Long-Lived Individuals Show a Lower Burden of Variants Predisposing to Age-Related Diseases and a Higher Polygenic Longevity Score.

Guillermo G Torres, Janina Dose, Tim P Hasenbein, Marianne Nygaard, Ben Krause-Kyora, Jonas Mengel-From, Kaare Christensen, Karen Andersen-Ranberg, Daniel Kolbe, Wolfgang Lieb, Matthias Laudes, Siegfried Görg, Stefan Schreiber, Andre Franke, Amke Caliebe, Gregor Kuhlenbäumer, Almut Nebel
Year of publication:
Journal title abbreviated:
Journal title long:
International journal of molecular sciences
Impact factor:
Longevity is a complex phenotype influenced by both environmental and genetic factors. The genetic contribution is estimated at about 25%. Despite extensive research efforts, only a few longevity genes have been validated across populations. Long-lived individuals (LLI) reach extreme ages with a relative low prevalence of chronic disability and major age-related diseases (ARDs). We tested whether the protection from ARDs in LLI can partly be attributed to genetic factors by calculating polygenic risk scores (PRSs) for seven common late-life diseases (Alzheimer's disease (AD), atrial fibrillation (AF), coronary artery disease (CAD), colorectal cancer (CRC), ischemic stroke (ISS), Parkinson's disease (PD) and type 2 diabetes (T2D)). The examined sample comprised 1351 German LLI (≥94 years, including 643 centenarians) and 4680 German younger controls. For all ARD-PRSs tested, the LLI had significantly lower scores than the younger control individuals (areas under the curve (AUCs): ISS = 0.59, <i>p</i> = 2.84 × 10<sup>-35</sup>; AD = 0.59, <i>p</i> = 3.16 × 10<sup>-25</sup>; AF = 0.57, <i>p</i> = 1.07 × 10<sup>-16</sup>; CAD = 0.56, <i>p</i> = 1.88 × 10<sup>-12</sup>; CRC = 0.52, <i>p</i> = 5.85 × 10<sup>-3</sup>; PD = 0.52, <i>p</i> = 1.91 × 10<sup>-3</sup>; T2D = 0.51, <i>p</i> = 2.61 × 10<sup>-3</sup>). We combined the individual ARD-PRSs into a meta-PRS (AUC = 0.64, <i>p</i> = 6.45 × 10<sup>-15</sup>). We also generated two genome-wide polygenic scores for longevity, one with and one without the <i>TOMM40</i>/<i>APOE</i>/<i>APOC1</i> gene region (AUC (incl. <i>TOMM40</i>/<i>APOE</i>/<i>APOC1</i>) = 0.56, <i>p</i> = 1.45 × 10<sup>-5</sup>, seven variants; AUC (excl. <i>TOMM40</i>/<i>APOE</i>/<i>APOC1</i>) = 0.55, <i>p</i> = 9.85 × 10<sup>-3</sup>, 10,361 variants). Furthermore, the inclusion of nine markers from the excluded region (not in LD with each other) plus the <i>APOE</i> haplotype into the model raised the AUC from 0.55 to 0.61. Thus, our results highlight the importance of <i>TOMM40</i>/<i>APOE</i>/<i>APOC1</i> as a longevity hub.