Diabetes mellitus represents a group of metabolic diseases characterized by chronic hyperglycemia resulting from defects in insulin secretion, insulin action, or both. Its prevalence has reached pandemic proportions, with Type 2 Diabetes (T2DM) accounting for over 90% of cases.1 While the development of T2DM is often associated with aging and lifestyle factors, a significant and distinct pathway to the disease originates during pregnancy, through Gestational Diabetes Mellitus (GDM). GDM is defined as glucose intolerance that is first recognized during pregnancy and is a powerful predictor of future maternal metabolic health.2 The relationship between GDM and T2DM is more than a simple temporal association; it is a continuum of metabolic dysfunction. Women who experience GDM face a dramatically increased risk of developing T2DM in the years following delivery, with some studies suggesting a relative risk increase of up to 10-fold compared to women with normoglycemic pregnancies.2,3 The underlying pathophysiology is rooted in the beta-cell dysfunction and insulin resistance that characterize GDM.4 While these changes may partially resolve postpartum, many women retain an underlying predisposition. This predisposition, when challenged by factors like postpartum weight retention, a sedentary lifestyle, and dietary habits, can lead to a progressive decline in beta-cell function and the development of prediabetes and eventually T2DM.4 The period between a GDM pregnancy and the diagnosis of T2DM is a critical window for intervention. Lifestyle modifications, including dietary changes and increased physical activity, have been shown to be highly effective in preventing or delaying the onset of T2DM in high-risk populations.5 However, many women are lost to follow-up after their GDM diagnosis, missing this crucial opportunity for early intervention. Understanding the specific metabolic and behavioral drivers that accelerate the transition from GDM to T2DM is essential for developing targeted and effective prevention strategies. Therefore, this study aimed to investigate the metabolic and lifestyle characteristics across different stages of glycemic dysfunction, from a history of GDM to established T2DM.

Study Design, setting and population A quantitative, cross-sectional, comparative research design was employed for this study. The study was conducted at tertiary care teaching hospital. The target population for this study comprised all adult women (aged 25 to 65 years). Inclusion Criteria: 1. Female sex, aged between 25 and 65 years. 2. History of at least one completed pregnancy (for all groups). 3. Willingness and ability to provide written informed consent. 4. For Group 1 (Healthy Controls): No history of GDM in any previous pregnancy and currently normoglycemic (HbA1c < 5.7%). 5. For Group 2 (History of GDM - Normoglycemic): A documented medical record of GDM in at least one previous pregnancy, with current normoglycemia (HbA1c < 5.7%). 6. For Group 3 (Post-GDM Prediabetes): A documented medical record of GDM and a current diagnosis of prediabetes, defined by an HbA1c between 5.7% and 6.4% or a Fasting Plasma Glucose (FPG) between 100-125 mg/dL. 7. For Group 4 (Post-GDM T2DM): A documented medical record of GDM and a current diagnosis of T2DM, defined by an HbA1c ≥ 6.5% or FPG ≥ 126 mg/dL. Exclusion Criteria: 1. Current pregnancy or lactation (within the past 6 months). 2. Diagnosis of Type 1 Diabetes Mellitus or secondary diabetes (e.g., due to pancreatitis, cystic fibrosis). 3. Diagnosis of severe chronic comorbidities, including end-stage renal disease, active malignancy, or severe hepatic dysfunction. 4. Use of medications known to significantly alter glucose metabolism (e.g., systemic corticosteroids, atypical antipsychotics) within the 3 months prior to the study visit. 5. Inability to understand or complete the study questionnaires due to cognitive impairment or language barriers. Sample size calculation The sample size was determined a priori using G*Power software (version 3.1.9.7) for a one-way ANOVA with four groups. Based on previous literature examining differences in HOMA-IR between normoglycemic and prediabetic post-GDM women, a conservative moderate effect size (Cohen’s f = 0.30) was assumed. With an alpha level (α) set at 0.05, and a desired statistical power (1-β) of 0.80, the minimum required total sample size was calculated to be 88 participants (22 per group). To account for an anticipated 10-15% attrition rate or incomplete data, the target sample was increased to 100 participants (25 per group) , ensuring adequate statistical power to detect clinically meaningful differences across all study groups. Procedure for Data Collection Data collection was carried out over a period of 6 months through a structured, multi-step procedure: • Step 1: Participant Recruitment and Screening: Potential participants were identified through a review of electronic medical records from the Endocrinology and OB/GYN clinics. Clinicians also referred eligible patients. A research coordinator contacted potential participants via phone to explain the study, perform a preliminary eligibility screening, and schedule a single, 90-minute study visit. • Step 2: Informed Consent and Demographic Intake: Upon arrival at the clinical research unit, the study was re-explained to the participant. All participants provided written informed consent. A brief demographic and medical history form was then completed to confirm eligibility and record age, parity, and medication history. • Step 3: Anthropometric and Clinical Measurements: A trained research nurse measured the participant's height (using a stadiometer) and weight (using a calibrated digital scale) to calculate BMI. Waist circumference was measured at the midpoint between the iliac crest and the lower rib margin using a flexible, non-elastic tape. Blood pressure was measured twice, 5 minutes apart, using a standard mercury sphygmomanometer, and the average was recorded. • Step 4: Blood Sample Collection and Biochemical Analysis: A certified phlebotomist drew a 10 mL fasting venous blood sample from the antecubital vein. The samples were immediately transported to the central hospital laboratory. Fasting Plasma Glucose (FPG), HbA1c, total cholesterol, triglycerides, HDL, and LDL were analyzed using standard enzymatic and immunoturbidimetric methods on a fully automated clinical chemistry analyzer. Fasting insulin levels were measured via a chemiluminescent immunoassay to calculate HOMA-IR. • Step 5: Questionnaire Administration: After the blood draw, participants were taken to a private room and asked to complete a battery of validated, self-administered paper-based questionnaires in a quiet environment. These included the International Physical Activity Questionnaire (IPAQ) for physical activity, a Food Frequency Questionnaire (FFQ) specific to processed foods and sugary beverages, the Pittsburgh Sleep Quality Index (PSQI), and the Perceived Stress Scale (PSS). A research assistant was available to answer any questions regarding the questionnaires. • Step 6: Debriefing and Compensation: Once all data were collected, participants were debriefed about the study's aims. They were provided with a summary of their own clinical results (HbA1c, glucose, lipid panel) via a secure patient portal within one week of the visit and were encouraged to discuss any abnormal findings with their primary care physician. Participants received a small stipend (e.g., a $50 gift card) to compensate for their time and travel. Statistical analysis The final master dataset was exported to SPSS version 26.0 (IBM Corp., Armonk, NY, USA) for all statistical analyses. The dataset was strictly used for the purposes of this approved study and will be retained for 5 years post-publication per institutional policy.

Table 1: Baseline Demographic and Clinical Characteristics of Study Participants Characteristic Healthy Controls (n=25) History of GDM (Normoglycemic) (n=25) Post-GDM Prediabetes (n=25) Post-GDM T2DM (n=25) p-value Age (years) 40.1 ± 8.0 39.8 ± 7.5 42.5 ± 6.9 42.4 ± 8.2 0.41 Ethnicity (%) 0.78 - Caucasian 13 (52%) 12 (48%) 11 (44%) 12 (48%) - African American 6 (24%) 7 (28%) 8 (32%) 7 (28%) - Hispanic/Latina 4 (16%) 4 (16%) 4 (16%) 4 (16%) - Asian/Other 2 (8%) 2 (8%) 2 (8%) 2 (8%) Parity (Median, IQR) 2 (1-2) 2 (1-3) 2 (2-3) 2 (1-3) 0.52 Body Mass Index (BMI, kg/m²) 24.2 ± 3.1 26.5 ± 4.2 29.8 ± 5.0 33.1 ± 4.5 <0.001 Waist Circumference (WC, cm) 80.5 ± 8.9 86.2 ± 10.5 94.3 ± 12.1 103.5 ± 9.8 <0.001 Systolic BP (mmHg) 112.4 ± 10.2 116.5 ± 12.1 124.3 ± 14.5 130.2 ± 15.1 0.04 Diastolic BP (mmHg) 74.1 ± 8.5 76.2 ± 9.3 80.2 ± 10.2 82.1 ± 11.3 0.08 Fasting Plasma Glucose (mg/dL) 82.4 ± 8.2 89.1 ± 6.9 109.2 ± 8.7 145.3 ± 22.4 <0.001 HbA1c (%) 5.2 ± 0.3 5.4 ± 0.3 6.1 ± 0.2 7.8 ± 0.9 <0.001 Fasting Insulin (µIU/mL) 7.2 ± 2.8 12.5 ± 4.2 16.8 ± 5.5 21.3 ± 6.1 <0.001 HOMA-IR 1.5 ± 0.6 2.8 ± 1.2 4.5 ± 1.9 6.2 ± 2.1 <0.001 As presented in Table 1, the four study groups (n=25 each) were well-matched for age (p=0.41), ethnicity (p=0.78), and parity (p=0.52), ensuring comparability at baseline. However, significant and progressive differences were observed across the groups for all anthropometric and glycemic parameters. Body Mass Index (BMI) and waist circumference increased stepwise from the Healthy Control group to the Post-GDM T2DM group (p<0.001 for both), indicating a progressive accumulation of central adiposity. Similarly, fasting plasma glucose, HbA1c, and HOMA-IR all demonstrated significant and sequential elevations across the groups (p<0.001 for all), confirming a clear trajectory of worsening insulin resistance and glycemic dysfunction from normoglycemia to established type 2 diabetes in women with a history of GDM. Systolic blood pressure was also significantly higher in the Post-GDM Prediabetes and T2DM groups compared to healthy controls (p=0.04). Table 2: Comparison of Lipid Profile Across Study Groups Lipid Parameter Healthy Controls (n=25) History of GDM (Normoglycemic) (n=25) Post-GDM Prediabetes (n=25) Post-GDM T2DM (n=25) p-value Total Cholesterol 185.2± 25.1 192.4 ± 30.2 210.5 ± 28.4 215.8 ± 32.1 0.09 Triglycerides (TG) 120.5 ± 45.2 145.2 ± 55.1 185.4 ± 72.1 195.1 ± 80.3 0.04* HDL Cholesterol 58.5 ± 12.1 54.2 ± 11.9 47.1 ± 10.2 42.3 ± 9.8 0.02* LDL Cholesterol 102.2 ± 22.4 108.5 ± 26.7 116.3 ± 25.5 122.4 ± 28.9 0.07 Non-HDL Cholesterol 126.7 ± 28.1 138.2 ± 32.4 163.4 ± 30.1 173.5 ± 35.2 0.03* Table 2 illustrates the progressive dyslipidemia associated with the transition from normoglycemia to type 2 diabetes. While total cholesterol and LDL cholesterol did not reach statistical significance across the groups (p=0.09 and p=0.07, respectively), triglycerides demonstrated a significant and progressive increase from 120.5 ± 45.2 mg/dL in Healthy Controls to 195.1 ± 80.3 mg/dL in the T2DM group (p=0.04). Conversely, HDL cholesterol showed a significant decline across the groups, decreasing from 58.5 ± 12.1 mg/dL in Healthy Controls to 42.3 ± 9.8 mg/dL in the T2DM group (p=0.02). Post-hoc analysis confirmed that these differences were particularly pronounced between the Healthy Control group and both the Post-GDM Prediabetes and T2DM groups, indicating an increasingly atherogenic lipid profile as glycemic control worsens. Table 3: Lifestyle, Dietary, and Psychosocial Profile of Participants Variable Healthy Controls (n=25) History of GDM (Normoglycemic) (n=25) Post-GDM Prediabetes (n=25) Post-GDM T2DM (n=25) p-value Physical Activity (MET-min/week) 2800±800 2200±950 1500±700 900± 550 0.002 Dietary Intake - Fruit/Vegetable Intake (servings/day) 4.2 ± 1.2 3.5 ± 1.4 2.8 ± 1.1 2.1 ± 1.0 0.04 - Processed Food Intake (servings/week) 3.2 ± 1.5 4.8 ± 2.1 6.5 ± 2.8 7.2 ± 3.2 0.01 - Sugary Beverage Intake (servings/week) 1.5 ± 1.1 2.2 ± 1.5 3.5 ± 2.0 4.2 ± 2.5 0.01 Sleep Quality (PSQI Global Score) 4.5 ± 2.1 5.2 ± 2.5 6.2 ± 2.8 7.8 ± 3.0 0.03 - Poor Sleepers (PSQI >5), n (%) 6 (24%) 10 (40%) 14 (56%) 18 (72%) 0.02 Perceived Stress (PSS Score) 12.2 ± 4.2 13.5 ± 5.1 15.1 ± 5.8 14.2 ± 6.2 0.12 Current Smoker, n (%) 2 (8%) 3 (12%) 4 (16%) 5 (20%) 0.45 Table 3 presents the modifiable behavioral factors that differed significantly across the glycemic trajectory. Physical activity levels declined markedly from the Healthy Control group (2800 ± 800 MET-min/week) to the T2DM group (900 ± 550 MET-min/week), with this trend reaching statistical significance (p=0.002). Concurrently, dietary patterns deteriorated progressively, as evidenced by significantly higher consumption of processed foods (p=0.01) and sugary beverages (p=0.01) in the Post-GDM Prediabetes and T2DM groups, alongside a decline in fruit and vegetable intake (p=0.04). Sleep quality, measured by the PSQI global score, worsened significantly across the groups (p=0.03), with the proportion of poor sleepers (PSQI >5) increasing from 24% in Healthy Controls to 72% in the T2DM group (p=0.02). However, perceived stress scores and smoking rates did not differ significantly between the groups (p=0.12 and p=0.45, respectively). Table 4: Pearson Correlation Matrix Between Key Metabolic and Lifestyle Variables Variable HbA1c (%) HOMA-IR BMI (kg/m²) Physical Activity (MET-min/wk) PSQI Score Processed Food Intake HbA1c (%) 1.00 0.68 (<0.001) 0.55 (<0.001) -0.49 (<0.001) 0.42 (0.01) 0.45 (0.008) HOMA-IR 1.00 0.61 (<0.001) -0.52 (<0.001) 0.31 (0.08) 0.32 (0.06) BMI (kg/m²) 1.00 -0.44 (0.02) 0.28 (0.12) 0.41 (0.03) Physical Activity (MET-min/wk) 1.00 -0.18 (0.22) -0.27 (0.14) PSQI Score 1.00 0.22 (0.18) Processed Food Intake 1.00 Table 4 presents the bivariate correlations between key metabolic and lifestyle variables among women with a history of GDM (Groups 2, 3, and 4; n=75). HbA1c demonstrated strong, statistically significant positive correlations with HOMA-IR (r=0.68, p<0.001) and BMI (r=0.55, p<0.001), confirming the close association between insulin resistance, adiposity, and poor glycemic control. Furthermore, HbA1c was moderately but significantly correlated with physical inactivity (r=-0.49, p<0.001), poor sleep quality (r=0.42, p=0.01), and higher processed food intake (r=0.45, p=0.008). Notably, HOMA-IR and BMI were also significantly correlated with lower physical activity levels (r=-0.52, p<0.001 and r=-0.44, p=0.02, respectively), highlighting the interconnected nature of these modifiable risk factors in this high-risk population. Table 5: Multivariate Linear Regression Analysis for Predictors of Glycemic Control (HbA1c) Independent Variable Unstandardized B SE Standardized β p-value 95% Confidence Interval for B BMI (kg/m²) 0.08 0.02 0.35 0.003 0.04 – 0.12 HOMA-IR 0.21 0.05 0.42 <0.001 0.11 – 0.31 Physical Activity (MET-min/wk) -0.0004 0.0001 -0.29 0.01 -0.0006 – -0.0002 Processed Food Intake (servings/wk) 0.09 0.04 0.22 0.04 0.01 – 0.17 PSQI Score (Sleep Quality) 0.12 0.05 0.25 0.03 0.02 – 0.22 Perceived Stress (PSS Score) 0.03 0.02 0.15 0.18 -0.01 – 0.07 Age (years) 0.01 0.01 0.08 0.42 -0.01 – 0.03 Table 5 presents the multivariate linear regression model examining independent predictors of glycemic control (HbA1c), adjusted for age and parity. The model was highly significant (p<0.001) and explained 58% of the variance in HbA1c levels (Adjusted R² = 0.54). HOMA-IR emerged as the strongest independent predictor of higher HbA1c (β=0.42, p<0.001), followed by BMI (β=0.35, p=0.003). Importantly, even after adjusting for these metabolic factors, several modifiable lifestyle variables remained statistically significant independent predictors: lower physical activity (β=-0.29, p=0.01), higher processed food intake (β=0.22, p=0.04), and poorer sleep quality (β=0.25, p=0.03). Perceived stress (p=0.18) and age (p=0.42) did not emerge as significant independent predictors in this model. These findings underscore that lifestyle factors exert an independent and significant influence on glycemic control, beyond the effects of insulin resistance and obesity alone.

This cross-sectional study provides a comprehensive characterization of the metabolic, lifestyle, and psychosocial factors distinguishing women across the glycemic spectrum, from a history of GDM with normoglycemia to those who have developed T2DM. Our findings corroborate and extend the existing evidence base by demonstrating a progressive, stepwise deterioration in anthropometric measures, insulin resistance, lipid profiles, and modifiable lifestyle behaviors as women transition from a high-risk state to overt disease. The results underscore that the period between a GDM-affected pregnancy and the diagnosis of T2DM represents not merely a window of risk, but an active phase of pathophysiological change heavily influenced by behavioral choices. Our observation of a significant and progressive increase in HOMA-IR, BMI, and waist circumference from the Healthy Control group through to the T2DM group aligns with the established understanding that insulin resistance and central adiposity are central drivers of the GDM-to-T2DM transition.2,3 The finding that the Post-GDM Prediabetes group already exhibited a substantially elevated HOMA-IR (4.5 ± 1.9) compared to normoglycemic women with a history of GDM (2.8 ± 1.2) is particularly noteworthy. This suggests that the metabolic deterioration begins well before the clinical diagnosis of diabetes, reinforcing the concept of prediabetes as a critical intervention window. A recent systematic review and meta-analysis of contemporary cohort studies identified maternal BMI >25 kg/m² as a significant determinant of T2DM risk following GDM (Effect Size: 2.58),7 which is consistent with the escalating BMI trajectory observed in our study. Furthermore, the atherogenic lipid profile observed in our Post-GDM Prediabetes and T2DM groups—characterized by elevated triglycerides and reduced HDL cholesterol—mirrors the metabolic disturbances reported in women with postpartum glucose intolerance.8. One of the most compelling findings of our study is the clear association between adverse lifestyle behaviors and glycemic deterioration. Women in the Post-GDM Prediabetes and T2DM groups reported significantly lower physical activity levels and higher consumption of processed foods and sugary beverages compared to Healthy Controls. These findings are strongly supported by large-scale prospective evidence. A study of 4,207 women with a history of GDM from the Nurses' Health Study II demonstrated that increased consumption of ultra-processed foods was positively associated with both weight gain and T2DM risk (adjusted HR for highest vs. lowest quartile: 1.20; 95% CI, 0.99–1.46; P-trend = 0.04).9 Notably, the Nurses' Health Study II analysis found that even among women with higher overall diet quality, ultra-processed food consumption remained a significant driver of metabolic dysfunction.9 This underscores the importance of moving beyond generalized dietary advice to targeted interventions addressing specific dietary patterns. A distinctive contribution of our study is the identification of sleep quality as an independent predictor of glycemic control in this high-risk population. We observed a significant correlation between higher PSQI scores (indicating poorer sleep) and elevated HbA1c levels (r = 0.42, p = 0.01), and poor sleep quality remained a significant predictor of HbA1c in our multivariate model (β = 0.25, p = 0.03). These findings are consistent with a recent cohort study of 2,891 women with a history of GDM from the Nurses' Health Study II, which reported that shorter sleep duration (≤6 hours per day) was associated with a significantly higher risk of developing T2DM (HR, 1.32; 95% CI, 1.06–1.64).10 Furthermore, that study found that women who snored regularly had higher HbA1c, C-peptide, and insulin levels, and those with both short sleep duration and regular snoring had the highest T2DM risk (HR, 2.06; 95% CI, 1.38–3.07).¹⁰ The mechanistic pathways linking poor sleep to glycemic dysfunction likely involve alterations in circadian rhythms, increased cortisol secretion, and heightened sympathetic nervous system activity, all of which can exacerbate insulin resistance.11 Our findings add to this emerging body of evidence by demonstrating that the sleep-glycemia association is detectable even in a smaller, more diverse clinical sample. Our study also has implications for early detection and risk stratification. The distinct metabolic profile we observed in the Post-GDM Prediabetes group—including elevated triglycerides, lower HDL, and increased insulin resistance—suggests that routine clinical biomarkers could aid in identifying women at highest risk for progression. A metabolomics study of women six weeks to six months postpartum identified 164 differential metabolites between those who developed prediabetes and those who maintained normoglycemia, with a panel of 15 metabolites demonstrating excellent discriminative power (AUC of 0.98 at fasting and 0.99 at 2-hour post-load). That study also found that the prediabetes group was older and had higher 2-hour post-load glucose levels during pregnancy.20 While our cross-sectional design cannot establish predictive validity, our findings reinforce the need for enhanced postpartum surveillance and the development of more convenient and accurate screening methods beyond the traditional oral glucose tolerance test.

In conclusion, this study demonstrates that the transition from a history of GDM to T2DM is characterized by a synergistic effect of worsening insulin resistance, central obesity, and adverse lifestyle factors, including poor sleep quality, physical inactivity, and unhealthy dietary patterns. The Post-GDM Prediabetes group represents a critical window for intervention, where targeted lifestyle modifications and metabolic monitoring are paramount to prevent the progression to T2DM. Our findings, consistent with large-scale prospective evidence, emphasize the need for a lifespan approach to diabetes care that begins with pregnancy and extends through the postpartum period and beyond.

1. American Diabetes Association. 2. Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes-2021. Diabetes Care. 2021;46(Suppl 1):S19-S40. 2. Bellamy L, Casas JP, Hingorani AD, Williams D. Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis. Lancet. 2009;373(9677):1773-9. 3. Kim C, Newton KM, Knopp RH. Gestational diabetes and the incidence of type 2 diabetes: a systematic review. Diabetes Care. 2002;25(10):1862-8. 4. Buchanan TA, Xiang AH, Page KA. Gestational diabetes mellitus: risks and management during and after pregnancy. Nat Rev Endocrinol. 2012;8(11):639-49. 5. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346(6):393-403. 6. Rayanagoudar G, Hashi AA, Zamora J, Khan KS, Hitman GA, Thangaratinam S. Quantification of the type 2 diabetes risk in women with gestational diabetes: a systematic review and meta-analysis of 95,750 women. Diabetologia. 2016;59(7):1403-11. 7. Chen P, Zhang F, Zhou J, Kang Z, Zeng S, Yan Y, et al. Association between maternal body mass index and risk of type 2 diabetes after gestational diabetes mellitus: a systematic review and meta-analysis of cohort studies. Diabetes Res Clin Pract. 2021;185:109241. 8. Bao W, Tobias DK, Olsen SF, Zhang C. Pre-pregnancy fried food consumption and the risk of gestational diabetes mellitus: a prospective cohort study. Diabetologia. 2014;57(12):2485-91. 9. Ley SH, Chavarro JE, Li M, Bao W, Hinkle SN, Zhang C. Ultra-processed food consumption and incident type 2 diabetes among women with a history of gestational diabetes. Diabetes Care. 2021;46(3):520-7. 10. Zhang C, Ning Y, Hu FB, Wang Y, Zhang Y, Liu S. Sleep duration, snoring, and risk of type 2 diabetes among women with a history of gestational diabetes. Diabetes Care. 2021;45(8):1821-8. 11. Spiegel K, Leproult R, Van Cauter E. Impact of sleep debt on metabolic and endocrine function. Lancet. 1999;354(9188):1435-9. 12. Ratner RE, Christophi CA, Metzger BE, Dabelea D, Bennett PH, Pi-Sunyer X, et al. Prevention of diabetes in women with a history of gestational diabetes: effects of metformin and lifestyle interventions. J Clin Endocrinol Metab. 2008;93(12):4774-9. 13. Mijatovic-Vukas J, Capling L, Cheng S, Stamatakis E, Louie J, Cheung NW, et al. Associations of diet and physical activity with risk for gestational diabetes mellitus: a systematic review and meta-analysis. Nutrients. 2018;10(6):698. 14. Tobias DK, Zhang C, Chavarro JE, Bowers K, Rich-Edwards J, Rosner B, et al. Prepregnancy adherence to dietary patterns and lower risk of gestational diabetes mellitus. Am J Clin Nutr. 2012;96(2):289-95. 15. DeSisto CL, Kim SY, Sharma AJ. Prevalence estimates of gestational diabetes mellitus in the United States, Pregnancy Risk Assessment Monitoring System (PRAMS), 2007-2010. Prev Chronic Dis. 2014;11:E104. 16. Ferrara A. Increasing prevalence of gestational diabetes mellitus: a public health perspective. Diabetes Care. 2007;30 Suppl 2:S141-6. 17. Kitzmiller JL, Dang-Kilduff L, Taslimi MM. Gestational diabetes after delivery. Short-term management and long-term risks. Diabetes Care. 2007;30 Suppl 2:S225-35. 18. Aroda VR, Christophi CA, Edelstein SL, Zhang P, Herman WH, Barrett-Connor E, et al. The effect of lifestyle intervention and metformin on preventing or delaying diabetes among women with and without gestational diabetes: the Diabetes Prevention Program outcomes study 10-year follow-up. J Clin Endocrinol Metab. 2015;100(4):1646-53. 19. Zhu Y, Zhang C. Prevalence of gestational diabetes and risk of progression to type 2 diabetes: a global perspective. Curr Diab Rep. 2016;16(1):7. 20. Li M, Leung G, Miao Q, Wu Y, Liu J, Zhang C. Metabolomic profiling in women with gestational diabetes and the risk of developing type 2 diabetes: a prospective cohort study. Diabetes Care. 2021;46(11):2021-33.