Heritability of Intelligence

Defining Heritability

Heritability, in quantitative genetics, is a statistical measure that estimates the proportion of phenotypic variation in a population that is attributable to genetic variation among individuals. It is often expressed as a coefficient, h², ranging from 0 to 1. A heritability of 0 indicates that all phenotypic variation is due to environmental factors, while a heritability of 1 indicates that all phenotypic variation is due to genetic factors. For intelligence, heritability estimates typically fall between 0.5 and 0.8 in adult populations in developed countries, meaning that 50% to 80% of the observed variation in IQ scores can be accounted for by genetic differences.

It is crucial to understand what heritability does not mean. Heritability is a population-level statistic; it does not indicate the degree to which an individual's trait is determined by their genes. An individual's intelligence is always a product of both their genes and their environment. Heritability also does not imply immutability. Highly heritable traits can still be modified by environmental interventions. For example, height is highly heritable, yet average heights have increased significantly over the past century due to improved nutrition. Similarly, heritability estimates are specific to the population and environment in which they are measured; they are not universal constants. Different environments can lead to different heritability estimates for the same trait.

Methods of Estimation

Heritability estimates for intelligence primarily derive from twin and adoption studies. Twin studies compare the similarity of monozygotic (MZ, identical) twins, who share nearly 100% of their genes, with dizygotic (DZ, fraternal) twins, who share, on average, 50% of their segregating genes, much like other siblings. If MZ twins are significantly more similar in intelligence than DZ twins, this difference is attributed to genetic factors. The classic formula by Falconer (1960) for broad-sense heritability is h² = 2 * (rMZ - rDZ), where rMZ and rDZ are the correlations of intelligence scores for MZ and DZ twins, respectively.*

Adoption studies compare the intelligence of adopted children with their biological and adoptive parents or siblings. If adopted children's intelligence correlates more strongly with their biological relatives than with their adoptive relatives, this provides evidence for genetic influence. Studies combining both twin and adoption designs, such as those involving MZ twins reared apart, offer particularly strong evidence, as they directly assess genetic influence across different environments.

More recently, molecular genetic methods, such as Genome-Wide Association Studies (GWAS) and genomic relatedness matrix restricted maximum likelihood (GREML) analyses, have begun to identify specific genetic variants associated with intelligence and to estimate heritability directly from DNA markers. These methods generally confirm the heritability estimates derived from classical twin and adoption studies, though the proportion of variance explained by identified common genetic variants (known as 'chip heritability') is often lower than estimates from twin studies, a phenomenon sometimes referred to as 'missing heritability' (Maher, 2008).

Gene-Environment Interaction and Development

The heritability of intelligence is not constant across the lifespan. Studies consistently show that heritability estimates for intelligence tend to increase from childhood through adolescence into adulthood. In early childhood, shared environmental factors (e.g., family upbringing, socioeconomic status) often play a substantial role, but their influence diminishes with age, while genetic influences become more pronounced (Plomin & Deary, 2015). This phenomenon, known as the developmental amplification of heritability, suggests that as individuals mature, they increasingly select or create environments that are correlated with their genetic predispositions (Scarr & McCartney, 1983). This is an example of active gene-environment correlation.

Gene-environment interaction (G x E) also plays a critical role. This refers to situations where the effect of a gene depends on the environment, or the effect of an environment depends on an individual's genes. For instance, some research suggests that the heritability of intelligence may be lower in impoverished environments and higher in enriched environments (Turkheimer et al., 2003). This implies that in environments where basic needs are not met and opportunities are limited, environmental factors exert a stronger influence on the expression of genetic potential. Conversely, in resource-rich environments, genetic differences may have more room to manifest, leading to higher heritability estimates.

Common Misunderstandings and Critiques

One common misunderstanding is that heritability implies genetic determinism or that intelligence cannot be changed. As noted, heritability is a population-level statistic and does not preclude environmental modification of an individual's trait. Another error is to conflate heritability within a group with differences between groups. High heritability of intelligence within a particular population does not mean that average differences in intelligence between different populations are necessarily due to genetic factors. Such group differences could be entirely environmental, as environments often differ systematically between groups.

Critics of heritability research, such as Lewontin (1970), have emphasized that heritability estimates are highly dependent on the range of environments sampled. If environments are very uniform, genetic differences will account for a larger proportion of variance, leading to higher heritability. If environments are highly varied, environmental factors will account for more variance, potentially lowering heritability. Therefore, heritability estimates cannot be generalized beyond the specific population and environmental context in which they were derived.

Furthermore, the concept of intelligence itself, typically measured by IQ tests, is a subject of ongoing debate (Gould, 1981). Critics argue that IQ tests may not fully capture the multifaceted nature of human cognitive abilities and can be culturally biased, which could influence heritability estimates. Despite these critiques, the consensus in behavioral genetics is that genetic factors contribute substantially to individual differences in intelligence within populations, and that these genetic influences interact dynamically with environmental factors throughout development.