INTRODUCTION: Plasma biomarkers have

demonstrated clinical

utility in the differential

diagnosis of Alzheimer’s

disease and are of

increasing importance in

other neurocognitive

disorders.

• AIM: to study the effect of

metabolic comorbidities

on plasma biomarker

concentrations in a

transdiagnostic research

cohort.

METHODS: 

N=376 participants

recruited from the Oxford

Memory Clinic and the FAST

study (AD=90, LBD=16,

CBS=7, FTD=14, MCI=24,

SCD=36, controls=189).

• p-tau217, Aβ42/40 ratio,

GFAP, NFL, PRDX6 and

ACHE measured using the

Alamar NULISA CNS Panel.

• Hypertension, diabetes,

hyperlipidemia, chronic

kidney disease and

concurrent medication

coded from medical

records. BMI, blood

pressure, HbA1c, serum

creatinine and HDL

cholesterol were available

or imputed in a subset of

participants.

• Linear regression used to

study associations between

biomarkers and

comorbidities/clinical

parameters, controlling for

age, sex and education.

Sensitivity analyses

controlled for BMI and

concurrent medication

respectively.

RESULTS: 

Kidney disease and serum

Cr associated with ↑p-

tau217, while serum

creatinine only associated

with ↑ACHE (Figure 1).

• Hyperlipidemia inversely

associated with NFL and

ACHE, but sensitivity

analyses including BMI

these were attenuated [NFL

(βHyperlipidemia+=-0.31,p=0.20),

ACHE (βHyperlipidemia+=-

0.064,p=0.84)].

• ↑ HDLC inversely associated

with PRDX6. This remained

significant when adjusting

for BMI (βHDLC=-0.31,p=0.20)

but attenuated after

excluding participants on

concurrent statin therapy

(βHDLC=-0.051,p=0.13).

• Diabetes and HbA1c

inversely associated with

GFAP; this association

remained significant when

the model adjusted for BMI

(βHbA1c=-0.097,p=0.004) but

attenuated after excluding

participants with glucose-

lowering therapy (βHbA1c=-

0.053,p=0.08). CONCLUSIONS: 

 Kidney impairment was associated with ↑ p-tau217 and ACHE,

consistent with the effect of impaired renal clearance.

• Paradoxically, inverse relationships between neurodegenerative

biomarkers and dyslipidemia, and between GFAP and diabetes were

seen. Confounding by higher BMI (leading to a dilutional effect on

plasma biomarkers) or medication use may account for this.