Polygenic risk scores and what it means for the genetic testing moratorium

At the 2022 All-Actuaries Summit, Scott McKay and Richard Russell delivered a session on genetic testing, including its recent history and developments, key concepts, relevant research and implications for the life insurance industry.

Since the early 2000s, the cost of genome sequencing has decreased by orders of magnitude in only a couple of decades, from $2.7 billion in 2003 to under $1,000 in 2022. In simplified terms, genome sequencing refers to the mapping of the four basic building blocks of DNA: adenine (A), thymine (T), cytosine (C) and guanine (G). Once mapped, it is possible to investigate abnormalities in the DNA to determine whether an individual has a higher risk of a particular disease, hence the natural interest among life insurers in exploring the value of genetic testing.

The enormous decrease in the cost of genetic testing, along with advancement in the sequencing process itself, has led to greater accessibility of genetic testing and the growth of genetic data. This is linked to the broader trend of the increasing availability of personal and health data, both within and beyond the realm of insurance. For example, the UK Biobank is a large national database consisting of the detailed personal data, including genetic risk information, of half a million volunteers throughout the UK.

Such trends are expected to continue into the future.

In Australia over the next decade, the Federal Government has committed $500 million in funding – through the Genomics Health Futures Mission (GHFM) – to support genomic research.

The overall aim of bolstering such research is to improve health care outcomes for Australians through more personalised diagnosis and treatment. In addition, January 2024 will see the establishment of Genomics Australia, a new advisory government agency responsible for operationalising the research emerging from the GHFM.

These Australian developments are mirrored on a global scale, with China in particular investing heavily (close to US$10 billion) in genetic treatment. Such developments have also resulted in increasing regulatory attention in recent years, including the Council of Europe’s 2016 recommendation to mitigate discrimination by insurers based on genetic data. For the Australian life insurance industry, the Financial Services Council (FSC) implemented a 2019 moratorium on the use of genetic test results for life insurance policies up to specified limits.

There is a natural hesitancy among prospective policyholders to take genetic tests, despite their relative accessibility, given the potential requirement to divulge the results during the underwriting process, which may result in a higher premium or even denial of coverage. At the same time, from the perspective of insurers, the increased risk of anti-selection needs to be managed, as prospective policyholders may withhold knowledge of adverse genetic test results to the detriment of insurers.

Before considering the risks to either party, an important question to ask is whether information gained from genetic testing provides significant additional value to insurers in assessing the risk of prospective policyholders beyond the more traditional underwriting techniques.

This is related to the issue of determining whether a genetic predisposition to a given disease poses a material risk to the target insured population.

It is useful to assess each disease along two dimensions: prevalence and penetrance. Prevalence measures how common a predisposition is within a population, while penetrance refers to the likelihood that an individual with a predisposition will actually contract the disease. While it is uncommon for a disease to have both high prevalence and high penetrance, this does not discount the riskiness of the disease to the insured, nor the usefulness of genetic testing for the insurer.

For example, some diseases are very rare but have high penetrance. Under such circumstances, genetic test results would have quite a strong reliability in indicating that a prospective policyholder has high risk. Of course, only a small cohort of individuals in the overall pool of policyholders would be affected, given the very low prevalence of such diseases. However, as genetic testing becomes more commonplace and the results of such testing are more readily shared, this disease category would be expected to become a more material consideration for insurers.

Conversely, other diseases are quite common (i.e. high prevalence), but there may not be a very high likelihood of contraction. This disease category has been gaining increasing attention in recent years in academia. The susceptibility to such diseases is measured using so-called polygenic risk scores, which are aggregate scores derived from the risk information produced by genome sequencing. Although there is variation by disease, the polygenic risk scores are generally quite effective in distinguishing high risk from low risk of contraction.

A potential objection to the value of polygenic risk scores, and genetic testing more generally, is that more conventional factors such as family history and biometric information may be sufficient in identifying whether an individual has elevated risk of contracting a particular disease. However, key findings from relevant research presented in the session suggest that genetic test results can provide useful risk information that is not adequately captured by traditional underwriting, especially as techniques such as polygenic risk scoring continue to improve.

An open question that may also challenge the usefulness of genetic testing is the relative contribution of genetic versus lifestyle and environmental factors towards contracting a particular disease. That is, even a strong genetic predisposition to a particular disease may be mitigated by a host of factors, such as diet and exercise, that an individual has control over.

In future decades, as the technology continues to develop, there is interest in observing how the growth of genetic testing and genetic data will be controlled, supported or perhaps accelerated across different jurisdictions, and whether it will ultimately be possible for both insurer and insured to reap the benefits if genetic testing is allowed to play a lead rather than supporting role in the life insurance industry and wider society.

CPD: Actuaries Institute Members can claim two CPD points for every hour of reading articles on Actuaries Digital.