BackgroundGrowth associated protein-43 (GAP-43) and neurofilaments light (NFL) are quantitative indicators of synaptic and axonal injury, are pathological characteristics associated with cognitive decline in Alzheimer's disease (AD). We used Polygenic Hazard Score (PHS) to predict cognitive decline and its relationship with synaptic and axonal degeneration biomarkers. MethodWe enrolled 646 subjects: 93 with AD, 350 with mild cognitive impairment (MCI), and 203 controls. Variables included GAP-43, Plasma NFL, and PHS. 190 MCI with four years of follow-up (FU) cognitive assessments were applied for subgroup analysis of the relationship between baseline AD biomarkers and FU cognitive tests with and without considering the PHS cut-off. Finally, the impact of PHS on the incidence of MCI conversion to AD was assessed. ResultsWe introduced PHS greater than 0.21 as positive PHS and below this threshold as negative PHS. Contrary to MCI PHS- patients, GAP-43 correlated with all the cognitive tests in four years of FU among PHS+ ones (negative correlations with MMSE and MoCA, while positive ones with ADAS11, ADAS13, and ADASQ4). At FU time points, PHS+ MCI scored higher compared to PHS- ones in ADAS11, ADAS13, ADASQ4, and CDR-SB tests, while MMSE and MoCA assessments were lower in PHS+. Moreover, MCI patients with positive PHS progressed to dementia more rapidly compared with the group with negative PHS. ConclusionGenetic hazard scores are detrimental in understanding the relationship between neurodegenerative biomarkers and cognitive decline in AD. PHS can enhance the prediction of cognitive outcomes and disease progression.