Omitted familial extrinsic risk inflates inferred intrinsic lifespan heritability

Shenhar et al. (2026) report 50% “intrinsic” lifespan heritability after calibrating a one-component correlated-frailty survival model to Scandinavian twin lifespans. Their framework is mathematically coherent, but the intrinsic component is not identified if heritable, mortality-relevant extrinsic susceptibility is omitted at calibration. We show that one-component calibration absorbs omitted familial extrinsic structure into the intrinsic frailty scale parameter σ _θ , and that this variance absorption is visible through separate diagnostics (1) Variance absorption . Under misspecification, σ _θ is inflated by +22.1% (95% CI: 21.5–22.7%), corresponding to +49% inflation in . Falconer h ² is downstream of calibration and inherits a +9.2 pp bias (95% CI: 8.7–9.7). The σ _θ inflation is model-general: +22% (GM), +18% (MGG), +14% (SR); any dependence summary that is strictly increasing in σ _θ inherits this inflation, so Falconer h ² is one affected downstream quantity among many (Corollary B3). (2) Structural fingerprint . In the joint twin survival surface S ( t ₁ , t ₂ ), misspecification produces systematic dependence errors (ISE 48× that of the recovery model). Conditional twin dependence is inflated at all ages, peaking at age 80 (Δ r = 0.048). (3) Specificity . The bias requires an omitted component that is both heritable and mortality-relevant. Three negative controls, a boundary check ( ρ = 0), and a two-component recovery refit ( σ _θ restored to within −3.2%) establish specificity. ACE decomposition yields C ≈ 0 throughout: the omitted extrinsic component loads onto A (because it is shared 1.0/0.5 in MZ/DZ), so switching summary statistics does not restore identification. (4) Sensitivity and falsifiability . Over an empirically anchored regime ( σ _γ ∈ [0.30, 0.65], ρ ∈ [0.20, 0.50]), Falconer bias ranges from +2.8 to +18.9 pp (mean 9 pp). If ρ is sufficiently negative, the bias reverses sign in all three model families (Corollary B4). A full-likelihood robustness check shows that this upward pull is partly structural and partly estimator-specific: in the same misspecified one-component model, ML still inflates σ _θ (+3%), whereas matching only r _MZ inflates it much more (+21%). These results do not resolve true intrinsic heritability but establish that Shenhar’s 50% estimate carries a structured, model-general upward bias originating in the fitted latent variance σ _θ .