Gunderson, Erica P. PhD; Crites, Yvonne MD; Chiang, Vicky MS; Walton, David MD; Azevedo, Robert A. MD; Fox, Gary MD; Elmasian, Cathie MD; Young, Stephen MD; Salvador, Nora MD; Lum, Michael MD; Hedderson, Monique M. PhD; Quesenberry, Charles P. PhD; Lo, Joan C. MD; Ferrara, Assiamira MD, PhD; Sternfeld, Barbara PhD
The fifth International Workshop–Conference on Gestational Diabetes recommended screening of women diagnosed with gestational diabetes mellitus (GDM) for glucose intolerance between 6 and 12 weeks postpartum,1 although 70% of women are still breastfeeding their infants.2 Lactating compared with nonlactating women have lower blood glucose concentrations, both fasting and 2-hour post-Glucola.3,4 Protocols advise no smoking, lying down, eating, drinking, or physical activity during the oral glucose tolerance test (OGTT), but there is no restriction regarding breastfeeding an infant during the test. Furthermore, previous studies assessing postpartum screening for diabetes have not queried postpartum women about their breastfeeding practices during the OGTT.
Consequently, it remains unclear whether breastfeeding during the OGTT, a routine clinical test for assessment of postpartum glucose tolerance, affects maternal metabolic parameters and test results. This study's objective is to examine the effect of breastfeeding (ie, putting the infant to the breast) during the 2-hour 75-g OGTT at 6–9 weeks postpartum on plasma glucose and insulin concentrations among lactating women enrolled in the Study of Women, Infant Feeding, and Type 2 Diabetes, a Kaiser Permanente Northern California postpartum GDM cohort.
MATERIALS AND METHODS
Study participants were recruited and enrolled between September 2008 and December 2011 into the Study of Women, Infant Feeding, and type 2 Diabetes, an ongoing prospective observational cohort study of 1,035 Kaiser Permanente Northern California members diagnosed with GDM through the 3-hour 100-g OGTT at 24–32 weeks of gestation,5 delivered of a singleton, live birth 35 weeks of gestation or more at a Kaiser Permanente Northern California hospital, and without any known major medical conditions. For this analysis, we included a subset of Study of Women, Infant Feeding, and type 2 Diabetes participants who reported any level of lactation intensity at 6–9 weeks postpartum (ie, excludes exclusive formula feeding) and who completed the 2-hour 75-g OGTT for the study baseline examination. The Kaiser Permanente institutional review board approved the study protocol. Written, informed consent was obtained from participants for all study procedures.
Research assistants interviewed women about infant-feeding intentions and practices from late pregnancy through 9 weeks postpartum by conducting two telephone interviews, including participants' infant-feeding diaries mailed for recording formula supplementation and breastfeeding (duration and frequency per 24 hours and amount of formula) and one in-person study interview at 6–9 weeks postpartum. At 3–5 weeks postpartum, potentially eligible women were classified as: 1) intensively lactating (6 oz formula or less per 24 hours); or 2) intensively formula feeding (14 oz formula or more per 24 hours). Women who stated the intention to maintain the infant-feeding practices for at least 4 months postpartum were scheduled for an in-person study examination and OGTT at 6–9 weeks postpartum. Research assistants mailed instructions to women on how to prepare for the 2-hour 75-g OGTT (eg, consume adequate carbohydrate for 3 days, fast for at least 10 hours the night before) and were advised to express their breast milk a few days before the test so they could feed expressed breast milk to their infants during the OGTT.
At the baseline study examination (6–9 weeks postpartum), women provided written, informed consent before the collection of the blood specimens, administration of the 2-hour 75-g OGTT, completion of medical and lifestyle questionnaires, and obtaining the anthropometric measurements as previously described.6 During the OGTT, a research assistant observed and reported any protocol modifications and also observed the number of episodes and duration (minutes) that the woman put the infant to her breast (ie, breastfeeding) during the OGTT to calculate total duration (minutes). We categorized women as having “breastfed” or “did not breastfeed” their infants during the 2-hour 75-g OGTT.
Research staff queried women about their frequency of breastfeeding, number of expressed breast milk feedings by bottle, and formula supplementation (quantity) per 24 hours during the past 7 days. Data collected from birth through 6–9 weeks postpartum were used to classify participants into mutually exclusive groups: 1) exclusive lactation, 0 oz of formula or other milk feeds; 2) mostly lactation, 6 oz of formula or less per 24 hours; 3) mixed (breast milk and formula, more than 6 oz per 24 hours) or inconsistent feeding method; 4) mostly formula feeding, more than 17 oz per 24 hours; and 5) exclusive formula feeding.7 We excluded women exclusively formula feeding from the analytic sample used to assess the influence of breastfeeding during the OGTT on maternal plasma glucose and insulin.
Trained research assistants obtained measurements of weight and height through standardized methods and equipment. Interviewer and self-administered questionnaires collected information on sociodemographics, medical conditions, contraception, and lifestyle behaviors including physical activity. Kaiser Permanente Northern California electronic medical record databases were used to retrieve prenatal laboratory results for GDM diagnosis, dates of delivery, and length of gestation.
Plasma glucose was analyzed enzymatically with the Hitachi 917 Autoanalyzer using the combined catalytic activities of hexokinase and glucose-6-phosphate-dehydrogenase by the University of Washington, Northwest Lipid Metabolism and Diabetes Research Laboratory. The Diabetes Endocrinology Research Center Immunoassay Core Laboratory performed the assay of total immunoreactive insulin (microunits/mL) using a double-antibody radioimmunoassay with high precision. The assay is a 48-hour polyethylene glycol-accelerated assay involving a primary antibody, guinea pig antihuman insulin, and a secondary antibody, goat antiguinea pig immunoglobulin.
We calculated indices of insulin resistance (homeostasis model assessment of insulin resistance, HOMA-IR) and insulin sensitivity (insulin sensitivity index, ISI0,120) using measures obtained from 0 and 120 minutes during the OGTT.8
Equation (Uncited)Image Tools
where G0=fasting glucose, I0=fasting insulin, G120=glucose post-2-hour OGTT, and I120=insulin post-2-hour OGTT;
Equation (Uncited)Image Tools
where m=[75,000 mg+(G0−I120)×0.19×body weight]/120 min; and
Equation (Uncited)Image Tools
MPG is mean of fasting and 2-h glucose concentrations (mg/dL) from the OGTT, and MSI is mean of fasting and 2-h plasma insulin concentrations (mU/L) from the OGTT.
Women were classified by glucose tolerance as follows: normal, glucose-intolerant (eg, prediabetes defined as impaired fasting glucose between 100 and 125 mg/dL, impaired glucose tolerance for 2-hour 75-g post-Glucola load between 140 and 199 mg/d, or both), or diabetes based on the American Diabetes Association diagnostic criteria for the 75-g OGTT (fasting 126 mg/dL or greater, 2-hour 200 mg/dL or greater, or both), which included a repeat test on a separate occasion for women with elevated values.9
Differences in characteristics between the groups (breastfeeding compared with not breastfeeding during the OGTT) were assessed using χ2 statistics categorical variables (race, education, glucose tolerance groups) and by comparison of means for continuous variables (blood glucose, insulin, homeostatic model assessment of insulin resistance, insulin sensitivity index0,120) using t-statistics. Insulin concentrations for the 2-hour 75-g OGTT were transformed by taking the natural log as a result of skewness of the distribution (5% of values greater than 199 microunits/mL). All P values are for two-sided tests with statistical significance at P<.05.
Unadjusted and adjusted means (95% confidence interval) and group differences in mean blood glucose and insulin concentrations and insulin resistance or sensitivity indices were estimated from multivariable linear regression models using SAS for Windows 9.1.3. Covariates were evaluated as potential confounders based on a priori hypotheses and based on 10% change in the breastfeeding during OGTT regression coefficient.10 Correlates of breastfeeding, duration of fasting, number of feedings per day, and amount of formula per day were examined as mediators. Age, race and ethnicity, and parity were assessed as effect modifiers with significance at P<.10. Pearson correlation coefficients were calculated to assess associations between duration of breastfeeding during the OGTT (minutes) and continuous biochemical measures (glucose, insulin, homeostatic model assessment of insulin resistance, and insulin sensitivity index0,120).
Participants' ages ranged from 21 to 45 years with a mean (standard deviation) age of 33.5 (4.8) years (median 33.5 years). Overall, 74% were nonwhite: 38% Asian, 6% black, and 30% Hispanic. Of the 835 participants, 205 (25%) had breastfed their infants during the OGTT. Of these 205, 149 (73%) were classified as exclusively lactating. Women who had breastfed during the OGTT were more likely to be of white race, multiparous, and slightly older, have fewer abnormal values from the 100-g 3-hour OGTT during pregnancy, a longer fasting period, breastfed during the fasting period, and fed less formula (Table 1). Mean (standard deviation) duration of breastfeeding during the OGTT was 15.3 (8.1) minutes (range, 2–35 minutes).
Compared to those who did not breastfeed during the OGTT, having breastfed during the OGTT was associated with lower maternal 2-hour post-glucola plasma glucose and insulin concentrations in unadjusted and fully adjusted models (Table 2). The fully adjusted mean (95% confidence interval) group difference for 2-hour plasma glucose (mg/dL) was −5.8 (−10.9 to −0.7; P=.03) and was strengthened to −6.2 (−11.5 to −1.0; P=.02) after controlling for covariates associated with having breastfed during the OGTT such as the intensity of lactation (ie, amount of formula fed per day), number of feedings per day, and duration of the fasting period. The unadjusted and adjusted mean (95% confidence interval) for 2-hour insulin concentrations (microunits/mL) were lower for those who breastfed compared with those who did not breastfeed during the OGTT, but the difference was attenuated slightly after adjustment for severity of GDM (number of elevated values for the 3-hour prenatal OGTT; −21.6 [−33.1 to −10.1]; P<.001; Table 2). When covariates that influenced whether women breastfed during the OGTT such as the intensity of lactation (ie, amount of formula fed per day) and duration of the fasting period were added to the models, the group difference was attenuated to −15.1 (−26.8 to −3.5; P=.01) but remained significant (model 3). Findings were stronger, but similar, for natural log-transformed plasma 2-hour insulin concentrations. The addition of the frequency of feedings to the models did not alter the results. Our sensitivity analysis, in which we excluded 80 mostly formula feeding women, yielded similar findings and P values.
The two groups did not differ in plasma fasting glucose, fasting insulin, or insulin resistance assessed by homeostasis model assessment of insulin resistance. The insulin sensitivity index0,120 was significantly higher among women who breastfed during the test by 0.14 (0.07–0.21; P<.001) but was attenuated to 0.08 (0.02–0.15; P=.02) after controlling for the severity of GDM and other covariates.
Fasting and 2-hour plasma glucose and insulin concentrations (Fig. 1) were uncorrelated with minutes of breastfeeding during the OGTT (linear correlations ranging from 0.06 to 0.11; all P≥.09). Insulin sensitivity index0,120 was weakly correlated with breastfeeding duration (r=−0.17; P=.02).
Published studies of lactation after GDM pregnancy have not reported whether women breastfed during the postpartum OGTT,3,11–15 except for our previous study in the Study of Women, Infant Feeding, and type 2 Diabetes cohort.7 Because lactation is associated with lower maternal blood glucose concentrations, the present study sought to examine whether breastfeeding an infant during the 2-hour 75-g OGTT altered the maternal blood glucose and insulin test results. We found no differences in mean fasting glucose or insulin, but women who breastfed during the OGTT had lower mean plasma glucose and insulin levels at 2-hour post-glucola that persisted after controlling for multiple confounders, including parity, age, race and ethnicity, GDM severity, the amount of formula per 24 hours, frequency of feedings per day, and duration of fasting. Insulin sensitivity index during the OGTT was higher in women who had breastfed during the test compared with those who did not breastfeed during the test as expected from the noninsulin-mediated uptake of glucose by the mammary gland.11,12
In the Study of Women, Infant Feeding, and type 2 Diabetes cohort, we previously reported a dose–response relationship between increasing intensity of lactation (exclusive lactation compared with mixed or inconsistent and exclusive or mostly formula feeding) and decreasing fasting plasma glucose (4–5 mg/dL) as well as lower fasting and 2-hour insulin at 6–9 weeks postpartum.7 However, we did not find significant differences in 2-hour glucose concentrations by infant-feeding methods, possibly because intensively formula feeding women achieved sufficient insulin secretion to control blood glucose levels. Thus, women with recent GDM have the most favorable blood glucose and insulin concentrations with highest levels of lactation intensity (ie, exclusive lactation without formula supplementation).7 These findings persisted in the Study of Women, Infant Feeding, and type 2 Diabetes cohort after controlling for several covariates associated with lactation, including whether or not a woman breastfed during the 2-hour 75-g OGTT.
Previous studies comparing any lactation versus no lactation consistently found improved insulin response, lower insulin levels, and lower glucose levels, including lower 2-hour blood glucose concentrations for the OGTT.3,4,13 A study of Latinas with recent GDM reported that any lactation at 4–12 weeks postpartum was associated with lower fasting serum glucose (93±13 compared with 98±17 mg/dL) and 2-hour OGTT glucose levels (124±41 compared with 134±49 mg/dL) controlling for body mass index (calculated as weight (kg)/[height (m)]2), age, and insulin use during pregnancy.3
Studies using more direct measures of glucose metabolism in postpartum women such as frequent blood sampling during the OGTT reported reduced pancreatic β-cell secretory responses in lactating women. Hubinont et al13 reported a 50% reduction in the corrected insulin response at 30 minutes postglucose ingestion for 14 lactating compared with 14 nonlactating women (0.67±0.11 compared with 1.24±0.26 microunits · mg−2 · 102).13 These results imply a lower basal and glucose-stimulated β-cell secretory activity for a standardized glucose load in lactating compared with nonlactating women that could preserve pancreatic β-cell function.
McManus et al14 examined the effects of lactation on glucose and insulin metabolism in women with recent GDM using a frequently sampled intravenous glucose tolerance test. Compared with nonlactating women, lactating women had higher insulin sensitivity, glucose effectiveness, and first-phase insulin response to glucose assessed by Bergman's minimal model that was reflected in the 2.5-fold higher disposition index (insulin sensitivity multiplied by first-phase insulin response to glucose) of 129.9±26.0 compared with 53.4±18.0×10(−4) min(−1) (P<.05).13 These findings demonstrate that lactating women with prior GDM have better β-cell compensation for insulin resistance than their nonlactating counterparts.
In our present study, lower 2-hour plasma glucose concentrations among women who breastfed during the OGTT may be the result of the uptake of glucose (approximately 50 g per day) from maternal plasma through noninsulin-mediated pathways into the mammary gland for milk production.15 The lower maternal glucose and insulin concentrations at the 2-hour post-glucola blood draw may have resulted from lactogenesis initiated in response to infant suckling and emptying of the mammary glands during the OGTT.
Our current study strengths include the large diverse sample of women diagnosed with GDM through standardized criteria, consistent treatment within a large, integrated health care system, the detailed prospective assessment of infant-feeding practices, and observation of frequency and duration of women breastfeeding an infant during the OGTT. The two groups did not differ in mean body mass index or education level. Limitations include the lack of direct assessments of insulin sensitivity and glucose metabolism to elucidate the mechanisms. We also noted that women who did not breastfeed during the OGTT had worse glucose intolerance during pregnancy. However, our findings of lower 2-hour insulin and glucose concentrations remained after adjustment for covariates and a sensitivity analysis excluding mostly formula-feeding women.
In conclusion, breastfeeding for an average of 15 minutes during the OGTT may have a modest effect on 2-hour glucose levels (5% lower on average) for the 75-g OGTT used to classify postpartum glucose intolerance. However, longer duration of breastfeeding had no relationship to the 2-hour plasma glucose and insulin concentrations, although there was a weak inverse correlation with insulin sensitivity. Currently, only 50% of postpartum women with prior GDM undergo the OGTT within 6 months postpartum.16 Therefore, standardized screening of women for glucose tolerance after GDM pregnancy has not reached recommended targets.1 Postpartum women with recent GDM are focused on caring for the newborn and attending their pediatric visits. Moreover, lactating women may decide to breastfeed their infants during the OGTT because they do not wish to introduce formula or the infant is not accustomed to the artificial nipple. For women, being able to breastfeed their infants during the OGTT may alleviate part of the burden in performing the test and potentially improve postpartum screening rates. Such practice is in concert with the Surgeon General's Call to Action to Support Breastfeeding because of the numerous health benefits for women and children.16 Breastfeeding during the OGTT has minor effects on the test results for glucose tolerance after GDM pregnancy. Clinicians should query women about whether they breastfed their infants during the OGTT and consider retesting those whose results fall within 6 mg/dL of the diagnostic cut points.
1. Metzger BE, Buchanan TA, Coustan DR, De Levia A, Dunger DB, Hadden DR, et al.. Summary and recommendations of the Fifth International Workshop–Conference on Gestational Diabetes Mellitus. Diabetes Care 2007;30:S251–60.
3. Kjos SL, Henry O, Lee RM, Buchanan TA, Mishell DR Jr. The effect of lactation on glucose and lipid metabolism in women with recent gestational diabetes. Obstet Gynecol 1993;82:451–5.
4. Lenz S, Kuhl C, Hornnes PJ, Hagen C. Influence of lactation on oral glucose tolerance in the puerperium. Acta Endocrinol (Copenh) 1981;98:428–31.
5. American Diabetes Association. Gestational diabetes mellitus. Diabetes Care 2000;23(Suppl. 1):S77–9.
6. Gunderson EP, Matias SL, Hurston SR, Dewey KG, Ferrara A, Quesenberry CP Jr, et al.. Study of Women, Infant Feeding, and Type 2 diabetes mellitus after GDM pregnancy (SWIFT), a prospective cohort study: methodology and design. BMC Public Health 2011;11:952.
7. Gunderson EP, Hedderson MM, Chiang V, Crites Y, Walton D, Azevedo RA, et al.. Lactation intensity and postpartum maternal glucose tolerance and insulin resistance in women with recent GDM: the SWIFT cohort. Diabetes Care 2012;35:50–6.
8. Hanley AJ, Williams K, Gonzalez C, D'Agostino RB Jr, Wagenknecht LE, Stern MP, et al.. Prediction of type 2 diabetes using simple measures of insulin resistance: combined results from the San Antonio Heart Study, the Mexico City Diabetes Study, and the Insulin Resistance Atherosclerosis Study. Diabetes 2003;52:463–9.
9. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1997;20:1183–97.
10. Mickey RM, Greenland S. The impact of confounder selection criteria on effect estimation. Am J Epidemiol 1989;129:125–37.
11. Ley SH, O'Connor DL, Retnakaran R, Hamilton JK, Sermer M, Zinman B, et al.. Impact of maternal metabolic abnormalities in pregnancy on human milk and subsequent infant metabolic development: methodology and design. BMC Public Health 2010;10:590.
12. O'Reilly MW, Avalos G, Dennedy MC, O'Sullivan EP, Dunne F. Atlantic DIP: high prevalence of abnormal glucose tolerance postpartum is reduced by breast-feeding in women with prior gestational diabetes mellitus. Eur J Endocrinol 2011;165:953–9.
13. Hubinont CJ, Balasse H, Dufrane SP, Leclercq-Meyer V, Sugar J, Schwers J, et al.. Changes in pancreatic B cell function during late pregnancy, early lactation and postlactation. Gynecol Obstet Invest 1988;25:89–95.
14. McManus RM, Cunningham I, Watson A, Harker L, Finegood DT. Beta-cell function and visceral fat in lactating women with a history of gestational diabetes. Metabolism 2001;50:715–9.
15. Xiang AH, Kawakubo M, Kjos SL, Buchanan TA. Long-acting injectable progestin contraception and risk of type 2 diabetes in Latino women with prior gestational diabetes mellitus. Diabetes Care 2006;29:613–7.
16. U.S. Preventive Services Task Force. Primary care interventions to promote breastfeeding: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2008;149:560–4.