Although many human traits and conditions are caused by genetics, many are not.  Environmental factors, diet and lifestyle, and even randomness can play a role.  And, for many conditions, there are both genetic and non-genetic factors.

Some diseases are caused by a mutation in a single gene: examples are cystic fibrosis (the CFTR gene), sickle-cell anemia (the HBB gene), and Huntingdon’s disease (the HTT gene) are all highly penetrant, caused by genetic mutations in a single gene.  Other traits are controlled solely by genetics, but are influenced by multiple genes; for example, eye color is controlled by several genes working together, which makes inheritance patterns less straightforward than we’d expect from single-gene inheritance.

On the other hand, diseases like scarlet fever, COVID-19, influenza, and botulism are caused by exposure to bacteria or viruses.  Exposure to heavy metals like lead can cause neurodevelopmental problems in children.  These are solely non-genetic causes.

And, to make things more complicated, some traits have both a genetic and a non-genetic component.  Examples of this include certain hereditary cancer syndromes (discussed in the Cancer Genetics Module) and phenylketonuria.

People with phenylketonuria have a homozygous loss of function mutation in the gene encoding phenylalanine hydroxylase (PAH).  This enzyme is responsible for the metabolism of phenylalanine.  Exposure to phenylalanine – an amino acid that is common in many foods – can cause severe neurological and other health problems.  A diet low in phenylalanine can largely prevent the most extreme conditions associated with phenylketonuria. Thus, both genetic (PAH mutations) and nongenetic (diet) factors influence the condition.

In some cases, non-genetic influences can affect the penetrance or expressivity of a trait; in the example of phenylketonuria, changes in diet can lessen or even prevent the neurological symptoms of the disease, making the neurological symptoms variably expressive or incompletely penetrant.

Because of shared environment, factors like infectious disease, diet, and exposure to toxins can be shared in families, which might make it appear as if there is a genetic cause even if there isn’t.  So, when studying a condition, how can geneticists distinguish between genetic and non-genetic causes? 

Twin Studies

Studies of twins and adoptive families (where families share an environment and culture, but not DNA) can help geneticists distinguish between traits that have a genetic component (even if multifactorial) and traits that are shared among family members because of a shared environment.

Twin studies typically compare the concordance among monozygotic (identical) twins to the concordance among dizygotic (fraternal) twins.  “Concordance” means agreement – it is the percentage of twin pairs that match each other.

For a fully genetic trait or disease, identical twins match in phenotype 100% of the time. They share all of their DNA, and their concordance would be 100. Dizygotic twins, on the other hand, share no more DNA than any other siblings – about 50%.  Concordance in dizygotic twins for a genetic trait is lower than for monozygotic twins.

For a trait or disease that is wholly environmental – symptoms due to exposure to a toxin or a virus, for example – both monozygotic twins and dizygotic twins are exposed to similar environments and sometimes may both be affected, although sometimes one twin might not be affected.  Concordance of less than 100 in monozygotic twins thus indicates an environmental component.

For many diseases, though, there might be both a genetic and environmental component.  One might inherit a predisposition to disease, for example, that is mitigated by diet (like phenylketonuria).  For traits with both an environmental and genetic cause, both conditions would be met: monozygotic concordance less than 100, and monozygotic concordance greater than dizygotic concordance.  Some examples are given in Table 1 below.

Note that for many traits with both genetic and non-genetic components, monozygotic concordance may actually be quite low.  It’s not the monozygotic concordance by itself that shows a genetic component: it is the monozygotic concordance relative to dizygotic concordance.  We see that for colon cancer in Table 1.

For colon cancer, the low rates of monozygotic concordance are due to the fact that colon cancer is a heterogeneous disease, caused by different factors in different people.  Sometimes colon cancer occurs randomly, with no previous family history of the disease.  This can be influenced somewhat by diet and lifestyle.  In other cases, individuals may inherit an allele that predisposes them to colon cancer.  This can be due to heterozygous loss of function mutations in tumor suppressor genes like MSH2 or MLH3. The families of such patients may have a high number of relatives who also develop colon cancer. But whether or not an individual with a mutant allele of MSH2 or MLH3 actually develops cancer depends on the acquisition of additional somatic mutations over the course of their lifetime.  Not all such individuals will acquire the additional mutation, and so not all will develop cancer (it is incompletely penetrant).  This is discussed more in the Cancer Genetics module.

Test Your Understanding

Twin studies are limited in how they can be interpreted.  Sometimes this is because only incomplete data is available: there are many case studies in the literature where only one or two sets of twins are studied.  In these cases, if a set of monozygotic twins are not concordant, it’s possible to conclude that there are non-genetic factors at play.  Several case studies on celiac disease in twins from the 1970’s-1990’s fall in this category, with monozygotic twins occasionally showing discordance.  But without a large enough sample size to compare monozygotic twins and dizygotic twins, it’s not possible to conclude whether there is also a genetic component to the trait.  A large-scale twin study of celiac disease was published in 2001.

Test Your Understanding

Another example of this is relatively recent work on gender, comparing cisgender and transgender twins. You’ll recall from the module on Sex the difference between sex and gender. Sex refers to measurable features associated with maleness or femaleness, such as gonads, external genitalia, or chromosomes.  Gender identity refers to an innate sense of self.  An individual might self-identify as male, female, or another gender.  Sometimes sex is referred to as “biological sex”, suggesting that there is no biological component to gender.  But is that true? Some twin studies have suggested otherwise.

One comparison of concordance^[4] in trans- and cisgender twins shows that for male twin pairs, MZ =33 and DZ = 5.  For female twin pairs, MZ = 23 and DZ = 0.  Because MZ < 100, there does appear to be a nongenetic component to gender.  But because MZ > DZ, there is also strong evidence for a genetic component to gender.  These studies also thus strongly suggest a biological component to gender as well as to sex.

Some interpretations of this type of work wrongly suggest that because there is a nongenetic component to gender, there cannot also be a genetic component.  But that is contrary to our understanding of many traits influenced by both genetic and nongenetic factors.  This is similar to the example of colon cancer: As shown in Table 1, monozygotic concordance for colon cancer is relatively low (4.7) but is still higher than dizygotic concordance (2.6).  Colon cancer is linked to both genetic factors (mutations in tumor suppressor genes like MSH2 and MLH3) and non-genetic factors.

It is important to recognize the limitations of twin studies.  Misinterpretation can happen when the dizygotic control is ignored, or if a small number of twin pairs makes a good comparison difficult.  Remember, it’s not the actual value of the monozygotic concordance that matters but the value relative to the dizygotic concordance.  Sometimes a low monozygotic concordance might be misinterpreted as “no genetic component” if it is not compared to the dizygotic control.  Also, keep in mind that identical twins may be treated more similarly than fraternal twins, so the monozygotic/dizygotic comparison is not a perfect control.

As a summary: if monozygotic concordance is less than 100 (MZ<100), there is a non-genetic or environmental cause for a trait.  Monozygotic concordance greater than dizygotic concordance (MZ>DZ) suggests there is a genetic cause for a trait.  If both conditions are met (MZ<100 and MZ>DZ) there is likely both a genetic component and a non-genetic component to the trait.

Adoption studies

Twin studies can identify conditions for which a genetic component exists by comparing two groups of people with shared environment but differing amounts of shared DNA: monozygotic twins (with 100% shared DNA) and dizygotic twins (with about 50% shared DNA).  Adoption studies are similar, comparing individuals with either shared DNA (biological relationships) or shared environment (adoptive relationships).  If a trait is shared more often between biological family members than adoptive family members, this suggests that there is a genetic component for the trait.

Adoption studies have shown that body mass index (BMI) and obesity are correlated for biological parents and children but not adoptive children^[5], suggesting a strong genetic component for obesity.  Adoption studies have also suggested a genetic component to alcohol use disorder and drug dependency^[6].