Hypothesis

Chronic mitochondrial reactive oxygen species (mtROS) release alters nuclear DNA methylation patterns through TET enzyme inhibition, thereby driving the acceleration of epigenetic clocks such as GrimAge and PhenoAge. Consequently, genetic or pharmacological enhancement of mitochondrial quality control (e.g., via UBA7‑mediated ubiquitination of damaged mitochondria) will causally slow epigenetic aging and extend lifespan.

Rationale

• Mendelian randomization (MR) shows no causal effect of leukocyte telomere length or epigenetic clocks on sex‑specific lifespan in UK Biobank (n≈500 k) [1]. This positions clocks as downstream readouts rather than upstream drivers.
• Two‑sample MR across 26 152 expression QTL instruments identified 47 gene‑aging links, with UBA7, PLA2G4B, and ATP8B4 downregulated with age in mice and validated as longevity GWAS hits [2]. UBA7 encodes an E2 ubiquitin‑conjugating enzyme implicated in mitophagy, suggesting a mitochondrial link.
• g‑methods (IPW, g‑computation, g‑estimation) struggle with model misspecification, censoring, and non‑collapsible outcomes [3], limiting their ability to disentangle time‑dependent confounding in aging observational data.
• Hybrid RMST estimators (AIPCW‑AIPTW, causal survival forests) offer better finite‑sample stability for survival outcomes but remain sensitive to misspecification [4]. These methodological constraints underscore the need for genetically anchored, mechanistic hypotheses that bypass complex time‑varying confounding.

Biologically, mtROS can oxidize Fe²⁺ in the TET catalytic domain, reducing 5‑hmC formation and leading to hypermethylation at CpG sites captured by epigenetic clocks. Improved mitophagy lowers mtROS, restores TET activity, and thereby decelerates clock‑associated methylation drift.

Testable Predictions

1. Genetic instrument test: Variants that increase UBA7 expression (or activity) will be associated with slower GrimAge acceleration in longitudinal cohorts, independent of telomere length and traditional risk factors.
2. Intervention test: Pharmacologic mitophagy inducers (e.g., urolithin A, spermidine) will reduce mtROS in murine leukocytes, increase TET‑mediated demethylation, and produce a measurable decrease in GrimAge‑predicted mortality risk over 12 months.
3. Mediation test: The effect of UBA7‑boosting alleles on lifespan will be mediated (≈30‑40 %) by changes in GrimAge, detectable via multivariable MR using GrimAge as an intermediate phenotype.
4. Falsification: If mtROS scavenging fails to alter epigenetic clock rates despite improving other aging phenotypes, the hypothesis is refuted.

Experimental Design

• Human MR: Use GWAS summary statistics for UBA7‑eQTLs as instruments, GrimAge change over time (e.g., from the Framingham Heart Study) as outcome, and adjust for platelet count, smoking, and BMI. Perform multivariable MR to exclude pleiotropic pathways via inflammation markers (IL‑6, CRP).
• Mouse Study: Cohorts of C57BL/6J mice receive lifelong urolithin A (50 mg/kg diet) vs. control. Measure mtROS (MitoSOX), TET activity, leukocyte GrimAge (using the mouse epigenetic clock), and survival. Sample size n=150 per arm provides 80 % power to detect a 15 % lifespan extension (α=0.05).
• Mediation Analysis: Apply causal mediation MR (two‑step) to estimate the proportion of UBA7‑lifespan effect explained by GrimAge change.

Potential Confounds and Mitigations

• Population stratification: Restrict MR to European ancestry and include principal components.
• Horizontal pleiotropy: Use MR‑Egger, weighted median, and MR‑PRESSO to detect and correct for pleiotropic instruments.
• Survival bias: Exclude prevalent cases of major disease at baseline and apply inverse‑probability‑of‑censoring weighting in longitudinal analyses.
• Off‑target drug effects: Include a second mitophagy inducer with distinct chemical structure (e.g., nicotinamide riboside) to corroborate findings.

If validated, this hypothesis would reposition mitochondrial quality control as a causal lever upstream of epigenetic aging, offering a concrete target for interventions that meaningfully modify biological age rather than merely reflecting it.