Most gene variants that have been labeled "pathogenic" may make only a small difference in a person's risk of actually developing disease, a new study suggests.

Scouring genetic data on more than 72,000 individuals, researchers found that most of the gene variants believed to contribute to disease actually have a minor impact: On average, those so-called pathogenic variants boosted the risk of developing disease by about 7%.

Some gene variants, including ones well-known to predispose people to disease, did show a substantial impact, the researchers said.

They included certain alterations in the BRCA1 and BRCA2 genes, which raise the risks of breast and ovarian cancers, and some variants in the LDLR gene, which cause an inherited form of very high cholesterol.

But most of the time, the odds that a pathogenic gene variant would lead to disease were low, according to findings published Jan. 25 in the Journal of the American Medical Association.

It all highlights a need to better quantify the risks linked to gene variants that are dubbed pathogenic, said researchers Ron Do and Iain Forrest of Mount Sinai's Icahn School of Medicine in New York City.

They said past studies may have overestimated the risks of some gene variants because the research involved smaller groups of people who either had a given disease or had a family history of it.

Those types of studies can be biased toward inflating risks, explained Do, an associate professor of genetics and genomic sciences.

So for their study, the researchers used data on more than 72,000 participants in two "biobanks" that link people's genetic data with their electronic health records: the BioMe project, which includes patients in the Mount Sinai Health System; and the U.K. Biobank, which involves British adults aged 40 to 69 from the general population.

All biobank participants give written informed consent to have their data used for research purposes.

Do's team focused on 5,360 gene variants that are considered pathogenic or to confer "loss of function." The variants were tied to 157 diseases in their study group, ranging from common afflictions like type 2 diabetes, stroke and various cancers, to rarer disorders known to be caused by certain gene mutations.

On average, the researchers found, the "penetrance" of the variants — the likelihood that they would cause disease — was roughly 7%. And for 89% of the variants, the difference in disease risk between carriers and non-carriers was 5% or less.

Some gene variants, not surprisingly, had a much stronger impact.

On average, pathogenic variants in BRCA1 or 2 carried a 38% likelihood of leading to breast cancer. And some variants in the LDLR gene were tied to a nearly 75% chance of familial hypercholesterolemia (FH) — an inherited disorder that causes very high levels of LDL ("bad") cholesterol and can lead to early heart disease.

Even in those genes, though, different pathogenic variants had different odds of causing disease, said Forrest, an MD/PhD candidate at Mount Sinai.

In the real world, most people would not be getting information on the vast array of gene variants they carry — unless they have the cash and motivation to pay for direct-to-consumer DNA-testing kits.

Usually, people undergo genetic testing for a specific reason, said Heather Zierhut, president of the National Society of Genetic Counselors. People are screened for LDLR variants, for example, to confirm that their very high LDL is caused by FH, or because a family member has the disorder.

Taking that gene as an example, Zierhut said it's been recognized that the different pathogenic variants in LDLR seem to carry a range of risk.

"We're learning more and more, and things are still evolving, and risk estimates will change," she said.

With a condition like FH, any "clinical diagnosis" — based on a person's LDL levels and family history of early heart disease — needs to be addressed, regardless of genetic test results, according to the FH Foundation.

Do said much more research is needed to better define the disease risks linked to individual gene variants. But "personalizing" people's health care will always be more than a matter of quantifying genetic risks.

While some disorders have a very strong genetic component, the odds of developing a disease is usually far more complicated than the risk assigned to any one gene variant, all three stressed.

There is the interaction between that gene variant and other gene variants a person carries — not to mention an array of non-genetic factors, from age to diet and exercise to environmental exposures.

That is especially the case, Zierhut said, with common, complex diseases like type 2 diabetes, where it is not surprising that any single pathogenic gene variant would make a small difference in disease risk.