Metabolic Effects of Breastfeeding in Women with Previous Gestational Diabetes Mellitus: A Meta-Analysis

Abstract

Objective:

To investigate the effects of breastfeeding (BF) on metabolic-related outcomes in women with previous gestational diabetes mellitus.

Methods:

Databases, including PubMed, Embase, Cochrane Library, and Web of Science, were searched until March 5, 2020. Finally, 14 high-quality articles were included. Relative risk (RR) and weighted mean difference (WMD) with 95% confidence interval (CI) were pooled using Stata14.0 Software.

Results:

Subjects in the BF group had a lower incidence of diabetes (RR: 0.611, 95% CI: 0.452–0.826, p < 0.001) and lower fasting plasma glucose level (WMD: −4.762, 95% CI: −5.552 to −3.973, p < 0.001), fasting insulin level (WMD: −21.513, 95% CI: −37.594 to −5.431, p = 0.009), homeostasis model assessment of insulin resistance (HOMA-IR) (WMD: −1.107, 95% CI: −1.683 to −0.532, p < 0.001), and triglyceride level (WMD: −33.951, 95% CI: −50.714 to −17.189, p < 0.001) than those in the non-BF group. The high-density lipoprotein level (WMD: 3.855, 95% CI: 2.629–5.081, p < 0.001), low-density lipoprotein level (WMD: 4.223, 95% CI: 0.6712–7.774, p = 0.020), and insulin sensitivity index (WMD: 1.503, 95% CI: 0.857–2.160, p < 0.001) in the BF group were higher than that in the non-BF group. No difference was found in the 2-hour postprandial blood glucose (WMD: −3.804, 95% CI: −8.237 to 0.630, p = 0.093), incidence of prediabetes mellitus (RR: 0.870, 95% CI: 0.750–1.009, p = 0.065), or cholesterol level (WMD: 1.377, 95% CI: −8.178 to 10.931, p = 0.778) between the two groups.

Conclusion:

BF may improve several metabolic markers and decrease the risk of developing diabetes.

Introduction

Gestational diabetes mellitus (GDM), defined as any degree of glucose intolerance that presents for the first time during a pregnancy, is a frequent metabolic complication of pregnancy.¹ It is an important public health issue with an occurrence rate of 9.8–25.5%, which is increasing annually.^2,3 Although most GDM women return to normal glucose tolerance after delivery, this history brings long-term health risks^4,5 as accumulating studies reveal that the GDM is correlated with a higher prevalence of diabetes later in life.^6,7 A systematic review of 20 studies demonstrated that GDM women were 7 times more likely to develop type 2 diabetes mellitus (T2DM) than those without GDM.⁸ Therefore, women with a history of GDM are a critical target group for diabetes prevention.

Marketing messages regarding the positive effects of breastfeeding (BF) typically focus on the child's health,⁹ but numerous studies showed that mothers also benefit from BF.^10–12 BF, a modifiable postpartum behavior, is generally considered to improve glucose or lipid metabolism and enhance insulin sensitivity, and the advantageous metabolic effects still persist after weaning.^13–15 Some studies have found that BF for more than 4 weeks can reduce the risk of developing T2DM by 20–50% in women with previous GDM,^16,17 although others failed to find these positive effects.^18,19

Research on the effects of BF on metabolic-related outcomes in women with prior GDM has yet to be comprehensively reviewed. To address this, we collected related articles from databases for a more thorough exploration of its impact.

Materials and Methods

Search strategy

Literature was retrieved from PubMed, Embase, the Cochrane Library, and Web of Science up to March 5, 2020. The English search terms included “Breast Feeding [Mesh]” OR “Feeding, Breast” OR “Breastfeeding” OR “Breast Feeding, Exclusive” OR “Exclusive Breast Feeding” OR “Breastfeeding, Exclusive” OR “Exclusive Breastfeeding” OR “Lactation [Mesh]” OR “Milk Secretion” OR “Milk Secretions” OR “Lactation, Prolonged” OR “Lactations, Prolonged” OR “Prolonged Lactation” OR “Prolonged Lactations” AND “Diabetes, Gestational [Mesh]” OR “Diabetes, Pregnancy-Induced” OR “Diabetes, Pregnancy Induced” OR “Pregnancy-Induced Diabetes” OR “Gestational Diabetes” OR “Diabetes Mellitus, Gestational” OR “Gestational Diabetes Mellitus” OR “GDM” OR “Gestational Diabetic” OR “Diabetic Pregnancy.”

Selection criteria

Cohort studies were included in this meta-analysis. The criteria for inclusion were (1) subjects with GDM; (2) BF group: GDM women with BF; (3) non-BF group: GDM women without BF; (4) studies on the relationship between BF and diabetes and/or metabolic parameters; (5) English language; and (6) balanced and comparable baseline data.

Exclusion criteria included (1) inappropriate and uncorrectable statistical method; (2) duplicated reports; (3) literature with incomplete data, ambiguous outcome or unpublished treatment results; (4) reports, reviews, and meta-analyses; and (5) poor research design.

Diagnostic criteria

Patients who met the World Health Organization (WHO) or International Association of Diabetes and Pregnancy Study Group (IADPSG) criteria were diagnosed with GDM. WHO diagnostic criteria were as follows: fasting plasma glucose (FPG) ≥7.0 mmol/L or 2-hour postprandial blood glucose ≥7.8 mmol/L. IADPSG diagnostic criteria were FPG ≥5.1 mmol/L or 1-hour postprandial blood glucose ≥10.0 mmol/L, or postprandial 2-hour blood glucose ≥8.5 mmol/L. Diagnostic criteria for diabetes were as follows: FPG ≥7.0 mmol/L; random blood glucose ≥11.1 mmol/L with typical symptoms of diabetes (polydipsia, polyuria, polyphagia, and weight loss); and oral glucose tolerance test 2-hour blood glucose ≥11.1 mmol/L, and the diagnosis would be confirmed with one of three points. Prediabetes mellitus diagnostic criteria were as follows: impaired FPG (FPG: 6.1–7.0 mmol/L) and/or impaired glucose tolerance (2-hour postprandial blood glucose: 7.8–11.1 mmol/L).

Methodological quality appraisal and data extraction

The quality of included articles was assessed by Newcastle-Ottawa scale (NOS).²⁰ The total possible score was 10, with <5 being low quality and ≥5 considered high quality.

Extracted data included first author, publication year, country, number of included patients, duration of follow-up, age, quality assessment score, and outcomes. Outcomes included incidence of diabetes, FPG, 2-hour postprandial blood glucose, incidence of prediabetes mellitus, fasting insulin, homeostasis model assessment of insulin resistance (HOMA-IR), triglycerides, cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, and insulin sensitivity index (ISI).

Statistical analysis

This meta-analysis was performed using Stata 14.0 software (Stata Corporation, College Station, TX, USA). Relative risk (RR) and weighted mean difference (WMD) were calculated to assess the effects of BF on metabolic-related outcomes in women with previous GDM. Each effect size was expressed by 95% confidence interval (CI). The I² statistic, which is a quantitative measure of inconsistency across studies, was also calculated, and the random effect model was performed if the heterogeneity statistics I² ≥ 50%, otherwise the fixed effect model was used for analysis. Sensitivity analysis was performed to examine the influence of various exclusion criteria on the overall risk estimate. Subgroup analysis and meta-regression were used for FPG level, fasting insulin level, HOMA-IR, and triglycerides. Publication bias was assessed by using the Begg's test. p < 0.05 meant that the difference was statistically significant.

Results

Literature selection and quality evaluation

From 2,299 citations searched from the database, 14 articles were ultimately accepted for meta-analysis according to the search strategy shown in Figure 1. They included a total of 4,020 GDM subjects with 2,540 in the BF group and 1,480 in the non-BF group. The characteristics and quality rating of the 14 studies are listed in Table 1.

FIG. 1.

The flow diagram of the meta-analysis.

Table 1.

Characteristics of Included Articles

Author	Year	Country	Total (BF/non-BF)	Follow-up	BF; age, years	Non-BF; age, years	Score	Outcomes
Kjos²²	1993	USA	809 (404/405)	4–12 weeks	31.6 ± 5.9	30.5 ± 6.3	7	ABCGHIJ
Buchanan²³	1998	USA	122 (70/52)	1–6 months	30.8 ± 5.2		6	A
McManus²⁴	2001	Canada	26 (14/12)	3 months	30.9 ± 1.3	30.7 ± 1.3	7	ABEGHK
Chouinard²⁵	2013	Canada	144 (116/28)	4.0 ± 1.9 years	36.9 ± 4.9	34.8 ± 5.2	6	BCE
Gunderson¹⁵	2014	USA	696 (437/259)	6–9 weeks	33.4 ± 4.7	32.7 ± 5.0	6	BCEFGHIJ
Mattei²⁶	2014	Italy	81 (62/19)	36 months	37.1 ± 4.6	37.2 ± 5.1	6	ABCFGHIJK
Capula²⁷	2014	Italy	454 (212/242)	6–12 weeks	35.0 ± 6.8		6	AD
Bentley²⁸	2014	USA	38 (35/3)	12–30 weeks			7	D
Benhalima²⁹	2014	Belgium	169 (128/41)	8–13 weeks			5	D
Gunderson³⁰	2015	USA	358 (205/153)	6–9 weeks			6	BCF
Yasuhi³¹	2017	Japan	87 (69/18)	12–14 months	33.3 ± 4.9		8	BEF
Corrado³²	2017	Italy	97 (81/16)	12–16 weeks	33.4 ± 4.6	31.9 ± 5.7	8	BCDFGH
Shen³³	2018	China	717 (541/176)	3.65 years			7	AD
Yasuhi³⁴	2019	Japan	222 (166/56)	6–12 weeks	34.0 ± 4.7	34.6 ± 5.1	8	BCDEF

BF, gestational diabetes mellitus women with breastfeeding; non-BF, gestational diabetes mellitus women without breastfeeding; BMI, body mass index; A: incidence of diabetes; B: fasting plasma glucose; C: 2-hour postprandial blood glucose; D: incidence of prediabetes mellitus; E: fasting insulin; F: homeostasis model assessment of insulin resistance; G: triglycerides; H: cholesterol; I: high-density lipoprotein cholesterol; J: low-density lipoprotein cholesterol; K: insulin sensitivity index.

Incidence of diabetes

The incidence of diabetes was reported in six studies.^{21–23,25,26,32} Results of pooled analysis indicate that the subjects in the BF group had a lower overall incidence of diabetes compared to those in the non-BF group (RR: 0.611, 95% CI: 0.452–0.826, p < 0.001; I² = 54.4%) (Table 2 and Fig. 2a). Subgroup analysis showed a significant difference of fasting insulin level between the group that was followed for ≤6 months and the group that was followed for >6 months, both p < 0.05 (Fig. 3a). Meta-regression analysis indicated that the follow-up time was not the source of heterogeneity, p > 0.05.

FIG. 2.

Forest plots for meta-analysis of the effects of BF on metabolic-related outcomes in women with GDM: (a) incidence of diabetes; (b) FPG level; (c) 2-hour postprandial blood glucose; (d) fasting insulin level; (e) HOMA-IR; (f) triglyceride level; (g) HDL level; (h) LDL level; (i) ISI; (j) prediabetes mellitus; (k) cholesterol level. BF, breastfeeding; FPG, fasting plasma glucose; GDM, gestational diabetes mellitus; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; ISI, insulin sensitivity index; LDL, low-density lipoprotein.

FIG. 3.

Forest plots for subgroup analysis of the effects of BF on metabolic-related outcomes in women with GDM: (a) incidence of diabetes; (b) fasting insulin level; (c) HOMA-IR; (d) triglyceride level.

Table 2.

Overall Meta-Analysis of Included Studies

Indicators	RR/WMD (95% CI)	p	I²
Incidence of diabetes
Overall	0.611 (0.452 to 0.826)	<0.001	54.4
Sensitivity analysis	0.611 (0.452 to 0.826)
Follow-up time
≤6 months	0.638 (0.424 to 0.960)	0.031	66.5
>6 months	0.574 (0.368 to 0.896)	0.014	45.9
Prediabetes mellitus
Overall	0.870 (0.750 to 1.009)	0.065	18.0
Sensitivity analysis	0.870 (0.750 to 1.009)
FPG level
Overall	−4.762 (−5.552 to −3.973)	<0.001	1.5
Sensitivity analysis	−4.762 (−5.552 to −3.973)
2-Hour postprandial blood glucose
Overall	−3.804 (−8.237 to 0.630)	0.093	56.8
Sensitivity analysis	−3.804 (−8.237 to 0.630)
Fasting insulin level
Overall	−21.513 (−37.594 to −5.431)	0.009	90.4
Sensitivity analysis	−21.513 (−37.594 to −5.431)
Follow-up time
≤6 months	−22.576 (−50.340 to 5.187)	0.111	95.0
>6 months	−19.621 (−27.755 to −11.488)	<0.001	0.0
HOMA-IR
Overall	−1.107 (−1.683 to −0.532)	<0.001	89.4
Sensitivity analysis	−1.107 (−1.683 to −0.532)
Follow-up time
≤6 months	−1.338 (−2.225 to −0.452)	0.003	93.8
>6 months	−0.743 (−1.067 to −0.419)	<0.001	0.0
Triglyceride level
Overall	−33.951 (−50.714 to −17.189)	<0.001	69.1
Sensitivity analysis	−33.951 (−50.714 to −17.189)
Follow-up time
≤6 months	−33.398 (−52.182 to −14.614)	<0.001	76.6
>6 months	−39.400 (−83.236 to −4.436)	0.078	NA
Cholesterol level
Overall	1.377 (−8.178 to 10.931)	0.778	81.3
Sensitivity analysis	1.377 (−8.178 to 10.931)
HDL cholesterol level
Overall	3.855 (2.629 to 5.081)	<0.001	0.0
Sensitivity analysis	3.855 (2.629 to 5.081)
LDL cholesterol level
Overall	4.223 (0.6712 to 7.774)	0.020	0.0
Sensitivity analysis	4.223 (0.6712 to 7.774)
Insulin sensitivity index
Overall	1.503 (0.857 to 2.160)	<0.001	0.0
Sensitivity analysis	1.503 (0.857 to 2.160)

CI, confidence interval; FPG, fasting plasma glucose; HOMA-IR, homeostasis model assessment of insulin resistance; HDL, high-density lipoprotein; LDL, low-density lipoprotein; RR, relative risk; WMD, weighted mean difference.

FPG level

FPG level was reported in nine studies^{14,21,23–25,29–31,33}; pooled results showed lower FPG levels in the BF group compared to the non-BF group (WMD: −4.762, 95% CI: −5.552 to −3.973, p < 0.001) with low heterogeneity (I² = 1.5%) (Table 2 and Fig. 2b).

Two-hour postprandial blood glucose

A total of seven articles evaluated the 2-hour postprandial blood glucose.^{14,21,24,25,29,31,33} No statistical difference was found in the 2-hour postprandial blood glucose between the BF and non-BF groups (WMD: −3.804, 95% CI: −8.237 to 0.630, p = 0.093; I² = 56.8%). Meta-regression analysis indicated that the follow-up time was not the source of heterogeneity, p > 0.05 (Table 2 and Fig. 2c).

Fasting insulin level

Five articles reported the fasting insulin levels^{14,23,24,30,33}; pooled analysis revealed lower fasting insulin levels in the BF group than in the non-BF group (WMD: −21.513, 95% CI: −37.594 to −5.431, p = 0.009; I² = 90.4%) (Table 2 and Fig. 2d). Subgroup analysis showed a significant difference in the fasting insulin level between the two groups for follow-up time >6 months, p < 0.001 (Fig. 3b). Meta-regression analysis indicated that the follow-up time was not the source of heterogeneity, p > 0.05.

Homeostasis model assessment of insulin resistance

Pooled analysis of six articles^{14,25,29–31,33} indicated that the BF group had lower HOMA-IR than the non-BF group (WMD: −1.107, 95% CI: −1.683 to −0.532, p < 0.001; I² = 89.4%) (Table 2 and Fig. 2e). Subgroup analysis was performed according to follow-up times. The results showed that BF women had significant lower HOMA-IR than non-BF women for the follow-up time >6 and ≤6 months, p < 0.05 (Fig. 3c). Meta-regression indicated that the follow-up time was not the source of heterogeneity, p > 0.05.

Triglyceride level

Five articles studied triglyceride levels^{14,21,23,25,31}; in pooled results, women in the BF group had lower triglyceride levels than those in the non-BF group (WMD: −33.951, 95% CI: −50.714 to −17.189, p < 0.001; I² = 69.1%) (Table 2 and Fig. 2f). In the subgroup analysis, BF women had lower triglyceride levels than non-BF women for the follow-up time ≤6 months, p < 0.001 (Fig. 3d). No relationship between the follow-up time and heterogeneity was found in meta-regression, p > 0.05.

HDL and LDL cholesterol level

Three articles looked specifically at HDL and LDLC cholesterol levels.^14,21,25 Combined analysis showed higher HDL cholesterol (WMD: 3.855, 95% CI: 2.629–5.081, p < 0.001; I² = 0.0%) and LDL cholesterol level (WMD: 4.223, 95% CI: 0.6712–7.774, p = 0.020; I² = 0.0%) in the BF group compared to the non-BF group (Table 2 and Fig. 2g, h).

Insulin sensitivity index

ISI was reported in only two articles,^23,25 both of which found a higher ISI for the BF group over the non-BF group (WMD: 1.503, 95% CI: 0.857–2.160, p < 0.001; I² = 0.0%) (Table 2 and Fig. 2i).

Prediabetes mellitus

Prediabetes mellitus was tracked in six articles.^{26–28,31–33} No difference was found in the incidence of prediabetes mellitus between the BF and non-BF groups (RR: 0.870, 95% CI: 0.750–1.009, p = 0.065; I² = 18.0%) (Table 2 and Fig. 2j).

Cholesterol level

A total of 5 articles evaluated cholesterol levels.^{14,21,23,25,31} No difference was found between the two groups in our analysis (WMD: 1.377, 95% CI: −8.178 to 10.931, p = 0.778; I² = 81.3%) (Table 2 and Fig. 2k).

Discussion

A meta-analysis conducted by Ma et al. confirmed the positive effects of BF in protection against the development of T2DM-related outcomes in midlife for women with a prior history of GDM.³⁴ In this review, the outcomes HDL and LDL cholesterol levels were specifically chosen to provide a more comprehensive analysis of the global effects of BF on metabolic-related outcomes in women with previous GDM. Results showed a decreased incidence of diabetes, FPG level, fasting insulin level, HOMA-IR, and triglyceride levels in women with previous GDM in the BF group compared with those in the non-BF group, and BF was significantly related to higher HDL cholesterol levels, LDL cholesterol levels, and the ISI. No difference was found in the incidence of 2-hour postprandial blood glucose, prediabetes mellitu or cholesterol levels between the two groups.

It is reported that BF may offer a practical, low-cost intervention that helps reduce or delay the risk of subsequent diabetes in women with prior GDM by Kjos in 1993.²¹ Ziegler et al. found a relationship between BF and a decreased diabetes risk among women with previous GDM after delivery.¹⁶ Although no difference was found in the incidence of 2-hour postprandial blood glucose and prediabetes mellitus between two groups in our review, the incidence of postpartum diabetes decreased in the BF group.

The ISI is closely related to diabetes. In T2DM women with sufficient insulin secretion, improving insulin sensitivity is the preferred treatment strategy. In this review, a higher ISI was observed in the BF group. Increased insulin sensitivity helps control blood sugar levels, resulting in a lower likelihood of developing diabetes. Butte et al. found a lower fasting insulin level in BF women compared to non-BF women at 6 months after delivery, while no difference in the fasting glucose level was found between two groups.³⁵ Others observed a trend for increased fasting insulin levels and HOMA-IR in non-BF women compared with BF mothers.³⁶ All the above findings are in line with this review, which confirmed that BF can reduce the incidence of diabetes, FPG level, fasting insulin level, and HOMA-IR in women with previous GDM.

An alternate hypothesis of an insulin-independent mechanism has also been proposed to explain the beneficial metabolic effects of BF in women with previous GDM.³⁵ The glucose values in plasma decrease due to the preferential diversion of glucose to the mammary glands during lactation. This occurs in all BF women and plays an especially critical role for those with a high risk of metabolic disorders.

BF may also positively influence lipoprotein profiles. Kjos et al. found an increased level of HDL cholesterol in the lactating group, while no difference in the triglyceride, HDL cholesterol, and total cholesterol levels was reported.²¹ A higher HDL cholesterol level in BF women also has been observed by Gunderson et al.³⁶ In this meta-analysis, lactation was significantly related to higher HDL and LDL cholesterol levels, and we also found lower triglyceride levels in the BF group, while no difference was found in total cholesterol.

Strengths of this meta-analysis include the following: (1) the outcomes of HDL and LDL were included to provide more details than previous meta-analyses³⁴ and (2) all enrolled articles were of high quality, strengthening the validity of our results. Several potential limitations of our study should be considered. The main limitation of our review is that it only includes 4,020 subjects, which may reduce the statistical power. More well-designed and larger sample size studies are needed to verify our results.

Conclusion

This study investigated the effects of BF on metabolic-related outcomes in women with previous GDM. Results indicate that BF may reduce the development of diabetes-related outcomes in at-risk women with previous GDM. These findings should be publicized and promoted to both mothers and their health care providers.

Ethical Approval

This article does not contain any study with human participants or animals performed by any of the authors.

Footnotes

Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by the 2018 Shenzhen Science and Technology Commission International Cooperation Project [Grant No. GJHZ20180410144638700].

References

Zhu

, Zhang

. Prevalence of gestational diabetes and risk of progression to type 2 diabetes: A global perspective. Curr Diab Rep, 2016; 16:7.

Langer

, Umans

, Miodovnik

. The proposed GDM diagnostic criteria: A difference, to be a difference, must make a difference. J Matern Fetal Neonatal Med, 2013; 26:111–115.

Olsen

, Houshmand-Oeregaard

, Granström

, et al. Diagnosing gestational diabetes mellitus in the Danish National Birth Cohort. Acta Obstet Gynecol Scand, 2017; 96:563–569.

Much

, Beyerlein

, Roßbauer

, et al. Beneficial effects of breastfeeding in women with gestational diabetes mellitus. Mol Metab, 2014; 3:284–292.

Manerkar

, Harding

, Conlon

, et al. Maternal gestational diabetes and infant feeding, nutrition and growth: A systematic review and meta-analysis. Br J Nutr, 2020; 123:1201–1215.

Bernstein

, Quinn

, Ameli

, et al. Onset of T2DM after gestational diabetes: What the prevention paradox tells us about risk. Prev Med, 2018; 113:1–6.

Zhang

, Olsen

, Hinkle

, et al. Diabetes & Women's Health (DWH) Study: An observational study of long-term health consequences of gestational diabetes, their determinants and underlying mechanisms in the USA and Denmark. BMJ Open, 2019; 9:e025517.

Bellamy

, Casas

, Hingorani

, et al. Type 2 diabetes mellitus after gestational diabetes: A systematic review and meta-analysis. Lancet, 2009; 373:1773–1779.

Kramer

, Kakuma

. Optimal duration of exclusive breastfeeding. Cochrane Database Syst Rev, 2012; 15:CD003517.

10.

Stuebe

, Schwarz

. The risks and benefits of infant feeding practices for women and their children. J Perinatol, 2010; 30:155–162.

11.

, Chung

, Raman

, et al. Breastfeeding and maternal and infant health outcomes in developed countries. Evid Rep Technol Assess (Full Rep), 2007; 153:1–186.

12.

James

, Lessen

, American Dietetic

Association

. Position of the American Dietetic Association: Promoting and supporting breastfeeding. J Am Diet Assoc, 2009; 109:1926–1942.

13.

Gunderson

, Hedderson

, Chiang

, et al. Lactation intensity and postpartum maternal glucose tolerance and insulin resistance in women with recent GDM: The SWIFT cohort. Diabetes Care, 2012; 35:50–56.

14.

Gunderson

, Kim

, Quesenberry

Jr , et al. Lactation intensity and fasting plasma lipids, lipoproteins, non-esterified free fatty acids, leptin and adiponectin in postpartum women with recent gestational diabetes mellitus: The SWIFT cohort. Metabolism, 2014; 63:941–950.

15.

Gunderson

, Jacobs

Jr , Chiang

, et al. Duration of lactation and incidence of the metabolic syndrome in women of reproductive age according to gestational diabetes mellitus status: A 20-Year prospective study in CARDIA (Coronary Artery Risk Development in Young Adults). Diabetes, 2010; 59:495–504.

16.

Ziegler

, Wallner

, Kaiser

, et al. Long-term protective effect of lactation on the development of type 2 diabetes in women with recent gestational diabetes mellitus. Diabetes, 2012; 61:3167–3171.

17.

Martens

, Shafer

, Dean

, et al. Breastfeeding initiation associated with reduced incidence of diabetes in mothers and offspring. Obstet Gynecol, 2016; 128:1095–1104.

18.

Kjos

, Peters

, Xiang

, et al. Contraception and the risk of type 2 diabetes mellitus in Latina women with prior gestational diabetes mellitus. JAMA, 1998; 280:533–538.

19.

Buchanan

, Xiang

, Kjos

, et al. Antepartum predictors of the development of type 2 diabetes in Latino women 11–26 months after pregnancies complicated by gestational diabetes. Diabetes, 1999; 48:2430–2436.

20.

Wells

, Shea

, O'Connell

, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality if nonrandomized studies in meta-analyses. 2012; Available at http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp (accessed April 10, 2020).

21.

Kjos

, Henry

, Lee

, et al. The effect of lactation on glucose and lipid metabolism in women with recent gestational diabetes. Obstet Gynecol, 1993; 82:451–455.

22.

Buchanan

, Xiang

, Kjos

, et al. Gestational diabetes: Antepartum characteristics that predict postpartum glucose intolerance and type 2 diabetes in Latino women. Diabetes, 1998; 47:1302–1310.

23.

McManus

, Cunningham

, Watson

, et al. Beta-cell function and visceral fat in lactating women with a history of gestational diabetes. Metabolism, 2001; 50:715–719.

24.

Chouinard-Castonguay

, Weisnagel

, Tchernof

, et al. Relationship between lactation duration and insulin and glucose response among women with prior gestational diabetes. Eur J Endocrinol, 2013; 168:515–523.

25.

Mattei

, Colatrella

, Bitterman

, et al. Long lasting effects of breastfeeding on metabolism in women with prior gestational diabetes. J Diab Mellitus, 2014; 4:257–263.

26.

Capula

, Chiefari

, Vero

, et al. Prevalence and predictors of postpartum glucose intolerance in Italian women with gestational diabetes mellitus. Diabetes Res Clin Pract, 2014; 105:223–230.

27.

Bentley-Lewis

, Xiong

, Lee

, et al. Metabolomic analysis reveals amino acid responses to an oral glucose tolerance test in women with prior history of gestational diabetes mellitus. J Clin Transl Endocrinol, 2014; 1:38–43.

28.

Benhalima

, Leuridan

, Calewaert

, et al. Glucose intolerance after a recent history of gestational diabetes. Int J Endocrinol, 2014; 2014:727652.

29.

Gunderson

, Hurston

, Ning

, et al. Lactation and progression to type 2 diabetes mellitus after gestational diabetes mellitus: A prospective cohort study. Ann Intern Med, 2015; 163:889–898.

30.

Yasuhi

, Soda

, Yamashita

, et al. The effect of high-intensity breastfeeding on postpartum glucose tolerance in women with recent gestational diabetes. Int Breastfeed J, 2017; 12:32.

31.

Corrado

, Giunta

, Granese

, et al. Metabolic effects of breastfeeding in women with previous gestational diabetes diagnosed according to the IADPSG criteria. J Matern Fetal Neonatal Med, 2019; 32:225–228.

32.

Shen

, Leng

, Li

, et al. Lactation intensity and duration to postpartum diabetes and prediabetes risk in women with gestational diabetes. Diabetes Metab Res Rev, 2019; 35:e3115.

33.

Yasuhi

, Yamashita

, Maeda

, et al. High-intensity breastfeeding improves insulin sensitivity during early post-partum period in obese women with gestational diabetes. Diabetes Metab Res Rev, 2019; 35:e3127.

34.

, Hu

, Liang

, et al. Metabolic effects of breastfeed in women with prior gestational diabetes mellitus: A systematic review and meta-analysis. Diabetes Metab Res Rev, 2019; 35:e3108.

35.

Butte

, Hopkinson

, Mehta

, et al. Adjustments in energy expenditure and substrate utilization during late pregnancy and lactation. Am J Clin Nutr, 1999; 69:299–307.

36.

Gunderson

, Lewis

, Wei

, et al. Lactation and changes in maternal metabolic risk factors. Obstet Gynecol, 2007; 109:729–738.