Zinc-Associated Variant in SLC30A8 Gene Interacts With Gestational Weight Gain on Postpartum Glycemic Changes: A Longitudinal Study in Women With Prior Gestational Diabetes Mellitus.

Zinc transporter 8 genetic variant SLC30A8 has been associated with postpartum risk of type 2 diabetes among women with gestational diabetes mellitus (GDM). Gestational weight gain is one of the strongest risk factors for postpartum hyperglycemia. We assessed the interaction between type 2 diabetes-associated SLC30A8 rs13266634 and gestational weight gain on 1-5 years of postpartum glycemic changes in 1,071 women with prior GDM in a longitudinal study. Compared with gestation of 26-30 weeks, postpartum levels of fasting glucose, oral glucose tolerance test 2-h glucose, and hemoglobin A1c (HbA1c) increased across rs13266634 TT, CT, and CC genotypes in women with excessive gestational weight gain, whereas opposite genetic associations were found in women with inadequate or adequate gestational weight gain. Postpartum changes in fasting glucose per additional copy of the C allele were -0.18, -0.04, and 0.12 mmol/L in women with inadequate, adequate, and excessive gestational weight gain, respectively (P for interaction = 0.002). We also found similar interactions for changes in 2-h glucose and HbA1c (P for interaction = 0.003 and 0.005, respectively). Our data indicate that gestational weight gain may modify SLC30A8 variant on long-term glycemic changes, highlighting the importance of gestational weight control in the prevention of postpartum hyperglycemia in women with GDM.

Introduction

Women with prior gestational diabetes mellitus (GDM) are at substantially increased risk of developing type 2 diabetes (1). Gestational weight gain has been considered as one of the strongest risk factors for GDM and has a long-term effect on postpartum development of hyperglycemia and type 2 diabetes (2,3).

Zinc is an essential micronutrient that is necessary in the human β-cell for insulin homeostasis. Both dietary zinc intake and serum levels of zinc have been inversely associated with gestational hyperglycemia, especially among women with GDM (4,5). Zinc homeostasis in the pancreatic β-cell is mainly mediated by zinc transporter 8 (ZnT8), the product of zinc transporter 8 gene SLC30A8 (6). The genetic variant rs13266634 in SLC30A8 has been associated with fasting glucose levels, type 2 diabetes, and GDM in genome-wide association studies and genetic association studies (7–9). Management of gestational weight gain prevents hyperglycemia (2,3,10). In addition, our previous study suggested that weight change might interact with the genetic factors in relation to glycemic control in women with a history of GDM (11). On the basis of the evidence above, we hypothesized that gestational weight gain might modify the genetic association of the SLC30A8 variant with postpartum glycemic changes.

In one of the largest cohorts of women with prior GDM, we took advantage of the longitudinally collected data on changes in body weight and glycemic traits during and after pregnancy and aimed to examine the association of SLC30A8 rs13266634 with postpartum changes in measures of glucose metabolism and particularly assessed the interactions of gestational weight gain with the SLC30A8 genotype on glycemic changes.

Research Design and Methods

Study Population

The study population included a retrospective cohort from the Tianjin Gestational Diabetes Mellitus Prevention Program (12–15). The details of the program have been described previously (12). All pregnant women living in the six central urban districts in Tianjin have participated in a GDM screening since 1999; the average proportion of screened pregnancies was >91% between 1999 and 2008 (15). All pregnant women at 26–30 weeks’ gestation participated in a 1-h 50-g glucose screening test, and those who had a glucose ≥7.8 mmol/L were invited to undergo a 75-g glucose 2-h oral glucose tolerance test (OGTT) at Tianjin Women and Children’s Health Center (15). In accordance with the World Health Organization (WHO) criteria (16), women with a 75-g glucose 2-h OGTT result confirming either diabetes (fasting glucose ≥7 mmol/L or 2-h glucose ≥11.1 mmol/L) or impaired glucose tolerance (2-h glucose ≥7.8 and <11.1 mmol/L) were diagnosed as having GDM. All pregnant women who were diagnosed with GDM between 2005 and 2009 (N = 4,644) were recruited 1–5 years after delivery from 2009 to 2011. Participants who met any of the following criteria were excluded: 1) diagnosed postpartum diabetes, 2) taking medicines known to alter OGTT, 3) presence of any chronic diseases that could seriously reduce the life expectancy or the ability to participate in the study, and 4) currently pregnant or planning to become pregnant in the next 2 years. Finally, 1,263 women with GDM returned and completed the survey. Between the returned and unreturned women, there were no differences with regard to age (28.9 vs. 28.7 years), fasting glucose (5.34 vs. 5.34 mmol/L), 2-h glucose (9.23 vs. 9.16 mmol/L), and prevalence of impaired glucose tolerance (90.9% vs. 91.8%) and diabetes (9.1% vs. 8.2%) at the 26–30 gestational weeks (13). The final analysis included 1,071 women with available data of genotyping, weight, and glycemic traits. The study was approved by the Human Subjects Committee of the Tianjin Women and Children’s Health Center, and informed consent was obtained from each participant.

Assessment of Weight Change, Glycemic Traits, and Covariates

At the postpartum survey, a questionnaire was used to collect information on socio-demographics, family history of diabetes, history of GDM (values of fasting and 2-h glucose in the 26–30 gestational weeks’ OGTT were copied from Central Lab and the treatment of GDM during the pregnancy), pregnancy outcomes (prepregnancy weight and gestational weight gain), diets, current alcohol drinking, current smoking, and physical activity (the frequency and duration of leisure time and sedentary activities). All participants completed a 3-day 24-h food record using dietary record collection method taught by a dietitian. The performance of the questionnaire and 3-day 24-h food record has been validated in the China National Nutrition and Health Survey (17,18).

Body weight and height were measured by trained researchers according to a standardized protocol at the postpartum survey. BMI was calculated by dividing weight in kilograms by the square of height in meters. Gestational weight gain was categorized as inadequate, adequate, and excessive according to the 2009 Institute of Medicine guideline (19). Adequate gestational weight gain was defined according to prepregnancy BMI as 12.5–18.0 kg (prepregnancy BMI <18.5 kg/m2), 11.5–16.0 kg (prepregnancy BMI 18.5–24.9 kg/m2), 7.0–11.5 kg (prepregnancy BMI 25.0–29.9 kg/m2), and 5.0–9.0 kg (prepregnancy BMI ≥ 30 kg/m2). We also defined adequate gestational weight gain according to the Chinese maternal prepregnancy BMI classification (20) and the 2009 Institute of Medicine guideline as 12.5–18.0 kg (prepregnancy BMI <18.5 kg/m2), 11.5–16.0 kg (prepregnancy BMI 18.5–23.9 kg/m2), 7.0–11.5 kg (prepregnancy BMI 24.0–27.9 kg/m2), and 5.0–9.0 kg (prepregnancy BMI ≥ 28 kg/m2). Gestational weight gain below or above the recommendation was defined as inadequate or excessive, respectively. Annual postpartum weight change was calculated by dividing the difference between postpartum weight and predelivery weight (prepregnancy weight plus gestational weight gain) by follow-up years.

Postpartum blood samples were collected in all participants after an overnight fast of at least 12 h. Participants were given a 2-h 75-g OGTT, and fasting and 2-h glucose were measured on an automatic analyzer (Toshiba TBA 120FR, Japan). Glycemic hemoglobin A1c (HbA1c) was measured using an automatic glycohemoglobin analyzer (ADAMS A1c, HA-8160; ARKRAY, Japan). Changes in fasting glucose, 2-h glucose, and HbA1c were calculated as differences in glucose and HbA1c levels between postpartum survey and pregnancy (at GDM diagnosis).

Genotyping

DNA was extracted from the buffy coat fraction of centrifuged blood using a QIAamp DNA Blood Maxi Kit (Qiagen, Chatsworth, CA). Single nucleotide polymorphism rs13266634 in the SLC30A8 gene was genotyped by quantitative real-time TaqMan PCR (Applied Biosystems, Foster City, CA). The genotyping success rate was over 98%. For quality control, 10% of the samples were genotyped, and the concordance rate was more than 99%.

Statistical Analysis

Statistical analyses were performed in SAS 9.4 (SAS Institute, Cary, NC). We applied χ2 test for categorical variables or general linear models for continuous variables to compare proportions or means of characteristics across categories of gestational weight gain (inadequate, adequate, and excessive). Changes in glycemic traits associated with each additional copy of the SLC30A8 rs13266634 C allele according to gestational weight gain categories were estimated using general linear models. To test for interactions, we examined the genotype, gestational weight gain (inadequate, adequate, and excessive), and genotype by gestational weight gain interactions as independent predictors of postpartum changes in glycemic traits, adjusted for age, follow-up time, prepregnancy BMI, total energy intake, zinc intake, sitting time, postpartum weight change, and the previous value for the respective glucose trait (continuous variables for above variables) and family history of diabetes, current smoking, current alcohol drinking, leisure time physical activity, and GDM therapy (categorical variables for above variables). We also calculated the multivariable-adjusted mean values of changes in glycemic traits according to gestational weight gain and rs13266634 genotype by using general linear models. We further performed two sensitivity analyses to examine the interaction effect: the first analysis was conducted in women within 1.5 years after delivery to control for potential bias dependent on follow-up period, and the second analysis was performed using gestational weight gain defined by the Chinese maternal prepregnancy BMI classification and the 2009 Institute of Medicine guideline. P values are two-sided and a P < 0.05 was considered statistically significant.

Results

Table 1 presents the characteristics of women with prior GDM according to categories of gestational weight gain. Compared with women with inadequate gestational weight gain, women with greater gestational weight gain were younger and had a higher predelivery BMI; higher weight at prepregnancy, predelivery, and postpartum; lower 2-h glucose levels at both GDM diagnosis and postpartum; and a longer sitting time (all P < 0.05). The frequency of SLC30A8 rs13266634 CC genotype was not different among the three categories of gestational weight gain.

Table 1

Characteristics of women with prior GDM by gestational weight gain

	Gestational weight gain			P
	Inadequate	Adequate	Excessive
N	150	372	549
Age (years)	33.0 ± 3.6	32.4 ± 3.5	31.9 ± 3.4	<0.001
Follow-up time (years)	2.0 ± 0.6	2.0 ± 0.6	2.1 ± 0.7	0.007
Prepregnancy BMI (kg/m2)	22.1 ± 2.9	22.4 ± 3.0	24.0 ± 3.4	<0.001
Weight (kg)
Prepregnancy	56.6 ± 8.2	57.1 ± 8.3	62.0 ± 9.4	<0.001
Predelivery	65.4 ± 7.4	70.6 ± 7.3	82.9 ± 9.3	<0.001
Postpartum	57.9 ± 8.9	58.7 ± 8.7	65.9 ± 11.4	<0.001
Fasting glucose (mmol/L)
At GDM diagnosis	5.3 ± 0.8	5.3 ± 0.8	5.4 ± 0.7	0.05
Postpartum	5.6 ± 1.3	5.3 ± 0.8	5.4 ± 0.9	0.06
2-h glucose (mmol/L)
At GDM diagnosis	9.3 ± 1.4	9.2 ± 1.3	9.0 ± 1.1	0.004
Postpartum	7.5 ± 3.1	7.0 ± 2.3	6.9 ± 2.2	0.04
HbA1c (%)
At GDM diagnosis	5.8 ± 0.6	5.8 ± 0.6	5.8 ± 0.6	0.37
Postpartum	5.7 ± 0.9	5.6 ± 0.6	5.6 ± 0.8	0.66
HbA1c (mmol/mol)
At GDM diagnosis	40 ± 4	40 ± 4	40 ± 4	0.37
Postpartum	39 ± 5	38 ± 4	38 ± 5	0.66
Family history of diabetes (%)	32.7	32.5	30.8	0.82
Current smoking (%)	4.0	0.8	2.4	0.05
Current alcohol drinking (%)	21.3	19.4	21.5	0.72
Leisure time physical activity (%)
0 min/day	76.7	80.7	80.5	0.27
1 to <30 min/day	23.3	17.7	17.7
≥30 min/day	0.0	1.6	1.8
Sitting time (h/day)	2.9 ± 2.0	3.1 ± 2.0	3.5 ± 2.3	0.002
Total energy intake (kcal/day)	1,712 ± 465	1,670 ± 424	1,679 ± 447	0.59
Dietary zinc intake (mg)	10.5 ± 3.4	10.1 ± 3.1	10.2 ± 3.3	0.66
rs13266634 CC genotype (%)	38.7	37.9	38.6	0.89

Data are mean ± SD for continuous variables or percentage for categorical variables. P values were calculated by general linear models for continuous variables or χ2 test for categorical variables.

Overall, there were no statistically significant associations between the rs13266634 genotype and postpartum glycemic changes (all P ≥ 0.48). We then tested the associations of each additional copy of the rs13266634 C allele with postpartum changes in glycemic traits according to gestational weight gain categories (Table 2). Gestational weight gain significantly interacted with the rs13266634 genotype on postpartum fasting glucose change. After full adjustment, changes in fasting glucose associated with each additional copy of the C allele were −0.18 (SE 0.15), −0.04 (0.05), and 0.12 mmol/L (0.05) in women with inadequate, adequate, and excessive gestational weight, respectively (P for interaction = 0.002). Similar interactions were observed for changes in 2-h glucose and HbA1c. Each additional copy of the C allele was associated with changes in 2-h glucose of −0.68 (0.31), −0.04 (0.16), and 0.13 mmol/L (0.12) across categories of inadequate, adequate, and excessive gestational weight gain, respectively (P for interaction = 0.003). Corresponding changes in HbA1c were −0.23 (0.11), −0.08 (0.05), and 0.05% (0.05) across these three categories of gestational weight gain, respectively (P for interaction = 0.005). We also performed similar analyses using gestational weight gain defined by the Chinese maternal prepregnancy BMI classification and the 2009 Institute of Medicine guideline (Supplementary Table 1), and the findings were not substantially changed.

Table 2

Changes in glycemic traits associated with each additional copy of the SLC30A8 rs13266634 C allele by gestational weight gain

	Inadequate		Adequate		Excessive		P for interaction
	β (SE)	P	β (SE)	P	β (SE)	P
Fasting glucose (mmol/L)
Age-adjusted	−0.25 (0.16)	0.12	−0.10 (0.07)	0.16	0.12 (0.06)	0.04	0.003
Multivariable-adjusted*	−0.18 (0.16)	0.25	−0.04 (0.05)	0.46	0.13 (0.05)	0.01	0.002
+ Zinc intake	−0.18 (0.15)	0.25	−0.04 (0.05)	0.44	0.12 (0.05)	0.01	0.002
2-h glucose (mmol/L)
Age-adjusted	−0.93 (0.33)	0.01	−0.05 (0.18)	0.79	0.11 (0.13)	0.41	0.005
Multivariable-adjusted*	−0.68 (0.31)	0.03	−0.03 (0.15)	0.85	0.13 (0.12)	0.26	0.003
+ Zinc intake	−0.68 (0.31)	0.03	−0.04 (0.16)	0.82	0.13 (0.12)	0.27	0.003
HbA1c (%)
Age-adjusted	−0.29 (0.12)	0.02	−0.10 (0.06)	0.11	0.08 (0.06)	0.16	0.002
Multivariable-adjusted*	−0.23 (0.11)	0.04	−0.08 (0.05)	0.08	0.06 (0.05)	0.21	0.004
+ Zinc intake	−0.23 (0.11)	0.05	−0.08 (0.05)	0.08	0.05 (0.05)	0.28	0.005

β-Coefficients (SE) represent changes in each glycemic trait per additional copy of the rs13266634 C allele.

*Results were adjusted for age, follow-up time, prepregnancy BMI, total energy intake, sitting time, postpartum weight change, and the previous value for the respective glucose trait (continuous variables for above variables) and family history of diabetes, current smoking, current alcohol drinking, leisure time physical activity, and GDM therapy (categorical variables for above variables).

Figure 1 shows multivariable-adjusted postpartum changes of fasting glucose, 2-h glucose and HbA1c according to joint categories of gestational weight gain and the rs13266634 genotype. In all participants, mean values of postpartum changes in fasting glucose, 2-h glucose, and HbA1c were 0.05 mmol/L, −2.13 mmol/L, and −0.18%, respectively. In category of inadequate gestational weight gain, women carrying the TT, CT, and CC genotypes showed change in fasting glucose of 0.49, 0.46, and 0.13 mmol/L, respectively; change in 2-h glucose of −1.11, −1.58, and −2.50 mmol/L, respectively; and change in HbA1c of 0.08, −0.01, and −0.31%, respectively. However, in category of excessive gestational weight gain, women carrying the TT, CT, and CC genotypes showed change in fasting glucose of −0.16, −0.05, and 0.09 mmol/L, respectively; change in 2-h glucose of −2.39, −2.05, and −2.04 mmol/L, respectively; and change in HbA1c of −0.25, −0.20, and −0.15%, respectively.

Figure 1

Changes in glycemic traits according to gestational weight gain and SLC30A8 rs13266634 genotype. Fasting glucose (A), 2-h glucose (B), and HbA1c (C). Data are mean (SE), adjusted for age, follow-up time, prepregnancy BMI, total energy intake, zinc intake, sitting time, postpartum weight change, and the previous value for the respective glucose trait (continuous variables for above variables) and family history of diabetes, current smoking, current alcohol drinking, leisure time physical activity, and GDM therapy (categorical variables for above variables).

We observed similar interactions between the rs13266634 genotype and gestational weight gain on changes in fasting glucose (P for interaction < 0.001), 2-h glucose (P for interaction < 0.001), and HbA1c (P for interaction = 0.059) in women with a postpartum follow-up <1.5 years (Fig. 2).

Figure 2

Changes in glycemic traits associated with each additional copy of the SLC30A8 rs13266634 C allele by gestational weight gain in women within 1.5-year postpartum. Plots (error bars) are β-coefficients (SE) for changes in each glycemic trait per additional copy of the C allele. There were 217 women with follow-up years less than 1.5 years, including 37 with inadequate gestational weight gain, 78 with adequate gestational weight gain, and 102 with excessive gestational weight gain. Fasting glucose (A), 2-h glucose (B), and HbA1c (C). Model 1: adjusted for age; model 2: adjusted for age, follow-up time, prepregnancy BMI, total energy intake, sitting time, postpartum weight change, and the previous value for the respective glucose trait (continuous variables for above variables) and family history of diabetes, current smoking, current alcohol drinking, leisure time physical activity, and GDM therapy (categorical variables for above variables); model 3: further adjusted for zinc intake based on model 2.

Discussion

In this cohort of Chinese women with prior GDM, we showed bidirectional interactions between the SLC30A8 rs13266634 genotype and gestational weight gain on postpartum glycemic changes. Fasting glucose, 2-h glucose, and HbA1c levels increased across the rs13266634 TT, CT, and CC genotypes in women with excessive gestational weight gain, whereas opposite genetic associations were found in women with inadequate or adequate gestational weight gain. These results suggest that in women with a history of GDM gestational weight gain could modify the effect of SLC30A8 rs13266634 on long-term glycemic changes and women carrying the rs13266634 C allele might particularly benefit by controlling gestational weight gain to avoid postpartum hyperglycemia.

A history of GDM is one of the strongest risk factors for subsequent type 2 diabetes. According to a large meta-analysis, women with GDM had a relative risk of 7.45 (95% CI 4.79–11.51) for type 2 diabetes compared with women who had a normoglycemic pregnancy (1). In addition, gestational weight gain has been suggested to considerably affect women’s long-term weight trajectory and predisposition to postpartum hyperglycemia and type 2 diabetes, especially in those with prior GDM (21,22). Emerging evidence has shown that in women with impaired glucose regulation or GDM, type 2 diabetes is preventable through effective lifestyle interventions on weight management (10,23,24). One feasibility randomized controlled trial particularly showed that lifestyle intervention was more effective among women who did not exceed the recommended gestational weight gain (10). In the current study, we found that the genetic effect of SLC30A8 variant on postpartum glycemic changes varied among women with different categories of gestational weight gain. Interestingly, the SLC30A8 rs13266634 C allele, which was associated with susceptibility to impaired glucose regulation in genome-wide association studies (7,8), was related to increased fasting glucose, 2-h glucose, and HbA1c levels among women with excessive gestational weight gain but also was related to decreased levels of these glycemic markers among women with inadequate or adequate gestational weight gain. When we restricted the analysis in women who had shorter follow-up years, the interaction patterns were not substantially changed. Moreover, the interaction was similar for gestational weight gain defined by the Chinese maternal prepregnancy BMI classification and the 2009 Institute of Medicine guideline, suggesting that the observed interaction was robust regardless of follow-up years and definitions of gestational weight gain.

The SLC30A8 gene encodes ZnT8, which is a zinc efflux transporter that is primarily expressed in β-cells and is essential for the homeostasis of zinc in β-cell granules. ZnT8 expression strongly modulates insulin secretion and storage and the formation of insulin crystals in β-cells (25,26). Intriguingly, it has been suggested that the rs13266634 genotype might interact with total zinc intake and plasma zinc levels in relation to fasting glucose levels and type 2 diabetes (27,28). Although SLC30A8 rs13266634 has also been related to GDM and body weight (9,25), no study has examined whether the association between this genetic variant and glycemia in women with GDM may be modified by weight change. Our findings are in line with the previous observations that the effect of type 2 diabetes–susceptible genes on hyperglycemia and type 2 diabetes was more prominent in obese individuals (29,30), suggesting that avoiding excessive gestational weight gain might protect women with GDM from a genetic susceptibility to hyperglycemia.

The current study provides a novel perspective to understand the association between gestational weight management and postpartum glycemia on a genetic background. The strength of this study is that the findings are based on longitudinal measures of weight and glycemic traits from one of the largest cohorts of women with a history of GDM. However, several limitations need to be acknowledged. First, although it has been reported that the SLC30A8 rs13266634 variant might interact with zinc intake on fasting glucose levels (27), we did not observe an apparent interaction between the genetic variant and zinc intake on postpartum glycemic changes in our study. The information of zinc intake was collected at postpartum survey in this study, which might not represent the exact intake during pregnancy, and the differences in health status (general vs. GDM), sex compositions (mixed vs. women), and ethnicities (Europeans vs. Chinese) among the studied populations may also partly explain the discrepant results. Second, we did not collect sufficient reproductive data, such as postpartum depression, which may influence weight change and glucose metabolism. Third, the study participants were restricted to Chinese women, and further studies in other demographic or ethnic populations are warranted to verify our findings.

In conclusion, we found that gestational weight gain significantly interacted with the SLC30A8 rs13266634 variant on postpartum glycemic changes in Chinese women with a history of GDM. Our findings suggest that among women with a history of GDM, the glucose-rising effect of the rs13266634 variant could be prevented by a better weight control during pregnancy, highlighting the importance of gestational weight management particularly in women genetically predisposed to impaired glucose regulation.