The Heritability of General Cognitive Ability Increases Linearly from Childhood to Young Adulthood
Haworth, C. M. A., Wright, M. J., Luciano, M., Martin, N. G., Boomsma, D. I., Bartels, M., Posthuma, D., et al. · 2010
grade Btwin-adoptionindependentreplicatednumbers spot-checked
Sample
~11,000 twin pairs across six studies in four countries
Population
Twins assessed at ages ~9, ~12 and ~17.
Design
Pooled ACE variance decomposition of general cognitive ability at three ages. Note the rearing range is that of ordinary twin-registry families — severe deprivation is rare.
Key findings
The Wilson effect: as children age, heritability of cognitive ability RISES and the shared-environment (family) component FALLS. This is the structural reason early-childhood ability gains are hard to sustain — the family channel that interventions load onto narrows with age. But note the honest ceiling: c2 lands near 0.10, not zero.
Genetic confound
This IS the genetically-sensitive design.
Replication notes
Extended by Bouchard (2013): c2 falls to ~0.10 by ages 18-20 and REMAINS ~0.10 thereafter while h2 asymptotes near 0.80. IMPORTANT: c2 is small but NOT zero — it leaves room for roughly 4-5 IQ points of shared-environment influence.
DOI / URL
10.1038/mp.2009.55
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Heritability of g, age 9 -> 12 -> 17 | h2 | 41% (34-49) -> 55% (49-61) -> 66% (58-73) | mixed | grade-6+ | none | not-applicable | g |
| Shared environment for g, age 9 -> 12 -> 17 | c2 | 33% (26-39) -> 18% (13-24) -> 16% (8-23) | mixed | grade-6+ | none | not-applicable | g |
Cited by
- Fadeout — why early gains disappear, and what actually persistsstrong supportconf: highgc: low
- The early home environment — what parents can and cannot causally movemixedconf: highgc: low