Gestational Diabetes Is at Least Three Different Conditions Wearing One Name
Two women get the same diagnosis at the same 28-week glucose test. One adjusts her breakfast and barely thinks about gestational diabetes again. The other cuts carbohydrates hard, walks after every meal, checks her numbers four times a day, and still needs insulin by week 32. Both leave the appointment with the identical label. Neither is told the label may be covering up two different diseases.
This article is for educational and organizational purposes only. It is not medical advice.
That is not a metaphor. A 2016 study of the Gen3G pregnancy cohort found that gestational diabetes splits into at least two physiologically distinct patterns — one a problem with insulin resistance, the other a problem with the pancreas not producing enough insulin in the first place. Same glucose test, same three words on the chart, different underlying failure (Powe et al., Diabetes Care, 2016).
"Gestational diabetes" was never built to describe a mechanism. It is a screening category — anyone whose glucose crosses a threshold gets the same diagnosis and usually the same diet-first instructions. What is happening underneath that threshold differs by person, and it is not a minor difference.
One Name, at Least Two Different Failures
Picture insulin as a key, and each cell in the body as a room that keeps its door locked. Sugar arriving in the bloodstream cannot be used for energy until insulin turns the lock and lets it in. The pancreas is the factory that makes the keys.
Late in pregnancy, the placenta releases hormones that make the locks harder to turn — a normal shift that keeps a little more fuel circulating for the baby. A healthy pancreas answers by running the key factory faster. Gestational diabetes appears when the factory cannot keep up with how stiff the locks have become, and that shortfall happens in two different ways (Powe, Hivert & Udler, Diabetes, 2020).
In the larger group, the locks turn unusually hard, and the factory, even at full output, cannot make enough keys to compensate. That is the insulin-resistant pattern, and it matches the usual picture of gestational diabetes: sugar cannot get in because the doors are unusually hard to open. In a smaller but substantial group, the locks stiffen at an entirely ordinary rate. The problem is the factory itself, which cannot raise output no matter how much is needed. That is the insulin-deficient pattern, and it does not match the usual picture at all — the doors are not unusually stubborn, there simply are not enough keys being made. A remaining group shows features of both, or does not sort cleanly into either.
The Split Isn't Even, and the Two Groups Don't Look Alike
In the original 2016 findings, roughly half the women fell into the insulin-resistant group and about a third into the insulin-deficient group, the remainder mixed. The insulin-resistant group had higher infant birth weight and a substantially higher rate of adverse outcomes. The insulin-deficient group's outcomes looked close to women who never developed gestational diabetes at all.
A diagnosis built to catch anyone whose glucose crosses a line has turned out to contain at least two different diseases — and routine care does not currently ask which one a person has.
A larger 2025 follow-up using the HAPO cohort tracked roughly 8,000 newborns and more than 4,000 of those children into early adolescence. All three subtypes were linked to higher birth weight, but the insulin-resistant and mixed groups also carried significantly higher odds of childhood obesity and impaired glucose tolerance years later; the insulin-deficient group did not (Osmulski et al., Diabetes Care, 2025). The same diagnosis has been carrying two different risk profiles forward, with no way for current care to tell them apart.
What the Split Does Not Tell You
The honest complication: subtype does not predict everything. A 2021 study following women a year postpartum found the risk of prediabetes or type 2 diabetes was essentially the same regardless of subtype — roughly 28 to 31 percent in both groups, versus about 10 percent in women who never had gestational diabetes (Retnakaran et al., EClinicalMedicine, 2021). Subtype shows up in pregnancy outcomes and, per the newer data, a child's later metabolic risk — but not yet in a mother's own future risk.
Not every researcher agrees this framework is ready to change anything. A 2024 critique in Diabetologia argued the standard measurement does not reliably predict who needs more intensive treatment, that body mass index predicts outcomes about as well for far less cost, and that roughly a quarter of people do not sort cleanly into either category (Jones, Kusinski, Gillies & Meek, Diabetologia, 2024). This is genuinely unsettled science; no clinic tests for subtype as part of routine care.
Why This Still Matters, Even Before It Changes Anything at the Clinic
None of this is a reason to ask for a test that does not exist in routine care yet. But it is worth knowing, because of what it says about "failing" a diet-first plan. If the underlying problem is that the pancreas cannot raise insulin output enough, no amount of carbohydrate discipline recruits pancreatic function that was never going to compensate on its own. A plan not working is not the same thing as effort not working.
The diagnosis has always arrived as one name. The biology underneath it was never one thing.
Does this mean gestational diabetes isn't really diabetes?
No. It means the diagnosis groups together more than one physiological pattern. All of them involve glucose regulation problems during pregnancy that warrant monitoring and care — this research is about what is happening underneath the diagnosis, not whether the diagnosis is real.
Can I find out which subtype I have?
Not through routine clinical care. Subtyping relies on research-study measurements of insulin sensitivity and secretion that are not part of standard prenatal testing, and researchers disagree about whether current methods are precise enough for use outside a study.
Does the subtype change my risk of type 2 diabetes after pregnancy?
So far, no meaningful difference has been found. One study following women a year postpartum found similar prediabetes and diabetes rates across subtypes, both markedly higher than in women without gestational diabetes.
If diet changes aren't controlling my glucose, does that mean I'm doing something wrong?
Not necessarily. For some people the underlying issue is limited insulin production rather than insulin resistance — a difference in physiology, not effort. Needing medication sooner is not evidence of a diet failure.
Is this settled science that should change how gestational diabetes is treated?
No. It is an active area of research with real disagreement among scientists about how useful subtyping is in practice. Current guidelines do not incorporate subtype, and treatment recommendations have not changed.
What should I do if my glucose numbers stay high despite following my care plan?
Contact your healthcare team. Rising or persistently high numbers despite following a prescribed plan are a reason to be seen, not a reason to try harder on your own — your care team may need to adjust monitoring or treatment.
Sources
- Powe CE, Allard C, Battista MC, et al. "Heterogeneous Contribution of Insulin Sensitivity and Secretion Defects to Gestational Diabetes Mellitus." Diabetes Care. 2016;39(6):1052–1055. doi.org/10.2337/dc15-2672
- Powe CE, Hivert MF, Udler MS. "Defining Heterogeneity Among Women With Gestational Diabetes Mellitus." Diabetes. 2020;69(10):2064–2074. doi.org/10.2337/dbi20-0004
- Osmulski ME, Yu Y, Kuang A, et al. "Subtypes of Gestational Diabetes Mellitus Are Differentially Associated With Newborn and Childhood Metabolic Outcomes." Diabetes Care. 2025;48(3):390–399. doi.org/10.2337/dc24-1735
- Jones DL, Kusinski LC, Gillies C, Meek CL. "A critique of measurement of defective insulin secretion and insulin sensitivity as a precision approach to gestational diabetes." Diabetologia. 2024;68(4):752–765. doi.org/10.1007/s00125-024-06334-x
- Retnakaran R, Ye C, Hanley AJ, et al. "Subtypes of gestational diabetes and future risk of pre-diabetes or diabetes." EClinicalMedicine. 2021;40:101087.
