Is Gestational Diabetes Genetic? What the Largest Gene Study Reveals About Polygenic Risk

Is Gestational Diabetes Genetic? What the Largest Gene Study Reveals About Polygenic Risk

This article is for educational purposes only. It does not replace medical care, diagnosis, or genetic counseling. It is not a substitute for advice from a licensed physician, board-certified medical geneticist, or board-certified genetic counselor. For any decisions about testing, treatment, or care, consult a qualified clinician. In emergencies, call 911.

Bottom line: Research shows gestational diabetes (GDM) is partly genetic, but in a polygenic way. The largest genetic study so far found two kinds of risk: ordinary type 2 diabetes risk and a pregnancy-specific part. No genetic result, high or typical, replaces the prenatal glucose test. The bigger long-term story is a higher later risk of type 2 diabetes, which lifestyle programs can lower.

What you’ll learn

• How genes, weight, age and ancestry combine to shape gestational diabetes risk, according to large studies.
• What the 2024 FinnGen study found, and what a 23andMe polygenic report can and cannot say.
• How screening differs between the US and Canada, and what treatment trials actually showed.
• Why the years after delivery matter most, and which legal protections apply to genetic results.

What gestational diabetes is and why every pregnancy is screened

Gestational diabetes is a type of diabetes that occurs only during pregnancy. It often has no symptoms. When symptoms appear, they are mild, such as feeling thirstier or urinating more often.

According to the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), pregnancy hormones can make the body less able to use insulin. Insulin is the hormone that moves sugar from the blood into cells. NIDDK adds that “genes and extra weight may also play a role”.

Think of the placenta as a second engine running on the same fuel line. Its hormones make the body work harder to keep blood sugar steady. Some bodies fall a little behind.

Why mild high blood sugar still matters

The landmark HAPO study tested 25,505 pregnant people in 9 countries with a glucose tolerance test. The study found “no obvious thresholds at which risks increased”. Risks rose steadily as maternal glucose rose, even below diabetic levels.

Outcome measured in HAPO	Change per 1-SD rise in fasting glucose
Birth weight above the 90th percentile	Odds 1.38 times higher
Cord-blood C-peptide above the 90th percentile	Odds 1.55 times higher
Primary cesarean delivery	Odds 1.11 times higher

Source: HAPO Study Cooperative Research Group, NEJM 2008. One standard deviation (SD) was 6.9 mg/dL. These are odds ratios, not absolute “X in 100” chances.

How common it is in the US and Canada

The CDC states that 5% to 9% of US pregnancies are affected each year. That means about 5 to 9 out of every 100 pregnancies. National birth data show the share of mothers with a GDM diagnosis rose from 6.0% in 2016 to 8.3% in 2021.

In Canada, a Public Health Agency of Canada analysis covered 4.3 million hospital deliveries outside Quebec. GDM rose from 41.2 to 104.3 per 1,000 deliveries between 2005 and 2019, including British Columbia. That is roughly 10 out of 100 deliveries by 2019. British Columbia had the highest provincial rate.

Because GDM is common, often silent, and harmful along a smooth curve, guidelines screen everyone. A prenatal clinician can explain where a given pregnancy sits on that curve.

Section recap: Gestational diabetes is common, often symptom-free, and risks rise steadily with glucose. That is why every pregnancy is screened, regardless of genes.

Is gestational diabetes genetic? Family history, ancestry and the polygenic picture

The short answer is “partly.” Gestational diabetes is polygenic. That means thousands of genetic variants each add a very small push, rather than one gene causing it. These small pushes combine with weight, age, ancestry and pregnancy hormones.

This differs from single-gene conditions, where one specific variant can strongly predict disease. For typical GDM, no single variant is enough on its own.

Established risk factors

Major agencies list similar risk factors.

Risk factor	Listed by	Can it change?
Previous gestational diabetes	CDC, ACOG	No
Family history of type 2 diabetes	CDC	No
Overweight or obesity	CDC, ACOG	Partly, before pregnancy
Physical inactivity	ACOG	Yes
Polycystic ovary syndrome (PCOS)	CDC, ACOG	Managed, not removed
Previous baby over 9 pounds	CDC, ACOG	No
African American, Hispanic/Latino, American Indian, Alaska Native, Native Hawaiian or Pacific Islander ancestry	CDC	No

Sources: CDC and ACOG.

ACOG adds an important note. “GD also can develop in women who have no risk factors”. A clean risk-factor list is never a free pass.

How strong is family history? Numbers by study type

Genetic numbers mean different things depending on study design. The table keeps each number with its study type.

Study type	Finding	Population	Source
Twin heritability	No verified figure in the sources reviewed	Not applicable	Not available
Family-history relative risk	Family history of diabetes: odds ratio 2.77	Asian cross-sectional studies	Lee 2018
Family-history relative risk	Prior GDM: odds ratio 8.42	Asian cross-sectional studies	Lee 2018
GWAS single-variant effect	CDKAL1 odds ratio 1.52; variant near MTNR1B odds ratio 1.45	Korean women	Kwak 2012
GWAS loci count	13 loci, 9 of them new	Finnish	Elliott 2024

The Lee meta-analysis pooled 84 Asian studies, with an average GDM prevalence of 11.5%. Its cross-sectional odds ratios cannot be converted directly into a US or Canadian “X out of 100” risk.

A broader umbrella review checked 61 suggested risk-factor links. Only four had “convincing” evidence: three involved body weight, and one was hypothyroidism. Weight is among the best-supported factors that can change before pregnancy.

Shared roots with type 2 diabetes

The Korean study by Kwak and colleagues was an early genome-wide association study (GWAS). A GWAS scans DNA markers across thousands of people for small differences. It found that known type 2 diabetes variants were also linked to GDM. Each variant’s effect was modest, with odds ratios near 1.5.

A rare single-gene exception: GCK-MODY

A small group of people have monogenic diabetes, caused by one gene. One form, GCK-MODY (maturity-onset diabetes of the young caused by the GCK gene), causes stable, mild fasting high blood sugar. Because it is often silent, it may first be found on routine pregnancy glucose testing. The American Diabetes Association (ADA) notes a small 2-hour glucose rise on testing, under 54 mg/dL. It is diagnosed by clinical genetic testing that a clinician orders, not by a consumer polygenic report.

Section recap: GDM risk is partly genetic but spread across many small-effect variants. Weight, age, ancestry and prior GDM matter alongside genes.

What the 2024 FinnGen study found: two kinds of genetic risk

In 2024, Elliott and colleagues published the largest GDM genetic study to date in Nature Genetics. They compared 12,332 people with GDM against 131,109 people who had given birth without it. All were of Finnish ancestry, from the FinnGen biobank.

The team identified 13 genetic regions (loci) linked to GDM, 9 of them new. Earlier GDM studies combined had found only five loci. In a later group from Finland and Estonia, 11 of the 13 loci replicated.

Two classes of loci

In the authors’ words, GDM genetic risk “falls into two distinct categories”. One part is ordinary type 2 diabetes risk. The other mainly affects mechanisms disrupted in pregnancy.

Feature	Class T (type 2 diabetes-predominant)	Class G (GDM-predominant)
Example genes	TCF7L2, CDKAL1	MTNR1B, PCSK1, MAP3K15
What it resembles	Ordinary type 2 diabetes risk	Risk that shows up mainly in pregnancy
Linked biology	General glucose control	Islet cells, central glucose control, steroid hormones, placental expression
Effect size example	Not quoted here	MTNR1B rs10830963, beta 0.403

Source: Elliott et al., Nat Genet 2024. “Beta” is a per-variant effect size on a statistical scale. It is not a percentage chance of disease.

One finding runs the other way. A MAP3K15 variant on the X chromosome was protective, with a beta of −0.404. Some variants nudge risk down.

What the overlap number means

The genetic correlation between GDM and type 2 diabetes was 0.71. A value of 1 would mean identical genetic risk, and 0 would mean none. So the overlap is substantial but incomplete.

As an analogy, Class T risk is a bridge that strains under any heavy load. Class G risk holds fine until one specific heavy truck, pregnancy, crosses it.

A second study in East Asian pregnancies

Gu and colleagues studied 116,144 Chinese pregnancies in 2025, finding 13 new GDM loci. About half of all loci were specific to GDM and pregnancy glucose levels.

Limits to keep in mind

• The FinnGen discovery and replication groups were European (Finnish and Estonian) only.
• Whether these findings transfer to Hispanic/Latina, Black or Indigenous readers is not yet established.
• Each variant’s effect is small, so single variants cannot predict an individual pregnancy.

These findings are research results, not personal predictions. An obstetric clinician remains the right person to weigh them.

Section recap: The 2024 FinnGen study split GDM genetic risk into type 2 diabetes-like and pregnancy-specific parts. Results are strong, but mostly from European ancestry.

From relative to absolute risk: reading a 23andMe polygenic report

23andMe offers a “Gestational Diabetes” report. The company describes it as a 23andMe+ Premium report for members who report female birth sex. It is not described as part of the basic Health + Ancestry service.

According to 23andMe, the report uses a person’s ethnicity and more than 6,000 genetic variants. A statistical model then estimates the likelihood of developing GDM. The catalog lists covered ancestry groups, including European, Hispanic/Latino, Asian and African American.

What the report itself says about its limits

Per 23andMe, the report “does not account for lifestyle or family history”. Its research-based reports “have not been reviewed by the FDA”. They are not meant “to be used to make medical decisions”.

Regulator	Status of the 23andMe GDM report
FDA (US)	Not reviewed by the FDA, per 23andMe
Health Canada	No status statement found in the sources reviewed

How to read “increased” versus “typical”

An “increased likelihood” result shifts odds compared with a baseline. It does not diagnose anything. A “typical likelihood” result does not mean low risk.

Think of humidity in a weather forecast. Humidity alone cannot say whether it will rain.

No verified accuracy figures for the 23andMe GDM score appear in the sources reviewed. So no “X out of 100” translation is possible. The closest research benchmark comes from the Chinese study.

Model	Performance	Population	Source
Polygenic score combined with prenatal clinical records	AUC 0.729 before 20 weeks	Chinese pregnancies	Gu 2025
23andMe GDM polygenic report alone	No verified figure published in sources reviewed	Not applicable	Not available

The AUC (area under the curve) measures how well a model ranks people by risk. A value of 0.5 equals a coin toss, and 1.0 is perfect. A value near 0.73 is moderate, useful for research but not a diagnostic test. It also came from genes plus clinical records, not genes alone.

The clear rule

No polygenic score replaces the prenatal glucose test. Guidelines recommend screening every pregnant person regardless of genetic results. A genetic result should never be used to skip or delay screening.

23andMe’s own blog notes that telling a health care provider about risk factors “may suggest testing earlier during pregnancy”. Bringing the report to a prenatal visit can start that conversation.

Three kinds of genetic results

Result type	What it means	Example
Diagnostic (pathogenic) variant	A single variant that explains a condition	GCK-MODY confirmed by clinical testing
Polygenic risk score	A summed estimate from thousands of small variants	23andMe GDM report
Variant of uncertain significance (VUS)	A change whose meaning is not yet known	Reported only in clinical gene testing

For most polygenic GDM reports, an obstetrician or primary care clinician can explain the result. If a single-gene form such as MODY is suspected, a genetic counselor can help. The National Society of Genetic Counselors (NSGC) is the professional society, with a directory at NSGC.org. Certification is granted by the American Board of Genetic Counseling (ABGC), shown as “CGC.”

Section recap: A polygenic report shifts odds and cannot diagnose GDM or rule it out. It is a reason to talk with the prenatal team, never a reason to skip testing.

How GDM is diagnosed in the US and Canada

Diagnosis depends on a glucose test, not on DNA. A person drinks a measured sugar drink, and blood sugar is checked over time.

Timing of screening

The US Preventive Services Task Force (USPSTF) gives a grade B recommendation. It advises screening pregnant people without symptoms at 24 weeks or later. Before 24 weeks, it says evidence is insufficient (an “I statement”). This 2021 statement remained current as of October 2026.

The ADA Standards of Care 2026 go further for early testing. They recommend testing people with risk factors before 15 weeks for undiagnosed diabetes. They also recommend GDM screening at 24-28 weeks for everyone.

ACOG’s 2024 Clinical Practice Update replaced parts of its older Practice Bulletin No. 190. ACOG says people with risk factors are tested early, and others at 24-28 weeks.

Diabetes Canada recommends screening all pregnant women at 24-28 weeks. For those at high risk of undiagnosed type 2 diabetes, it recommends an early A1C test before 20 weeks.

One-step versus two-step testing

Approach	United States	Canada
Two-step, first step	50 g one-hour glucose challenge	50 g one-hour glucose challenge, preferred approach
Two-step, second step	100 g three-hour OGTT	75 g two-hour OGTT
Two-step diagnosis rule	Two or more abnormal values	One abnormal value
Two-step cutoffs	Fasting 95, 1-h 180, 2-h 155, 3-h 140 mg/dL	Fasting 5.3, 1-h 10.6, 2-h 9.0 mmol/L
One-step	75 g OGTT; fasting 92, 1-h 180, 2-h 153 mg/dL	Alternative 75 g OGTT; 5.1, 10.0, 8.5 mmol/L
Most common	Two-step, per USPSTF	Two-step is preferred

Sources: ADA Standards of Care 2026, Diabetes Canada 2018, USPSTF 2021.

OGTT stands for oral glucose tolerance test. The US and Canadian one-step cutoffs are the same, in different units. They came from HAPO data.

The ADA notes a trade-off. Switching to one-step testing was expected to raise GDM rates from about 5-6 in 100 to 15-20 in 100. A practice chooses its protocol, and a clinician can explain the choice.

Regulatory status of the tests

The OGTT, glucose challenge test and A1C are standard lab tests recommended by guidelines. The sources reviewed did not confirm FDA clearance or Health Canada licensing for specific assays. So no regulatory approval is claimed here for these tests.

Treating GDM found early

The TOBOGM trial included 802 women with a risk factor who were diagnosed with GDM before 20 weeks. They were randomized to immediate treatment or deferred or no treatment. A composite of adverse newborn outcomes occurred in 24.9% versus 30.5%. That is about 25 versus 31 out of 100 babies, a modest reduction.

There was no clear difference in pregnancy-related high blood pressure. The trial found no between-group differences in “serious adverse events associated with screening and treatment”. Adverse events in a trial are not the same as drug side effects.

Coverage

In the US, Marketplace plans and many others must cover GDM screening without a copay. This applies at 24 weeks or later and for those at high risk. A clinic billing office can confirm details for a specific plan.

Section recap: Diagnosis uses glucose testing, with two-step testing common in both countries but different second steps. People with risk factors can ask about earlier testing.

Managing GDM during pregnancy: what the treatment trials show

Two large trials compared GDM treatment with usual care.

ACHOIS (Australia, 2005)

ACHOIS randomized 1,000 women with GDM at 24-34 weeks. Treatment included dietary advice, glucose monitoring and insulin as needed. Serious perinatal complications fell from 4% to 1%. That is about 4 versus 1 out of 100 babies.

There were trade-offs. Nursery admissions rose from 61% to 71%, and inductions from 29% to 39%. Cesarean rates were similar. At 3 months after birth, the treated group reported less depression.

NICHD MFMU mild-GDM trial (US, 2009)

This US trial randomized 958 women with mild GDM. Its main combined outcome did not differ significantly, at 32.4% versus 37.0%. Several other outcomes did improve.

Outcome	Treatment	Usual care
Large-for-gestational-age baby	7.1 in 100	14.5 in 100
Birth weight over 4,000 g	5.9 in 100	14.3 in 100
Shoulder dystocia	1.5 in 100	4.0 in 100
Cesarean delivery	26.9 in 100	33.8 in 100

Source: Landon et al., NEJM 2009. Shoulder dystocia means the baby’s shoulder gets stuck during birth.

First-line care

Care usually starts with nutrition changes and physical activity. Glucose monitoring tracks whether targets are met. A registered dietitian can build a meal plan that still feeds the pregnancy well.

The CDC advises against trying to lose weight while pregnant. Extreme carbohydrate cutting or fasting is not part of guideline care. Social-media “reversal” diets are not a substitute for the care team’s plan.

Medicines and their regulatory status

When targets are not met, medicine may be added. The US and Canada differ here.

Medicine	FDA status	Health Canada status	Guideline role in GDM
Insulin (human insulin, HUMULIN R)	BLA 018780 approved 1982-10-28; indicated for diabetes mellitus	Cartridge and KwikPen marketed; vial cancelled post-market 2026-05-25	First-line in the US (ADA)
Insulin aspart (NOVORAPID)	Not covered in sources reviewed	Marketed since 2001-10-31	Insulin option in Canada
Metformin (GLUCOPHAGE)	NDA 020357 approved 1995-03-03; labeled for type 2 diabetes only	Marketed; monograph contraindicates use in pregnancy	Not first-line in US; alternative to insulin in Canada
Glyburide (DIABETA, MICRONASE)	Approved 1984-05-01; labeled for type 2 diabetes only	TEVA-GLYBURIDE marketed; monograph contraindicates use in pregnancy	Not first-line in US; last option in Canada

Guideline roles: ADA 2026 and Diabetes Canada 2018.

Use of metformin or glyburide for GDM is off-label in both countries. “Off-label” means a use not listed on the official product label.

The US metformin label says pregnancy data are “not sufficient to determine a drug-associated risk” for major birth defects. The Canadian monograph, dated 2026-04-07, says the drug “is contraindicated during pregnancy”. Diabetes Canada still allows it as an alternative to insulin. So Canadian guideline use departs from the Canadian label.

The ADA explains why it prefers insulin. Metformin and glyburide both cross the placenta. Glyburide cord-blood levels reach about 50-70% of the mother’s levels. The FDA glyburide label notes prolonged severe low blood sugar in some newborns.

The MiG trial compared metformin with insulin in 751 women. The main perinatal outcome was nearly identical, at 32.0% versus 32.2%. Nearly half (46.3%) of the metformin group also needed insulin. The trial reported “no serious adverse events associated with the use of metformin”. It did not assess long-term effects on children.

Insulin has decades of pregnancy data without an established link to major birth defects. Medicine choices belong to the obstetric and diabetes care team.

Section recap: Treatment trials show fewer large babies and birth injuries, with some trade-offs. Insulin is first-line in the US, while Canada also allows metformin.

After delivery: long-term type 2 diabetes risk and prevention

For many people, the most important part begins after the baby arrives. GDM is a strong signal of later type 2 diabetes risk.

How big is the risk?

Sources report different numbers because they measure different things.

Measure	Figure	Source
Relative risk vs no GDM	7.43 times higher	Bellamy 2009, 20 cohort studies
Relative risk vs no GDM	9.51 times higher	Vounzoulaki 2020, 20 studies
Cumulative incidence during follow-up	About 10 to 16 out of 100	Vounzoulaki 2020
Prediabetes or type 2 diabetes after about 11 years	52 out of 100 vs 20 out of 100 without GDM	HAPO Follow-up Study
Estimated lifetime risk	50 to 60 out of 100	ADA 2026
Agency summary	“About half” develop type 2 diabetes	CDC

A “nearly 10-fold” relative risk is not a 10-fold absolute chance. In the Vounzoulaki cohorts, most women had not developed type 2 diabetes during follow-up. Longer follow-up and broader definitions explain the higher HAPO, ADA and CDC figures.

Bellamy and colleagues found the higher relative risk was consistent across ethnicity, age and weight groups. The largest single study within that review reported a relative risk of 12.6.

Testing after birth: US vs Canada

Step	United States	Canada
First test	75 g OGTT at 4-12 weeks postpartum	75 g OGTT between 6 weeks and 6 months
Preferred test	OGTT over A1C early postpartum	75 g OGTT
Ongoing	Testing every 1-3 years for life	Periodic testing

Sources: ADA 2026, ACOG, Diabetes Canada.

The ADA prefers the OGTT because A1C can read falsely low soon after birth. In the US, plans must cover diabetes screening for women with past GDM who are not pregnant.

Prevention that works regardless of genes

The Diabetes Prevention Program (DPP) was an NIH-sponsored trial at 27 US centers. Ratner and colleagues studied 350 women with prior GDM in this trial. All had impaired glucose tolerance, a prediabetes stage, at entry.

On placebo, women with prior GDM developed diabetes 71% more often than other mothers. Intensive lifestyle change and metformin each cut diabetes by about 50% in this group. These results apply to women who already had prediabetes. Metformin here was used outside pregnancy.

The ADA recommends lifestyle change and/or metformin for prediabetes after GDM, a decision for a primary care clinician.

The CDC’s National DPP offers a year-long lifestyle program. In the original trial, 5-7% weight loss plus 150 minutes of weekly activity cut type 2 diabetes risk by 58%. Ten years later, participants were still about one-third less likely to develop it. The CDC keeps a registry of recognized programs, and coverage varies by insurer.

Think of a smoke detector plus fire prevention. Testing catches problems early, and lifestyle change lowers the chance of a fire.

Breastfeeding

Diabetes Canada recommends breastfeeding right after birth and for at least 4 months. The link to lower later diabetes risk is observational and guideline-based. It is not proven to be a direct cause.

The sources reviewed give no verified recurrence percentage for a future pregnancy. Lee 2018 reported an odds ratio of 8.42 for prior GDM, but only in Asian cross-sectional data. A clinician can discuss personal recurrence risk before the next pregnancy.

Section recap: GDM raises later type 2 diabetes risk several-fold, though not to certainty. Postpartum testing and DPP-style lifestyle change help regardless of genetic results.

Family implications, insurance and emotional weight

What GDM means for the baby

GDM does not mean the baby has diabetes. The HAPO Follow-up Study tracked 4,832 children for about 11 years. Among children of mothers with GDM, 19.1% had obesity, versus 9.9% of other children. That is about 19 versus 10 out of 100 children.

After adjusting for the mother’s weight, the odds ratio for child obesity was 1.58. The link with overweight or obesity combined was borderline. A companion study found higher maternal glucose linked to higher child glucose at ages 10-14. These are associations, not proof that GDM causes childhood diabetes.

NIDDK notes that healthy food choices and activity may lower risk for both parent and child. A pediatrician can track growth during routine visits.

Sisters, daughters and family history

Family history is useful information for relatives. A sister or daughter can tell her own clinician about a family history of GDM or type 2 diabetes. That history is a recognized risk factor and may support asking about early testing.

Cascade testing, offering tests to relatives after one positive result, suits single-gene conditions. For polygenic GDM risk, sharing family history with clinicians matters more than a consumer DNA kit.

Insurance and genetic discrimination

Protection	What it covers	What it does not cover
GINA (US, 2008)	Health insurance (Title I) and employment (Title II)	Life, disability or long-term-care insurance
ACA preventive coverage (US)	GDM screening without copay	Not a genetic privacy law
Genetic Non-Discrimination Act (Canada, 2017)	Bans requiring genetic tests or results as a condition of contracts or services	Limited to genetic tests as defined in the Act

GINA’s health insurance protections “do not cover long-term care insurance, life insurance, or disability insurance”. Some US states add further protections. People considering life insurance may want to learn their state’s rules first.

Canada’s law defines a genetic test as one analyzing DNA, RNA or chromosomes. It bars collecting, using or disclosing results in these contexts without written consent. It also bars refusing service because a person declines to share results.

Emotional weight

GDM is common and is not a personal failing. Pregnancy hormones and genes play roles outside anyone’s control.

ACHOIS found lower rates of depression at 3 months after birth in the treated group. Anyone feeling persistent low mood should tell a prenatal or primary care clinician. Perinatal mental health support is part of good care.

Section recap: GDM does not mean the baby has diabetes, but healthy family habits help. GINA and Canada’s genetic law protect against key types of misuse, with limits.

Frequently asked questions

Does an “increased likelihood” report mean a person will get gestational diabetes?

No. A polygenic report shifts odds relative to a baseline and is not a diagnosis. Weight, age, prior GDM and family history also matter. Every pregnant person is still screened, typically at 24-28 weeks.

Can a person with a “typical” genetic result skip the glucose test?

No. The USPSTF, ADA and Diabetes Canada recommend screening regardless of genetic results. GDM can also develop in people with no risk factors.

Will a child inherit gestational diabetes?

GDM itself is not passed down like a single-gene condition. Children of mothers with GDM had higher obesity rates in follow-up studies. Healthy family habits and routine pediatric care are the practical response.

Could a genetic result affect health or life insurance?

In the US, GINA protects against misuse in health insurance and employment. It does not cover life, disability or long-term-care insurance. Canada’s Genetic Non-Discrimination Act restricts requiring or using genetic test results.

Should a person get a second opinion or see a genetic counselor?

For a polygenic GDM report, an obstetrician or primary care clinician can usually interpret it. If a single-gene form such as GCK-MODY is suspected, an ABGC-certified genetic counselor can help. A second opinion is reasonable whenever a plan feels unclear.

Summary

Research shows gestational diabetes is partly genetic, through many small-effect variants. The 2024 FinnGen study found two parts: shared type 2 diabetes risk and a pregnancy-specific part. A Chinese study of 116,144 pregnancies found a similar pregnancy-specific pattern.

Genes are only part of the picture. Weight, age, ancestry and prior GDM matter too. A consumer polygenic report cannot diagnose GDM or rule it out.

The most useful step is a conversation. Bring the report, family history and questions to the next prenatal visit.

This article is for educational purposes only. It does not replace medical care, diagnosis, or genetic counseling. It is not a substitute for advice from a licensed physician, board-certified medical geneticist, or board-certified genetic counselor. For any decisions about testing, treatment, or care, consult a qualified clinician. In emergencies, call 911.

References

1. Elliott A, et al. Distinct and shared genetic architectures of gestational diabetes mellitus and type 2 diabetes. Nat Genet. 2024. https://pubmed.ncbi.nlm.nih.gov/38182742/
2. Kwak SH, et al. A genome-wide association study of gestational diabetes mellitus in Korean women. Diabetes. 2012. https://pubmed.ncbi.nlm.nih.gov/22233651/
3. Gu Y, et al. Genetic architecture and risk prediction of gestational diabetes mellitus in Chinese pregnancies. Nat Commun. 2025. https://pubmed.ncbi.nlm.nih.gov/40325049/
4. Lee KW, et al. Prevalence and risk factors of gestational diabetes mellitus in Asia. BMC Pregnancy Childbirth. 2018. https://pubmed.ncbi.nlm.nih.gov/30547769/
5. Giannakou K, et al. Risk factors for gestational diabetes: an umbrella review. PLoS One. 2019. https://pubmed.ncbi.nlm.nih.gov/31002708/
6. HAPO Study Cooperative Research Group. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med. 2008. https://pubmed.ncbi.nlm.nih.gov/18463375/
7. Lowe WL Jr, et al. HAPO Follow-up Study. JAMA. 2018; Scholtens DM, et al. Diabetes Care. 2019. https://pubmed.ncbi.nlm.nih.gov/30208453/
8. Crowther CA, et al. Effect of treatment of gestational diabetes mellitus on pregnancy outcomes (ACHOIS). N Engl J Med. 2005. https://pubmed.ncbi.nlm.nih.gov/15951574/
9. Landon MB, et al. A multicenter, randomized trial of treatment for mild gestational diabetes. N Engl J Med. 2009. https://pubmed.ncbi.nlm.nih.gov/19797280/
10. Simmons D, et al. Treatment of gestational diabetes mellitus diagnosed early in pregnancy (TOBOGM). N Engl J Med. 2023. https://pubmed.ncbi.nlm.nih.gov/37144983/
11. Rowan JA, et al. Metformin versus insulin for the treatment of gestational diabetes (MiG). N Engl J Med. 2008. https://pubmed.ncbi.nlm.nih.gov/18463376/
12. Ratner RE, et al. Prevention of diabetes in women with a history of gestational diabetes. J Clin Endocrinol Metab. 2008. https://pubmed.ncbi.nlm.nih.gov/18826999/
13. Vounzoulaki E, et al. Progression to type 2 diabetes in women with a known history of gestational diabetes. BMJ. 2020. https://pubmed.ncbi.nlm.nih.gov/32404325/
14. Bellamy L, et al. Type 2 diabetes mellitus after gestational diabetes. Lancet. 2009. https://pubmed.ncbi.nlm.nih.gov/19465232/
15. US Preventive Services Task Force. Screening for Gestational Diabetes. JAMA. 2021. https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/gestational-diabetes-screening
16. American College of Obstetricians and Gynecologists. Clinical Practice Update: Screening for Gestational and Pregestational Diabetes. 2024. https://pubmed.ncbi.nlm.nih.gov/42131962/ ; patient FAQ: https://www.acog.org/womens-health/faqs/gestational-diabetes
17. American Diabetes Association. Standards of Care in Diabetes-2026, Sections 2 and 15. Diabetes Care. 2026. https://pmc.ncbi.nlm.nih.gov/articles/PMC12690181/ ; https://pmc.ncbi.nlm.nih.gov/articles/PMC12690183/
18. Diabetes Canada Clinical Practice Guidelines Expert Committee. Diabetes and Pregnancy. Can J Diabetes. 2018. https://guidelines.diabetes.ca/cpg/chapter36
19. NIDDK, National Institutes of Health. Gestational Diabetes. https://www.niddk.nih.gov/health-information/diabetes/overview/what-is-diabetes/gestational
20. CDC. Gestational Diabetes. 2024; NCHS QuickStats. 2023. https://www.cdc.gov/diabetes/about/gestational-diabetes.html ; https://www.cdc.gov/nchs/blog/posts/2023/01/quickstats-percentage-of-mothers-with-gestational-diabetes-by-maternal-age-national-vital-statistics-system-united-states-2016-and-2021.html
21. Nelson C, et al. Diabetes mellitus in pregnancy across Canada. BMC Pregnancy Childbirth. 2024. https://pubmed.ncbi.nlm.nih.gov/38714923/
22. FDA (Drugs@FDA) and Health Canada (Drug Product Database). Metformin (GLUCOPHAGE) labeling and product monograph. https://pdf.hres.ca/dpd_pm/00084167.PDF ; https://api.fda.gov/drug/drugsfda.json?search=application_number:%22NDA020357%22
23. FDA (Drugs@FDA) and Health Canada (Drug Product Database). Insulin (HUMULIN R, NOVORAPID) labeling and product status. https://pdf.hres.ca/dpd_pm/00084670.PDF ; https://api.fda.gov/drug/drugsfda.json?search=application_number:%22BLA018780%22
24. FDA (Drugs@FDA) and Health Canada (Drug Product Database). Glyburide labeling and product monograph. https://pdf.hres.ca/dpd_pm/00060647.PDF ; https://api.fda.gov/drug/drugsfda.json?search=application_number:%22NDA017532%22
25. CDC. About the National Diabetes Prevention Program. 2024. https://www.cdc.gov/diabetes-prevention/programs/what-is-the-national-dpp.html
26. NHGRI. Genetic Discrimination (GINA); HealthCare.gov. Preventive care benefits for women. https://www.genome.gov/about-genomics/policy-issues/Genetic-Discrimination ; https://www.healthcare.gov/preventive-care-women/
27. Government of Canada, Justice Laws. Genetic Non-Discrimination Act (S.C. 2017, c. 3). https://laws-lois.justice.gc.ca/eng/acts/g-2.5/page-1.html
28. 23andMe. Gestational Diabetes report announcement and clinician catalog (company product description). https://www.23andme.org/blog/articles/gestational-diabetes/ ; https://www.23andme.org/clinicians/genetic-reports/health-predispositions/

---

Last updated: 2026-10-15

Author: Yu Mizuno (Editor-in-Chief, non-physician), GeneLumen editorial team.

This article aggregates 28 sources. Of these, 19 are tier 1 and 8 are tier 2. One company product page describes the 23andMe report only. Sources include NIH, CDC, FDA, Health Canada, USPSTF, ACOG, ADA and Diabetes Canada guidance. They also include PubMed-indexed studies.

This article is for educational purposes only. It is not a substitute for medical advice from a licensed physician, board-certified medical geneticist, or board-certified genetic counselor. It does not replace medical care, diagnosis, or genetic counseling. In emergencies, call 911 (US/Canada).

Related: Metabolic and Hematologic Genetic Diseases

🇯🇵 For readers in Japan — a separate Japanese edition written for Japan’s healthcare system (not a translation): https://genelumen.com/ja/ja-metabolic-hematologic-genetic/gestational-diabetes-genetic-risk