By Dr. Asnat Raziel, bariatric surgeon specializing in obesity treatment, metabolic health and healthy longevity; Director of the Multidisciplinary Center for Obesity Treatment, Assia Medical, Assuta Ramat HaChayal, Tel Aviv
CGM in pregnancy: what the new research found
Using a continuous glucose monitor (CGM) in pregnancy may bring meaningful benefits when diabetes is part of the picture. That is the message of a new study. The researchers saw fewer babies with a high birth weight and fewer admissions to the neonatal intensive care unit (NICU). They also recorded fewer episodes of low blood sugar in newborns.
However, the researchers stress that the evidence for gestational diabetes is still incomplete. We need much larger studies to learn who benefits most from this technology. We also need them to set the right glucose targets.
A continuous picture instead of a few snapshots a day
Managing diabetes in pregnancy calls for an especially delicate balance. High blood sugar can affect how the baby grows and raise the risk of complications. On the other hand, pushing blood sugar down too hard can increase the mother’s risk of hypoglycemia (low blood sugar).
Mayo Clinic’s page on gestational diabetes explains one such risk. High blood sugar may cause the baby to grow too large, which can make delivery more difficult.
Fingerstick testing goes by the abbreviation BGM (blood glucose monitoring). It is a one-time reading of blood sugar using a lancet and a meter. As a result, it captures only a single moment. It can therefore miss spikes after meals, dips during the night and swings between readings.
CGM stands for continuous glucose monitoring. It uses a tiny sensor worn on the skin. The sensor measures glucose in the fluid between the cells, around the clock.
This way, you can see which direction glucose is heading and how long it stays in the target range. You can also see when it strays out of range. With this information, you can fine-tune diet, physical activity and insulin doses precisely.
The new study: 17 trials and 2,672 pregnancies
The study is a systematic review and meta-analysis that appeared in September 2026 in a journal of the American Diabetes Association. The researchers analyzed 17 randomized controlled trials. Together, these trials included 2,672 pregnancies with diabetes.
The researchers compared two approaches. In one, women used CGM alongside fingerstick testing. In the other, monitoring relied on fingerstick testing alone.
The review included women whose diabetes existed before pregnancy, mainly type 1 and type 2. It also included women who developed diabetes during pregnancy (gestational diabetes).
What sets this review apart
This review has three notable methodological features.
First, the researchers harmonized how the trials defined newborn size. They defined macrosomia as a birth weight of 4,000 grams or more. Large for gestational age meant above the 90th percentile. Small for gestational age meant below the 10th percentile.
Second, they systematically graded the certainty of the evidence.
Third, they ran a trial sequential analysis. This method checks whether the data collected so far are enough to reach a firm conclusion.
Separating the types of diabetes matters. There are important differences between type 1, type 2 and gestational diabetes. They differ in how long the baby faces high blood sugar and in the need for insulin. They also differ in whether women can reach good control with diet alone.
A 45% lower risk of macrosomia
The most striking finding came from women with gestational diabetes. In this group, CGM in pregnancy came with a 45% relative reduction in the risk of macrosomia, that is, giving birth to a particularly large baby.
The relative risk was 0.55, with a confidence interval of 0.35 to 0.86. The researchers rated the certainty of this evidence as high.
Across the whole study population, macrosomia fell by about 33%. Here the relative risk was 0.67, with a confidence interval of 0.48 to 0.94. In this case, the researchers rated the certainty of the evidence as moderate.
The rate of babies who were large for gestational age also dropped. Across the whole population, the decline was about 23%, with a relative risk of 0.77. In gestational diabetes, the decline was about 35%, with a relative risk of 0.65.
Still, the researchers rated the certainty of this evidence as very low. That is because the upper limit of the confidence interval sits very close to the line of no effect. So the statistical significance could change as more trials appear.
Why macrosomia matters
Reducing macrosomia may have real clinical value. A high birth weight increases the risk of a difficult delivery and shoulder dystocia. It also raises the risk of birth injuries and cesarean delivery. In addition, it goes hand in hand with metabolic complications in the newborn.
Keep in mind that this figure describes a relative reduction in risk. The absolute benefit for each woman depends on her baseline risk. It also depends on her level of glucose control and other features of her pregnancy.
CGM in pregnancy: fewer NICU admissions and less newborn hypoglycemia
CGM use also went along with a 26% lower risk of NICU admission for the newborn. The relative risk was 0.74, with a confidence interval of 0.60 to 0.91. The researchers rated the certainty of the evidence as moderate. This finding appeared in the analysis of all pregnancies.
Among women whose diabetes existed before pregnancy, there was an additional finding. The risk of hypoglycemia in the newborn fell by 29%. The relative risk was 0.71, with a confidence interval of 0.51 to 0.98.
Moreover, in this group the time the mother’s glucose spent in the target range rose by 5.68%. Time below range also went down. Among women with gestational diabetes, fasting glucose dropped by 3.56 milligrams per deciliter.
The likely explanation is simple. A CGM reveals swings that one-off readings miss. When you catch after-meal spikes or nighttime dips early, small and precise adjustments become possible.
Which outcomes did not change?
The study found no significant difference in rates of cesarean delivery or preterm birth. Similarly, there was no difference in preeclampsia or in high blood pressure disorders of pregnancy. Nor was there a difference in the mother’s weight gain.
For some of these outcomes, the certainty of the evidence is actually high. In other words, this is a solid finding of no effect, not simply a lack of data.
This means a CGM in pregnancy may improve several important outcomes. Even so, it does not remove all the risks of a pregnancy with diabetes on its own. Care still relies on a tailored diet, physical activity, medical follow-up and sometimes insulin.
One more point deserves attention. In some trials, by the end of follow-up, more women in the sensor group needed insulin than women in the control group. The data were too sparse for a formal analysis. Yet the direction makes sense: a sensor that reveals out-of-range readings sometimes leads to treatment that the team would not otherwise have offered.
The signal to watch: babies small for gestational age
The analysis showed a trend toward more babies who were small for gestational age among sensor users. The relative risk was 37% higher. However, the difference did not reach statistical significance. The confidence interval ranged from 0.96 to 1.95, so it also included the possibility of no real difference.
The researchers rated this finding as moderate certainty of no effect. Nevertheless, the sequential analysis revealed a trend. The cumulative data curve was moving toward more small babies in the sensor group. In other words, we cannot yet rule out a clinically meaningful increase.
This finding does not prove that the sensor slows the baby’s growth. But it does raise an important question. Could the effort to reach very low glucose values tip the balance too far the other way?
The practical message is clear. Aim for good, personalized control and keep track of the baby’s growth. An especially low number is not necessarily the right target.
Why are the researchers still cautious?
The sequential analysis checks whether the total number of participants is already large enough for a stable conclusion. It answers a simple question: could a result that looks significant today change tomorrow?
The answer shows that a lot of information is still missing. For macrosomia, the data reach only about 16% of the required sample size. In gestational diabetes, they reach only about 12%. For small for gestational age, the figure is about 15%. The most mature outcome is, in fact, large for gestational age, at about 55%.
The researchers also estimated how many pregnancies with gestational diabetes a study would need. Macrosomia would require about 10,700 pregnancies. Large for gestational age would require about 6,600. Small for gestational age would require about 6,160.
This does not contradict the high certainty rating for macrosomia. The certainty rating looks at the quality and consistency of the existing data. The sequential analysis, by contrast, asks whether enough events have accumulated. That is why the researchers recommend looking at the overall trends rather than fixating on p-values.
In gestational diabetes, further improvement is hard to show
One interesting finding concerned time in range. In gestational diabetes, even the control groups were in the target range at least 87% of the time.
When the starting point is already very good, there is little room left to show an advantage. The researchers believe this is a ceiling effect, not a lack of benefit. Their conclusion is blunt: time in range is a poor measure of effectiveness in gestational diabetes.
Other measures may better reflect the link to the baby’s growth. Examples include average glucose and time above a certain level. Nighttime glucose levels and after-meal swings are also candidates.
CGM in pregnancy: what do current guidelines say?
The American Diabetes Association Standards of Care for 2026 recommend CGM in pregnancy with type 1 diabetes. Using it helps women reach glucose targets, above all time in range. In addition, it may lower the risk of a large-for-gestational-age baby and of newborn hypoglycemia.
This recommendation rests in part on the CONCEPTT trial. In that trial, pregnant women with type 1 diabetes who used a sensor spent more time in the target range than women in the control group. Their babies also had fewer complications.
For type 2 diabetes and gestational diabetes, doctors still make the decision case by case. The new study strengthens the case for benefit, mainly in reducing macrosomia. Still, it does not yet justify using a sensor for every woman with gestational diabetes. Indeed, randomized trials such as GRACE and DipGluMo reached different results in gestational diabetes.
The considerations include the level of control, the need for insulin and glucose swings. The risk of hypoglycemia and the baby’s growth rate also matter. Finally, the care team weighs the woman’s ability to use the data and the cost of the system.
The sensor is a treatment tool, not just a measuring device
The main value of a CGM is not just the numbers on the screen. The benefit comes when the data turn into the right treatment decisions.
The system can show how a particular meal affects glucose. It shows whether a light walk after eating blunts the rise. In addition, it reveals nighttime dips and shows whether the insulin dose needs to change. This is the same principle that guides personalized medicine: first understand the mechanism, and only then intervene.
How to choose a CGM for pregnancy
The study did not show a clear advantage for any particular type of sensor. For this reason, the researchers recommend choosing based on availability, cost and the patient’s preference.
In Israel, several real-time continuous monitoring systems are available today. One of them is the SUGAR360 sensor from Trupharm, based on SIBIONICS GS1 technology. It is a thin sensor that you wear on the upper arm for up to 14 days.
According to the manufacturer, its measurement accuracy is 8.7% on the MARD scale. MARD (mean absolute relative difference) is the average gap between the sensor reading and the blood glucose value. The sensor provides real-time alerts for rises and drops. In addition, the app lets you share reports with your care team. The ability to share data with your doctor is what turns the sensor into a treatment tool.
Despite these advantages, the sensor does not replace medical judgment. Readings may lag slightly behind blood glucose. Therefore, you sometimes still need a fingerstick test. For example, test when how you feel does not match the reading, or when glucose is changing quickly.
Choosing a system, and above all setting glucose targets in pregnancy, is something to do together with your medical team. If you are planning a pregnancy after treatment with weight loss medication, you may also want to read about incretins, fertility and pregnancy. It is also worth reading about what happens to your weight when you stop GLP-1 drugs before pregnancy.
The bottom line on CGM in pregnancy
The new study paints an encouraging picture. Adding CGM to pregnancy monitoring in women with diabetes may reduce macrosomia. It may also reduce NICU admissions for newborns. In addition, it may reduce hypoglycemia in babies born to mothers with pre-existing diabetes.
The evidence for benefit is strongest in pregnancy with type 1 diabetes. In gestational diabetes, the findings are promising and strengthen the sensor’s place as a treatment option. At the same time, they underline the need for a personalized approach, close tracking of the baby’s growth and further large studies.
This technology lets us see what one-off readings miss. The next challenge is to use that information to reach precise and safe control, for both mother and baby. After delivery, my guide to postpartum weight loss can help with the next step.
The information in this article is general medical information. It is not personal medical advice, a diagnosis or a treatment recommendation. Any decision about glucose monitoring in pregnancy and about glucose targets belongs with your medical team.
Disclosure: this article mentions a commercial product available in Israel.
Sources
7. Diabetes Technology: Standards of Care in Diabetes-2026
15. Management of Diabetes in Pregnancy: Standards of Care in Diabetes-2026
