Impact of cardiovascular risk factors on plasma biomarkers in prediction of Alzheimer's and cerebrovascular neuropathology.

backgroundPlasma biomarkers for Alzheimer's disease and neurodegeneration have shown accurate prediction of underlying neuropathology. However, chronic cardiovascular risk factors such as diabetes and hypertension are associated with plasma biomarker levels and can influence the accurate prediction of underlying neuropathologic changes.

objectiveTo understand the interaction between plasma biomarkers of Alzheimer's disease and neurodegeneration with cardiovascular risk factors in relation to neuropathologic change in a heterogenous population to ascertain a more accurate utilization of these biomarkers.

designRetrospective, case-control study.

settingPopulation-based, Olmstead county, Minnesota, USA.

participantsThree-hundred and fifty-one participants (aged 87.4 ± 7.5 years) with brain autopsy and antemortem plasma biomarker testing. MEASUREMENTS: Plasma biomarker testing for Aβ42/40, p-tau181, GFAP, and NfL using Quanterix Simoa assays. Cardiovascular risk factors were quantified by a composite score of cardiovascular metabolic conditions (CMC) consisting of a binary history of diabetes, congestive heart failure, stroke, coronary artery disease, atrial fibrillation, hypertension, or dyslipidemia. Plasma biomarkers and cardiovascular metabolic conditions score were Z-scored and neuropathologic scales were binarized into high and low categories. Outcomes included elevated microvascular (Kalaria) and macrovascular (Strozyk) neuropathologic scales as well as Alzheimer's disease neuropathologic change (ADNC), Thal phase, Braak stage, and neuritic plaque score. Multivariate logistic regression models incorporated interaction terms between plasma biomarkers and CMC while controlling for age, sex, cognitive impairment, and BMI.

resultsWe observed that at higher cardiovascular metabolic conditions score, the association between GFAP and overall ADNC (OR = 0.61 [0.42, 0.89]), Thal phase (OR = 0.48 [0.33, 0.71]), and Braak Stage (OR = 0.56 [0.37, 0.84]), became weaker, while the association with Strozyk score (OR = 1.65 [1.11, 2.46]) was stronger with higher CMC. Meanwhile, at higher CMC Aβ42/40 became more strongly negative with high Braak stage (OR = 0.63 [0.47, 0.85]), neuritic plaque score (OR 0.70 [0.52, 0.95]), Kalaria score (OR = 0.71 [0.57, 0.88]), and Strozyk score (OR = 0.60 [0.43, 0.83]). The association between p-tau181 and Thal phase (OR = 1.43 [1.00, 2.04]) was stronger at higher CMC while the association between p-tau181 and Strozyk score (OR = 0.47 [0.31, 0.71]) was weaker at higher CMC. There was no interaction between NfL and CMC score for any metric of neuropathologic change.

conclusionUnderstanding how cardiovascular risk factors can modulate plasma biomarkers is important for their interpretation with respect to underlying pathology and their clinical application in screening, diagnosis, and prognosis of neurodegenerative diseases.