Abstract
The association between hypertensive disorders of pregnancy (HDP) and the subsequent development of type 2 diabetes (T2D) in Japanese general population remains unclear. To investigate the influence of HDP on long-term postpartum development of metabolic disorders and T2D, we conducted a population-based cross-sectional study using the 75 g oral glucose tolerance test (75g-OGTT) in 978 parous Japanese women (median age: 66 years). We further evaluated the combined effect of HDP and T2D susceptibility genes on developing T2D. HDP history was identified in 101 participants (10.3%) and associated with high blood pressure, hypertriglyceridemia, increased homeostasis model assessment 2 for insulin resistance, decreased Matsuda index and insulin secretion-sensitivity index-2, increased postprandial glucose levels during the 75g-OGTT, and increased prevalence of hypertension, metabolic syndrome and T2D. The age and BMI adjusted odds ratio (OR) for T2D was 1.74 (95%CI: 1.04–2.93) in individuals with HDP as compared to those without HDP. Stratified analyses demonstrated an increased OR of T2D prevalence for individuals with HDP history harboring the cyclin-dependent kinase 5 regulatory subunit associated protein 1-like 1 (CDKAL1) C/C genotype compared with that of the reference group after adjusting for age and current BMI (OR = 5.18 vs. reference group; 95% CI: 1.99–13.50). Further, validation analyses using bootstrap method showed high reproducibility. HDP history was associated with postpartum prevalence of hypertension, insulin resistance and T2D later in life in Japanese general population. Further, the simultaneous assessment of an HDP history and CDKAL1 genotype is valuable to predict the future T2D development (UMIN000036074).
Introduction
Hypertensive disorders of pregnancy (HDP) are major complications of pregnancy, with an estimated prevalence of 5.2–8.2% [1]. Globally, they are the leading cause of serious complications and mortality in mothers and infants during the perinatal period [2]. Moreover, women with HDP are prone to non-communicable diseases such as hypertension and cardiovascular disorders later in life [3, 4]. It is theorized that pregnancy is a “metabolic stress test” with the potential to identify individuals at high risk for such diseases [5]. Previous studies have suggested a link between HDP and the risk of subsequent insulin resistance (IR) and type 2 diabetes (T2D) development; however, most reports on the association between HDP and T2D were limited to short-to-medium-term postpartum assessments, and there is currently no Japanese cohort investigating the prevalence of HDP with assessments by 75g-OGTT. Therefore, the relation remains unclear [6-16]. Furthermore, gene–environment interactions have been considered as a crucial aspect of T2D development [17-19]. The exposure to HDP is considered as one of the environmental stresses. However, no studies have evaluated the interaction effects between HDP history and T2D susceptibility genes on the long-term postpartum risk of diabetes. We hypothesized that HDP is associated with T2D in long-term postpartum, and combined effect of HDP and T2D susceptibility genes may enhance the onset of these conditions. Therefore, to investigate these questions, we conducted a population-based cross-sectional study using the 75 g oral glucose tolerance test (75g-OGTT) to assess the risk of developing T2D in parous Japanese women. Further, to assess the combined effect between a history of HDP and genes related to the risk of T2D development, we analyzed the risk genotypes based on SNPs in candidate loci that were reported to be associated with T2D identified in our prior replication study based on genome wide association study (GWAS) of a Japanese population [20].
Materials and Methods
Study overview
Toon Health Study is a prospective cohort study initiated in 2009 targeting general community residents of Toon City, Ehime Prefecture, Japan [21]. The primary aim of this study was to establish risk factors for preventing diabetes and cardiovascular diseases. A total of 2,504 individuals aged 30–79 years have participated (888 men and 1,616 women), and follow-up surveys are conducted every five years. All participants are Japanese. Patients with type 1 diabetes mellitus and pregnant women, including those with pregestational diabetes, are not included. The participants underwent detailed interviews, physical examinations, and blood tests, including a 75g-OGTT.
To investigate a history of complications of pregnancy, including HDP, we mailed a self-administered questionnaire to female participants in April 2021 and December 2022. Those who did not respond to the questionnaire, provided ambiguous responses about HDP history, or were nulliparous were excluded.
From the initial 2,504 participants, 1,526 were excluded (888 men; 503 non-responders; 99 nulliparous; 36 with unknown HDP history). Subsequently, 978 parous women were analyzed based on HDP history. Moreover, 22 were excluded from single nucleotide polymorphism (SNP) analysis due to unavailable DNA samples. Ultimately, 956 parous women were included in the HDP and T2D susceptibility gene analyses (Fig. 1).
Fig. 1
Flowchart for participant selection.
A history of HDP was determined using a self-administered questionnaire. Specifically, individuals who had been clinically diagnosed with toxemia or high blood pressure during pregnancy were defined as having a history of HDP. The diagnosis of hypertension was based on the criteria of the Japanese Society of Hypertension Guidelines 2019: office blood pressure >140/90 mmHg, home blood pressure >135/85 mmHg, or the current use of antihypertensive medication [22]. T2D was diagnosed according to the American Diabetes Association criteria: fasting blood glucose ≥7.0 mmol/L, or 2 h-postprandial glucose ≥11.1 mmol/L, HbA1c ≥48 mmol/mol (6.5%), or the current use of glucose-lowering drugs [23]. The diagnosis of dyslipidemia was based on the Japan Atherosclerosis Society 2022 diagnostic criteria or the current use of hyperlipidemia medications [24]. Obesity was defined as body mass index (BMI) ≥25 kg/m2. Metabolic syndrome was diagnosed using the 2005 International Diabetes Federation criteria for Asian women [25].
Calculations
Areas under the plasma glucose curve (AUC0–2 h PG) and insulin curve (AUC0–2 h Ins) were calculated using the trapezoidal rule applied to the glucose and insulin curves obtained during the 75g-OGTT. The Matsuda index (MI) was defined as previously reported [26]. Insulin secretion-sensitivity index-2 (ISSI-2) was defined as follows: (AUC0–2 h Ins/AUC0–2 h PG) × MI [27-31].
Genotyping
To assess the combined effect between a history of HDP and genes related to the risk of T2D development, we analyzed the risk genotypes based on SNPs in candidate [20]. The following SNPs were considered: rs7754840 (CDKAL1), rs10811661 (CDKN2A/2B), rs1111875 (HHEX), rs1470579 (IGF2BP2), rs5219 (KCNJ11), rs13266634 (SLC30A8), and rs12255372 (TCF7L2). Genotyping was conducted using TaqMan analysis with a primer–probe mix from the Assay-on-Demand system (Applied Biosystems, Waltham, MA, USA). SNPs that did not significantly deviate from Hardy–Weinberg equilibrium (p > 0.05) and had a genotyping call-rate ≥95% were included.
Statistical analysis
First, the participants were divided into two groups based on the presence or absence of a history of HDP. Continuous variables were analyzed using the Mann–Whitney U test, and categorical variables were analyzed using the chi-square test. We calculated the odds ratios (ORs) and 95% confidence intervals (CIs) for the prevalence of non-communicable diseases between the HDP history group and the non-HDP group by multivariate logistic regression analysis to control for potential confounders of age and current BMI. Second, participants were divided into four groups based on the presence or absence of HDP history and T2D. A Kruskal–Wallis test was used for continuous variables and the chi-square test for categorical variables. A post-hoc analysis of intergroup differences was conducted using Steel’s multiple comparison test. Third, a multivariate logistic regression analysis was performed to calculated ORs and 95% CIs for T2D of the HDP history group compared to the non-HDP group based on the presence or absence of SNP risk genotypes of known T2D susceptibility genes. Additionally, we calculated the polygenic risk score (PRS) as the overall influence of the T2D susceptibility genes and conducted a similar analysis by dividing the individuals into two groups whether their PRS was in the upper quintile or below. The PRS was derived for each individual using the formula: PRS = β1χ1 + β2χ2 + … βkχk … + βnχn [32], where βk is the per-allele log OR for T2D associated with the risk allele homogenous for SNP k, and xk the number of alleles for the same SNP (0, 1, or 2), and n = 6 is the total number of SNPs. The OR values were taken from our prior replication study based on GWAS [20]. The following OR values were used: rs7754840 (CDKAL1), 1.290; rs10811661 (CDKN2A/2B), 1.234; rs1111875 (HHEX), 1.258; rs1470579 (IGF2BP2), 1.185; rs5219 (KCNJ11), 1.126; and rs12255372 (TCF7L2), 1.714. Likewise, we assessed the multiplicative interaction between HDP history and risk allele homozygotes using logistic regression analysis. A history of HDP (non-HDP = 0, HDP = 1), risk allele homozygotes (absence of risk allele homozygote = 0, presence of risk allele homozygote = 1), and a combination of HDP history and risk allele homozygote (HDP × risk allele homozygote) were used as independent variables. Additionally, similar analysis was conducted by using the upper quintile of the PRS as the risk group, instead of the homozygous risk allele (upper quintile = 1, lower four quintile = 0). Age and BMI were adopted as covariates for adjustment. The MOVER method was used to calculate 95% CIs for relative excess risks due to interaction (RERIOR) for testing additive interactions because the prevalence of the dependent variables for T2D was 14.0% [33, 34]. The independent variables and covariates were the same as described above. Subsequently, participants were divided into four groups based on the presence or absence of HDP history and the C/C genotype of cyclin-dependent kinase 5 regulatory subunit associated protein 1-like 1 gene rs7754840 (CDKAL1 C/C genotype). We then calculated the age- and BMI-adjusted ORs for T2D according to HDP history and CDKAL1 C/C genotype using multivariate logistic regression analysis. Further, bootstrap analyses with 2000 iterations were applied for internal validation of results and bias-corrected 95% CIs were calculated. All p values were two-sided, and statistical significance was set at p < 0.05. Potential confounders of age and current BMI were used as continuous variables. All statistical analyses were performed using SAS v9.4 and JMP version 18 (SAS Institute Inc., NC, USA).
Results
Clinical characteristics of participants with a history of HDP
The characteristics of the participants are summarized in Table 1. Briefly, the median (interquartile range) age and BMI of the participants was 66 (57–73) years and 22.4 (20.3–24.6), respectively. A history of HDP was identified in 101 (10.3%) participants. Compared to those in the non-HDP group, values for waist circumference, blood pressure, triglyceride, HbA1c, fasting insulin, homeostasis model assessment 2 for insulin resistance (HOMA2-IR), homeostasis model assessment 2 for beta cell function (HOMA2-beta), 1 h- and 2 h-PG levels, and AUC0–2 h PG were increased and Δ0–1 h Ins/Δ0–1 h PG, MI and ISSI-2 were decreased in the HDP group. The prevalence of T2D, hypertension, dyslipidemia, obesity, and metabolic syndrome was higher in the HDP group than in the non-HDP group (Table 1). The age- and current BMI-adjusted OR for T2D was 1.74 (95% CI: 1.04–2.93; p = 0.04), and that for hypertension was 1.98 (95% CI: 1.24–3.17; p < 0.01) in individuals with a history of HDP, as compared to those without a history of HDP (Table 2).
Table 1 Characteristics of the women with or without a history of HDP (
n = 978)
|
Women with
HDP (n = 101) |
Women without
HDP (n = 877) |
p value |
| Age (years) |
68 (60–74) |
66 (57–72) |
0.10 |
| BMI (kg/m2) |
22.7 (20.7–26.4) |
22.3 (20.3–24.5) |
0.06 |
| Waist (cm) |
85.0 (75.0–93.5) |
82.0 (75.0–88.5) |
0.03 |
| SBP (mmHg) |
129 (118–142) |
124 (109–137) |
0.01 |
| DBP (mmHg) |
77 (70–85) |
74 (67–82) |
<0.01 |
| HbA1c (%) |
5.7 (5.4–6.0) |
5.6 (5.4–5.8) |
0.04 |
| (mmol/mol) |
38.8 (35.5–42.1) |
37.7 (35.5–39.9) |
0.04 |
| Fasting PG (mmol/L) |
5.28 (4.94–5.67) |
5.17 (4.83–5.50) |
0.046 |
| Fasting IRI (pmol/L) |
47.4 (30.9–67.8) |
36.7 (26.9–52.6) |
<0.01 |
| HOMA2-beta |
75.5 (59.7–92.5) |
69.1 (55.7–84.6) |
0.01 |
| HOMA2-S |
115.4 (95.6–176.8) |
144.8 (101.8–199.0) |
<0.01 |
| HOMA2-IR |
0.87 (0.57–1.15) |
0.69 (0.50–0.98) |
<0.01 |
| Total cholesterol (mmol/L) |
5.48 (5.02–5.90) |
5.48 (4.94–6.05) |
0.69 |
| HDL-cholesterol (mmol/L) |
1.71 (1.42–1.99) |
1.68 (1.45–1.97) |
0.75 |
| LDL-cholesterol (mmol/L) |
3.13 (2.66–3.44) |
3.15 (2.69–3.67) |
0.16 |
| Triglyceride (mmol/L) |
1.11 (0.78–1.55) |
0.97 (0.72–1.31) |
0.02 |
| Hs-CRP (mg/dL) |
0.051 (0.021–0.084) |
0.041 (0.019–0.087) |
0.30 |
| eGFR (mL/min/1.73 m2) |
68.3 (57.5–76.8) |
68.1 (60.7–76.5) |
0.39 |
| 75g-OGTTa |
| Fasting PG (mmol/L) |
5.22 (4.91–5.61) |
5.17 (4.88–5.50) |
0.31 |
| 1 h PG (mmol/L) |
9.69 (7.19–11.61) |
8.22 (6.56–10.17) |
<0.01 |
| 2 h PG (mmol/L) |
7.22 (5.89–9.11) |
6.83 (5.78–8.50) |
0.07 |
| Fasting IRI (pmol/L) |
46.8 (30.5–64.1) |
36.7 (27.0–52.5) |
<0.001 |
| 1 h IRI (pmol/L) |
409.1 (228.2–644.0) |
348.6 (243.6–504.4) |
0.22 |
| 2 h IRI (pmol/L) |
382.0 (265.9–567.3) |
335.7 (238.6–496.6) |
0.06 |
| AUC0–2 h PG (mmol/L h) |
15.9 (13.0–18.9) |
14.4 (11.9–17.0) |
<0.01 |
| AUC0–2 h Ins (pmol/L h) |
619.2 (397.4–1,000.8) |
545.7 (393.9–776.0) |
0.12 |
| AUC0–2 h Ins/AUC0–2 h PG |
38.0 (26.6–63.3) |
39.0 (28.4–52.0) |
0.70 |
| Δ0–1 h Ins/Δ0–1 h PG |
96.9 (51.5–177.0) |
109.8 (65.1–217.7) |
0.05 |
| Matsuda index |
5.12 (3.28–8.18) |
6.23 (4.33–8.72) |
<0.01 |
| ISSI-2 |
1.75 (1.19–2.23) |
2.02 (1.51–2.72) |
<0.001 |
| Urinary ACR (mg/gCr) |
8.25 (5.57–14.01) |
7.58 (5.08–11.86) |
0.08 |
| Prevalence of disease, n (%) |
| T2D |
23 (22.8) |
114 (13.0) |
0.01 |
| Hypertension |
61 (60.4) |
356 (40.6) |
<0.001 |
| Dyslipidemia |
67 (66.3) |
488 (55.6) |
0.04 |
| Obesity |
32 (31.7) |
178 (20.3) |
<0.01 |
| MetS |
41 (40.6) |
262 (29.9) |
0.03 |
| CKD |
35 (34.7) |
226 (25.8) |
0.06 |
Continuous variables were analyzed using the Mann–Whitney U test, and categorical variables were analyzed using the chi-square test. Data are presented as median (IQR) unless otherwise specified. a 75g-OGTT data were obtained from 918 individuals without a history of diabetes on the day of the study. HDP, hypertensive disorders of pregnancy; SBP, systolic blood pressure; DBP, diastolic blood pressure; Hs-CRP, high-sensitivity C-reactive protein; PG, plasma glucose; IRI, immunoreactive insulin; AUC0–2 h PG, glucose AUC; AUC0–2 h Ins, insulin AUC; AUC0–2 h Ins/AUC0–2 h PG, insulin AUC/glucose AUC; Δ0–1 h Ins/Δ0–1 h PG, (1 h insulin – fasting insulin)/(1 h plasma glucose – fasting plasma glucose); ISSI-2, insulin secretion-sensitivity index-2; ACR, albumin-to-creatinine ratio; MetS, metabolic syndrome; T2D, type 2 diabetes mellitus; CKD, chronic kidney disease.
Table 2 Odds ratios of non-communicable diseases in individuals with a history of HDP
|
Odds ratios (vs. non-HDP) |
| T2D |
Hypertension |
Dyslipidemia |
Obesity |
MetS |
CKD |
| Unadjusted ORs (95% CIs) |
1.97
(1.19, 3.27)
p = 0.01 |
2.23
(1.47, 3.40)
p < 0.001 |
1.57
(1.02, 2.42)
p = 0.04 |
1.82
(1.16, 2.86)
p < 0.01 |
1.60
(1.05, 2.45)
p = 0.03 |
1.53
(0.99, 2.37)
p = 0.06 |
| Adjusted ORs (95% CIs) |
| Age |
1.85
(1.10, 3.09)
p = 0.02 |
2.17
(1.37, 3.42)
p < 0.001 |
1.44
(0.92, 2.27)
p = 0.11 |
1.75
(1.11, 2.76)
p = 0.02 |
1.48
(0.94, 2.31)
p = 0.09 |
1.39
(0.88, 2.22)
p = 0.16 |
| Age and current BMI |
1.74
(1.04, 2.93)
p = 0.04 |
1.98
(1.24, 3.17)
p < 0.01 |
1.32
(0.83, 2.09)
p = 0.25 |
N/A |
N/A |
1.34
(0.84, 2.14)
p = 0.22 |
ORs, 95% CIs, and p values were calculated using uni- and multivariable logistic regression analyses (n = 978). Metabolic syndrome was not adjusted for BMI. HDP, hypertensive disorders of pregnancy; T2D, type 2 diabetes mellitus; MetS, metabolic syndrome; CKD, chronic kidney disease; N/A, not applicable.
Next, we assessed the characteristics of participants with HDP history and T2D. Compared to the HDP history without T2D group, the HDP history with T2D group exhibited higher HOMA2-IR and lower MI, AUC0–2 h Ins/AUC0–2 h PG, Δ0–1 h Ins/Δ0–1 h PG, and ISSI-2 values (Supplementary Table 1).
The combined effect of HDP history and the CDKAL1 C/C genotype on the future development of T2D
We next investigated the interaction between the HDP history and genotypes of six T2D susceptibility genes identified in our prior replication study based on GWAS of Japanese [27]. SLC30A8 (rs13266634) was excluded for deviating from Hardy-Weinberg equilibrium. There were significant multiplicative and additive interaction effects between HDP history and the CDKAL1 C/C genotype on the development of T2D (p for multiplicative interaction = 0.03, RERIOR = 3.85; 95% CI: 0.50–12.11). However, risk genotypes of other T2D susceptibility genes and upper quintile of PRS did not interact with HDP in the risk of diabetes (Supplementary Table 2). Therefore, we focused on CDKAL1, which is reported to decrease insulin secretion and increase the risk of T2D including Japanese population [20, 35, 36]. Among the subjects included in the current analysis, the rate of risk allele homozygosity for CDKAL1 stood at 17.6%. There was no difference in the prevalence of HDP depending on the presence or absence of this risk allele homozygosity of the CDKAL1. The individuals with the CDKAL1 C/C genotype group exhibited decreased Δ0–1 h Ins/Δ0–1 h PG and ISSI-2 values and elevated 1 h-, 2 h-PG, and AUC0–2 h PG levels compared to those with the CDKAL1 non-C/C genotype group (Supplementary Table 3).
Next, we evaluated the clinical characteristics of subjects with a history of HDP harboring the CDKAL1 C/C genotype. Stratified analyses according to HDP history and the CDKAL1 C/C genotype revealed that values for HbA1c; 0 h-, 1 h-, and 2 h-PG levels; AUC0–2 h PG; Δ0–1 h Ins/Δ0–1 h PG during the 75g-OGTT; MI, and ISSI-2 were exaggerated in the HDP history and CDKAL1 C/C genotype group compared to those in the other groups (Supplementary Table 4). Furthermore, stratified analyses confirmed an increased OR of T2D prevalence for individuals with HDP history harboring the CDKAL1 C/C genotype compared with that of the reference group (non-HDP history with the C/G or G/G genotype) after adjusting for age and current BMI (OR = 5.18 vs. reference group; 95% CI: 1.99–13.50, Table 3). Further, bootstrap analyses with 2000 iterations for internal validation of results showed high reproducibility.
Table 3 Odds ratios of type 2 diabetes mellitus according to history of HDP and genotype of the
CDKAL1 SNP rs7754840
| Variable |
Overall
(n = 956) |
Non-HDP and non-C/C
(n = 708) |
Non-HDP and C/C
(n = 149) |
HDP and non-C/C
(n = 80) |
HDP and C/C
(n = 19) |
| Cases of T2D, n (%) |
128 (13.4) |
88 (12.4) |
18 (12.1) |
14 (17.5) |
8 (42.1) |
| Unadjusted ORs (95% CIs) |
|
1.00 (Ref) |
0.97 (0.56, 1.66) |
1.49 (0.81, 2.77) |
5.12 (2.00, 13.09) |
| Adjusted ORs (95% CIs) |
| Age |
|
1.00 (Ref) |
1.03 (0.60, 1.79) |
1.36 (0.72, 2.56) |
5.07 (1.94, 13.29) |
| Age and current BMI |
|
1.00 (Ref) |
1.08 (0.62, 1.89) |
1.25 (0.66, 2.37) |
5.18 (1.99, 13.50) |
ORs, 95% CIs, and p values were calculated using a logistic regression analysis. Further, validation analyses by bootstrap method with 2000 iterations showed reproducibility; bias-corrected 95% CI was 1.89 to 13.92 in HDP and C/C group compared with the reference group after adjusting for age and current BMI.
HDP, hypertensive disorders of pregnancy; T2D, type 2 diabetes mellitus.
Discussion
The present study of 978 parous women using a 75g-OGTT provides robust evidence that a history of HDP is associated with increased HOMA2-IR, decreased MI, ISSI-2, and increased the prevalence of T2D. Some studies have reported the association between HDP and increased risk of IR and diabetes; however, the actual relation remains controversial because most of these studies were limited to short- to medium-term postpartum assessments, and glucose intolerance and diabetes were not defined by the 75g-OGTT [13-16]. In the present study, the median age of the study population was 66 (IQR, 57–73) years. Considering the average maternal age at birth in Japan and comparing with other studies examining childbirth ages, this study provides long-term data for approximately three decades postpartum. Furthermore, the prevalence of T2D defined by the 75g-OGTT was 14.0%, which is higher than that in previous reports; for example, a prevalence of 7.9% was reported by Mannisto et al. (mean age 60 years), and a prevalence of 10.4% reported by Libby et al. (median age 72 years) [9, 10]. Therefore, the current study employing the 75g-OGTT offers a more precise assessment of glucose intolerance and facilitates the diagnosis of latent diabetes in individuals with a history of HDP during the long-term postpartum period. The mechanistic pathways underlying the induction of postpartum IR and T2D by HDP remain unclear. There has been considerable speculation as to whether (1) HDP and IR share common risk factors such as obesity, (2) damage such as endothelial disfunction and chronic inflammation caused by HDP during pregnancy could potentially influence the development of IR [8, 37]. Our results demonstrated that HDP, as one of the environmental stresses, is associated with long-term postpartum IR (Table 1 and Graphical Abstract). However, it was noteworthy that the HDP and T2D group demonstrated not only IR (increased HOMA2-IR and decreased MI) but also significantly decreased early phase insulin secretion (decreased Δ0–1 h Ins/Δ0–1 h PG) and ISSI-2 compared to the HDP and non-T2D group, indicating that insulin secretion factors were also involved as potential contributors to the development of T2D in HDP (Supplementary Table 1).
Graphical Abstract
A considerable number of genes associated with various diseases have been identified through GWAS, however, they account for a small portion of the genetic background. Polygenic model could substantially improve predictive ability; however, even this approach has been insufficient and fails to explain missing heritability [38]. Additionally, such polygenic model approach for common disease including T2D remains challenging due to costs. Further, it is well known that along with aging, environmental factors have a greater influence on the development of multifactorial diseases like T2D than genetic factors. Therefore, gene–environment interaction has recently attracted attention [39]. In the present study, we focused on the interaction between HDP history and T2D susceptibility gene variants identified in our prior replication study based on GWAS of a Japanese population. We found that the statistically significant interaction effect between HDP history and the CDKAL1 C/C genotype on future onset of T2D.
There are no reports on the association between CDKAL1 and the onset of HDP, and consequently, no universally relevant genes to HDP have been identified [40]. CDKAL1 has been reported to play roles in post-translational modification, insulin folding, and insulin synthesis through CDK5 inhibition. Mutations in CDKAL1 lead to impaired pancreatic beta cell function and reduced insulin secretion [36]. The rs7754840 SNP is located on intron 5 of CDKAL1. Consistent with our results, the prevalence of the C/C genotype at this locus was reported to be around 20% [41], and carrying the C allele at rs7754840 was associated with reduced early-phase insulin secretion during the 75g-OGTT, increased glucose AUC, and an increased risk of T2D [42-44]. Therefore, together with our results, it suggests that despite HDP-associated IR, insulin secretion was not compensated, at least in part, due to a CDKAL1 C/C genotype. As a result, ISSI-2, which served as a composite index of insulin secretion and insulin sensitivity, was significantly decreased (Supplementary Table 4 and Graphical Abstract).
Notably, the prevalence of the CDKAL1 C/C genotype stood at 19.2% among subjects with a history of HDP. Therefore, the simultaneous assessment of HDP history and the CDKAL1 genotype can be used to predict the risk of T2D. Consequently, understanding the CDKAL1 genotype might pave the way for an effective postpartum management strategy for individuals with a history of HDP. From the standpoint of precision medicine, these results indicate the potential for future diabetes prevention by evaluating the CDKAL1 genotype in individuals with HDP. Furthermore, our findings could serve as epidemiological evidence for future research to elucidate the mechanisms underlying the development of T2D in individuals with a history of HDP.
Finally, we performed additional analysis on the characteristics of individuals with HDP harboring CDKAL1 C/C, comparing those with and without T2D (Supplementary Table 5). Although the sample size may not be sufficient (n = 8 vs. 11, respectively), the HDP with CDKAL1 C/C developing T2D group showed a notable relative insulin secretion deficiency (decrease in AUC0–2 h Ins/AUC0–2 h PG, Δ0–1 h Ins/Δ0–1 h PG, and ISSI-2). Although the difference was not statistically significant, age in HDP, CDKAL1 C/C, and T2D group was older than those without T2DM group (71 [IQR 65–76] vs. 66 [IQR 55–70], respectively). It suggests that the decreased insulin secretion ability maybe, at least in part, due to aging. Further prospective studies with larger populations are necessary to validate the influence of aging on the development of T2D in subjects with HDP harboring CDKAL1 C/C.
Our study is subject to certain limitations. First, to confirm the reproducibility of the results of our study, a validation analysis using the other cohort may be necessary. However, there is currently no Japanese cohort investigating both HDP history and T2D development with assessments by 75g-OGTT. Therefore, we applied the bootstrap method for internal validation of results and that demonstrated excellent reproducibility. Second, the diagnosis of HDP relied on self-reported information obtained through questionnaires, and the accuracy of this diagnosis has not been independently verified or confirmed by additional tests. However, our study holds a certain level of reliability because self-reported HDP data have a sensitivity of 80% and a specificity of 96%, even when collected more than 20 years after the postpartum period [45]. Third, some confounding factors of HDP and T2D, such as the presence of pre-pregnancy obesity, have not been collected. Fourth, contrary to previous reports [36], the CDKAL1 C/C genotype was not related to T2D prevalence in this study, which involved 956 parous women (Supplementary Table 3) and a non-HDP and C/C genotype group (Table 3). Through the overall analysis of this cohort (n = 2,449), T2D prevalence was increased in participants with the CDKAL1 C/C genotype compared to the non-C/C genotype (OR = 1.77; 95% CI: 1.38–2.28, p < 0.0001). Thus, we initially estimated that surveying over 950 women would yield an 80% statistical power (α = 0.05). However, in the present study the prevalence of T2D was much lower than we expected, even though it was drawn from the same cohort. In the overall analysis of this cohort, the impact of the CDKAL1 C/C genotype on the prevalence of T2D was more significant in men (n = 874; OR 2.20 [1.48–3.25]) than women (n = 1,575; OR 1.62 [1.16–2.26]); therefore, insufficient statistical power derived from sex differences may account for this result. Fifth, our study only included Japanese; hence, there was no additional information on ethnicity. Therefore, further prospective studies with larger populations and multi-ethnic cohorts are necessary to validate the interaction effect between HDP history and CDKAL1 genotype on the development of T2D.
In conclusion, HDP history was associated with long-term postpartum IR and the prevalence of T2D defined by 75g-OGTT. And the interaction between HDP history and the CDKAL1 C/C genotype further increased the OR for T2D. Our study underscores the value of the simultaneous assessment of an HDP history and CDKAL1 genotype for predicting the future T2D development.
Acknowledgements
We thank Maki Yokoyama, Ryoichi Kawamura, Madoka Sano, Yosuke Ikeda, Toshimi Hadate, and Misaki Takakado of Ehime University for assistance with collecting samples and data, as well as Akiko Otaki and Risa Kagawa of Ehime University for their technical assistance.
Disclosure
The protocol for this research project has been approved by a suitably constituted Ethics Committee of the institution and it conforms to the provisions of the Declaration of Helsinki. Committee of Ehime University Graduate School of Medicine, Approval No. 29-K3. All informed consent was obtained from the subjects.
Takashi Sugiyama is a member of Endocrine Journal’s Editorial Board.
None of the authors have any potential conflicts of interest associated with this research.
Author Contribution
S.I., Y. Takata, H.O., and S.T. were involved in the conception and design of this study. S.I., Y. Takata, K.M., and A.Y. collected the samples. S.I., Y. Takata, and Y. Tabara analysed the data. S.I. and Y. Takata interpreted the data. S.I. and Y. Takata wrote the first draft of the manuscript. All authors edited, reviewed, and approved the final version of the manuscript. Y. Takata was the guarantor of this study.
Funding
This work was supported by Japan Society for the Promotion of Science KAKENHI [grant number JP17K08985] and the Ehime University Research Unit System. The study sponsors were not involved in the design of the study; the collection, analysis, and interpretation of data; writing the report; and did not impose any restrictions regarding the publication of the report.
Supplementary Table 1 Characteristics of participants in four groups divided by history of HDP and type 2 diabetes mellitus
| Variable |
Non-HDP and non-T2D
(n = 763) |
Non-HDP and T2D
(n = 114) |
HDP and non-T2D
(n = 78) |
HDP and T2D
(n = 23) |
p value |
| Age (years) |
65.0 (55.0–72.0) |
70.0 (64.0–75.0) |
66.0 (59.0–73.0) |
72.0 (67.0–76.0)* |
<0.0001 |
| BMI (kg/m2) |
22.2 (20.2–24.3) |
23.7 (21.4–25.7) |
23.0 (20.7–26.4) |
21.9 (20.5–27.4) |
<0.01 |
| HbA1c (%) |
5.5 (5.4–5.7) |
6.2 (5.8–6.5) |
5.6 (5.4–5.8) |
6.2 (5.8–6.6)*† |
<0.0001 |
| (mmol/mol) |
36.6 (35.5–38.8) |
43.7 (39.9–47.5) |
37.7 (35.5–39.9) |
44.3 (39.9–48.6)*† |
<0.0001 |
| Fasting PG (mmol/L) |
5.1 (4.8–5.4) |
6.0 (5.5–6.6) |
5.1 (4.8–5.5) |
6.0 (5.4–6.7)*† |
<0.0001 |
| Fasting IRI (pmol/L) |
36.0 (26.7–50.7) |
45.4 (30.2–67.7) |
46.6 (28.8–62.8) |
50.8 (36.0–89.1) |
<0.0001 |
| HOMA2-B |
70.2 (57.0–85.6) |
57.2 (39.8–76.3) |
76.5 (63.2–97.7) |
66.1 (54.2–88.2) |
<0.0001 |
| HOMA2-S |
149.0 (105.8–201.8) |
114.2 (81.0–173.7) |
119.3 (89.8–181.2) |
100.0 (61.9–153.4)* |
<0.0001 |
| HOMA2-IR |
0.67 (0.50–0.95) |
0.88 (0.58–1.23) |
0.84 (0.55–1.11) |
1.00 (0.65–1.62)* |
<0.0001 |
| 75g-OGTTa |
| Fasting PG (mmol/L) |
5.1 (4.8–5.4) |
5.9 (5.4–6.5) |
5.1 (4.8–5.4) |
5.8 (5.3–6.2)*† |
<0.0001 |
| 1 h PG (mmol/L) |
7.9 (6.4–9.6) |
12.8 (10.7–14.3) |
8.7 (6.9–10.6) |
12.4 (10.8–13.6)*† |
<0.0001 |
| 2 h PG (mmol/L) |
6.6 (5.7–7.8) |
11.7 (9.9–13.9) |
6.8 (5.7–8.2) |
11.2 (8.8–13.4)*† |
<0.0001 |
| Fasting IRI (pmol/L) |
36.0 (26.7–50.6) |
44.4 (29.9–62.2) |
46.1(28.7–62.8) |
47.4 (36.0–89.1)* |
<0.001 |
| 1 h IRI (pmol/L) |
357.1 (251.8–516.2) |
287.2 (198.7–433.9) |
424.4 (240.7–714.6) |
316.5 (162.1–559.8) |
<0.01 |
| 2 h IRI (pmol/L) |
324.3 (232.1–476.2) |
410.0 (302.1–626.4) |
370.0 (265.9–547.5) |
490.7 (255.5–722.3)* |
<0.0001 |
| AUC0–2 h PG (mmol/L h) |
13.9 (11.7–16.0) |
21.6 (18.5–24.3) |
14.0 (12.4–17.2) |
21.4 (16.8–22.8)*† |
<0.0001 |
| AUC0–2 h Ins (pmol/L h) |
550.9 (398.4–767.8) |
526.0 (384.8–789.8) |
622.9 (397.5–1031.3) |
561.6 (304.5–973.8) |
0.38 |
| AUC0–2 h Ins/AUC0–2 h PG |
40.4 (29.7–54.0) |
25.1 (18.8–36.4) |
40.9 (27.4–67.2) |
29.3 (16.1–45.3)*† |
<0.0001 |
| Δ0–1 h ins/Δ0–1 h PG |
125.3 (77.3–231.1) |
41.9 (25.4–60.9) |
122.3 (65.1–232.6) |
43.8 (26.2–66.3)*† |
<0.0001 |
| Matsuda index |
6.5 (4.7–8.9) |
4.3 (3.1–6.9) |
5.8 (3.5–8.5) |
4.2 (2.4–6.8)* |
<0.0001 |
| ISSI-2 |
2.1 (1.7–2.8) |
1.0 (0.7–1.3) |
2.0 (1.5–2.5) |
0.9 (0.8–1.2)*† |
<0.0001 |
Data were analyzed using the Kruskal–Wallis test. Intergroup differences were analyzed using the Steel’s multiple comparison test. Data are presented as median (interquartile range). a 75g-OGTT data were obtained from 918 individuals without a history of diabetes. *p < 0.05 vs. reference, †p < 0.05 vs. HDP and non-T2D group. HDP, hypertensive disorders of pregnancy; PG, plasma glucose; IRI, immunoreactive insulin; AUC0–2 h PG, glucose AUC; AUC0–2 h Ins, insulin AUC; AUC0–2 h Ins/AUC0–2 h PG, insulin AUC/glucose AUC; Δ0–1 h Ins/Δ0–1 h PG, (1 h insulin – fasting insulin) over (1 h plasma glucose – fasting plasma glucose); ISSI-2, insulin secretion-sensitivity index-2.
Supplementary Table 2 Interaction between HDP history and type 2 diabetes mellitus susceptibility genes and in the incidence of type 2 diabetes mellitus
|
aOR (95% CI) for T2D
(HDP vs. non-HDP) |
p value |
p for multiplicative interaction |
RERIOR (95%CI) |
| Overall |
1.75 (1.02, 2.97) |
0.04 |
|
|
| CDKAL1 (rs7754840) |
|
|
0.03 |
3.85 (0.50, 12.11) |
| Non-C/C genotype |
1.25 (0.66, 2.37) |
0.49 |
|
|
| C/C genotype |
4.79 (1.66, 13.83) |
<0.01 |
|
|
| KCNJ11 (rs5219) |
|
|
0.09 |
–2.31 (–4.97, 1.54) |
| Non-T/T genotype |
2.22 (1.20, 4.12) |
0.01 |
|
|
| T/T genotype |
0.19 (0.02, 2.15) |
0.18 |
|
|
| HHEX (rs1111875) |
|
|
0.41 |
–1.41 (–4.15, 8.03) |
| Non-G/G genotype |
1.97 (1.09, 3.56) |
0.02 |
|
|
| G/G genotype |
0.64 (0.07, 6.36) |
0.70 |
|
|
| IGF2BP2 (rs1470579) |
|
|
0.41 |
2.79 (–1.40, 17.19) |
| Non-C/C genotype |
1.84 (0.97, 3.47) |
0.06 |
|
|
| C/C genotype |
3.37 (0.79, 14.35) |
0.10 |
|
|
| CDKN2A/B (rs10811661) |
|
|
0.93 |
–0.12 (–2.08, 2.65) |
| Non-T/T genotype |
1.80 (0.96, 3.39) |
0.07 |
|
|
| T/T genotype |
1.92 (0.69, 5.33) |
0.21 |
|
|
| TCF7L2 (rs12255372)a |
|
|
0.84 |
0.52 (–7.10, 8.14) |
| G/G genotype |
1.84 (1.01, 3.33) |
<0.05 |
|
|
| T/G genotype |
2.00 (0.23, 17.29) |
0.53 |
|
|
| Polygenic risk score |
|
|
0.59 |
1.31 (–1.52, 6.90) |
| Lower four quintile |
1.64 (0.80, 3.39) |
0.18 |
|
|
| Upper quintile |
2.32 (0.80, 6.75) |
0.12 |
|
|
Age and BMI adjusted odds ratio (aOR)s, 95% CIs, and p values for multiplicative interactions between a history of HDP and homozygous T2D susceptibility risk alleles or upper quintile of polygenic risk score (PRS) were calculated by using a multivariate logistic regression analysis. RERIOR for additive interaction was calculated by the MOVER method. These SNPs and ORs used to calculate PRS were selected from our prior replication study based on genome wide association study of a Japanese population. SLC30A8 (rs13266634) was excluded for deviating from Hardy-Weinberg equilibrium. a As there were no individual homozygotes of the TCF7L2 (rs12255372) risk allele (T/T), individual heterozygotes (T/G) were considered as the risk genotype. HDP, hypertensive disorders of pregnancy; T2D, type 2 diabetes mellitus; RERI, relative excess risks due to interaction; N/A, not applicable.
Supplementary Table 3 Characteristics of parous women with or without the
CDKAL1 C/C genotype
|
With CDKAL1 C/C
(n = 168) |
With CDKAL1 C/G or G/G
(n = 788) |
p value |
| Age (years) |
65.0 (57.0–71.0) |
66.0 (57.0–73.0) |
0.24 |
| BMI (kg/m2) |
22.0 (20.3–24.2) |
22.5 (20.4–24.7) |
0.33 |
| HbA1c (%) |
5.6 (5.4–5.9) |
5.6 (5.4–5.8) |
0.68 |
| (mmol/mol) |
37.7 (35.5–41.0) |
37.7 (35.5–39.9) |
0.68 |
| Fasting PG (mmol/L) |
5.22 (4.94–5.61) |
5.11 (4.83–5.50) |
0.02 |
| Fasting IRI (pmol/L) |
38.2 (27.5–53.9) |
37.4 (27.3–54.8) |
0.94 |
| HOMA2-B |
68.5 (52.8–83.0) |
69.9 (56.3–86.4) |
0.18 |
| HOMA2-S |
140.2 (96.9–204.7) |
143.1 (98.8–197.9) |
0.88 |
| HOMA2-IR |
0.71 (0.49–1.03) |
0.70 (0.51–1.01) |
0.88 |
| Total cholesterol (mmol/L) |
5.6 (5.1–6.1) |
5.5 (4.9–6.0) |
<0.05 |
| HDL-cholesterol (mmol/L) |
1.7 (1.5–2.0) |
1.7 (1.5–2.0) |
0.73 |
| LDL-cholesterol (mmol/L) |
3.2 (2.8–3.6) |
3.1 (2.7–3.7) |
0.19 |
| Triglyceride (mmol/L) |
1.0 (0.8–1.4) |
1.0 (0.7–1.3) |
0.32 |
| 75g-OGTTa |
| Fasting PG (mmol/L) |
5.2 (4.9–5.6) |
5.1 (4.8–5.5) |
0.31 |
| 1 h PG (mmol/L) |
9.2 (7.2–11.1) |
8.2 (6.6–10.2) |
<0.01 |
| 2 h PG (mmol/L) |
7.3 (6.1–9.2) |
6.8 (5.8–8.3) |
0.02 |
| Fasting IRI (pmol/L) |
38.4 (26.3–53.9) |
37.1 (27.2–53.8) |
0.97 |
| 1 h IRI (pmol/L) |
353.0 (259.6–490.8) |
351.4 (242.7–525.7) |
0.93 |
| 2 h IRI (pmol/L) |
343.6 (252.2–504.8) |
336.8 (238.9–501.4) |
0.06 |
| AUC0–2 h PG (mmol/L h) |
15.4 (12.9–18.3) |
14.2 (11.9–16.9) |
<0.001 |
| AUC0–2 h Ins (pmol/L h) |
570.5 (419.5–749.3) |
549.2 (390.2–802.5) |
0.81 |
| AUC0–2 h Ins/AUC0–2 h PG |
37.0 (28.8–49.4) |
39.4 (28.3–53.6) |
0.15 |
| Δ0–1 h Ins/Δ0–1 h PG |
91.6 (62.9–139.9) |
117.1 (65.7–222.8) |
<0.001 |
| Matsuda index |
5.9 (4.0–8.1) |
6.2 (4.2–8.7) |
0.17 |
| ISSI-2 |
1.8 (1.3–2.5) |
2.0 (1.5–2.7) |
<0.01 |
| Prevalence of HDP, n (%) |
19 (11.3) |
80 (10.2) |
0.66 |
| Prevalence of T2D, n (%) |
26 (15.5) |
102 (12.9) |
0.38 |
Continuous variables were analyzed using the Mann–Whitney U test, and categorical variables were analyzed using the chi-square test. Data are presented as median (interquartile range) unless otherwise specified.
a 75g-OGTT data were obtained from 898 individuals without a history of diabetes on the day of the study. HDP, hypertensive disorders of pregnancy; T2D, type 2 diabetes mellitus; PG, plasma glucose; IRI, immunoreactive insulin; AUC0–2 h PG, glucose AUC; AUC0–2 h Ins, insulin AUC; AUC0–2 h Ins/AUC0–2 h PG, insulin AUC/glucose AUC; Δ0–1 h Ins/Δ0–1 h PG, (1 h insulin – fasting insulin) over (1 h plasma glucose – fasting plasma glucose); ISSI-2, insulin secretion-sensitivity index-2.
Supplementary Table 4 Characteristics of participants in four groups divided by history of HDP and genotype of the
CDKAL1 SNP rs7754840
| Variable |
Non-HDP and non-C/C
(n = 708) |
Non-HDP and C/C
(n = 149) |
HDP and non-C/C
(n = 80) |
HDP and C/C
(n = 19) |
p value |
| Age (years) |
66 (57–73) |
65 (56–70) |
68 (61–74) |
68 (60–74) |
0.16 |
| BMI (kg/m2) |
22.3 (20.3–24.6) |
22.0 (20.0–24.2) |
23.4 (20.7–26.8) |
21.9 (20.7–24.2) |
0.15 |
| Waist (cm) |
82.0 (76.0–89.0) |
81.3 (74.0–87.0) |
85.8 (75.8–93.8) |
83.0 (75.0–89.0) |
0.06 |
| SBP (mmHg) |
124 (110–137) |
126 (109–138) |
130 (117–142) |
124 (118–139) |
0.09 |
| DBP (mmHg) |
74 (66–82) |
74 (67–82) |
78 (70–85) |
76 (69–80) |
0.06 |
| HbA1c (%) |
5.6 (5.4–5.8) |
5.5 (5.4–5.8) |
5.6 (5.4–5.9) |
6.0 (5.6–6.3)*†‡ |
<0.01 |
| (mmol/mol) |
37.7 (35.5–39.9) |
37.7 (35.5–39.9) |
37.7 (35.5–41.0) |
42.1 (37.7–45.4)*†‡ |
<0.01 |
| Fasting PG (mmol/L) |
5.1 (4.8–5.5) |
5.2 (4.9–5.6) |
5.2 (4.9–5.6) |
5.7 (5.2–6.2)*†‡ |
<0.01 |
| Fasting IRI (pmol/L) |
36.5 (27.0–52.6) |
37.9 (24.7–53.9) |
48.9 (31.1–68.4) |
43.3 (28.8–52.9) |
<0.01 |
| HOMA2-beta |
69.0 (56.1–85.1) |
70.2 (54.6–84.4) |
77.8 (63.0–96.0) |
66.1 (52.7–75.7) |
0.01 |
| HOMA2-S |
147.0 (102.4–198.7) |
140.2 (95.4–214.0) |
112.4 (78.5–173.5) |
138.1 (96.9–181.2) |
0.02 |
| HOMA2-IR |
0.68 (0.50–0.98) |
0.71 (0.47–1.05) |
0.89 (0.58–1.28) |
0.72 (0.55–1.03) |
0.02 |
| Total cholesterol (mmol/L) |
5.46 (4.91–6.00) |
5.59 (5.15–6.15) |
5.50 (5.09–5.90) |
5.33 (4.86–5.92) |
0.15 |
| HDL-cholesterol (mmol/L) |
1.71 (1.45–1.97) |
1.68 (1.45–1.99) |
1.72 (1.44–2.02) |
1.66 (1.22–1.97) |
0.94 |
| LDL-cholesterol (mmol/L) |
3.13 (2.66–3.67) |
3.23 (2.87–3.65) |
3.13 (2.73–3.50) |
2.84 (2.40–3.39) |
0.08 |
| Triglyceride (mmol/L) |
0.97 (0.72–1.31) |
1.03 (0.76–1.37) |
1.10 (0.77–1.54) |
1.11 (0.79–2.00) |
0.10 |
| Hs-CRP (mg/dL) |
0.03 (0.02–0.07) |
0.05 (0.03–0.10) |
0.04 (0.02–0.07) |
0.07 (0.04–0.10) |
0.34 |
| 75g-OGTTa |
| Fasting PG (mmol/L) |
5.2 (4.9–5.5) |
5.2 (4.9–5.6) |
5.2 (4.9–5.6) |
5.7 (5.2–6.2)*†‡ |
<0.01 |
| 1 h PG (mmol/L) |
8.1 (6.5–10.1) |
8.8 (7.1–10.3) |
8.8 (6.9–10.8) |
11.9 (10.0–13.1)*†‡ |
<0.0001 |
| 2 h PG (mmol/L) |
6.7 (5.8–8.3) |
7.2 (6.0–9.1) |
7.1 (5.8–8.6) |
8.8 (7.1–10.9)*†‡ |
<0.01 |
| Fasting IRI (pmol/L) |
36.4 (27.0–52.6) |
37.8 (24.4–53.7) |
48.9 (31.1–68.4) |
43.3 (29.5–73.5) |
<0.01 |
| 1 h IRI (pmol/L) |
346.2 (242.7–511.4) |
368.5 (271.5–490.8) |
452.3 (254.2–644.0) |
279.6 (194.4–726.4) |
0.19 |
| 2 h IRI (pmol/L) |
330.4 (237.0–494.5) |
347.0 (249.7–504.8) |
389.2 (266.2–564.4) |
399.7 (270.0–596.5) |
0.13 |
| AUC0–2 h PG (mmol/L h) |
14.2 (11.8–16.9) |
15.0 (12.6–17.2) |
14.5 (12.5–17.3) |
18.7 (17.4–20.4)*†‡ |
<0.0001 |
| AUC0–2 h Ins (pmol/L h) |
543.3 (389.1–780.2) |
573.9 (435.8–749.3) |
664.7 (422.1–1000.8) |
501.0 (382.5–1046.7) |
0.16 |
| AUC0–2 h Ins/AUC0–2 h PG |
39.3 (28.4–53.3) |
38.3 (29.5–49.5) |
41.0 (29.5–65.6) |
30.8 (17.0–59.2) |
0.14 |
| Δ0–1 h Ins/Δ0–1 h PG |
117.5 (67.6–225.6) |
95.1 (66.1–147.6) |
113.9 (61.9–219.0) |
61.0 (26.2–108.3)*†‡ |
<0.001 |
| Matsuda index |
6.39 (4.36–8.88) |
6.01 (4.16–8.12) |
5.08 (3.08–8.41) |
4.32 (3.31–7.30) |
0.01 |
| ISSI-2 |
2.05 (1.53–2.73) |
1.89 (1.41–2.60) |
1.95 (1.42–2.47) |
1.21 (0.85–1.39)*†‡ |
<0.0001 |
Continuous variables were analyzed using the Kruskal–Wallis test, and categorical variables were analyzed using the chi-square test.
Intergroup differences were analyzed using the Steel’s multiple comparison test. Data are presented as median (IQR) unless otherwise specified. a 75g-OGTT data were obtained from 898 individuals without a history of diabetes on the day of the study.
*p < 0.01 vs. reference, †p < 0.01 vs. non-HDP and C/C group, ‡p < 0.05 vs. HDP and non-C/C group. HDP, hypertensive disorders of pregnancy; SBP, systolic blood pressure; DBP, diastolic blood pressure; Hs-CRP, high-sensitivity C-reactive protein; PG, plasma glucose; IRI, immunoreactive insulin; AUC0–2 h PG, glucose AUC; AUC0–2 h Ins, insulin AUC; AUC0–2 h Ins/AUC0–2 h PG, insulin AUC/glucose AUC; Δ0–1 h Ins/Δ0–1 h PG, (1 h insulin – fasting insulin)/(1 h plasma glucose – fasting plasma glucose); ISSI-2, insulin secretion-sensitivity index-2.
Supplementary Table 5 Characteristics of women with a history of HDP and carrying the
CDKAL1 C/C genotype in relation to T2D
|
HDP and CDKAL1 C/C
with T2D (n = 8) |
HDP and CDKAL1 C/C
without T2D (n = 11) |
p value |
| Age (years) |
71.0 (65.0–76.0) |
66.0 (55.0–70.0) |
0.11 |
| BMI (kg/m2) |
21.5 (20.4–22.3) |
23.7 (21.5–25.5) |
0.07 |
| Waist (cm) |
78.5 (73.0–85.0) |
85.0 (83.0–95.0) |
0.08 |
| SBP (mmHg) |
124 (114–134) |
127 (118–139) |
0.52 |
| DBP (mmHg) |
73 (68–80) |
76 (71–80) |
0.56 |
| HbA1c (%) |
6.3 (5.8–6.7) |
5.9 (5.6–6.1) |
0.11 |
| (mmol/mol) |
44.8 (39.9–49.2) |
41.0 (37.7–43.2) |
0.11 |
| Fasting PG (mmol/L) |
5.22 (4.94–5.61) |
5.11 (4.83–5.50) |
0.02 |
| Fasting IRI (pmol/L) |
38.2 (27.5–53.9) |
37.4 (27.3–54.8) |
0.94 |
| HOMA2-B |
59.7 (50.5–67.2) |
70.2 (52.7–119.7) |
0.16 |
| HOMA2-S |
148.1 (109.4–211.4) |
134.9 (53.4–178.1) |
0.28 |
| HOMA2-IR |
0.68 (0.47–0.93) |
0.74 (0.56–1.87) |
0.28 |
| Total cholesterol (mmol/L) |
5.8 (4.6–6.4) |
5.3 (5.0–5.8) |
0.43 |
| HDL-cholesterol (mmol/L) |
1.6 (1.2–1.9) |
1.7 (1.6–2.0) |
0.56 |
| LDL-cholesterol (mmol/L) |
3.1 (2.6–3.6) |
2.8 (2.4–3.2) |
0.32 |
| Triglyceride (mmol/L) |
1.2 (0.9–2.1) |
1.1 (0.8–2.0) |
0.93 |
| Hs-CRP (mg/dL) |
0.050 (0.020–0.073) |
0.058 (0.035–0.093) |
0.32 |
| eGFR (mL/min/1.73 m2) |
66.4 (57.8–75.9) |
70.3 (58.3–77.3) |
0.41 |
| 75g-OGTTa |
| Fasting PG (mmol/L) |
5.8 (5.2–6.4) |
5.5 (5.3–5.8) |
0.53 |
| 1 h PG (mmol/L) |
12.5 (10.4–15.1) |
11.2 (8.6–12.4) |
0.20 |
| 2 h PG (mmol/L) |
11.6 (8.9–15.7) |
7.2 (5.3–9.1) |
<0.01 |
| Fasting IRI (pmol/L) |
39.7 (30.3–47.1) |
47.1 (31.1–68.4) |
0.25 |
| 1 h IRI (pmol/L) |
177.5 (139.6–288.2) |
551.9 (247.2–789.6) |
0.01 |
| 2 h IRI (pmol/L) |
357.9 (248.8–529.0) |
429.1 (273.6–795.9) |
0.28 |
| AUC0–2 h PG (mmol/L h) |
19.8 (18.9–26.3) |
18.2 (15.5–18.7) |
0.02 |
| AUC0–2 h Ins (pmol/L h) |
421.4 (274.1–531.3) |
748.7 (401.1–1228.9) |
0.04 |
| AUC0–2 h Ins/AUC0–2 h PG |
19.6 (12.7–26.9) |
43.0 (30.8–67.4) |
0.01 |
| Δ0–1 h Ins/Δ0–1 h PG |
30.4 (14.0–40.3) |
105.6 (69.1–154.3) |
<0.01 |
| Matsuda index |
5.8 (4.3–7.7) |
3.8 (2.3–7.0) |
0.12 |
| ISSI-2 |
1.0 (0.7–1.3) |
1.3 (1.0–1.7) |
0.07 |
| Urinary ACR (mg/gCr) |
7.68 (4.13–9.24) |
6.97 (4.40–11.43) |
0.85 |
| Prevalence of disease, n (%) |
| Hypertension |
6 (75.0) |
7 (63.6) |
0.61 |
| Dyslipidemia |
8 (100.0) |
8 (72.7) |
0.12 |
| Obesity |
1 (12.5) |
3 (27.3) |
0.45 |
| MetS |
4 (50.0) |
7 (63.6) |
0.56 |
| CKD |
2 (25.0) |
3 (27.3) |
0.91 |
Continuous variables were analyzed using the Mann–Whitney U test, and categorical variables were analyzed using the chi-square test. Data are presented as median (IQR) unless otherwise specified.
a 75g-OGTT data were obtained from 91 individuals without a history of diabetes on the day of the study. HDP, hypertensive disorders of pregnancy; T2D, type 2 diabetes; SBP, systolic blood pressure; DBP, diastolic blood pressure; Hs-CRP, high-sensitivity C-reactive protein; PG, plasma glucose; IRI, immunoreactive insulin; AUC0–2 h PG, glucose AUC; AUC0–2 h Ins, insulin AUC; AUC0–2 h Ins/AUC0–2 h PG, insulin AUC/glucose AUC; Δ0–1 h Ins/Δ0–1 h PG, (1 h insulin – fasting insulin)/(1 h plasma glucose – fasting plasma glucose); ISSI-2, insulin secretion-sensitivity index-2; ACR, albumin-to-creatinine ratio; MetS, metabolic syndrome; CKD, chronic kidney disease.
References
- 1 Umesawa M, Kobashi G (2017) Epidemiology of hypertensive disorders in pregnancy: prevalence, risk factors, predictors and prognosis. Hypertens Res 40: 213–220.
- 2 (2020) Gestational hypertension and preeclampsia: ACOG practice bulletin summary, number 222. Obstet Gynecol 135: 1492–1495.
- 3 Bellamy L, Casas JP, Hingorani AD, Williams DJ (2007) Pre-eclampsia and risk of cardiovascular disease and cancer in later life: systematic review and meta-analysis. BMJ 335: 974.
- 4 McDonald SD, Malinowski A, Zhou Q, Yusuf S, Devereaux PJ (2008) Cardiovascular sequelae of preeclampsia/eclampsia: a systematic review and meta-analyses. Am Heart J 156: 918–930.
- 5 Sattar N, Greer IA (2002) Pregnancy complications and maternal cardiovascular risk: opportunities for intervention and screening? BMJ 325: 157–160.
- 6 Kurabayashi T, Mizunuma H, Kubota T, Kiyohara Y, Nagai K, et al. (2013) Pregnancy-induced hypertension is associated with maternal history and a risk of cardiovascular disease in later life: Japanese cross-sectional study. Maturitas 75: 227–231.
- 7 Lykke JA, Langhoff-Roos J, Sibai BM, Funai EF, Triche EW, et al. (2009) Hypertensive pregnancy disorders and subsequent cardiovascular morbidity and type 2 diabetes mellitus in the mother. Hypertension 53: 944–951.
- 8 Feig DS, Shah BR, Lipscombe LL, Wu CF, Ray JG, et al. (2013) Preeclampsia as a risk factor for diabetes: a population-based cohort study. PLoS Med 10: e1001425.
- 9 Männistö T, Mendola P, Vääräsmäki M, Järvelin MR, Hartikainen AL, et al. (2013) Elevated blood pressure in pregnancy and subsequent chronic disease risk. Circulation 127: 681–690.
- 10 Libby G, Murphy DJ, McEwan NF, Greene SA, Forsyth JS, et al. (2007) Pre-eclampsia and the later development of type 2 diabetes in mothers and their children: an intergenerational study from the Walker cohort. Diabetologia 50: 523–530.
- 11 Savitz DA, Danilack VA, Elston B, Lipkind HS (2014) Pregnancy-induced hypertension and diabetes and the risk of cardiovascular disease, stroke, and diabetes hospitalization in the year following delivery. Am J Epidemiol 180: 41–44.
- 12 Callaway LK, Lawlor DA, O’Callaghan M, Williams GM, Najman JM, et al. (2007) Diabetes mellitus in the 21 years after a pregnancy that was complicated by hypertension: findings from a prospective cohort study. Am J Obstet Gynecol 197: 492.e491–e497.
- 13 Laivuori H, Tikkanen MJ, Ylikorkala O (1996) Hyperinsulinemia 17 years after preeclamptic first pregnancy. J Clin Endocrinol Metab 81: 2908–2911.
- 14 Spaan JJ, Houben AJ, Musella A, Ekhart T, Spaanderman ME, et al. (2010) Insulin resistance relates to microvascular reactivity 23 years after preeclampsia. Microvasc Res 80: 417–421.
- 15 McDonald SD, Ray J, Teo K, Jung H, Salehian O, et al. (2013) Measures of cardiovascular risk and subclinical atherosclerosis in a cohort of women with a remote history of preeclampsia. Atherosclerosis 229: 234–239.
- 16 Alsnes IV, Janszky I, Forman MR, Vatten LJ, Økland I (2014) A population-based study of associations between preeclampsia and later cardiovascular risk factors. Am J Obstet Gynecol 211: 657.e651–e657.
- 17 Kido Y (2017) Gene-environment interaction in type 2 diabetes. Diabetol Int 8: 7–13.
- 18 Griffin S (2022) Diabetes precision medicine: plenty of potential, pitfalls and perils but not yet ready for prime time. Diabetologia 65: 1913–1921.
- 19 Sørensen TIA, Metz S, Kilpeläinen TO (2022) Do gene-environment interactions have implications for the precision prevention of type 2 diabetes? Diabetologia 65: 1804–1813.
- 20 Tabara Y, Osawa H, Kawamoto R, Onuma H, Shimizu I, et al. (2009) Replication study of candidate genes associated with type 2 diabetes based on genome-wide screening. Diabetes 58: 493–498.
- 21 Tabara Y, Saito I, Nishida W, Kohara K, Sakurai S, et al. (2011) Relatively lower central aortic pressure in patients with impaired insulin sensitivity and resistance: the toon health study. J Hypertens 29: 1948–1954.
- 22 Umemura S, Arima H, Arima S, Asayama K, Dohi Y, et al. (2019) The Japanese Society of Hypertension Guidelines for the Management of Hypertension (JSH 2019). Hypertens Res 42: 1235–1481.
- 23 American Diabetes Association Professional Practice Committee (2022) 2. Classification and diagnosis of diabetes: standards of medical care in diabetes-2022. Diabetes Care 45: S17–S38.
- 24 Okamura T, Tsukamoto K, Arai H, Fujioka Y, Ishigaki Y, et al. (2024) Japan Atherosclerosis Society (JAS) Guidelines for Prevention of Atherosclerotic Cardiovascular Diseases 2022. J Atheroscler Thromb 31: 641–853.
- 25 Alberti KG, Zimmet P, Shaw J (2005) The metabolic syndrome—a new worldwide definition. Lancet 366: 1059–1062.
- 26 Matsuda M, DeFronzo RA (1999) Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care 22: 1462–1470.
- 27 Retnakaran R, Qi Y, Goran MI, Hamilton JK (2009) Evaluation of proposed oral disposition index measures in relation to the actual disposition index. Diabet Med 26: 1198–1203.
- 28 Abdul-Ghani MA, Williams K, DeFronzo RA, Stern M (2007) What is the best predictor of future type 2 diabetes? Diabetes Care 30: 1544–1548.
- 29 Retnakaran R, Shen S, Hanley AJ, Vuksan V, Hamilton JK, et al. (2008) Hyperbolic relationship between insulin secretion and sensitivity on oral glucose tolerance test. Obesity (Silver Spring) 16: 1901–1907.
- 30 Stancáková A, Javorský M, Kuulasmaa T, Haffner SM, Kuusisto J, et al. (2009) Changes in insulin sensitivity and insulin release in relation to glycemia and glucose tolerance in 6,414 Finnish men. Diabetes 58: 1212–1221.
- 31 DeFronzo RA, Tripathy D, Abdul-Ghani M, Musi N, Gastaldelli A (2014) The disposition index does not reflect β-cell function in IGT subjects treated with pioglitazone. J Clin Endocrinol Metab 99: 3774–3781.
- 32 Dudbridge F (2013) Power and predictive accuracy of polygenic risk scores. PLoS Genet 9: e1003348.
- 33 VanderWeele T, Knol M (2014) A tutorial on interaction. Epidemiol Method 3: 33–72.
- 34 Zou GY (2008) On the estimation of additive interaction by use of the four-by-two table and beyond. Am J Epidemiol 168: 212–224.
- 35 Imamura M, Shigemizu D, Tsunoda T, Iwata M, Maegawa H, et al. (2013) Assessing the clinical utility of a genetic risk score constructed using 49 susceptibility alleles for type 2 diabetes in a Japanese population. J Clin Endocrinol Metab 98: E1667–E1673.
- 36 Ghosh C, Das N, Saha S, Kundu T, Sircar D, et al. (2022) Involvement of Cdkal1 in the etiology of type 2 diabetes mellitus and microvascular diabetic complications: a review. J Diabetes Metab Disord 21: 991–1001.
- 37 Calles-Escandon J, Cipolla M (2001) Diabetes and endothelial dysfunction: a clinical perspective. Endocr Rev 22: 36–52.
- 38 Torkamani A, Wineinger NE, Topol EJ (2018) The personal and clinical utility of polygenic risk scores. Nat Rev Genet 19: 581–590.
- 39 Virolainen SJ, VonHandorf A, Viel K, Weirauch MT, Kottyan LC (2023) Gene-environment interactions and their impact on human health. Genes Immun 24: 1–11.
- 40 Parada-Niño L, Castillo-León LF, Morel A (2022) Preeclampsia, natural history, genes, and miRNAs associated with the syndrome. J Pregnancy 2022: 3851225.
- 41 Tuerxunyiming M, Mohemaiti P, Wufuer H, Tuheti A (2015) Association of rs7754840 G/C polymorphisms in CDKAL1 with type 2 diabetes: a meta-analysis of 70141 subjects. Int J Clin Exp Med 8: 17392–17405.
- 42 Horikawa Y, Miyake K, Yasuda K, Enya M, Hirota Y, et al. (2008) Replication of genome-wide association studies of type 2 diabetes susceptibility in Japan. J Clin Endocrinol Metab 93: 3136–3141.
- 43 Lee YH, Kang ES, Kim SH, Han SJ, Kim CH, et al. (2008) Association between polymorphisms in SLC30A8, HHEX, CDKN2A/B, IGF2BP2, FTO, WFS1, CDKAL1, KCNQ1 and type 2 diabetes in the Korean population. J Hum Genet 53: 991–998.
- 44 Stancáková A, Pihlajamäki J, Kuusisto J, Stefan N, Fritsche A, et al. (2008) Single-nucleotide polymorphism rs7754840 of CDKAL1 is associated with impaired insulin secretion in nondiabetic offspring of type 2 diabetic subjects and in a large sample of men with normal glucose tolerance. J Clin Endocrinol Metab 93: 1924–1930.
- 45 Diehl CL, Brost BC, Hogan MC, Elesber AA, Offord KP, et al. (2008) Preeclampsia as a risk factor for cardiovascular disease later in life: validation of a preeclampsia questionnaire. Am J Obstet Gynecol 198: e11–e13.
Abbreviations
HDPHypertensive disorders of pregnancy
IRInsulin resistance
T2DType 2 diabetes
75g-OGTT75 g oral glucose tolerance test
GWASgenome wide association study
SNPSingle nucleotide polymorphism
BMIBody mass index
PGPlasma glucose
AUC0–2h PGAreas under the plasma glucose curve
AUC0–2h InsAreas under the insulin curve
MIMatsuda index
ISSI-2Insulin secretion-sensitivity index
OROdds ratio
CIConfidence interval
PRSPolygenic risk score
RERIRelative excess risks due to interaction
CDKAL1Cyclin-dependent kinase 5 regulatory subunit associated protein 1-like 1
HOMA2-IRHomeostasis model assessment 2 for insulin resistance
HOMA2-betaHomeostasis model assessment 2 for beta cell function
Δ0–1 h Ins/Δ0–1 h PGΔ0–1 h insulin/Δ0–1 h plasma glucose
