Maternal diabetes following gestational diabetes: Does multiple gestation multiply the risk?

To assess whether the risk for future diabetes is higher among women diagnosed with gestational diabetes (GD) during twin versus singleton gestations.

5][6] The lifetime risk for diabetes is up to seven times greater among women diagnosed with GD as compared with women without GD. 7e rates of multiple gestations, comprising mainly twin gestations, have also increased in recent decades, mainly due to the increasing rates of pregnancies conceived following infertility treatments and older maternal ages. 6,8Multiple gestations entail greater challenges to most maternal systems.[11] Multiple gestations are associated with greater risk for pregnancy complications, including pregnancy-related hypertension diseases, intrahepatic cholestasis, and postpartum hemorrhage. 12,13Regarding GD, some but not all studies have found that GD diagnoses are higher in women with multiple as compared with singleton gestations, varying between 5% and 9%, as compared with 10.1% in singleton and twin gestations, respectively, with higher rates in triplet pregnancies (12.8%). 14,15agnostic criteria for GD is indifferent in singleton and multiple gestation and usually based on a 50-g glucose challenge test (GCT) followed by the 100-g oral glucose tolerance test (OGTT), which is considered the gold standard for GD diagnosis. 6,16ile the association between GD and future maternal health has been widely studied, 6,17 less is known regarding this possible association among mothers diagnosed with GD during twin gestation. 18Since the body load is greater in twin gestation, the placenta is bigger, and insulin resistance increases, GD may be less indicative of maternal insufficient insulin tolerance or adequate beta cell reserve (pancreatic insufficiency) in the long term; therefore, the risk for future maternal morbidities may be lower if GD was diagnosed in twin gestation.

| OBJEC TIVE
To compare women with GD diagnosis during twin versus singleton gestations, in terms of risk for maternal long-term diabetes.

| ME THODS
A retrospective cohort study was conducted at the Soroka University Medical Center, the largest birth center in Israel, with >16 000 births per year in recent years.The study included all women who delivered between the years 1991 and 2021 and had at least one GD diagnosis.
Only women insured by Clalit, the largest health maintenance organization in Israel, were included, since their medical records and laboratory test results were available for analysis.
For each participating woman, the pregnancy with the first GD diagnosis was considered the index pregnancy.Women diagnosed with GD during multiple gestations were compared with women diagnosed with GD during singleton gestations.
In pregnancies complicated with GD, a 1-h 75-g OGTT post partum is recommended within 3 months following delivery, and an Hba1C test may also be performed. 19Due to low adherence with the 1-h 75-g post partum yielding high rates of missing values, the study focused on Hba1C test results.According to the American Diabetes Association, Hba1C levels 5.7% to 6.4% indicate prediabetes, and higher levels are indicative of diabetes. 20Hba1C levels represent blood sugar levels over the preceding 2 to 3 months.
The following outcomes were investigated, which were based on laboratory tests from the computerized medical records.The levels of the first postpartum and maximal (highest) postpartum results during the follow up period of up to 30 years, were categorized as follow: Hba1C levels 5.7% to 6.4% (prediabetes), >6.5% (diabetes), and Hba1C > 5.7% (prediabetes and diabetes).An additional analysis excluded Hba1C results, which were tested within 3 months post partum, since this most likely represents prenatal glucose levels.
Background characteristics were compared between the groups, including the mode of conception (spontaneous vs following infertility treatments), ethnicity (Bedouin Arab vs Jewish), immigration status (born in Israel or abroad), obesity (prepregnancy body mass index >30 kg/m 2 ), pregnancy-related hypertensive disorders including pre-eclampsia or eclampsia, and age.Characteristics that were statistically different between the study groups (P < 0.05) were adjusted for in the multivariable analysis.
The cumulative survival rate (time to Hba1C > 6.4) between the two study groups was compared using Kaplan-Meier and log-rank tests.
Multivariable logistic models (Table 2) for high first Hba1C level, according to the abovementioned cut points, as well as Cox proportional multivariable analyses (for highest Hba1C levels during the entire follow-up period, according to the abovementioned cut points), were used to compare the risk for any of the outcomes between the two groups, while adjusting for confounding variables.
The study protocol has been approved by the Soroka University Medical Center's institutional review board committee (approval #0438-15-SOR), and informed consent was exempt.
Statistical analysis was performed using SPSS version 29 software (IBM).
Results of Hba1C levels in the different times are presented in Table 2.The mean first Hba1C levels were 6.18% ± 1.7% and 5.95% ± 1.3%, among singleton and twin pregnancies, respectively (P = 0.007).The incidence of prediabetes (Hba1C 5.7-6.4%),diabetes (Hba1C > 6.4%), or either of them (Hba1C > 5.7%) was lower among women with GD diagnosis during twin pregnancy (although the finding regarding Hba1C 5.7% to 6.4% was with marginal significance).Results were similar in the multivariable logistic models, which adjusted for maternal age at index pregnancy and ethnicity.
Mean follow-up time starting at the index pregnancy was 12.25 ± 9.3 years, with a median follow-up of 12.34 years (range, 0-30 years).No differences were found in years of follow-up between the two study groups (12.25 ± 9.3 and 12.21 ± 9.1 for singleton and twin pregnancies, respectively; P = 0.99).
The mean highest Hba1C levels measured during the entire follow-up period was 7.21% ± 2.6% and 6.71% ± 2.2%, among singleton and twin pregnancies, respectively (P < 0.001).The incidence rates of Hba1C 5.7% to 6.4% (prediabetes), Hba1C > 5.7% (prediabetes or diabetes), or Hba1C > 6.5% (diabetes) were all lower among women with a GD diagnosis during a twin pregnancy as compared with a singleton pregnancy.The differences in survival between the study groups with Hba1C > 6.4 can be seen in the Figure 1, which presents a better survival rate among twins compared with singletons (log-rank, P = 0.016).
The risk remained significantly lower in the Cox survival model, which adjusted for maternal age at index pregnancy, number of pregnancies postindex pregnancy, fertility treatments, and ethnicity (adjusted hazard ratio, 0.77 [95% CI, 0.59-0.99).

| DISCUSS ION
In this large population-based cohort with a long follow-up period of up to 30 years, women diagnosed with GD during twin pregnancy were at lower risk for future diabetes, as compared with mothers diagnosed with GD during singleton gestation.The estimates changed only marginally after adjustment for confounding variables, and when focusing on the Hba1C tests performed >3 months post partum.
F I G U R E 1 Cumulative survival (in days) from the index pregnancy until first hemoglobin A1c (Hba1c) >6.4.GD, gestational diabetes.
The pathogenesis of GD involves failure of maternal pancreas and beta cells to compensate for the physiologic increase in insulin resistance during pregnancy. 13GD diagnosis is a well-established risk factor, or rather an early indicator of maternal diabetes, and they are genetically linked. 21Diabetes is associated with increased risk for many comorbidities, mainly cardiac, in addition to other health complications.Diabetes impairs quality of life and is a risk factor for mortality. 3,22,23Many diabetes complications are preventable with early detection and treatment, including lifestyle changes such as diet and physical activity and monitoring and maintaining blood glucose levels and Hba1C within normal range. 23men diagnosed with GD are otherwise relatively young and healthy, and the pregnancy is considered a window of opportunity to identify women at risk for future morbidities. 6GD diagnosis is usually made following an initial 50-g GCT, followed by a confirmative 100-g OGTT, which is considered the gold standard for GD diagnosis. 24Both tests evaluate how the body handles a load of glucose, in terms of the pace of the lowering of serum glucose levels following the single load.
According to American College of Obstetricians and Gynecologists guidelines, 25 high OGTT levels (on two of the four measurements) are indicative of GD, and this is indifferent to singleton or multiple gestations.It is possible that since the body load is greater in twin gestations, more women are diagnosed with GD during twin gestation; however, in these cases it does not necessarily reflect a greater risk for future diabetes.Based on current findings, GD diagnosis in women with twin gestation is less valid in predicting future maternal diabetes as compared with GD diagnosis in women with singleton gestation.The differences in the threshold for GD diagnosis in singleton versus multiple gestations has been a concern and studied in many settings, wherein some of the studies are summarized in a recent meta-analysis. 14st studies thus far have compared the results of the initial GCT screening cut points against the OGTT results, the gold standard for GD diagnosis.Some studies have supported a lower threshold for GD diagnosis in twin versus singleton gestations, 26 while others have suggested that the threshold should be higher in multiple gestations. 27,28rezowsky et al. have shown that among women with GD in twin gestation, good glycemic control is not associated with a reduction in the risk of GD-related complications, but may even increase the risk of a small-for-gestational age newborn. 29These findings differ from singleton gestations complicated with GD, in which good glycemic control reduces the risk for GD-related complications and suggests that there is an overdiagnosis of GD in twin gestation.GD in women with either singleton or multiple gestations has been associated with adverse pregnancy outcomes, 30 while other studies have found greater risk for complications in women with twin versus singleton GD pregnancies. 31,32e long follow-up period and the availability of Hba1C laboratory results, which limits the possibility of misclassification, are additional study strengths.This enabled a greater sample size, as compared with studying an association with postpartum 75 g OGTT, which is commonly recommended in many countries, since adherence with the OGTT is relatively low in Israel.
One of the limitations of the study is the unavailable data on glycemic control, insulin dependence, and levels of the OGTT results.
Although the type of GD was similar in twins and singletons, these variables may have enabled better identification of women at higher risk for long-term diabetes, and the identification of an optimal cut point for the GD diagnosis in twin gestation, yielding a higher predictive value based on maternal health in the long term.Future studies are recommended to research the optimal cut point for GD diagnosis among women with twin gestation.

| CON CLUS IONS
With the increasing incidence of both GD and multiple gestations, it is important to identify women with such pregnancy complications in twin gestation.Obviously, the aim of the diagnosis of GD is intended to improve fetal outcome, and not long-term maternal health.Still, differential cut points for GD diagnosis based on singleton or twins gestations will enable better secondary prevention for women diagnosed with GD during twin gestation, in addition to the possibility for unbiased studies on the long-term maternal and offspring long-term morbidities associated with GD diagnosis.
Findings from our study add support for the possible need for different diagnostic criteria for GD in singleton and twin gestations, as suggested by Berezowsky et al., as the currently used threshold is less valid among twin gestations when considering future maternal diabetes risk.The main strength of the current study is the large sample of women, specifically among twin gestations, which have not been adequately studied.The association between maternal GD and the risk for other pregnancy complications as well as long-term diabetes is clear and supported by many studies.Mothers of twins, however, have been less studied.The limited studies on GD in twin pregnancies and the risk for other pregnancy complications and long-term maternal diabetes have shown inconsistent results.In some studies,
TA B L E 1Abbreviations: CI, confidence interval; GD, gestational diabetes; OR, odds ratio; SD, standard deviation.TA B L E 2Hba1C results by study group.aOdds ratio (OR) adjusted for maternal age at index pregnancy and ethnicity.bHazard ratio (HR) adjusted for maternal age at index pregnancy, number of pregnancies postindex pregnancy, fertility treatments, and ethnicity.