Paternal age and offspring cognitive ability: 2023–2025 evidence summary

1. Effect magnitude

• N ≈ 14 000 U.S.–European families, polygenic-score-adjusted:

– −0.20 IQ points per additional year of paternal age (95 % CI −0.29 to −0.11).

– ≈ 2-point deficit for every 10-year increase after adjustment for parental genotype and socio-economic status.

• Genetic control attenuates the raw association by > 80 %, indicating that population stratification (higher-education males deferring reproduction) explains most of the uncorrected trend.

2. Functional form

• Inverted-U trajectory replicated across four national cohorts (United States, Sweden, Scotland, Poland).

• Optimum: paternal age 28–32 y; peak-to-trough amplitude ≈ 2–5 IQ points.

• Variance attributable to paternal age < 1 % after adjustment for parental education and household income (R² ≈ 0.003).

3. Candidate mechanisms

A. De-novo mutation load:

– 1.9–2.2 single-nucleotide mutations per paternal year (Kong et al. 2012; Jonsson et al. 2023).

– Enrichment in constrained neurodevelopmental genes (pLI > 0.9).

B. Epigenetic drift:

– Age-associated sperm DNA-methylation changes at > 1 500 CpGs (F0 → F1 transmission demonstrated in animal models; human evidence correlational).

C. Paternal psychopathology pathway:

– 2025 meta-analysis (k = 17, N ≈ 90 000): antenatal/paternal depression and anxiety predict 0.10 SD lower offspring cognitive score (β = −1.5 IQ points, 95 % CI −2.1 to −0.8).

4. Interpretation for reproductive decision-making

• Expected delta-IQ for offspring of a 38-year-old versus 28-year-old father: ≈ −2 points (range −1 to −3).

• Equivalent to one-third of the IQ difference associated with a four-year maternal education gap or a ±$10 000 annual household-income shift.

• Population attributable risk is negligible compared with modifiable post-natal factors (breast-feeding duration, home literacy, school quality).