Skip to main content
. 2020 Sep;93:69–77. doi: 10.1016/j.neurobiolaging.2020.04.024

Table 2.

Association results between parental AD and polygenic risk scores in the UK Biobank data

Phenotype Age N APOE (ε2 + ε4)
PRS (excl. APOE region) (pT ≤ 0.5)
APOE + PRS (pT ≤ 0.5)
β SE p AUC β SE p AUC p AUC
Parental AD All BP = 1554 OP = 38,417 NP = 284,110 0.242 0.004 <1 × 10−350 0.576 0.031 0.005 2.4 × 10−10 0.501 <1 × 10−350 0.573
<80 BP = 326 OP = 10,755
NP = 213,877
0.348 0.009 <1 × 10−350 0.591 0.037 0.009 0.00008 0.511 <1 × 10−350 0.598
≥80 BP = 1228 OP = 27,662
NP = 70,233
0.208 0.005 <1 × 10−350 0.569 0.030 0.006 1.7 × 10−7 0.510 6.0 × 10−314 0.576

Poisson regression was performed. The models were adjusted for PCs, parental age and participants' sex. Genetic variants with p-value pT ≤ 0.5 from Kunkle et al. (2019) summary statistics were used to calculate the genetic score. The parental age was either the age at death or the last recorded age if alive.

Key: N, sample size; β, beta coefficient (effect size); AD, Alzheimer's disease; APOE, apolipoprotein E; AUC, area under the curve; BP, both parents have AD; excl., excluding; NP, neither parent has AD; OP, one parent has AD; p, p-value; pT,p-value with threshold T ≤ 0.5; PRS, polygenic risk score; SE, standard error.