Variants in the apolipoprotein E (APOE) region in addition to e2, e3, and e4 may be associated with Alzheimer disease (AD) risk. These results, from a genetic association study, were published in JAMA Network Open.
Researchers from the Alzheimer’s Disease Genetics Consortium (ADGC) used data available at their website from 18,795 individuals of European ancestry, either with AD or controls, to associate genetic variants with risk for AD.
Participants were mostly women (59.4%) affected by AD (51.6%) who were aged median 76 (interquartile range [IQR], 70-82) years at onset.
Within the APOE region, 14,415 single nucleotide variants (SNVs) were identified. Frequencies of e2 (3.5% vs 8.0%) and e4 (37.9% vs 13.7%) differed between cases and controls, respectively. A total of 71 individuals were homozygous for e2, 8848 for e3, and 1503 for e4.
Without adjusting, rs2075650 (odds ratio [OR], 2.59; 95% CI, 2.45-2.75; P =3.19 x 10-228) and rs4420638 (OR, 2.77; 95% CI, 2.62-2.94; P =2.99 x 10-254) were associated with AD. rs4420638 no longer remained significant after adjusting for APOE (P =.24), among e3 homozygotes (P =.18), or among e4 homozygotes (P =.66). rs2075650 remained marginally significant among e4 homozygotes (OR, 1.33; 95% CI, 1.00-1.77; P =.047) but not after adjusting for APOE (P =.07) or among e3 homozygotes (P =.09).
The investigators identified the 12 SNVs which had the strongest association for all the 3 corrective models (APOE adjustment and among e3 or e4 homozygotes). rs192879175 was the only variant which remained significant after multiple test correction for its associating with e3 homozygotes (OR, 0.50; 95% CI, 0.37-0.68; P =8.30 x 10-6). None of these most strongly associated variants were common among the ADGC cohort or in linkage with rs429358, rs7412, or rs117737673.
A previous analysis found evidence of an association between AD and rs2075650, which these results partially replicated. However, the previous work was not an entirely independent dataset from this current study.
This study was limited by its imputation. The investigators reported a 1.3% and 2.3% mismatch between observed and imputed genotypes for e2 and e4, respectively. These mismatched imputations were observed to be correlated with genotyping strategies, in which SNV-based genotyping was more highly correlated with e2 (r2, 0.81) and e4 (r2, 0.90) imputation and high-throughput sequencing was less correlated with the imputation of e2 (r2, 0.47) and e4 (r2, 0.78). This variance in imputation quality may have led to spurious associations between SNVs and AD.
These data indicated additional variants in the APOE region, beyond previously identified e2, e3, and e4 alleles, may be associated with risk for AD. Future studies using alternative approaches are needed to replicate these results.
Blue EE, Cheng A, Chen S, Yu CE. Association of uncommon, noncoding variants in the APOE region with risk of Alzheimer disease in adults of European ancestry. JAMA Netw Open. 2020;3(10):e2017666. doi: 10.1001/jamanetworkopen.2020.17666.