-wide-significant genes associated with Alzheimer disease (AD) biomarkers.

AD affects tens of millions of people worldwide, however, at autopsy an estimated 25% of patients with AD do not show underlying pathology, indicating a potential inaccurate diagnosis. Despite difficulties with diagnosis, early detection is essential for combating the effects of the disease.

In order to evaluate the accuracy of biomarkers and to identify novel related genes, researchers at Cardiff University in the UK sourced data from the Alzheimer Disease Cardiff Cohort (ADCC). Plasma samples were collected between 2004 and 2020 from 1439 patients with early and late onset sporadic AD and 508 control individuals. A single nucleotide polymorphism (SNP)-based association analysis was performed for plasma biomarkers.

Plasma concentrations of 42 amino acid-long amyloid b fragments (Ab42) and 40 amino acid-long amyloid b fragments (Ab40) were correlated among both cases (r, 0.8) and controls (r, 0.7; P <1016).

Among cases, age of onset was strongly correlated with circulating levels of Ab40, Ab42, glial fibrillary acidic protein (GFAP), and neurofilament light (NfL; all P £4.2´1023), moderately correlated with tau phosphorylated at amino acid 181 (P-tau181; P =.0023), and negatively associated with the ratio of Ab42/Ab40 (P £4.8´104).

These biomarkers could predict disease status with an area under the curve (AUC) of 0.74. On their own, the prediction accuracies AUC for Ab42 and P-tau181 were 0.66 and 0.65, respectively.

The prediction model using the biomarkers combined with previously associated genes had an AUC of 0.81.

In the genetic analysis, relating genes with Ab40, Ab42, NfL, P-tau181, GFAP, and the ratio of Ab42/Ab40, the top 5 SNPs had a significance level of P £1´105 and 2 regions reached genome-wide significance.

The first genome-wide significant locus was on chromosome 7 and included the genes COPI coat complex subunit gamma 2 (COPG2) and testis specific 13 (TSGA13). This region has high linkage disequilibrium. The 2 lead SNPs in the region had high combined annotation-dependent deletion (CADD) scores (more than 12.37) which is indicative of being deleterious and one intergenic variant had a RegulomeDB (RDB) score (3a) indicating a functional impact on gene regulation.

The second genome-wide significant region was on chromosome 16 and contained the WW domain-containing oxidoreductase (WWOX) gene which has previously been linked with AD by genome-wide association study.

These findings should be confirmed among an independent dataset.

“The idea that biomarkers alone might provide more accurate prediction for Alzheimer’s disease remains to be fully validated,” the researchers concluded. Additional longitudinal study is needed to evaluate whether biomarkers alone may provide an accurate prediction of AD.

Disclosure: Multiple authors declared affiliations with industry. Please refer to the original article for a full list of disclosures.


Stevenson-Hoare J, Heslegrave A, Leonenko G, et al. Plasma biomarkers and genetics in the diagnosis and 1 prediction of Alzheimer’s disease. Brain. April 6, 2022 doi:10.1093/brain/awac128