Comparative risk of type 2 diabetes development between women with gestational diabetes and women with impaired glucose tolerance over two decades: a multiethnic prospective cohort in New Zealand Authors: Dahai Yu, Hang Fu, Zhanzheng Zhao, Karen Pickering, John Baker, Richard Cutfield, Brandon J Orr-Walker, Gerhard Sundborn, Yamei Cai, Zheng Wang, Chengzeng Wang, David Simmons Published: 2024 ABSTRACT AbstractABSTRACTIntroductionTo evaluate the long-term risk of developing type 2 diabetes (T2D) among women with a history of gestational diabetes mellitus (GDM) compared with those with impaired glucose tolerance (IGT).Research design and methodsUsing data from a primary care dataset linked with multiple health registries, this longitudinal study analyzed demographics, clinical data, and lifestyle factors of women diagnosed with GDM or IGT, assessing T2D incidence over 25 years, using Cox regression models.ResultsWomen with GDM, especially those over 35 years of Māori ethnicity, or socioeconomic deprivation, exhibited an elevated risk of T2D compared with those with IGT. The first 5 years post partum emerged as a critical window for intervention.ConclusionsThis study underscores the importance of early, targeted post-GDM interventions to mitigate T2D risk. It highlights the necessity of personalized post-GDM interventions to reduce T2D incidence which consider age, ethnicity, and socioeconomic status to maximize effectiveness.Keywords: Diabetes, Gestational; Prediabetic State; Diabetes Mellitus, Type 2; Healthcare Disparities WHAT IS ALREADY KNOWN ON THIS TOPIC Prior research has established that women with a history of gestational diabetes mellitus (GDM) and those with impaired glucose tolerance (IGT) are at an increased risk of developing type 2 diabetes (T2D). Prior research has established that women with a history of gestational diabetes mellitus (GDM) and those with impaired glucose tolerance (IGT) are at an increased risk of developing type 2 diabetes (T2D). However, the long-term comparative risk between these two groups, particularly in a multiethnic cohort, remained unclear, necessitating further investigation. However, the long-term comparative risk between these two groups, particularly in a multiethnic cohort, remained unclear, necessitating further investigation. WHAT THIS STUDY ADDS This study elucidates that women with a history of GDM, especially those over 35 years, of Māori ethnicity, or experiencing socioeconomic deprivation, exhibit a significantly higher risk of transitioning to T2D compared with those with IGT, across a 25-year period in New Zealand. This study elucidates that women with a history of GDM, especially those over 35 years, of Māori ethnicity, or experiencing socioeconomic deprivation, exhibit a significantly higher risk of transitioning to T2D compared with those with IGT, across a 25-year period in New Zealand. It highlights the critical window for intervention within the first 5 years post partum. It highlights the critical window for intervention within the first 5 years post partum. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY The findings advocate for early, targeted postpartum interventions and underscore the importance of considering ethnic and socioeconomic factors in T2D risk assessment and management. The findings advocate for early, targeted postpartum interventions and underscore the importance of considering ethnic and socioeconomic factors in T2D risk assessment and management. This study informs future research directions and supports the development of tailored prevention strategies to mitigate the rising global burden of T2D, especially in high-risk populations identified through GDM screening. This study informs future research directions and supports the development of tailored prevention strategies to mitigate the rising global burden of T2D, especially in high-risk populations identified through GDM screening. Introduction Type 2 diabetes (T2D) contributes significantly to global morbidity and mortality, presenting a complex challenge to public health systems worldwide.1 Given the escalating global burden of T2D, understanding the complex pathways leading to T2D development is crucial. Research, such as the Diabetes Prevention Program, has indicated that women with a history of gestational diabetes mellitus (GDM) are particularly susceptible to progressing to T2D.2 Both GDM and impaired glucose tolerance (IGT) are recognized as conditions that confer an elevated risk of progressing to T2D, compared with individuals without these conditions.3 4 Previous studies have demonstrated that women with a history of GDM are at a significantly higher risk of developing T2D compared with those with normal glucose levels during pregnancy.5 Similarly, women with IGT also face an increased risk of T2D, often comparable to or slightly lower than those with GDM.6 7 However, the comparative long-term risk between these two groups remains less clear. A systematic review found that the relative risk of developing T2D in women with GDM was 7.43 times higher compared with those without GDM, highlighting the critical need for targeted interventions.8 Despite these findings, there is limited research directly comparing the long-term risk of T2D between women with GDM and those with IGT, particularly in multiethnic populations. Studies have often focused on homogeneous cohorts, limiting the generalizability of the results. Furthermore, most research has not adequately addressed how sociodemographic factors, such as age, ethnicity, and socioeconomic status, influence these risks differentially in diverse populations. However, the long-term comparative risk of T2D between women with GDM and those with IGT, especially in a multiethnic population, remains underexplored. Addressing this research gap, our study investigates whether GDM history entails a higher T2D development risk compared with IGT in a diverse New Zealand (NZ) cohort. This study explores whether prior GDM entails a higher risk of developing T2D compared with a history of IGT in a diverse NZ cohort. Analyzing long-term risk trajectories from GDM and IGT to T2D, this paper aims to inform effective prevention and management strategies, addressing this critical public health issue. Research design and methods Established in 1991, the Diabetes Care Support Service (DCSS) was introduced to improve diabetes management standards in general practices across South, East, and West Auckland.6 The DCSS database was linked with national registries, including death, hospitalization, pharmaceutical claims, cancer registration, and socioeconomic status databases, to create a comprehensive cohort of patients with T2D. This linkage provides a longitudinal, anonymized dataset encompassing demographics, clinical measurements, and medication records, validated by previous studies.7 9 NZ offers universal health coverage through a predominantly public healthcare system, ensuring patients receive free or subsidized healthcare. GDM diagnosed during the study period was identified through primary care records and included in the DCSS audit program. The oral glucose tolerance test (OGTT) is administered based on national guidelines, and all patients with IGT identified are recorded in primary care records. This approach minimizes selection bias common in traditional cohort studies. The study used a comprehensive dataset encompassing demographics, clinical data, and lifestyle factors. It also included detailed medication records, validated through extensive quality checks and data verification processes.6 Data were anonymized as per Strengthening the Reporting of Observational Studies in Epidemiology, with all participating practices consenting. The IGT cohort included women aged ≥18 years with IGT, based on a 2-hour glucose between 7.8 and 11 mmol/L during an OGTT.6 The GDM cohort included women aged ≥18 years diagnosed with GDM. The GDM diagnosis was based on universal screening with a 50 g glucose challenge test, and if the glucose concentration at 1 hour post-load was ≥7.8 mmol/L (140 mg/dL), a 75 g OGTT was performed using the New Zealand Society for the Study of Diabetes (NZSSD) criteria (fasting glucose ≥5.5 (99 mg/dL) mmol/L and/or 2-hour glucose ≥9.0 (162 mg/dL) mmol/L).3 The identification of incident T2D as the study’s outcome incorporated data from both primary and secondary care, leveraging comprehensive datasets coded by general practitioners and validated across healthcare settings. This approach spanned from enrollment through to either the recording of a T2D diagnosis or until December 31, 2019, for those without T2D, with mortality data acting as a competing risk. It is crucial to note that the evolution of GDM or IGT post-enrollment was viewed as part of the natural history, not factoring into the T2D risk model. Diagnosis used precise International Classification of Diseases Ninth Revision (ICD-9) and ICD-10 codes from the linked datasets. To minimize information bias, people with a diagnosis of T2D recorded within 12 months of enrollment were removed from the final cohort(online supplemental file 1). To address multiparity and repeated diagnoses of GDM, we included only the first instance of GDM for each woman in our analysis. For individuals successively diagnosed with IGT and then GDM, we categorized them based on their initial diagnosis to avoid duplication and bias in risk assessment. At enrollment, we evaluated patient demographics, clinical metrics, and usage of antihypertensive, anticoagulant, and lipid-lowering medications as covariates. Ethnicity was based on self-identification and categorized into NZ European, Māori, and Pasifika, with remaining ethnic groups with small frequencies pooled into an ‘Other’ category. Socioeconomic status was assessed using the NZDep2013 Index, dividing the country into areas by deprivation level, based on IMD scores, into five categories for analysis: from least deprived (IMD-1) to most deprived (IMD-5), enhancing the study’s statistical robustness.9 Cox regression models accounting for competing risk of all-cause death were used to estimate the HR of developing into T2D between comparison cohorts, with IGT as the reference. The risks of incident T2D between comparable cohorts were also estimated based on the same model. Subgroup analyses were also processed by age group, ethnicity, socioeconomic status, smoking status, obesity status, blood pressure and renal function. There were no missing data on covariates in the GDM cohort. In the IGT cohort, less than 1% of the covariate data were missing. We generated five imputed datasets using multiple imputation with chained equations, and the estimations were consistent with those from the complete case analysis. Final adjusted estimations were derived from the imputed models. Stata MP V18.0 was used for data manipulation and analysis. The significance level was set as a two-tailed p value <0.05. Results In the GDM cohort, participants were younger and had fewer treatments for hypertension, statins, and anticoagulants than the IGT cohort, which showed a higher prevalence of low estimated glomerular filtration rate (eGFR), suggesting differing management approaches due to distinct diabetes risk profiles. Ethnic distributions were similar, though the IGT cohort had more individuals in the most deprived category. Both cohorts were comparable in body mass index (BMI) and blood pressure, although slightly higher averages in the IGT cohort. Additionally, the GDM cohort included a larger percentage of ‘never smokers’ (table 1). Table 1Characteristics and incidence of outcomes of comparison cohortsGDM cohortn=998IGT women cohortn=14 694Age, years36.6 (9.5)58.8 (14.8)Ethnicity, n (%) New Zealand European403 (40.4)7290 (49.6) Māori127 (12.7)2179 (14.8) Pasifika196 (19.6)2832 (19.3) Other272 (27.3)2393 (16.3)IMD group (NZDep2013 scale), n (%) IMD-1 (1 or 2) (least deprivation)160 (16.0)2606 (17.7) IMD-2 (3 or 4)189 (18.9)2125 (14.5) IMD-3 (5 or 6)136 (13.6)1646 (11.2) IMD-4 (7 or 8)186 (18.6)2635 (17.9) IMD-5 (9 or 10) (most deprivation)327 (32.8)5682 (38.7)Smoking status, n (%) Never smoker737 (73.9)9538 (64.9) Ex-smoker147 (14.7)3318 (22.6) Current smoker114 (11.4)1846 (12.6)Body mass index, kg/m231.7 (8.0)32.4 (8.4)Systolic blood pressure, mm Hg120 (15)130 (16)Diastolic blood pressure, mm Hg75 (10)79 (11)HbA1c, mmol/mol39.7 (9.1)40.6 (8.0)Total cholesterol, mmol/L5.1 (0.9)5.2 (1.0)Triglyceride, mmol/L1.6 (1.0)1.7 (1.0)Low-density lipoprotein cholesterol, mmol/L3.0 (0.8)3.0 (0.9)High-density lipoprotein cholesterol, mmol/L1.4 (0.4)1.4 (0.4)Estimated glomerular filtration rate <90 mL/min/1.73 m236 (3.6)3683 (25.1)Antihypertensive treatment, n (%)64 (6.4)1489 (10.1)Statin treatment, n (%)32 (3.2)1128 (7.7)Antiplatelet or anticoagulant treatment, n (%)0 (0.0)50 (0.3)GDM, gestational diabetes mellitus; IGT, impaired glucose tolerance; IMD, Index of Multiple Deprivation. Characteristics and incidence of outcomes of comparison cohorts GDM, gestational diabetes mellitus; IGT, impaired glucose tolerance; IMD, Index of Multiple Deprivation. The overall incidence of T2D development was 11.5 (95% CI 9.2 to 14.3) and 10.9 (95% CI 10.2 to 11.6) per 1000 person-years in the GDM and IGT cohorts, respectively (table 2). Figure 1 displays the adjusted cumulative risk of T2D development in the GDM and IGT cohorts over the study period. Table 2Incidence of development of type 2 diabetes in gestational diabetes and impaired glucose tolerance cohorts over the study periodGDM cohortn=998IGT women cohortn=14 694Outcomes85846Person-years of follow-up period7370.377 573.6Incidence rates per 1000 person-years (95% CI)11.5 (9.2 to 14.3)10.9 (10.2 to 11.6)GDM, gestational diabetes mellitus; IGT, impaired glucose tolerance. Incidence of development of type 2 diabetes in gestational diabetes and impaired glucose tolerance cohorts over the study period GDM, gestational diabetes mellitus; IGT, impaired glucose tolerance. Cumulative risk of development of type 2 diabetes (T2D) in gestational diabetes mellitus (GDM) and impaired glucose tolerance (IGT) women over 25 years. Red and blue lines indicate GDM and IGT women, respectively. Risks were estimated from the final model. During the study, the GDM cohort had a higher overall T2D risk than the IGT cohort, with an adjusted HR of 1.11 (0.87–1.41), indicating no statistically difference between the groups. Subgroup analysis showed that women over 35 years, Māori or other ethnicities, and those of lower socioeconomic status faced greater T2D risks. Similarly, non-smokers, those with shorter follow-up, and women without obesity, high blood pressure, or reduced renal function in the GDM cohort displayed elevated T2D risks (figure 2). In particular, the adjusted HR was 1.51 (1.01–2.25) among women with follow-up less or equal to 5 years while 0.53 (0.07–4.12) among women with follow-up more than 5 years. Among current or ex-smokers, the risk of developing T2D between GDM and IGT cohorts was not significantly different (HR 0.41, 95% CI 0.09 to 1.81). Adjusted HR (95% CIs) for risk of development of type 2 diabetes (T2D) between GDM and IGT women (IGT women as reference). eGFR, estimated glomerular filtration rate; GDM, gestational diabetes mellitus; IGT, impaired glucose tolerance; NZE, New Zealand European; SBP, systolic blood pressure. Discussion Our study analyzed the long-term T2D risk among women with prior GDM or IGT, aligning with the outcomes of the Diabetes Prevention Program.2 The study highlights the critical need for early, specific post-GDM interventions to curb T2D progression. Over 25 years, risk trajectories for T2D notably diverged between the cohorts, with particular variance in some demographics. Women aged over 35 years, of Māori ethnicity, or facing socioeconomic challenges in the GDM cohort showed increased T2D risk compared with their IGT counterparts. This aligns with prior studies emphasizing the elevated T2D risk post-GDM in specific populations, advocating for focused interventions.10 11 While previous studies have shown a uniformly elevated risk of T2D across all age groups post-GDM,5 our results indicate that this elevated risk is more pronounced in older women, particularly those over 35 years. This discrepancy suggests that age-specific factors may play a more significant role in T2D progression than previously thought. Our research highlights the significant impact of socioeconomic status on T2D risk, echoing findings of the heavier toll from diabetes on impoverished communities.12 Integrating social health determinants into diabetes prevention is essential, aligning with the WHO’s initiative against non-communicable disease inequities.13 Although our study found that socioeconomic deprivation was a more potent predictor of T2D risk among women with prior GDM compared with those with IGT, this contrasts with some studies suggesting an equal impact across different glucose intolerant conditions.14 While the current study used integrated socioeconomic status measurements, as opposed to the single social determinants employed in previous studies, further research with similar design and socioeconomic status metrics is warranted. The increased risk in the GDM cohort within the first 5 years post partum indicates a critical period for intervention as reported elsewhere.15 This time frame is ideal for implementing lifestyle and pharmacological measures to prevent T2D progression, supporting guidelines that recommend early postpartum glucose assessments.16 Early intervention can address modifiable risk factors effectively during this window, potentially altering the disease trajectory. The association between renal function, as indicated by eGFR rates, and T2D risk in our study resonates with a growing body of evidence linking early renal function decline with future T2D risk.17 18 This calls for renal function to be a focal point in the post-GDM follow-up, especially considering the potential for intervention at stages of renal impairment that precede overt kidney disease. This calls for integrating renal function monitoring into routine postpartum care for women with GDM, which could mitigate long-term renal and metabolic complications. Traditional risk models have linked smoking to T2D.19 Our study revealed inconsistency in T2D risk between GDM and IGT cohorts based on smoking status. Among non-smokers, the HR for T2D was significant, indicating a higher risk in the GDM cohort compared with the IGT cohort. However, among current or ex-smokers, the HR was not significant, suggesting no difference in T2D risk between the cohorts. This suggests GDM screening identifies at-risk individuals lacking conventional risk factors. In contrast, the IGT cohort may identify those with metabolic syndrome reflecting the indication for the OGTT. However, due to the small sample size of the smoking subgroup, further research with larger sample sizes is needed to confirm this finding. Ethnic disparities in diagnosis rate, particularly among Pasifika and Māori, might reflect differences in screening practices, with Māori possibly receiving more comprehensive T2D screening. A noted low retinopathy prevalence in newly diagnosed Māori diabetics suggests effective screening in certain areas.20 These paradoxes highlight T2D risk factors’ complexity, urging further study into post-GDM lifestyle changes and screening practice disparities’ impact on long-term outcomes. The slightly higher BMI and blood pressure in the IGT cohort reveal a subtle dynamic; these elevations, though not statistically linked to increased T2D risk, suggest metabolic syndrome’s presence without clear risk factors.21 This highlights the role of GDM screening in pinpointing T2D risk among women lacking traditional indicators. It advocates for a comprehensive clinical approach, emphasizing the importance of assessing metabolic syndrome in women post-GDM. A holistic strategy incorporating weight and blood pressure management is essential for preventing T2D, underscoring the necessity of focused interventions for this group. Such measures are in line with studies showing that managing metabolic syndrome elements can notably lower T2D risk.22 Our study has some limitations. The analysis was retrospective and observational, which precludes causal inferences. Additionally, the generalizability of our findings may be constrained by the distinct ethnic and socioeconomic composition of the cohort, the NZ setting, and the employment of the NZSSD criteria. This specific ethnic mix and criterion application might not fully translate to other populations or contexts, suggesting a need for cautious application of the study conclusions beyond NZ. The lack of statistical significance in the overall HR suggests that while a trend toward higher risk in the GDM group was observed, it does not definitively confirm a greater T2D risk. This limitation should be considered when interpreting subgroup analyses, which may reflect variability in risk factors rather than consistent, significant differences. Many subgroup analyses also yielded insignificant estimations, highlighting the variability and lack of definitive evidence for some observed differences in T2D risk. These wide CIs may be due to sample size limitations within subgroups, particularly among smokers and ethnic minorities. Additionally, the potential effects of unmeasured confounders, variability in clinical management, and differing health-seeking behaviors between groups may have contributed to the non-significant findings. Future studies with larger sample sizes and more comprehensive measurements (eg, physical activity and nutrition exposures) are needed to explore these relationships more thoroughly. In the current study, our primary objective was to compare the long-term risk of developing type 2 diabetes mellitus (T2DM) between women with GDM and those with IGT. Future research should consider including a subgroup of women who developed T2DM without prior diagnoses of GDM or IGT to allow for more thorough comparisons and to enhance the understanding of different pathways leading to T2DM. This additional analysis could provide valuable insights into the distinct risk profiles and progression patterns among these groups. However, it is important to note that our current study was limited to data on women diagnosed with GDM or IGT, and we did not have access to data for women who developed T2DM without these prior diagnoses. Conclusions In conclusion, our study provides substantial evidence that the risk of T2D in women with a history of GDM is greater than those with IGT in some, but not all population groups, and is influenced by age, ethnicity, and socioeconomic factors. The emphasis on early universal screening for GDM is pivotal in identifying women without traditional risk factors but at risk of progressing to T2D. This finding reinforces the necessity for interventions tailored to medical and social determinants of health. Future research should explore the effectiveness of such comprehensive care strategies, aiming to mitigate T2D’s global impact. Supplementary material Acknowledgements We thank the participating GPs, audit nurses and the DCSS executive group for their support. The funders had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. Footnotes Funding: The DCSS was funded by the New Zealand Ministry of Health through Counties Manukau District Health Board. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: The North Health Ethics Committee approved the DCSS for research purposes in 1992 and then as an ongoing audit in 1996 (92/006). Ethics review was waived by the New Zealand Health and Disability Ethics Committees on March 25, 2019. Anonymized data were used for this analysis. Signed consent to participate was provided by an authorized signatory for each general practice. Participants gave informed consent to participate in the study before taking part. Data availability free text: The datasets analyzed in the current study are not publicly available because of agreements with the primary care organizations and the Ministry of Health who provided the data. Data availability statement No data are available. References Associated Data Supplementary Materials Data Availability Statement No data are available.