South Asian women have a higher risk of type 2 diabetes after gestational diabetes mellitus (GDM) than Nordic women; however, the mechanisms behind this difference remain unclear. We investigated insulin sensitivity, β-cell function, and hepatic insulin clearance in 179 South Asian and 108 Nordic women ∼17 months after GDM (mean age 35.3 years, BMI 29.1 kg/m2) by oral glucose tolerance test using deconvolution of C-peptide kinetics. Thirty-one percent of South Asian and 53% of Nordic participants were normoglycemic at the time of measurement. South Asian women had higher areas under the curve (AUCs) for glucose, prehepatic insulin, and peripheral insulin and lower insulin sensitivity, disposition index, and fasting hepatic insulin clearance than Nordic women. In the group with prediabetes or diabetes, South Asian women had similar AUCs for glucose and prehepatic insulin but a higher AUC for peripheral insulin, lower disposition index, and lower fasting hepatic insulin clearance than Nordic women. The waist-to-height ratio mediated ∼25–40% of the ethnic differences in insulin sensitivity in participants with normoglycemia. Overall, our novel data revealed that South Asian women with normoglycemia after GDM showed lower insulin secretion for a given insulin resistance and lower hepatic insulin clearance than Nordic women. South Asian women are at high risk of developing type 2 diabetes after GDM, and preventive efforts should be prioritized.

The risk of type 2 diabetes after gestational diabetes mellitus (GDM) (1,2) is twice as high in South Asian compared with European women and develops at younger ages and at a lower BMI (3). The mechanisms behind this higher risk for impaired glucose tolerance are still highly debated (4).

During normal pregnancy, insulin secretion increases to compensate for pregnancy-induced insulin resistance, and GDM develops if the pancreatic β-cells cannot meet this increased demand (5). South Asian women’s increased susceptibility to developing diabetes after GDM may also reflect a failure of the insulin secretion capacity in response to the increased insulin resistance known to be present at early ages in South Asian populations in both liver and muscle (4,68). Hepatic glucose production is the main determinant of fasting glucose levels, whereas glucose uptake in muscle is more important in determining postprandial plasma glucose (9). The increased demand placed on the β-cells by this insulin resistance may, over time, lead to failure and a decline in insulin secretion adjusted for insulin resistance, estimated as the disposition index (10).

Peripheral insulin concentrations reflect the balance between insulin secretion and insulin clearance, with liver clearing ∼50% of newly secreted insulin (11). Emerging evidence suggests that lower hepatic insulin clearance (HIC) could contribute to increased peripheral insulin levels and acts as an early adaption to insulin resistance and hyperglycemia (12). On the other hand, higher insulin levels are also associated with higher insulin resistance and an increased risk of type 2 diabetes (13). Although, HIC is difficult to measure directly in humans, it can be estimated indirectly by prehepatic insulin levels based on C-peptide deconvolution kinetics, as C-peptide clearance in liver is negligible (13). Here, we estimate measures of 1) insulin sensitivity, 2) β-cell function, and 3) HIC in South Asian and Nordic women undergoing an oral glucose tolerance test (OGTT) 1–3 years after GDM.

The Diabetes in South Asians 1 (DIASA 1) study was approved by the South-Eastern Norway Regional Committee for Medical and Health Research Ethics (reference no. 2018/689). All participants provided written informed consent.

Design, Study Population, and Data Collection

Between 1 September 2018 and 31 December 2021, we recruited women with a history of GDM in their last pregnancy who delivered 12–36 (±3) months previously at one of three hospitals in the Oslo area of Norway. Because of changes in the GDM definition during past years, most women were included on the basis of modified International Association of Diabetes and Pregnancy Study Group criteria (fasting plasma glucose [FPG] 5.3–6.9 mmol/L or 2-h glucose 9.0–11.0 mmol/L, n = 267) (14), but a few were included according to the former World Health Organization 1999 criteria (FPG ≥7.0 mmol/L or 2-h plasma glucose ≥7.8 mmol/L, n = 16) (15). Additional inclusion criteria were age ≥18 years and both parents born in a South Asian (Pakistan, India, Bangladesh, or Sri Lanka) or Nordic (Norway, Sweden, Denmark, Finland, or Iceland) country. Exclusion criteria included new pregnancies after the index pregnancy, exclusive breastfeeding at the time of examination, known diabetes before the index pregnancy or at the time of examination, ongoing inflammatory or serious disease, or history of a major surgical procedure <3 months before inclusion. Eligible women were identified by searching medical records from the three hospitals and recruited through an invitation letter. The South Asian women also received a telephone invitation in their native language to address any potential issues with communication in Norwegian (as recommended by the regional ethics committee). Of the 1,220 eligible women with a GDM diagnosis (449 South Asian, 771 Nordic), 179 South Asian (110 Pakistani, 33 Indian, 5 Bangladeshi, 31 Sri Lankan) and 108 Nordic (101 Norwegian, 3 Swedish, 3 Danish, and 1 Icelandic) women participated. Among the invited South Asian women, 16 were excluded because of new-onset diabetes after the index pregnancy and 29 because of a new pregnancy, while 225 declined or were not contactable. Among the Nordic women, who were invited only by letter, reasons for nonparticipation were not available (Supplementary Fig. 1).

At the study visit, we measured height, weight, and waist and hip circumferences (16). Thereafter, all women underwent an OGTT between 0800–1000 h after at least 8 h of fasting. Before and 15, 30, 60, and 120 min after a 75-g oral glucose load, blood was collected in 1) cooled sodium fluoride tubes for glucose analysis and kept on ice until centrifugation at 4°C within 10 min and 2) serum-separating tubes for analyses of insulin and C-peptide and centrifuged after 30 min. Plasma glucose level was measured by enzymatic photometry (Roche Diagnostics, Mannheim, Germany), whole-blood HbA1c by high-performance liquid chromatography (G8 HPLC Analyzer; Tosoh Biosciences, Tokyo, Japan), and serum C-peptide and insulin by electrochemiluminescence immunoassay (cobas e601, Roche Diagnostics). All measurements were performed at Oslo University Hospital, Aker, Norway. The coefficients of variation were 2.5%, 7.0%, 4.0–5.0%, and 1.5–2.5% for glucose, insulin, C-peptide, and HbA1c, respectively. Clinical and biochemical data from the participants obtained during pregnancy were retrieved from medical records.

Definitions

Prediabetes was defined according to World Health Organization International Expert Committee criteria as follows: FPG 6.1–6.9 mmol/L and/or 2-h plasma glucose 7.8–11.0 mmol/L and/or HbA1c 6.0–6.4% (42–47 mmol/mol) (15,17). Diabetes was defined according to internationally agreed criteria (18,19).

Calculations

HOMA2 of β-cell function (HOMA2-B) and HOMA2 of insulin sensitivity (HOMA2-S) were calculated from fasting serum C-peptide (pmol/L) or insulin (pmol/L), respectively, together with FPG (mmol/L) using the HOMA calculator (20,21). HOMA2-S was considered to mainly represent hepatic insulin sensitivity. Muscle insulin sensitivity index (MISI) was calculated by the MISI calculator (22) on the basis of the reduction in plasma glucose (mmol/L) from peak to nadir and the mean plasma insulin concentration (pmol/L) during the OGTT (23). The whole-body insulin sensitivity was estimated by the Matsuda index of insulin secretion (ISI) as 10,000/√ (fasting serum insulin [μIU/mL] × FPG [mg/dL]) × (mean OGTT insulin [μIU/mL]) × (mean OGTT glucose [mg/dL]) (24).

Prehepatic insulin (pmol/L) was estimated from C-peptide deconvolution using the ISEC software program (25) with standard settings (subjects with obesity, coefficient of variation 5%, and basal function on). The program’s assumptions included that 1) the secretion of insulin and C-peptide are equimolar; 2) C-peptide kinetics are described by a two-compartment model; 3) the parameters in the model are estimated from age, sex, height, weight classified as normal or obese, and type 2 diabetes status; and 4) measurement errors are uncorrelated with the zero mean and with a constant SD.

Indexes of insulin secretion were calculated with both peripheral insulin measurements and estimated prehepatic insulin levels as follows: insulinogenic index (IGI) = Δinsulin0–30 min (μIU/mL)/Δglucose0–30 min (mg/dL) during the OGTT (26,27). β-Cell function was also estimated by calculating HOMA2-B, and insulin secretion adjusted for insulin resistance was estimated as the disposition index (IGI × Matsuda ISI) (10). The latter assumes a hyperbolic relationship between insulin secretion and insulin resistance and that the product of these two variables is constant for women with the same degree of glucose tolerance (28).

β-Cell glucose sensitivity was estimated as the relationship between glucose levels and calculated prehepatic insulin levels during the OGTT (29). It was derived as the slope of this linear relationship and reflects the picomolar increase in prehepatic insulin levels per millimolar increase in plasma glucose levels.

HIC was calculated from prehepatic and peripheral insulin levels in the fasting (HICfasting = prehepatic insulinfasting/peripheral insulinfasting) and postprandial state (HICOGTT = prehepatic insulinAUC/peripheral insulinAUC), where AUC is the area under the curve (13,30). Hepatic insulin extraction was defined as the percent increase in HIC from one time interval to the next during the OGTT. Total AUC was calculated by the trapezoid rule (31).

Statistical Analyses

Sample size was estimated for the primary outcomes of the study as described elsewhere (32). Participants with prediabetes or diabetes were grouped together because of a low number of women with diabetes. This was considered appropriate as South Asian and Nordic women showed no difference in the prevalence of diabetes (32 [18%] of 178 vs. 15 [14%] of 108, P = 0.366).

Characteristics are presented as mean (SD), median (interquartile range [IQR]), or n (%). Differences between groups were assessed with unpaired t tests for normally distributed data. Variables were log-transformed to approximate normality if necessary. Mann-Whitney U tests were used for nonnormally distributed data.

OGTT data were analyzed with linear mixed models for repeated measures using random intercepts for participants and an unstructured covariance matrix (fixed effect were ethnicity, time, and time-by-ethnicity interaction; random effect was the participants). Data were estimated using marginal means with corresponding 95% CIs.

The mediation analyses were conducted using the PROCESS macro in SPSS (33). A directed acyclic graph was used to visualize the relationship of the covariates, exposure, and outcome in the model (Supplementary Fig. 2). Several parallel mediators were visualized: age, time since index pregnancy, waist-to-height ratio (WHtR), parity, GDM before index pregnancy, first-degree relatives with diabetes, years of education (as a proxy for socioeconomic status), gestational weight retention (the difference between weight at visit and prepregnancy weight), glucose-lowering drugs during pregnancy, and duration of breastfeeding. Covariates with P ≤ 0.25 were included in multivariate regression analyses to find the most significant mediators. Statistical significance was considered at a two-tailed P < 0.05. We used SPSS version 27 and R version 4.1.3 statistical software for the analyses.

Data and Resource Availability

All data generated or analyzed during this study are included in the published article and its online supplementary files.

Baseline Characteristics

At a median (IQR) of 16.5 (12.1) months after delivery, 31% of the South Asian and 53% of the Nordic participants had a normal OGTT (P < 0.001). The South Asian participants had higher parity, more first-degree relatives with diabetes, and fewer years of education than comparable Nordic participants. BMI did not differ between these groups, but South Asian participants had a higher WHtR and were somewhat younger than the Nordic participants (Table 1).

Table 1

Participant characteristics by ethnicity and glucose tolerance

NGTPrediabetes or diabetes
South AsianNordicPSouth AsianNordicP
Participants, n (%) 55 (31) 57 (53)  123 (69) 51 (47)  
Age (years) 34.2 (4.0) 36.3 (4.9) 0.007 34.3 (4.2) 35.7 (4.7) 0.031 
Time since index pregnancy (months), median (IQR) 14.9 (12.4) 16.4 (9.6) 0.088 15.1 (11.0) 19.0 (13.0) 0.144 
Weight (kg), median (IQR) 69.5 (16.7) 71.8 (28.1) 0.117 71.6 (19.7) 86.6 (21.6) <0.001 
Height (cm) 158.4 (5.4) 167.7 (6.0) <0.001 160.1 (6.7) 165.5 (6.1) <0.001 
BMI (kg/m2), median (IQR) 27.3 (5.5) 25.4 (9.1) 0.147 28.7 (6.1) 32.2 (6.1) 0.005 
Waist circumference (cm) 94.2 (12.4) 91.8 (14.5) 0.586 97.6 (11.6) 101.2 (12.7) 0.185 
Waist-to-hip ratio 0.89 (0.00) 0.87 (0.10) 0.213 0.91 (0.07) 0.90 (0.09) 0.351 
WHtR 0.60 (0.08) 0.55 (0.09) 0.007 0.61 (0.07) 0.61 (0.08) 0.657 
Parity 2.1 (1.0) 1.7 (0.7) 0.028 2.2 (1.0) 1.7 (0.8) <0.001 
GDM before index pregnancy, n (%) 13 of 54 (24) 9 (16) 0.274 42 of 121 (35) 15 (29) 0.500 
First-degree relatives with diabetes, n (%) 38 of 51 (75) 10 of 46 (22) <0.001 87 of 117 (74) 12 of 45 (27) <0.001 
Years of education 14.3 (3.3) 17.0 (3.0) <0.001 15.0 (3.5) 16.2 (3.0) 0.022 
Gestational weight retention (kg), median (IQR) 2.5 (8.7) 1.0 (6.3) 0.654 2.9 (5.6) 2.4 (6.3) 0.312 
Insulin with or without metformin use in pregnancy 26 (47) 14 (25) 0.012 71 (58) 26 (51) 0.415 
Insulin use in pregnancy 19 (35) 11 (19) 0.069 53 (43) 23 (45) 0.808 
Breastfeeding (months) 10.0 (6.7) 10.2 (5.6) 0.821 9.3 (7.4) 8.6 (5.9) 0.453 
Breastfeeding (≥3 months) 48 (87) 50 (88) 0.943 95 (77) 39 (76) 0.913 
NGTPrediabetes or diabetes
South AsianNordicPSouth AsianNordicP
Participants, n (%) 55 (31) 57 (53)  123 (69) 51 (47)  
Age (years) 34.2 (4.0) 36.3 (4.9) 0.007 34.3 (4.2) 35.7 (4.7) 0.031 
Time since index pregnancy (months), median (IQR) 14.9 (12.4) 16.4 (9.6) 0.088 15.1 (11.0) 19.0 (13.0) 0.144 
Weight (kg), median (IQR) 69.5 (16.7) 71.8 (28.1) 0.117 71.6 (19.7) 86.6 (21.6) <0.001 
Height (cm) 158.4 (5.4) 167.7 (6.0) <0.001 160.1 (6.7) 165.5 (6.1) <0.001 
BMI (kg/m2), median (IQR) 27.3 (5.5) 25.4 (9.1) 0.147 28.7 (6.1) 32.2 (6.1) 0.005 
Waist circumference (cm) 94.2 (12.4) 91.8 (14.5) 0.586 97.6 (11.6) 101.2 (12.7) 0.185 
Waist-to-hip ratio 0.89 (0.00) 0.87 (0.10) 0.213 0.91 (0.07) 0.90 (0.09) 0.351 
WHtR 0.60 (0.08) 0.55 (0.09) 0.007 0.61 (0.07) 0.61 (0.08) 0.657 
Parity 2.1 (1.0) 1.7 (0.7) 0.028 2.2 (1.0) 1.7 (0.8) <0.001 
GDM before index pregnancy, n (%) 13 of 54 (24) 9 (16) 0.274 42 of 121 (35) 15 (29) 0.500 
First-degree relatives with diabetes, n (%) 38 of 51 (75) 10 of 46 (22) <0.001 87 of 117 (74) 12 of 45 (27) <0.001 
Years of education 14.3 (3.3) 17.0 (3.0) <0.001 15.0 (3.5) 16.2 (3.0) 0.022 
Gestational weight retention (kg), median (IQR) 2.5 (8.7) 1.0 (6.3) 0.654 2.9 (5.6) 2.4 (6.3) 0.312 
Insulin with or without metformin use in pregnancy 26 (47) 14 (25) 0.012 71 (58) 26 (51) 0.415 
Insulin use in pregnancy 19 (35) 11 (19) 0.069 53 (43) 23 (45) 0.808 
Breastfeeding (months) 10.0 (6.7) 10.2 (5.6) 0.821 9.3 (7.4) 8.6 (5.9) 0.453 
Breastfeeding (≥3 months) 48 (87) 50 (88) 0.943 95 (77) 39 (76) 0.913 

Data are mean (SD) unless otherwise indicated.

Plasma Glucose, Estimated Prehepatic Insulin, and Peripheral Insulin Levels

In the normal glucose tolerance (NGT), despite no ethnic difference in fasting glucose or 2-h OGTT findings, South Asian participants had a 7% higher AUC for glucose than Nordic women (P < 0.01) (Fig. 1A). South Asian participants also had a 23% higher AUC for prehepatic insulin (P < 0.01) and 67% higher AUC for peripheral insulin (P < 0.01) (Fig. 1B and C). AUCs for peripheral insulin levels were approximately twofold higher in South Asian than in Nordic participants (P < 0.01) (Fig. 1C).

Figure 1

Glucose (A and D), estimated prehepatic insulin (B and E), and peripheral insulin (C and F) levels in South Asian (red) and Nordic (blue) participants with NGT (dark color) and prediabetes or diabetes (light color). Data are mean ± 95% CI. *P < 0.05, **P < 0.01, ***P < 0.001 for South Asian vs. Nordic participants; †P < 0.05, ††P < 0.01, †††P < 0.001 for the group-by-time response.

Figure 1

Glucose (A and D), estimated prehepatic insulin (B and E), and peripheral insulin (C and F) levels in South Asian (red) and Nordic (blue) participants with NGT (dark color) and prediabetes or diabetes (light color). Data are mean ± 95% CI. *P < 0.05, **P < 0.01, ***P < 0.001 for South Asian vs. Nordic participants; †P < 0.05, ††P < 0.01, †††P < 0.001 for the group-by-time response.

Close modal

In prediabetes or diabetes group, no ethnic differences in AUCs for glucose and prehepatic insulin were found (Fig. 1D and E), but South Asian participants had a 34% higher AUC for peripheral insulin than comparable Nordic participants (P < 0.01) (Fig. 1F).

South Asian participants with NGT showed no difference in AUC for prehepatic insulin levels (β = 54 [95% CI -365, 474], P = 0.798) but higher AUC for peripheral insulin than Nordic participants with prediabetes or diabetes (β = 459 [95% CI 32, 886], P = 0.036).

Insulin Sensitivity

In the NGT group, all estimates for insulin sensitivity were 35% (±9) lower in South Asian than in Nordic women participants (Table 2). lower in South Asian than in Nordic participants (Table 2). In the prediabetes or diabetes group, HOMA2-S and Matsuda ISI were 30% and 31% lower in South Asian than in Nordic participants, respectively (Table 2). HOMA2-S and Matsuda ISI were substantially higher in the NGT versus prediabetes or diabetes group. No difference in MISI was found between the NGT and prediabetes or diabetes groups (Table 2).

Table 2

Ethnic differences in insulin sensitivity and secretion, β-cell function, and HIC by glucose tolerance categories

NGTPrediabetes or diabetesAll (N = 286)
South AsianNordicSouth AsianNordicP
Participants, n (%) 55 (31) 57 (53) 123 (69) 51 (47)  
Insulin sensitivity      
 HOMA2-S, median (IQR) 75 (41) 110 (54)*** 44 (37) 63 (42)** <0.001 
 MISI 0.17 (0.10) 0.23 (0.12)** 0.13 (0.12) 0.16 (0.20) 0.341 
 Matsuda ISI, median (IQR) 3.1 (1.8) 5.5 (3.7)*** 2.2 (1.6) 3.2 (2.1)*** <0.001 
Insulin secretion      
 HOMA2-B 126 (28) 114 (28)* 124 (29) 115 (40) 0.967 
 Prehepatic IGI, median (IQR) 1.8 (1.6) 2.0 (1.4) 1.4 (0.9) 1.3 (1.1) <0.001 
 Peripheral IGI, median (IQR) 1.4 (1.4) 1.0 (0.9)* 0.9 (0.7) 0.8 (1.0) <0.001 
β-Cell function      
 β-Cell glucose sensitivity 185 (57) 171 (63) 152 (55) 149 (62) <0.001 
 Prehepatic DI, median (IQR) 7.4 (7.2) 10.9 (9.8)*** 3.2 (2.8) 4.9 (3.0)** <0.001 
 Peripheral DI, median (IQR) 4.4 (4.1) 5.2 (5.3) 2.3 (1.8) 2.7 (1.6) <0.001 
HIC      
 Fasting 2.7 (0.9) 3.3 (0.8)*** 2.4 (0.8) 3.2 (1.1)*** <0.001 
 OGTT 1.9 (0.6) 1.7 (0.4) 1.7 (0.8) 1.6 (0.6) 0.205 
NGTPrediabetes or diabetesAll (N = 286)
South AsianNordicSouth AsianNordicP
Participants, n (%) 55 (31) 57 (53) 123 (69) 51 (47)  
Insulin sensitivity      
 HOMA2-S, median (IQR) 75 (41) 110 (54)*** 44 (37) 63 (42)** <0.001 
 MISI 0.17 (0.10) 0.23 (0.12)** 0.13 (0.12) 0.16 (0.20) 0.341 
 Matsuda ISI, median (IQR) 3.1 (1.8) 5.5 (3.7)*** 2.2 (1.6) 3.2 (2.1)*** <0.001 
Insulin secretion      
 HOMA2-B 126 (28) 114 (28)* 124 (29) 115 (40) 0.967 
 Prehepatic IGI, median (IQR) 1.8 (1.6) 2.0 (1.4) 1.4 (0.9) 1.3 (1.1) <0.001 
 Peripheral IGI, median (IQR) 1.4 (1.4) 1.0 (0.9)* 0.9 (0.7) 0.8 (1.0) <0.001 
β-Cell function      
 β-Cell glucose sensitivity 185 (57) 171 (63) 152 (55) 149 (62) <0.001 
 Prehepatic DI, median (IQR) 7.4 (7.2) 10.9 (9.8)*** 3.2 (2.8) 4.9 (3.0)** <0.001 
 Peripheral DI, median (IQR) 4.4 (4.1) 5.2 (5.3) 2.3 (1.8) 2.7 (1.6) <0.001 
HIC      
 Fasting 2.7 (0.9) 3.3 (0.8)*** 2.4 (0.8) 3.2 (1.1)*** <0.001 
 OGTT 1.9 (0.6) 1.7 (0.4) 1.7 (0.8) 1.6 (0.6) 0.205 

Data presented as mean (SD) unless otherwise indicated. Peripheral refers to measured peripheral insulin levels, and prehepatic refers to estimated insulin levels based on deconvolution of C-peptide kinetics. DI, disposition index.

*

P ≤ 0.05,

**

P ≤ 0.01,

***

P ≤ 0.001 for South Asian vs. Nordic women.

P ≤ 0.05,

††

P ≤ 0.01,

†††

P ≤ 0.001 for NGT vs. prediabetes or diabetes.

Median (IQR).

β-Cell Function

In participants with NGT, the median IGI calculated from peripheral insulin levels was 40% higher in South Asian than in Nordic participants (Table 2 and Fig. 2B). However, when calculating IGI with prehepatic insulin, no difference was seen between the ethnicities (Table 2 and Fig. 2A). In addition, when applying prehepatic insulin in estimating the disposition index, we observed a 32% lower disposition index in South Asian than in Nordic participants (Table 2 and Fig. 2C). In the prediabetes or diabetes group, disposition index estimates were 35% lower in South Asian than in Nordic participants when applying the prehepatic insulin levels (Table 2 and Fig. 2C).

Figure 2

Box plot of early prehepatic (A) vs. early peripheral insulin secretion (prehepatic IGI vs. peripheral IGI) (B), and prehepatic (C) vs. peripheral disposition index (DI) (D) in South Asian (red) and Nordic (blue) participants with NGT (dark color) and prediabetes or diabetes (light color). *P < 0.05, ***P < 0.001.

Figure 2

Box plot of early prehepatic (A) vs. early peripheral insulin secretion (prehepatic IGI vs. peripheral IGI) (B), and prehepatic (C) vs. peripheral disposition index (DI) (D) in South Asian (red) and Nordic (blue) participants with NGT (dark color) and prediabetes or diabetes (light color). *P < 0.05, ***P < 0.001.

Close modal

Using the Nordic participants with NGT as a reference, the hyperbolic relationship between prehepatic insulin secretion and insulin sensitivity showed that South Asian participants with NGT tended to “fall off the curve” and cluster to the lower left, approaching those with prediabetes or diabetes (Fig. 3 and Supplementary Fig. 3).

Figure 3

The disposition index curve. The relationship between early prehepatic (A) and peripheral insulin secretion (B) (prehepatic IGI vs. peripheral IGI) and insulin sensitivity (Matsuda ISI) in South Asian and Nordic participants with NGT and prediabetes or diabetes. The hyperbolic curve was regressed in Nordic participants with NGT (blue line). South Asian participants with NGT (red cross) tended to fall off the curve and cluster to the lower left, approaching South Asian (light red cross) and Nordic participants with prediabetes or diabetes (light blue cross). Data are mean ± 95% CI.

Figure 3

The disposition index curve. The relationship between early prehepatic (A) and peripheral insulin secretion (B) (prehepatic IGI vs. peripheral IGI) and insulin sensitivity (Matsuda ISI) in South Asian and Nordic participants with NGT and prediabetes or diabetes. The hyperbolic curve was regressed in Nordic participants with NGT (blue line). South Asian participants with NGT (red cross) tended to fall off the curve and cluster to the lower left, approaching South Asian (light red cross) and Nordic participants with prediabetes or diabetes (light blue cross). Data are mean ± 95% CI.

Close modal

We observed no ethnic differences in β-cell glucose sensitivity (Table 2). Participants with prediabetes or diabetes showed lower β-cell glucose sensitivity than those with NGT (Supplementary Fig. 4A). In addition, responses in estimated prehepatic insulin at different intervals of plasma glucose levels were also largely similar between the ethnicities (Supplementary Fig. 4B).

HIC

In the NGT and prediabetes or diabetes groups, fasting HIC was 18% and 25% lower in South Asian than in Nordic participants, respectively (Table 2 and Fig. 4A). In South Asian participants, we found that fasting HIC was lower in the prediabetes or diabetes group than in the NGT group (Table 2 and Fig. 4A). Postprandial HIC (during the OGTT) was, on average, one-half the level of fasting HIC for all groups (Fig. 4B). However, the decline in HIC from the fasting to the postprandial state was more pronounced in Nordic than in South Asian participants, independent of glucose tolerance (Fig. 4B and Supplementary Fig. 5A). The percent HIC per minute (i.e., the hepatic insulin extraction) during the OGTT did not differ between the ethnicities (Supplementary Fig. 5B). Comparison of South Asian participants with NGT with Nordic participants with prediabetes or diabetes revealed a substantially lower fasting HIC (β = −1.7 [95% CI −2.7, −0.7], P < 0.001). All significant differences shown in Table 2 remained significant in a sensitivity analysis adjusting for insulin sensitivity, β-cell function, and HIC indexes for time since index pregnancy (Supplementary Table 1) and BMI (Supplementary Table 2).

Figure 4

The ratio of prehepatic to peripheral insulin levels (HIC). A: Fasting HIC was lower in South Asian (red) vs. Nordic (blue) participants, in both the NGT (dark color) and prediabetes or diabetes (light color) groups. Fasting HIC declined from the NGT to prediabetes or diabetes group only in South Asian participants. B: The decline in HIC from fasting to postprandial levels were steeper in Nordic than in South Asian participants for both the NGT and prediabetes or diabetes groups. HIC was, on average, one-half in the postprandial vs. fasting state. *P < 0.05, ***P < 0.001. NOR, Nordic; SA, South Asian.

Figure 4

The ratio of prehepatic to peripheral insulin levels (HIC). A: Fasting HIC was lower in South Asian (red) vs. Nordic (blue) participants, in both the NGT (dark color) and prediabetes or diabetes (light color) groups. Fasting HIC declined from the NGT to prediabetes or diabetes group only in South Asian participants. B: The decline in HIC from fasting to postprandial levels were steeper in Nordic than in South Asian participants for both the NGT and prediabetes or diabetes groups. HIC was, on average, one-half in the postprandial vs. fasting state. *P < 0.05, ***P < 0.001. NOR, Nordic; SA, South Asian.

Close modal

Regression Analysis of β-Cell Function, HIC, and Insulin Sensitivity

In a multiple regression analysis in participants with NGT, ethnicity was the only significant predictor of prehepatic disposition index (P = 0.038) and fasting HIC (P = 0.007). With HOMA2-S as the outcome, ethnicity and WHtR were the only significant predictors (P = 0.017 and 0.002, respectively). With MISI as the outcome, WHtR was the most significant predictor (P = 0.013). With Matsuda ISI as the outcome, ethnicity and WHtR were the most significant predictors (P = 0.003 and 0.001, respectively). We tested whether the associated phenotypic traits could mediate the ethnic differences shown in insulin sensitivity in participants with NGT and found that WHtR mediated 25–29% of the ethnic differences in HOMA2-S and Matsuda ISI and 38% of the difference in MISI (Supplementary Fig. 6A–C). In the prediabetes or diabetes group, we did not perform the mediation analysis because no ethnic difference in WHtR was found (Table 1).

In the current study of women with previous GDM assessed at a median of ∼17 months after pregnancy, South Asian women with NGT presented with lower fasting HIC and lower insulin secretion adjusted for insulin resistance than Nordic women with NGT. These factors may suggest a more rapid course toward the development of type 2 diabetes.

An important observation was that calculating insulin secretion from estimated prehepatic insulin levels indicated a markedly lower β-cell function relative to insulin resistance in South Asian women. Hence, analyzing only peripheral insulin levels may mask an early β-cell dysfunction. Although reduced β-cell function could be expected in women with previous GDM (34), our findings that reduced β-cell function in South Asian women with NGT are novel. The lower β-cell function was also supported by no ethnic differences in β-cell glucose sensitivity. The current literature suggests a positive correlation between β-cell glucose sensitivity and insulin resistance to enable a limited increase in glucose levels (35). Our data, however, show a higher AUC for glucose without an increase in β-cell glucose sensitivity among South Asian compared with Nordic women with NGT, reflecting a lower β-cell function.

Of note, in the prediabetes or diabetes group, no ethnic differences in prehepatic insulin levels were found. However, after its first passage through the liver, we found significantly higher peripheral insulin levels in the South Asian compared with Nordic participants. This difference in peripheral hyperinsulinemia indicates lower fasting HIC among South Asian women. A similar pattern was found in the NGT group, but here the South Asian participants had higher prehepatic insulin levels than comparable Nordic participants.

Insulin has a major action in, and is extracted by, the liver (13). Our findings of lower HIC in South Asian participants in the fasted state may be a consequence of increased hepatic insulin resistance. However, it may also be regarded as an adaption within the HIC pathways to provide peripheral tissues with higher insulin levels. Increased insulin resistance has been demonstrated to be present several years before the diagnosis of prediabetes or type 2 diabetes among South Asian individuals (8,36,37). We confirmed these findings of higher insulin resistance in South Asian compared with Nordic women across categories of glucose tolerance. Despite no difference in fasting and 2-h OGTT glucose levels, South Asian participants with NGT showed slightly higher glucose levels during the 1st hour of the OGTT. This is in accordance with previous literature (4,8) and implies less suppression of hepatic glucose production during the OGTT. Both whole-body and MISI seemed to be lower in South Asian participants, perhaps partly because of more central fat accumulation and lower muscle mass (8,38). Notably, the MISI was similar in the NGT and prediabetes or diabetes groups, supporting that a gradual reduction in insulin secretory function is the main driver for a deteriorating glucose tolerance (39). There are reports, however, that have suggested a role of excess insulin in driving insulin resistance and that suppression of high plasma insulin levels enhances insulin sensitivity (9,12,40,41). We, in accordance with others (35,42), speculate that the increased peripheral insulin levels, following reduced HIC, may be an early and important adaption to developing insulin resistance. By reducing the toll of enhanced insulin secretion to compensate for insulin resistance, lower HIC may offload the β-cells (42,43). Notably, the HIC was downregulated from the fasting to postprandial state, followed by a hepatic insulin extraction that was precisely regulated throughout the OGTT, indicating a precise regulation according to the insulin demand independent of ethnicities. However, because the baseline level of HIC was lower in the South Asian NGT group, a metabolic inflexibility was seen that may explain South Asian women’s propensity to develop type 2 diabetes after GDM. This interpretation is at variance with a previous study that suggested a genetic defect in a main glycoprotein (CEACAM1) in the HIC pathways as a possible reason for ethnic divergence in diabetes prevalence (41).

Another important question is how HIC is associated with obesity and with remission of diabetes by a substantial weight loss, such as in the Diabetes Remission Clinical Trial (DiRECT) (44) or Diabetes Intervention Accentuating Diet and Enhancing Metabolism-I (DIADEM-I) study (45). Such weight loss is reported to improve hepatic insulin sensitivity and may improve β-cell function, but data on HIC pathways are scarce (46). A recent study showed improved β-cell function and fasting HIC with time after bariatric surgery (47), but the relative importance of improved hepatic insulin sensitivity versus HIC pathways is still unclear.

Our data in women with NGT indicate that WHtR, potentially capturing important ethnic differences in body composition and central fat accumulation, mediated significant ethnic differences in insulin sensitivity. We, in accordance with others (8,38,48), therefore suggest that central adiposity is instrumental for the lower insulin sensitivity in South Asian individuals. This is important, as the Diabetes Prevention Program reported a 50% decline in type 2 diabetes incidence after GDM if weight loss was obtained (49), while no effect on glucose deterioration was observed in a similar study in South Asian individuals without weight loss (50). Our findings thus lend support to initiatives that recommend strong preventive measures against overweight and obesity, particularly in South Asian women with a high risk of diabetes. Even though HIC is reported to be negatively associated with obesity (35), we did not find that our estimates of obesity mediated the ethnic differences in HIC.

The strengths of this study include well-characterized and sufficiently large groups of the two ethnicities with NGT and prediabetes or diabetes cared for in the same health care setting. All included women had previously been referred to the hospital with a GDM diagnosis, and hence, our findings are not valid for a population without GDM. Many women did not reply to the invitation letter or declined to participate because of time constraints and other reasons; hence, we cannot exclude a selection bias. We therefore compared key baseline characteristics in a sensitivity analysis of women who participated in the study versus a randomly selected subgroup of women who did not (100 South Asian women, 100 Nordic women) (Supplementary Table 3). Among the South Asian women, no differences in age, prepregnancy BMI, glucose values during pregnancy, use of glucose-lowering drugs, GDM before index pregnancy, or first-degree relatives with diabetes were found. The participating Nordic women were older than nonparticipants, but the other characteristics did not differ. The older age among participating Nordic women could have led to an overestimation of the proportion of women with prediabetes or diabetes in this group and, thereby, might have led to an underestimation of the ethnic differences in prevalence. Furthermore, differences in the recruitment procedures may have introduced a selection bias between the ethnic groups, as only one of the groups received a telephone reminder. Thus, we might have recruited a higher proportion of Nordic women with severe GDM because they only had the one invitation by letter. Conversely, a higher percentage of South Asian than Nordic women were using glucose-lowering drugs during pregnancy, and minimal differences existed in glucose levels during pregnancy between the ethnic groups (32). Moreover, there could be differences in lifestyle habits not picked up by our questionnaires and examinations. We did not control for menstrual cycle phases, although their effect on glucose metabolism is debated. Furthermore, prehepatic insulin levels are difficult to measure directly in humans and may be best estimated from modeling of C-peptide kinetics. We also acknowledge that our estimates of insulin sensitivity are indirect, and direct measurements of insulin sensitivity with euglycemic clamp were not available. Importantly, our study did not directly measure hepatic insulin resistance, which may, in addition to HIC, contribute to ethnic differences in peripheral insulin levels. This as well as analyses on ethnic differences in HIC pathways, such as a glycoprotein (CEACAM1) that promotes HIC (51), deserves further study. Finally, the cross-sectional nature of our data can only describe associations and cannot imply causality.

In conclusion, South Asian women with NGT investigated a few years after GDM showed lower β-cell function, lower HIC, and higher insulin resistance compared with Nordic women. Our novel observations accordingly add to our understanding of diabetes pathophysiology in South Asian and White people in general and in the context of previous GDM.

This article contains supplementary material online at https://doi.org/10.2337/figshare.21108910.

Acknowledgments. The authors dedicate this article to the memory of Dr. Cecilie Wium (Institute of Clinical Medicine, University of Oslo, and Oslo University Hospital), who died shortly before completion of this study. She conceptualized and designed the study, wrote the protocol, and obtained the funding. The study would never have been performed without her. She will be sorely missed by her colleagues, the patients she treated, and her family and friends. The authors also thank the women who participated in the study; study nurses Åshild Stavik (Akershus University Hospital), Åse Halsne (Oslo University Hospital), Jesini Anurathan (Oslo University Hospital), and Karin Pleym (Vestre Viken Health Trust, Drammen) and study coordinator Ellen Hillestad (Oslo University Hospital) for invaluable help in the recruitment and examination of participants; and librarian Åse Marit Hammersbøen (Akershus University Hospital) and statistician Ragnhild S. Falk (Oslo University Hospital) for assistance with various aspects of this study.

Funding. This study was funded by the Research Council of Norway (grant 273252).

The study funder was not involved in the design of the study; collection, analysis, and interpretation of data; or writing of the manuscript and did not impose any restrictions with regard to the publication of the article.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. A.S. researched the data and drafted the manuscript. A.S. and S.L.-Ø. performed the statistical analysis. A.S., S.L.-Ø., E.Q., C.S., N.S., J.M.R.G., H.L.G., S.T.S., I.N., and K.I.B. contributed to the analysis or interpretation of data for the work, reviewed the manuscript critically for intellectual content, and approved the final manuscript. E.Q., C.S., H.L.G., S.T.S., I.N., and K.I.B. contributed to the design. K.I.B. contributed to the study protocol, aided in data acquisition, and supervised the study performance. K.I.B. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented orally at the 58th Annual Meeting of the European Association for the Study of Diabetes, Stockholm, Sweden, and virtual, 19–23 September 2022.

1.
Vounzoulaki
E
,
Khunti
K
,
Abner
SC
,
Tan
BK
,
Davies
MJ
,
Gillies
CL
.
Progression to type 2 diabetes in women with a known history of gestational diabetes: systematic review and meta-analysis
.
BMJ
2020
;
369
:
m1361
2.
Dennison
RA
,
Chen
ES
,
Green
ME
, et al
.
The absolute and relative risk of type 2 diabetes after gestational diabetes: a systematic review and meta-analysis of 129 studies
.
Diabetes Res Clin Pract
2021
;
171
:
108625
3.
Goyal
A
,
Gupta
Y
,
Kalaivani
M
, et al
.
Long term (>1 year) postpartum glucose tolerance status among Indian women with history of gestational diabetes mellitus (GDM) diagnosed by IADPSG criteria
.
Diabetes Res Clin Pract
2018
;
142
:
154
161
4.
Narayan
KMV
,
Kanaya
AM
.
Why are South Asians prone to type 2 diabetes? A hypothesis based on underexplored pathways
.
Diabetologia
2020
;
63
:
1103
1109
5.
McIntyre
HD
,
Catalano
P
,
Zhang
C
,
Desoye
G
,
Mathiesen
ER
,
Damm
P
.
Gestational diabetes mellitus
.
Nat Rev Dis Primers
2019
;
5
:
47
6.
Siddiqui
MK
,
Anjana
RM
,
Dawed
AY
, et al
.
Young onset diabetes in Asian Indians is associated with lower measured and genetically determined beta-cell function: an INSPIRED study
.
Diabetologia
2022
;
65
:
973
983
7.
Ikehara
S
,
Tabák
AG
,
Akbaraly
TN
, et al
.
Age trajectories of glycaemic traits in non-diabetic South Asian and white individuals: the Whitehall II cohort study
.
Diabetologia
2015
;
58
:
534
542
8.
Sattar
N
,
Gill
JMR
.
Type 2 diabetes in migrant south Asians: mechanisms, mitigation, and management
.
Lancet Diabetes Endocrinol
2015
;
3
:
1004
1016
9.
DeFronzo
RA
,
Tripathy
D
.
Skeletal muscle insulin resistance is the primary defect in type 2 diabetes
.
Diabetes Care
2009
;
32
(
Suppl. 2
):
S157
S163
10.
Abdul-Ghani
MA
,
Williams
K
,
DeFronzo
RA
,
Stern
M
.
What is the best predictor of future type 2 diabetes?
Diabetes Care
2007
;
30
:
1544
1548
11.
Campioni
M
,
Toffolo
G
,
Basu
R
,
Rizza
RA
,
Cobelli
C
.
Minimal model assessment of hepatic insulin extraction during an oral test from standard insulin kinetic parameters
.
Am J Physiol Endocrinol Metab
2009
;
297
:
E941
E948
12.
Hamley
S
,
Kloosterman
D
,
Duthie
T
, et al
.
Mechanisms of hyperinsulinaemia in apparently healthy non-obese young adults: role of insulin secretion, clearance and action and associations with plasma amino acids
.
Diabetologia
2019
;
62
:
2310
2324
13.
Piccinini
F
,
Bergman
RN
.
The measurement of insulin clearance
.
Diabetes Care
2020
;
43
:
2296
2302
14.
Helsedirektoratet
.
Nasjonal faglig retningslinje: svangerskapsdiabetes [national guidelines: gestational diabetes mellitus], 2018
.
15.
World Health Organization, Institutional Repository for Information Sharing
.
Definition, diagnosis and classification of diabetes mellitus and its complications: report of a WHO consultation
.
Part 1, diagnosis and classification of diabetes mellitus, 1999. Accessed 30 November 2020. Available from https://apps.who.int/iris/handle/10665/66040
16.
World Health Organization
.
Waist circumference and waist-hip ratio: report of a WHO expert consultation, 2008
.
17.
International Expert Committee
.
International expert committee report on the role of the A1C assay in the diagnosis of diabetes
.
Diabetes Care
2009
;
32
:
1327
1334
18.
American Diabetes Association
.
2. Classification and diagnosis of diabetes: Standards of Medical Care in Diabetes-2021
.
Diabetes Care
2021
;
44
(
Suppl. 1
):
S15
S33
19.
World Health Organization; Institutional Repository for Information Sharing
.
Use of glycated haemoglobin (HbA1c) in diagnosis of diabetes mellitus: abbreviated report of a WHO consultation, 2011
.
Accessed 25 October 2021. Available from https://apps.who.int/iris/handle/10665/70523
20.
Wallace
TM
,
Levy
JC
,
Matthews
DR
.
Use and abuse of HOMA modeling
.
Diabetes Care
2004
;
27
:
1487
1495
21.
University of Oxford
.
HOMA2 calculator, 2019
.
Accessed 12 March 2022. Available from https://www.dtu.ox.ac.uk/homacalculator/
22.
O’Donovan
SD
,
Lenz
M
,
Goossens
GH
, et al
.
Improved quantification of muscle insulin sensitivity using oral glucose tolerance test data: the MISI calculator
.
Sci Rep
2019
;
9
:
9388
23.
Abdul-Ghani
MA
,
Matsuda
M
,
Balas
B
,
DeFronzo
RA
.
Muscle and liver insulin resistance indexes derived from the oral glucose tolerance test
.
Diabetes Care
2007
;
30
:
89
94
24.
Matsuda
M
,
DeFronzo
RA
.
Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp
.
Diabetes Care
1999
;
22
:
1462
1470
25.
Boston
RC
,
Pei
D
,
Moate
PJ
.
A numerical deconvolution method to estimate C-peptide secretion in humans after an intravenous glucose tolerance test
.
Metabolism
2009
;
58
:
891
900
26.
Haffner
SM
,
Miettinen
H
,
Gaskill
SP
,
Stern
MP
.
Decreased insulin secretion and increased insulin resistance are independently related to the 7-year risk of NIDDM in Mexican-Americans
.
Diabetes
1995
;
44
:
1386
1391
27.
Phillips
DI
,
Clark
PM
,
Hales
CN
,
Osmond
C
.
Understanding oral glucose tolerance: comparison of glucose or insulin measurements during the oral glucose tolerance test with specific measurements of insulin resistance and insulin secretion
.
Diabet Med
1994
;
11
:
286
292
28.
Kahn
SE
,
Prigeon
RL
,
McCulloch
DK
, et al
.
Quantification of the relationship between insulin sensitivity and β-cell function in human subjects. Evidence for a hyperbolic function
.
Diabetes
1993
;
42
:
1663
1672
29.
Hovorka
R
,
Soons
PA
,
Young
MA
.
ISEC: a program to calculate insulin secretion
.
Comput Methods Programs Biomed
1996
;
50
:
253
264
30.
Castillo
MJ
,
Scheen
AJ
,
Letiexhe
MR
,
Lefèbvre
PJ
.
How to measure insulin clearance
.
Diabetes Metab Rev
1994
;
10
:
119
150
31.
Allison
DB
,
Paultre
F
,
Maggio
C
,
Mezzitis
N
,
Pi-Sunyer
FX
.
The use of areas under curves in diabetes research
.
Diabetes Care
1995
;
18
:
245
250
32.
Sharma
A
,
Nermoen
I
,
Qvigstad
E
, et al
.
High prevalence, and significant ethnic differences in actionable HbA1c after gestational diabetes mellitus
.
BMC Med
2022
;
20
:
318
33.
Preacher
KJ
,
Hayes
AF
.
SPSS and SAS procedures for estimating indirect effects in simple mediation models
.
Behav Res Methods Instrum Comput
2004
;
36
:
717
731
34.
Ignell
C
,
Shaat
N
,
Ekelund
M
,
Berntorp
K
.
The impact of ethnicity on glucose homeostasis after gestational diabetes mellitus
.
Acta Diabetol
2013
;
50
:
927
934
35.
Jung
SH
,
Jung
CH
,
Reaven
GM
,
Kim
SH
.
Adapting to insulin resistance in obesity: role of insulin secretion and clearance
.
Diabetologia
2018
;
61
:
681
687
36.
Tillin
T
,
Hughes
AD
,
Godsland
IF
, et al
.
Insulin resistance and truncal obesity as important determinants of the greater incidence of diabetes in Indian Asians and African Caribbeans compared with Europeans: the Southall and Brent Revisited (SABRE) cohort
.
Diabetes Care
2013
;
36
:
383
393
37.
Kanaya
AM
,
Herrington
D
,
Vittinghoff
E
, et al
.
Understanding the high prevalence of diabetes in U.S. South Asians compared with four racial/ethnic groups: the MASALA and MESA studies
.
Diabetes Care
2014
;
37
:
1621
1628
38.
Bakker
LEH
,
Sleddering
MA
,
Schoones
JW
,
Meinders
AE
,
Jazet
IM
.
Pathogenesis of type 2 diabetes in South Asians
.
Eur J Endocrinol
2013
;
169
:
R99
R114
39.
Kahn
SE
.
Clinical review 135: the importance of β-cell failure in the development and progression of type 2 diabetes
.
J Clin Endocrinol Metab
2001
;
86
:
4047
4058
40.
Tricò
D
,
Natali
A
,
Arslanian
S
,
Mari
A
,
Ferrannini
E
.
Identification, pathophysiology, and clinical implications of primary insulin hypersecretion in nondiabetic adults and adolescents
.
JCI Insight
2018
;
3
:
e124912
41.
Piccinini
F
,
Polidori
DC
,
Gower
BA
,
Bergman
RN
.
Hepatic but not extrahepatic insulin clearance is lower in African American than in European American women
.
Diabetes
2017
;
66
:
2564
2570
42.
Kaga
H
,
Tamura
Y
,
Takeno
K
, et al
.
Correlates of insulin clearance in apparently healthy non-obese Japanese men
.
Sci Rep
2017
;
7
:
1462
43.
Kim
SP
,
Ellmerer
M
,
Kirkman
EL
,
Bergman
RN
.
β-Cell “rest” accompanies reduced first-pass hepatic insulin extraction in the insulin-resistant, fat-fed canine model
.
Am J Physiol Endocrinol Metab
2007
;
292
:
E1581
E1589
44.
Lean
MEJ
,
Leslie
WS
,
Barnes
AC
, et al
.
Primary care-led weight management for remission of type 2 diabetes (DiRECT): an open-label, cluster-randomised trial
.
Lancet
2018
;
391
:
541
551
45.
Taheri
S
,
Zaghloul
H
,
Chagoury
O
, et al
.
Effect of intensive lifestyle intervention on bodyweight and glycaemia in early type 2 diabetes (DIADEM-I): an open-label, parallel-group, randomised controlled trial
.
Lancet Diabetes Endocrinol
2020
;
8
:
477
489
46.
Taylor
R
,
Al-Mrabeh
A
,
Sattar
N
.
Understanding the mechanisms of reversal of type 2 diabetes
.
Lancet Diabetes Endocrinol
2019
;
7
:
726
736
47.
Fatima
F
,
Hjelmesæth
J
,
Birkeland
KI
, et al
.
Gastrointestinal hormones and β-cell function after gastric bypass and sleeve gastrectomy: a randomized controlled trial (Oseberg)
.
J Clin Endocrinol Metab
2022
;
107
:
e756
e766
48.
Chandalia
M
,
Lin
P
,
Seenivasan
T
, et al
.
Insulin resistance and body fat distribution in South Asian men compared to Caucasian men
.
PLoS One
2007
;
2
:
e812
49.
Aroda
VR
,
Christophi
CA
,
Edelstein
SL
, et al.;
Diabetes Prevention Program Research Group
.
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
:
1646
1653
50.
Tandon
N
,
Gupta
Y
,
Kapoor
D
, et al.;
LIVING Collaborative Group
.
Effects of a lifestyle intervention to prevent deterioration in glycemic status among South Asian women with recent gestational diabetes: a randomized clinical trial
.
JAMA Netw Open
2022
;
5
:
e220773
51.
Najjar
SM
,
Perdomo
G
.
Hepatic insulin clearance: mechanism and physiology
.
Physiology (Bethesda)
2019
;
34
:
198
215
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/journals/pages/license.