A complex trait is any characteristic influenced by many genes working together, each contributing a small effect, combined with environmental factors. Unlike traits controlled by a single gene (such as sickle cell disease or cystic fibrosis), complex traits don’t follow simple inheritance patterns. Height, weight, blood pressure, and susceptibility to diseases like cancer, diabetes, and heart disease are all complex traits.
How Complex Traits Differ From Single-Gene Traits
In classic genetics, some conditions follow a straightforward pattern: one gene, one outcome. If you inherit two copies of the sickle cell variant, you develop sickle cell disease. These single-gene (or Mendelian) traits are relatively predictable.
Complex traits work differently. Their heritability comes from the contribution of thousands of genetic locations across the genome, each with a small individual effect. No single variant “causes” the trait. Instead, the cumulative weight of many variants, layered on top of environmental influences, shapes the outcome. This is why two siblings can share most of their DNA yet differ noticeably in height, personality, or disease risk.
How Small Genetic Effects Add Up
The predominant model for complex traits is additive. Each genetic variant nudges the trait in one direction by a tiny amount, and these nudges stack on top of each other. For psychiatric disorders, diabetes, cardiovascular disease, obesity, and height, the data consistently show that a large number of locations across the genome contribute predominantly additively to the variation we see in the population. Because each individual variant’s contribution is small, you need very large studies to detect them at all.
This doesn’t mean genes interact in simple ways at the cellular level. Inside cells, gene products do interact with each other in complex networks. But at the population level, the statistical effect of any one variant averages out to a small, roughly additive contribution. That’s why predicting a complex trait from genetics alone remains difficult.
Height: The Best-Studied Example
Human height is the textbook case of a complex trait, and it illustrates just how many genes are involved. A genome-wide study of 5.4 million people identified 12,111 independent genetic variants significantly associated with height. These variants cluster within about 7,209 segments of the genome, covering roughly 21% of the entire human genome. Together, common genetic variants are predicted to explain 40 to 50% of the variation in height between people.
That means even after cataloging over twelve thousand variants, genetics only accounts for about half of why people differ in height. The rest comes from nutrition, childhood health, and other environmental factors. This split between genetic and non-genetic contributions is typical of complex traits.
The Role of Environment
Genes set a range of possibilities. Environment determines where within that range you land. For metabolic traits like obesity and type 2 diabetes, researchers have identified specific environmental triggers that interact with genetic risk. Factors affecting insulin resistance, glucose regulation, and metabolic function can activate or silence genes in muscle, fat, and liver cells. In studies that exposed these cell types to 21 different environmental perturbations, thousands of genes changed their activity, and many of those responsive genes overlapped with regions flagged in genetic studies of metabolic disease.
This means your genetic risk for a condition like type 2 diabetes isn’t fixed at birth. Diet, physical activity, stress, and other exposures can dial that risk up or down by influencing which genes are active and how strongly they operate. The same principle applies broadly: complex traits are, by definition, a conversation between your genome and your environment.
Epigenetic Modifications
One mechanism connecting environment to gene activity is epigenetics. Chemical tags can be added to DNA or to the proteins that package it, changing whether a gene is turned on or off without altering the DNA sequence itself. Two well-studied tags mark active regulatory regions of the genome, essentially flagging which stretches of DNA are currently in use in a given tissue.
Researchers studying human liver tissue found that genetic variants can influence these chemical tags, which in turn affect gene activity. In some cases, the same genetic variant flagged in a large disease study turned out to work by changing the epigenetic state of a nearby regulatory region. This adds another layer to complex trait biology: the effect of a genetic variant may depend on the epigenetic landscape of the tissue it operates in, which itself can shift with age, diet, or disease.
Penetrance and Expressivity
Two concepts help explain why complex traits are so variable between people, even those who share genetic risk variants. Penetrance refers to the proportion of people carrying a particular variant who actually show its effects. In familial cancer syndromes, for example, many people with a BRCA1 or BRCA2 variant will develop cancer during their lifetime, but some will not. This is called reduced or incomplete penetrance.
Variable expressivity describes how the same genetic condition can look different in different people. In Marfan syndrome, some individuals are simply tall and thin with long fingers, while others develop life-threatening heart and blood vessel complications. For complex traits, both reduced penetrance and variable expressivity are the norm rather than the exception, because so many genetic and environmental factors modify the final outcome.
Polygenic Risk Scores
Scientists have developed tools called polygenic risk scores that combine the effects of many genetic variants into a single number estimating your predisposition toward a trait or disease. These scores are becoming increasingly useful, though their accuracy varies by condition.
For distinguishing between type 1 and type 2 diabetes, a score based on just 30 genetic variants achieved strong accuracy on its own, and when combined with other clinical information, it correctly classified patients 96% of the time. For breast cancer, a score incorporating 313 variants accounts for an estimated 35% of familial risk, with more than half of people scoring either 1.5 times above or below the population average. When this genetic score was combined with traditional risk factors and breast density measurements, it substantially outperformed traditional risk assessment alone.
These scores are most useful for stratifying risk rather than making definitive predictions. Someone with a high polygenic risk score for heart disease isn’t guaranteed to develop it, and someone with a low score isn’t immune. The scores reflect probabilities shaped by the additive weight of many small genetic effects.
The Missing Heritability Problem
One of the persistent puzzles in complex trait genetics is that known genetic variants don’t account for all the heritability estimated from family studies. For intelligence, twin studies suggest heritability of about 50%, but the variants identified through genome-wide studies explain only about 10%. This gap is called the missing heritability problem.
Several explanations are on the table. Many common variants likely have effects too small to detect even in large studies. Rare variants with stronger effects may contribute but are harder to find. Structural differences in DNA, such as duplications or deletions of large segments, aren’t captured by standard genetic studies. Gene-gene interactions and gene-environment interactions could inflate heritability estimates from twin studies in ways that are difficult to untangle. The gap is narrowing as study sizes grow and methods improve, but it remains a reminder that the full genetic architecture of complex traits is not yet mapped.