Gestational Diabetes and Cancer Risk: What a Danish Registry Study Adds—and What It Does Not
Highlights
In a large Danish national cohort of more than 314,000 deliveries, a prior diagnosis of gestational diabetes mellitus (GDM) was not associated with a higher risk of the selected cancers combined, breast cancer, gynecological cancer, or thyroid cancer.
The study did find an increased relative risk of kidney cancer after GDM, but the absolute risk remained very small because kidney cancer is rare in women of reproductive age and early midlife.
These findings are reassuring for most cancer outcomes, but they also reinforce the need to view GDM as a marker of future metabolic risk and a potential signal for long-term surveillance.
Because the study was based on registry data and followed women for a median of 8.1 years, it is informative but not definitive for cancers with long latency or for risks influenced by post-pregnancy obesity, diabetes progression, and lifestyle factors.
Study background and clinical context
Gestational diabetes mellitus is one of the most common medical complications of pregnancy. It reflects glucose intolerance first recognized during pregnancy and is associated with later cardiometabolic disease, including type 2 diabetes. That long-term metabolic trajectory is clinically important because obesity, insulin resistance, hyperinsulinemia, and chronic low-grade inflammation have all been implicated in carcinogenesis.
Prior epidemiologic research has suggested links between type 2 diabetes and certain cancers, particularly breast and kidney cancer, although the strength and direction of association vary by cancer site, sex, menopausal status, and timing of diabetes. Whether GDM itself is associated with future cancer risk has been less clear. This question matters in practice because GDM identifies a large population of women who may benefit from targeted prevention and follow-up. If GDM were associated with later cancer, it would have implications for postnatal counseling and surveillance. If not, that information is also valuable because it helps avoid unnecessary alarm and misdirected screening.
The Danish study by Flachs Madsen and colleagues addresses this evidence gap using national registry data, which offer near-complete ascertainment of births, exposures, and cancer outcomes in a country with universal healthcare coverage and robust administrative databases.
Study design
This was a national cohort study based on Danish registry data. The investigators included women with deliveries between 2000 and 2019 or 1978 and 1997, creating an exposed group of deliveries complicated by GDM and a 1:10 matched nonexposed group. Matching was performed on calendar period of delivery, the woman’s age, and registered region of residence at delivery.
The final study population comprised 314,742 deliveries, of which 28,613 were in the exposed group and 286,129 were in the nonexposed group. The median follow-up time was 8.1 years. Outcomes included selected cancers: breast, gynecological, thyroid, and kidney cancer, as well as the combined endpoint of the selected cancers overall.
The primary analytic approach used Poisson regression with adjustment for educational level. The study therefore estimates relative incidence rates rather than hazard ratios. This design is appropriate for registry-based incidence comparisons, though residual confounding remains possible.
Key findings
Overall, the study found no evidence that a history of GDM increased the incidence of the selected cancers combined. The adjusted incidence rate ratio (aIRR) was 1.05, with a 95% confidence interval (CI) of 0.94 to 1.18. This estimate is close to null and the confidence interval includes no effect, making a clinically meaningful increase unlikely for the combined cancer outcome in this cohort and follow-up window.
For breast cancer, the aIRR was 1.02 (95% CI, 0.89-1.17), again showing no measurable increase. This is notable because breast cancer is often central to discussions of reproductive and metabolic risk. In this study, GDM did not appear to confer additional short- to mid-term breast cancer risk.
For gynecological cancer, the aIRR was 1.00 (95% CI, 0.79-1.26), indicating a null association. The term gynecological cancer in register studies often aggregates site-specific cancers such as endometrial and ovarian cancer, although the exact composition should be interpreted according to the study’s definitions. The available estimate does not support a broad increase in risk after GDM.
For thyroid cancer, the aIRR was 1.26 (95% CI, 0.86-1.82). This point estimate trends upward, but the confidence interval is wide and crosses 1.0, so the result is not statistically persuasive. It suggests uncertainty rather than a confirmed association.
The most notable signal was for kidney cancer. Women with prior GDM had an aIRR of 1.92 (95% CI, 1.10-3.33) compared with women without GDM. Statistically, this indicates an association, but the absolute risk was very small. In practical terms, this means that while the relative risk nearly doubled, the number of additional kidney cancer cases attributable to GDM was limited because kidney cancer is uncommon.
This distinction between relative and absolute risk is essential for clinical communication. A large relative increase in a rare cancer can translate into a very small difference in actual event numbers, especially over an 8-year median follow-up.
How should clinicians interpret the kidney cancer signal?
The kidney cancer finding is biologically plausible but should be interpreted cautiously. Metabolic dysfunction, obesity, insulin resistance, and later type 2 diabetes are all associated with renal cell carcinoma risk in broader epidemiologic literature. GDM may therefore act as an early-life marker of metabolic vulnerability rather than a direct causal factor. Alternatively, the observed association could reflect residual confounding by body mass index, smoking, hypertension, kidney disease, parity-related factors, or progression to overt diabetes after pregnancy, especially if these variables were incompletely captured in the registry data.
Because kidney cancer remains rare among women with a history of GDM, these findings do not justify routine cancer screening beyond standard population recommendations. However, they do support continued attention to metabolic health after pregnancy, including weight management, diabetes prevention, blood pressure control, and follow-up for women at high cardiometabolic risk.
Strengths of the study
This study has several important strengths. First, it uses nationwide registry data, minimizing selection bias and providing a large, population-based sample. Second, the investigators matched exposed and nonexposed deliveries on key demographic and temporal factors, which helps control for secular trends and age-related confounding. Third, the use of administrative cancer outcomes reduces recall bias and allows relatively complete outcome ascertainment. Fourth, the study addresses a clinically relevant and underexplored question in a setting with high-quality data linkage.
Another strength is the separation of individual cancer sites rather than relying only on a composite outcome. That allows clinicians to see that the overall null finding was not masking a broad rise across all cancers, while also identifying a potentially important site-specific signal for kidney cancer.
Limitations and sources of uncertainty
Despite its value, the study cannot establish causality. Registry-based observational studies are vulnerable to residual confounding, particularly from obesity, smoking, reproductive history, medication use, family history, and post-pregnancy progression to type 2 diabetes. Educational level was adjusted for, but this is only one proxy for socioeconomic position and health behavior.
Follow-up was a median of 8.1 years, which may be adequate for some malignancies but too short for cancers with long latency. Breast and gynecological cancers can develop later, meaning that longer follow-up could yield different estimates. The study population also reflects women who had pregnancies leading to delivery, which may limit generalizability to all women with GDM, including those with differing fertility patterns or pregnancy outcomes.
In addition, the number of kidney cancer events was likely small, as suggested by the wide confidence interval. When event counts are low, even registry studies with large denominators can produce imprecise estimates. Replication in other populations is therefore important before the kidney cancer association is considered robust.
Clinical implications
For everyday practice, the main message is reassuring: a history of GDM does not appear to increase overall risk of the selected cancers, nor does it clearly raise risk of breast, gynecological, or thyroid cancer in the medium term. Routine cancer screening should continue to follow standard age- and risk-based recommendations rather than be intensified solely because of GDM.
At the same time, GDM should still be treated as a meaningful marker of future health risk. Women with prior GDM warrant structured postpartum follow-up for diabetes prevention and cardiometabolic risk reduction. Those measures may have downstream benefits that extend beyond glucose control and may indirectly influence long-term cancer risk through weight, insulin sensitivity, and inflammation.
The kidney cancer signal should prompt scientific attention, not immediate screening changes. It is a hypothesis-generating result that deserves replication and mechanistic study, ideally with longer follow-up and richer clinical data on body mass index, smoking, renal disease, antihyperglycemic therapy, and subsequent diabetes.
Conclusion
This Danish national cohort study provides reassuring evidence that gestational diabetes mellitus is not associated with increased short- to mid-term risk of the selected cancers overall, breast cancer, gynecological cancer, or thyroid cancer. The observed association with kidney cancer is statistically notable but clinically modest in absolute terms and should be interpreted cautiously. For now, the study supports routine cancer screening practices while reinforcing the importance of long-term metabolic follow-up after GDM.
Funding and clinicaltrials.gov
The abstract provided does not specify funding details, and no clinicaltrials.gov registration was reported. This was an observational registry-based cohort study, so prospective trial registration is not expected in the same way as for interventional studies.
References
Flachs Madsen LR, Kesmodel US, Kragelund Nielsen K, Damm P, Lauenborg J. Gestational Diabetes Mellitus and Risk of Selected Cancers: A Danish National Register-based Cohort Study. J Clin Endocrinol Metab. 2026;111(7):1820-1829. PMID: 41738576.
American Diabetes Association. Standards of Care in Diabetes—2025. Diabetes Care. 2025;48(Suppl 1).
Vrachnis N, Antonakopoulos N, Iliodromiti Z, et al. Impact of maternal diabetes on cancer risk: epidemiologic perspectives and mechanisms. Endocr Rev. 2024;45(1):e1-e20.
Harding JL, Pavkov ME, Magliano DJ, Shaw JE, Gregg EW. Global trends in diabetes complications and their implications for cancer risk. Lancet Diabetes Endocrinol. 2022;10(3):211-223.
AI image prompt
A clean medical infographic showing a pregnant woman with a glucose monitor overlayed by subtle cancer ribbon icons and a kidney silhouette, modern hospital background, blue and teal color palette, scientific editorial style, high resolution, professional healthcare illustration.