Written and maintained by CASRAI Editorial Board
Last updated
Short answer
Genes and behaviour are linked through many small genetic influences that act together with the environment, not through single “behaviour genes”. Researchers measure the genetic share of differences between people with heritability, a statistic that describes one population at one time and does not say how fixed a trait is.
Definition and scope
The study of how inherited DNA variation relates to differences in behaviour is called behavioural genetics (US spelling, behavioral genetics). Searches for genetics and behaviour, genetics and behavior and genetic behaviour all point to the same body of work: estimating how much of the variation in a trait across people is associated with genetic variation, and testing what else shapes that trait. For the underlying biology see what is genetics; for how gene activity is regulated without changing the DNA sequence see what is epigenetics and the comparison genetics vs epigenetics.
How researchers study genes and behaviour
Twin and family designs
Twin studies compare how similar twin pairs are on a trait. Johnson, Turkheimer, Gottesman and Bouchard (2009) argue that the twin design remains useful, and that its main current use is testing environmental explanations. In their words, “the twin study is far from having outlived its usefulness”. They describe within-pair analyses that reversed an apparent link between early sexual debut and delinquency, and findings that income and perceived control moderated genetic influence on physical health. They also note the method is not fully rigorous: twins may differ from each other in other ways, and twin life may differ from life as a singleton.
Genome-wide association and heritability estimates
Genome-wide association studies (GWAS) test many genetic variants across the genome for association with a trait. Zaitlen and Kraft (2012) review how heritability is estimated in this era. Accurate GWAS-based estimates of narrow-sense heritability need samples of roughly 5,000 to 10,000 or more individuals, and population stratification and cryptic relatedness can bias estimates.
Reading heritability correctly
Heritability is a ratio, not a measure of how much of an individual’s trait is “caused by genes”. Zaitlen and Kraft define broad-sense heritability as the ratio of total genetic variance to phenotypic variance, including dominance and epistasis, and narrow-sense heritability as the additive contribution of genetic variation only. They also state that heritability is defined with respect to a population at a particular time.
| Question | What the source says |
|---|---|
| Is heritability a fixed property of a trait? | No. Estimates are “completely dependent on the specifics of the samples and environmental conditions”, so they vary widely across studies (Johnson et al., 2009). |
| Does high heritability mean a trait cannot be changed? | No. Heritability says little about whether a trait can be changed, and highly heritable traits can be strongly shaped by environment (Johnson et al., 2009). |
| Broad-sense vs narrow-sense? | Broad-sense includes all genetic variance; narrow-sense counts only the additive part (Zaitlen and Kraft, 2012). |
| Is a heritability number tied to one population? | Yes. It is defined for a population at a particular time (Zaitlen and Kraft, 2012). |
Zaitlen and Kraft give a hypothetical illustration: if the correlation between sibling heights is 0.4, doubling it gives a heritability of 0.8, but only if dominance, epistatic and shared-environment components are zero, an assumption they call questionable. The example shows why a single number hides its assumptions.
Missing heritability
Family and twin designs often estimate more genetic influence than the variants found by GWAS explain. Zaitlen and Kraft define missing heritability as the additive genetic variance not yet captured with GWAS or other methods of identifying associated variants.
Where it matters for labs and study design
- Gene-environment correlation. Johnson et al. describe how genetically influenced people select or are placed in environments (passive, active and evocative correlation), so an environmental measure can partly reflect genetic influence. A confounding variable check should consider this.
- Sample structure. Because population stratification and cryptic relatedness can bias estimates and large samples are needed, the sampling plan is part of the result, not a detail.
Behaviour genetics sits alongside clinical and brain-science fields; see what is neurology for a neighbouring discipline. Gene-expression measurement, a different question from heritability, is covered in differential gene expression analysis.
Common misunderstandings
- “Heritability of 50% means half of my trait is genetic.” It describes variation across a population, not the makeup of one person’s trait.
- “Heritable means fixed.” Highly heritable traits can be strongly shaped by environment.
- “Twin studies are obsolete.” Their current main use is testing environmental causes, though they are not fully rigorous.
- “Genes versus environment.” The two interact through gene-environment correlation, so they cannot always be separated cleanly.
FAQ
Is there a gene for a behaviour?
The evidence reviewed here points to many genetic influences acting with the environment rather than single genes. Zaitlen and Kraft describe missing heritability as the additive genetic variance not yet captured by variants identified with GWAS, which means identified variants do not yet account for the full genetic influence estimated from family and twin designs.
What is the difference between genetics and behavior and genetics and behaviour?
Only the spelling. Both refer to behavioural (behavioral) genetics, the study of how inherited DNA variation relates to differences in behaviour.
What does a heritability figure tell you?
It is a ratio for one population at a particular time (Zaitlen and Kraft, 2012). It does not say how much of an individual’s trait is caused by genes, and it says little about whether a trait can be changed (Johnson et al., 2009).
Are twin studies still useful?
Johnson et al. (2009) argue they are, mainly for testing environmental explanations, while noting the tests are not completely rigorous.
How large a sample does GWAS-based heritability estimation need?
Zaitlen and Kraft report that accurate narrow-sense estimates need samples on the order of 5,000 to 10,000 or more individuals.
Sources
- Johnson W, Turkheimer E, Gottesman II, Bouchard TJ. Beyond Heritability: Twin Studies in Behavioral Research. Current Directions in Psychological Science, 2009;18(4):217-220. https://pmc.ncbi.nlm.nih.gov/articles/PMC2899491. Checked 2026-10-29.
- Zaitlen N, Kraft P. Heritability in the genome-wide association era. Human Genetics, 2012;131(10):1655-1664. https://pmc.ncbi.nlm.nih.gov/articles/PMC3432754. Checked 2026-10-29.
