Review-Key Symposium

The review addresses how genetic factors contribute to human longevity and what conserved pathways underlie extended healthy lifespan. It frames ageing as a major risk factor for chronic diseases and notes the discrepancy between increasing lifespan and stagnating healthspan. Definitions of lifespan, healthspan, and longevity vary and complicate study comparisons. The authors argue that studying exceptionally long-lived, generally disease-resilient individuals can reveal protective mechanisms. They posit longevity is heritable and propose leveraging human genetic studies together with insights from model organisms to identify conserved mechanisms that may be targeted to extend healthspan.

The review synthesizes multiple lines of evidence: (1) Heritability and phenotype definition: Family studies support heritability of longevity; lifespan heritability is estimated at 10–30%. Rigorous longevity definitions (e.g., top 10% of birth cohort with familial clustering) improve genetic signal and inspire scores such as the longevity relatives count and family longevity selection score. (2) Human genetics of longevity: Candidate gene studies consistently implicate APOE (ε2 enriched, ε4 depleted in long-lived) and FOXO3; pathway-based candidate analyses highlight IIS, mTOR, DNA-damage/repair, and telomere maintenance (e.g., POT1). Linkage studies identified few, non-replicated loci due to small samples. GWAS across European, East Asian, and Ashkenazi Jewish cohorts largely replicate APOE; other hits are inconsistent. Polygenic risk analyses show depletion of Alzheimer’s disease PGS in long-lived individuals but otherwise mixed results, suggesting protective variants may dominate. (3) Functional studies: Common intronic FOXO3 variants alter FOXO3 expression and stress responses; rare IGF1R coding variants reduce IIS signalling consistent with lifespan extension in model organisms; SIRT6 variants show functional impact though not enriched in long-lived cohorts. (4) Model organism genetics: Cross-species manipulations (C. elegans, Drosophila, mice) repeatedly link lifespan to IIS, cellular stress responses (e.g., p53/HIF), apoptosis, endothelin, and immune pathways. (5) Comparative genomics: Across mammals (including bats, whales, naked mole rats, elephants) convergence highlights DNA repair, immune function, IIS, telomere maintenance, and DNA-independent cell cycle regulation; genes under positive selection and pathway conservation inform candidate prioritization. Overall, convergent evidence narrows key pathways while emphasizing the likely role of rare variants in humans.

As a narrative review, the authors collate evidence from human and animal studies using predefined inclusion preferences: human genetic studies with well-defined longevity phenotypes and sizeable cohorts (n ≥ 200 long-lived individuals), results from genetic manipulations in multicellular model organisms (mice, Drosophila, C. elegans), and comparative genomic analyses across mammals with differing lifespans. For cross-species pathway synthesis, they collected longevity-associated genes for mouse, fruit fly, and nematode from GenAge (accessed May 4, 2022), then performed pathway enrichment using Gene Ontology with the PANTHER Pathways annotation. They compared significantly enriched pathways across the three organisms and distilled those shared among them. They summarized human candidate gene/pathway, linkage, GWAS, and polygenic score studies, and integrated functional validation studies (in vitro and in vivo) of both common and rare variants. Finally, they intersected genes implicated by at least two approaches to prioritize 18 candidate genes mapped to five overarching pathways.

• Human genetics: Only two loci/genes consistently associate with longevity across multiple populations and designs: APOE (ε2 enriched, ε4 depleted) and FOXO3. Many other reported loci lack replication due to phenotype heterogeneity, limited sample sizes, and control cohort challenges.
• Polygenic architecture: Lifespan/longevity likely influenced by many rare variants with small-to-moderate effects acting synergistically rather than a few common high-penetrance variants. Long-lived individuals show depleted PGS for Alzheimer’s disease, potentially explaining preserved cognition; they may still carry many pathogenic variants overall.
• Candidate pathways conserved across evidence streams: Five pathways consistently linked to longevity: (1) insulin/IGF-1 signalling (IIS), (2) DNA-damage response and repair, (3) immune function, (4) cholesterol metabolism (driven in humans by APOE), and (5) telomere maintenance (e.g., POT1). IIS emerged as the most prominent, with multiple implicated genes (AKT, FOXO, IGF, INSR, IRS, PIK3C, RPS6KB/S6K, SHC) and strong support from model organisms and interventions (dietary restriction, rapamycin).
• Functional validation: Common FOXO3 intronic variants (e.g., rs2802292, rs12206094, rs4946935) modulate FOXO3 expression and stress responsiveness; rare IGF1R coding variants (Ala67Thr, Arg437His) decrease IIS signalling (reduced AKT phosphorylation) and downstream transcriptional programs, aligning with longevity in model organisms. SIRT6 variants show functional effects on genome stability, though not enriched in long-lived cohorts.
• Model organisms and comparative genomics: Cross-species analyses confirm enrichment of pathways related to cellular stress (p53, HIF), IIS, endothelin and apoptosis signalling, and immune function. Comparative mammalian genomics repeatedly implicates DNA repair, immune pathways, IIS, telomere maintenance, and DNA-independent cell cycle regulation as correlates of extended species lifespan.
• Quantitative context: Lifespan heritability estimated at 10–30%. Despite a dramatic rise in the number of sequenced long-lived individuals, replicated human longevity loci beyond APOE and FOXO3 remain scarce, underscoring power and phenotype-definition limitations.

The synthesis indicates that robust human longevity associations converge on APOE and FOXO3, while conserved biology across species centers on IIS, genome maintenance, immune regulation, cholesterol metabolism, and telomere dynamics. This addresses the key question by narrowing the mechanistic landscape to a handful of pathways with cross-validated support. Given GWAS limitations (small effects, rare variant architecture, control cohort issues), focusing on rare, likely functional variants within these pathways and establishing causality through functional genomics is the most promising route for mechanistic insight and translational targets. Integrating human genetic discovery with model organism validation can illuminate how perturbations in IIS, DNA repair, immunity, lipid handling, or telomere maintenance modulate healthy ageing. The evidence also supports personalized approaches, as genetic background may influence responsiveness to environmental and pharmacological interventions that target these pathways.

This review consolidates human and cross-species evidence to prioritize five major pathways for human longevity: IIS, DNA-damage response and repair, immune function, cholesterol metabolism, and telomere maintenance, with the IIS pathway most strongly supported. In humans, only APOE and FOXO3 consistently associate with longevity, implying a complex architecture driven largely by rare variants. The authors advocate for future research emphasizing: (1) discovery of rare variants in genes within the prioritized pathways, (2) functional validation in vitro and in vivo (including CRISPR-based modelling), (3) improved longevity phenotype definitions and family-based enrichment strategies, (4) broader ancestral diversity, larger sample sizes, and longitudinal cohorts, and (5) multi-omics integration to map variant-to-function mechanisms. Such efforts could enable interventions that extend healthspan rather than merely lifespan.

• Phenotype heterogeneity: Inconsistent definitions of longevity and healthspan hinder comparability and power across studies.
• Sample size and controls: Limited numbers of long-lived cases and lack of ideal age-matched controls (same birth cohorts) reduce statistical power, especially for rare variants.
• Population diversity: Most human cohorts are European, East Asian, or Ashkenazi Jewish; findings may not generalize globally.
• Replication: Many reported loci lack replication across studies and ancestries.
• Rare variant detection: Standard GWAS are underpowered for rare variants; enrichment analyses are challenging due to variant rarity.
• Functional interpretation: Intronic and regulatory variants are difficult to model; many studies stop short of functional validation, limiting causal inference.
• Model organism biases: Non-mammalian findings may not fully translate to humans; rodent studies often use one sex/strain, limiting generalizability.
• Comparative genomics: While less hypothesis-driven, it may implicate broad pathways without pinpointing causal human variants; positive selection signals are complex to interpret.
• Limited multi-omics: Sparse availability of integrated transcriptomic, metabolomic, epigenetic data in the same human cohorts constrains mechanism mapping.