AbstractBackgroundRecently, it was reported that as much as 45% of cognitively normal 70‐year olds from the general population had pathological CSF levels of either amyloid beta or tau, and the APOE e4 allele was significantly more common among these individuals. However, very few studies have investigated the effect of non‐APOE polygenic risk score (PRS) for Alzheimer’s disease (AD) on CSF biomarkers in representative population‐based samples. The aim of the present study was thus to investigate the relation between AD‐PRS and CSF biomarkers for AD and neurodegeneration in a representative sample of 70‐year‐olds.MethodA total of 303 individuals belonging to the Gothenburg H70 Birth Cohort Studies were included in the study. Genotyping was performed using the NeuroChip (Illumina) and imputation was done with the Sanger imputation service using the reference panel of Haplotype Reference Consortium data (HRC1.1). A PRS was calculated based on the allele counts of 23 AD risk‐variants multiplied with effect sizes (β‐coefficients) from IGAP. CSF levels of amyloid beta 42, total‐tau, phosphorylated tau, neurogranin and neurofilament light were measured with ELISA. Linear regression models, adjusted for sex and APOE e4 status, were used to analyze the association between AD‐PRS and CSF biomarkers.ResultAn association was found between the AD‐PRS and amyloid beta 42 (p = 0.003), showing a decreasing level of the biomarker with an increasing level of the PRS. No associations were found between the AD‐PRS and any of the other biomarkers. When only including cognitively healthy individuals (CDR = 0; n = 247), the association with amyloid beta 42 was still significant (p = 0.03). There was an interaction between the AD‐PRS and APOE e4 status (p = 0.04); the association between the PRS and amyloid beta 42 was present in e4 carriers only.ConclusionOur results show that genetic risk for AD beyond the APOE locus is associated with CSF‐amyloid beta 42, but not with any of the other investigated CSF‐biomarkers, among cognitively healthy 70‐year‐olds recruited from the general population. This finding is in line with the amyloid cascade hypothesis, showing that changed levels of CSF amyloid beta 42 is one of the earliest biomarkers related to AD.