Settings and Study Design
The primary studies on which this secondary data analysis is based were cross-sectional and were carried out at All India Institute of Medical Sciences (AIIMS), New Delhi, India, a tertiary care center dealing with a low- and middle-income population . The institutional ethics committee of AIIMS, New Delhi, India approved these studies (Ref. No. IECPG-177//27.01.2016, RT-4/24.02.2016, dated February 26, 2016 and IECPG-166/19.04.2018, dated April 23, 2018), and we obtained informed written consent from all participants. This study was performed in accordance with the Helsinki Declaration of 1964 and its later amendments.
Inclusion and Exclusion Criteria
This study included adult non-pregnant women who were diagnosed with GDM using International Association of Diabetes in Pregnancy Study Group (IADPSG) criteria (plasma glucose ≥ 5.1, 10.0, or 8.5 mmol/l at 0, 1, and 2 h, respectively, on a 75-g OGTT) during their index pregnancy (2012–2019) . They were registered in the Department of Endocrinology and Metabolism and/or Department of Obstetrics and Gynecology at AIIMS, New Delhi, India. The additional inclusion criteria were postpartum duration of at least 6 weeks and willingness to provide written informed consent. We excluded women with diabetes other than GDM, and those with current pregnancy. We also excluded women with known diabetes on glucose-lowering medications, or those using steroids.
Procedure on Day of Testing
We invited women in a fasting state and performed an OGTT using 83.3 g glucose monohydrate (equivalent to 75 g of anhydrous glucose) dissolved in 250–300 ml water and consumed over 5–10 min. We took samples for plasma glucose estimation at 0 and 120 min and collected blood in the fasting state for HbA1c measurement. A detailed questionnaire was completed for all participants, documenting relevant personal and medical history. Detail regarding the anthropometric and biochemical variables and their measurements were provided in our previous publications [15, 17].
Definitions Used in Study
Individuals were classified as having normoglycemia [FPG < 5.6 mmol/L, 2-h plasma glucose < 7.8 mmol/L and HbA1c < 5.7% (39 mmol/mol)], prediabetes by American Diabetes Association (ADA) criteria [FPG 5.6–6.9 mmol/L and/or 2-h plasma glucose 7.8–11.0 mmol/L and/or HbA1c 5.7–6.4% (39–47 mmol/mol)], prediabetes by World Health Organization (WHO) criteria [FPG 6.1–6.9 mmol/L and/or 2-h plasma glucose 7.8–11.0 mmol/L], diabetes mellitus [FPG ≥ 7.0 mmol/L and/or 2-h plasma glucose ≥ 11.1 mmol/L and/or HbA1c ≥ 6.5% (48 mmol/mol)] [5, 18]. Overweight/obesity and metabolic syndrome were defined as per WHO and International Diabetes Federation (IDF) criteria, respectively [19, 20]. Hypertension was defined as systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg, or use of drugs for lowering of blood pressure.
Blood sample for plasma glucose estimation was collected in a gray-top fluoride tube, centrifuged immediately, and transported to the laboratory under cold conditions. Plasma glucose was measured using the hexokinase method on an autoanalyzer. For HbA1c estimation, blood was collected in a purple-top ethylenediaminetetraacetic acid (EDTA) tube. HbA1c was measured using a high-performance liquid chromatography-based ion exchange chromatography method. The inter-assay coefficients of variation (CV) for HbA1c derived from low- and high-quality control samples were 2.0% (at a mean HbA1c of 4.9% or 30 mmol/mol) and 2.9% (at a mean HbA1c of 10.1% or 87 mmol/mol), respectively. Besides, the departmental laboratory also participates in an external quality assurance program for these two analytes (glucose and HbA1c) with an acceptable performance.
Sample Size Calculation
Considering the ADA definition as the gold standard for the diagnosis of diabetes, we calculated sample size for the predictive performance of FPG and HbA1c in the postpartum period. Considering 95% confidence interval, 10% precision, and a sensitivity of 90%, the number of events (subjects with diabetes) was estimated to be 35. Therefore, for a 10% prevalence of diabetes, the sample size was estimated to be 350.
STATA 15.0 (Stata Corp, College Station, TX, USA) was used for the statistical analyses. Data are presented as n (%), mean ± SD or median (interquartile range). We analyzed the data to calculate sensitivity, specificity, positive predictive value, and negative predictive value of FPG and HbA1c, either alone or in combination, at varying cutoffs for the diagnosis of diabetes. The cutoffs chosen for FPG were ≥ 5.6 mmol/L and ≥ 6.1 mmol/L, as these are thresholds used by ADA and WHO, respectively, to define impaired fasting glucose [5, 18]. For HbA1c, the cutoffs chosen were ≥ 5.7% (39 mmol/mol, used by ADA to define prediabetes) and ≥ 6.0% (42 mmol/mol, the threshold shown to be predictive of future diabetes) .