Approximately 7% of all pregnant women are diagnosed with gestational diabetes mellitus (GDM) and comprise a high-risk group for future development of type 2 diabetes mellitus. Women with GDM are 7 times more likely to develop type 2 diabetes after pregnancy [1
], although a 4-fold higher incidence of overt diabetes after GDM pregnancy was reported by Gunderson et al. after excluding women with hyperglycemia before pregnancy based on prepregnancy blood glucose measures [2
About 5-10% of women will be diagnosed with type 2 diabetes within the first 6 months after GDM pregnancy and another 10-15% will develop diabetes within the subsequent 1-2 years postpartum [3
]. Predictors of diabetes among women with a history of GDM include maternal antepartum and early postpartum glycemia, insulin use during pregnancy, pancreatic β-cell compensation for higher insulin resistance and GDM recurrence [6
] and family history of diabetes, especially having a mother with diabetes [7
]. Prepregnancy obesity, gestational weight gain, postpartum weight gain, and subsequent pregnancies have been associated with higher risk of diabetes years later [6
]. In cross-sectional studies, greater central obesity has been reported in women who developed type 2 diabetes after GDM pregnancy [11
Lactation intensity and duration have rarely been assessed in relation to type 2 diabetes after GDM pregnancy. Of 28 studies cited in a comprehensive review by Kim et al.[6
] and 5 subsequent studies [9
], only 5 of 33 studies examined lactation status (yes or no) in relation to incident diabetes, and the findings were inconclusive [16
]. Most studies examined "any" lactation versus none, have utilized primarily retrospective designs, involved Latinas, did not conduct standardized postpartum screening for diabetes, and had relatively small sample sizes. Of the only two prospective studies that examined lactation duration in relation to incident diabetes, a previous history of GDM was not ascertained [18
]. These two studies, including either white or Chinese women, reported that increasing lactation duration was associated with lower incidence of diabetes after pregnancy, which was ascertained via self-report in mid to late life. A retrospective cohort study of White women with a history of GDM found a null association between lactation duration and incident diabetes ascertained by self-report [18
]. A major limitation of these three previous studies is that they did not conduct periodic standardized screening of women to ascertain diabetes incidence after pregnancy.
One of the few prospective studies to examine women of reproductive age (50% Black and 50% White) during a 20 year period (1985-2005) is the U.S. multi-center study, the Coronary Artery Risk Development in Young Adults (CARDIA) Study. In CARDIA women, glycemia was measured both before pregnancy and post-weaning to assess the association between lactation duration and incidence of the metabolic syndrome in women with and without previous GDM pregnancies. Longer duration of lactation was associated with a 50-89% reduction in incident metabolic syndrome on average 8 years after pregnancy among women with a history of GDM as well as those with no history of GDM [23
]. To our knowledge, lactation intensity has not been evaluated in any previous studies that examined incident diabetes after GDM pregnancy.
We herein critically review the epidemiologic evidence and biological plausibility that lactation may protect women from developing type 2 diabetes in mid to late life. We examine findings from studies of lactation and persistent changes in biochemical risk factors, as well as incident metabolic disease, including type 2 diabetes, after pregnancy. The evidence for short-term changes in metabolic risk profiles and limited evidence from large epidemiologic studies were the impetus for the funding and design of the SWIFT study of postpartum women with recent GDM pregnancy.
The goal of the SWIFT study is to prospectively examine lactation intensity and duration in relation to incident diabetes after GDM pregnancy. The SWIFT study specific aims, design, and methodologies are presented as well as a description of the participant eligibility criteria. We also summarize the recruitment, in-person study assessments and retention protocols for SWIFT, a prospective postpartum cohort of women with recent GDM who delivered a term infant within the Kaiser Permanente Northern California (KPNC) integrated healthcare system.
Physiological effects of lactation on metabolic status: biological plausibility
Lactation has favorable effects on maternal metabolism including increased glucose-disposal rates, enhanced lipolysis and diverting glucose (> 50 g/d) for utilization in milk production [24
]. Data are less available regarding whether lactation protects β-cell function [27
], or has lasting effects on maternal glucose tolerance to ultimately influence the risk of diabetes after GDM pregnancy. Prospective studies are needed that assess lactation more precisely and completely in relation to changes in oral glucose tolerance and body adiposity to determine conclusively whether lactation may delay or prevent future diabetes.
Lactation and glucose homeostasis
Overall, alterations in the hormonal milieu and responsiveness during lactation are designed to favor lower insulin levels as a result of higher glucose utilization by the mammary gland [25
] and increased lipolysis to accommodate the metabolic demands of milk production. Lactation is characterized by increased maternal basal metabolic rates, greater energy needs for milk production, and mobilization of fat stores [25
]. Lactating women generally exhibit lower blood glucose and insulin concentrations and higher glucose production rates due to increased glycogenolysis (not gluconeogenesis or increased use of free fatty acids) [31
]. In other cross-sectional studies, lactating women had lower fasting plasma glucose and insulin levels [32
], and lower post-absorptive insulin levels than non-lactating women [26
Frequently sampled oral glucose tolerance (FSOGT) tests showed a higher corrected insulin response at 30 min (p
< 0.03) in non-lactating (1.24 + 0.26 μU· mg-2 ·102) than lactating women (0.67 + 0.11 μU· mg-2 ·102). Thus, basal and glucose-stimulated β-cell secretory activity for a standardized glucose load may be lower for lactating than non-lactating women [27
], indicating that lactation may reduce the load on the β-cells as reported in studies that examined women without glucose tolerance during pregnancy as outlined above.
Lactation's effects on maternal body weight
About 4-6 kg of body fat is stored during pregnancy partially in preparation for fetal growth during late gestation and lactation [33
]. In small clinical studies, average weight loss during the first 6 months of lactation in affluent populations is about -0.5 to -0.8 kg/month [35
]. Although lactation increases total energy expenditure by 15-25% for milk production [25
], evidence is inconsistent as to whether lactation promotes greater postpartum weight loss [37
Prospective studies that measured maternal weights before or during early pregnancy reported lower postpartum weight retention, more rapid return to prepregnancy weight or greater weight losses within 6 months to 1 year postpartum among lactating women [30
]. Greater frequency of lactation and higher breast milk energy output are associated with greater weight loss from 3 to 6 months. Higher intensity of breastfeeding from 2.5 to 6 months postpartum [39
] and for the first year [43
] resulted in 2 kg greater average maternal weight loss. Another study among women (n = 110) classified as fully breastfeeding, partly breastfeeding, or bottle-feeding at six different points in time found that women who lactated more than 1 year lost 2 kg more by 18 months postpartum than women who bottle-fed [43
]. Olson et al. in 2003 [38
] reported 1.2 kg lower postpartum weight retention in women still breastfeeding at 1 year postpartum, after controlling for first trimester weight and other confounders. More recently, an analysis with a large sample of mothers (n = 32, 920) enrolled in the Special Supplemental Nutrition Program for Women, Infants and Children (WIC) resulted on a modest but significant effect of lactation on weight retention from one pregnancy to the next [44
]. Thus, after adjusting for potential confounder effects, lactation for 20 or more weeks resulted on 0.39 kg less weight retention at the beginning of the second pregnancy compared to no lactation. Studies that carefully assessed lactation exposure demonstrate that greater intensity and duration of lactation promote greater postpartum weight loss.
Overall and regional adiposity during lactation
During lactation, fat stores are mobilized to a greater extent from the trunk and thighs [45
]. Lactating women show greater declines in suprailiac and subscapular regions but fat increased in the triceps region [35
]. However, skinfold thickness is a relatively imprecise measure for evaluating total body fat or regional fat, and does not precisely assess changes due to fluid losses [30
In studies using dual-energy X-ray absorptiometry (DXA) to measure changes in body composition and regional fat distribution in lactating and non-lactating women, total fat mass showed a greater linear 12-month decline in lactating women versus non-lactating women with the largest decline between 3 and 6 months [35
]. No differences in fat mobilization from leg, arm and trunk regions were found, but lactating women had a non-significant 2 kg greater decline in total fat mass [35
]. However, this study lacked statistical power to detect this clinically significant difference. Central adiposity is of greater importance metabolically than overall obesity since intra-abdominal (visceral) fat is associated with development of obesity-related insulin resistance and progression to type 2 diabetes [50
]. Visceral fat is more metabolically active and is thought to differ from sub-cutaneous fat in the production of adipocytokines that may regulate insulin sensitivity [51
Figure presents our theoretical model and summarizes the hypothesized relationships of lactation to immediate and long-term effects on glucose tolerance as discussed in the preceding sections.
Theoretical model of effects of lactation on glucose tolerance and Type 2 DM.
Lactation and postpartum metabolic status in women with a history of GDM
McManus et al. administered a frequently sampled intravenous glucose tolerance test (FSIGT) to 14 lactating and 12 non-lactating women with previous GDM at 3 months postpartum, who were matched for age, weight, postpartum weight loss and exercise habits. In lactating women, insulin sensitivity, glucose effectiveness and first phase insulin response to glucose (AIRg) assessed by Bergman's Minimal Model were higher, but statistical significance was not reached given the small sample size. However, the disposition index (DI = insulin sensitivity multiplied by AIRg) was 2.5 times higher (129.9 ± 26.0 vs 53.4 ± 18.0 × 10(-4) min(-1); p
< 0.05) in lactating versus non-lactating women [28
]. The higher DI supports the hypothesis that lactation promotes much better β-cell compensation for insulin resistance, which may help maintain β-cell function in the long-term. Because glucose is diverted for milk production, reduced plasma glucose levels "unload" the β-cells such that β-cell function is preserved; insulin response for given levels of resistance is improved. These physiologic changes may sustain glucose tolerance and protect against β-cell exhaustion leading to type 2 diabetes.
Lactation is associated with more favorable glucose metabolism at 2-3 months postpartum in some, but not all small clinical studies. Fasting plasma glucose and insulin levels and incremental responses of plasma glucose (at 30, 60, 90 and 120 min) to a test meal did not differ between lactating (n = 12) and non-lactating (n = 7) groups, but lactating women had lower plasma insulin response and higher rates of glucose utilization [52
]. No significant differences in plasma glucose and insulin levels after a 50-gram OGTT were found between lactating and non-lactating groups in two other studies [27
]. However, the β-cell secretory response assessed by the corrected insulin response at 30 min (CIR 30 min) during the OGTT was reduced by half for lactating versus non-lactating women [0.67 ± 0.11 vs. 1.24 ± 0.26 μU· mg-2 ·102] [27
A series of cross-sectional and follow-up studies of Latinas with previous GDM examined early postpartum lactation with conflicting findings. Lactation was associated with lower prevalence of diabetes and better glucose tolerance at 4-12 weeks postpartum; a lower total area under the glucose tolerance curve (AUC) (17.0 ± 4.2 vs. 17.9 ± 5.0 g.minute/dL), and lower fasting serum glucose (93 ± 13 vs. 98 ± 17 mg/dL) and 2-hour OGTT glucose levels (124 ± 41 vs. 134 ± 49 mg/dL) after controlling for body mass index (BMI), maternal age and insulin use during pregnancy [19
]. Lactation status at 4-16 weeks postpartum was not associated with risk of type 2 diabetes within 5 years [20
]. Buchanan et al. examined 122 Latinas with normal fasting glucose and no insulin use during GDM pregnancy and found that those diagnosed with diabetes within 6 months postpartum were less likely to have breastfed (42%) than those with normal glucose tolerance (71%) [16
]. Finally, among 91 Latinas receiving OGTT screening at 15-month intervals, lactation status (yes vs. no) at 11-26 months postpartum did not influence onset of type 2 diabetes [21
]. Other confounders such as employment, intentional dieting, dietary intake, and physical activity have not been examined in any of the studies. Weight gain and higher BMI during young adulthood are associated with insulin resistance and greater risk of type 2 diabetes [53
]. No studies have examined changes in waist girth or central adiposity after GDM pregnancy.
Inconclusive evidence for lasting effects on future disease
Clinical and epidemiologic evidence support the hypothesis that lactation has immediate favorable effects on maternal glucose tolerance, and may reduce the load on the β-cells by lowering plasma glucose levels through diversion of glucose for milk production, thereby lessening insulin demands. The Nurses Health Study (NHS) found a lower incidence of self-reported diabetes by 14-15% for each year of lactation, and a stronger risk reduction for exclusive lactation. This association was independent of current BMI and behavioral risk factors, and there was no interaction by parity. However, the association was null among women with history of GDM, in whom lactation and other lifestyle behaviors (i.e., diet or exercise) were unrelated to diabetes risk [18
]. By contrast, a subsequent 20-year longitudinal study based on biochemical measurements before pregnancy and after weaning, found that lactation for 2 or more months versus less than 1 month was associated with 2 to 8-fold lower incidence of the metabolic syndrome after GDM pregnancy controlling for pre-pregnancy risk, socio-demographics, behavioral changes, and race [23
Evidence is equivocal that lactation, except for exclusive lactation for several months, promotes greater weight loss during the first year postpartum. Few studies have included diverse racial and ethnic groups, assessed postpartum behaviors, weight loss, regional adiposity, or other factors in relation to long-term maternal glucose tolerance after GDM pregnancy.
Lactation has immediate favorable effects on glucose tolerance, but only limited evidence for any long-term effects. Studies have never prospectively assessed both intensity and duration of lactation as well as other major confounders in women with recent GDM. Well-controlled, prospective studies with more precise and complete measures of lactation are required to assess the impact on development of type 2 diabetes. Data are currently unavailable to conclude that lactation reduces the risk of type 2 diabetes after GDM pregnancy.
The Diabetes Prevention Program (DPP) showed a greater than 50% reduction in the incidence of diabetes with a weight management/exercise program for high risk adults, demonstrating that a relatively modest weight loss of 3-4 kg (equivalent to 5% of initial body weight) could prevent diabetes [56
]. A relatively modest 2 kg higher weight loss within 1 year has been attributed to intensive lactation in the studies that prospectively assessed lactation intensity and measured body weights. Based on the DPP findings, lactation may lower the incidence of diabetes by 30% after GDM pregnancy based on loss of fat mass. If lactation exerts other effects (e.g., preservation of β-cell function) that are independent of adiposity, then the reduction in type 2 diabetes may be even greater.
The American Academy of Pediatrics recommends breastfeeding as the preferred method of infant feeding for at least 1 year of age [57
]. Although 80% of U.S. women initiate lactation, only 45% report "any" level of breastfeeding at 6 months [57
]. Lactation is a modifiable behavior that may be translated into a practical, low-cost intervention, and has the potential to enhance postpartum interventions that have primarily relied on strategies to promote healthy diet and increase physical activity levels. Breastfeeding may prevent recurrence of GDM in a future pregnancy, and thereby also influence the risk of type 2 diabetes in the offspring [58
The SWIFT study design and aims: prospective GDM cohort
The overall objective of this study is to assess whether lactation prevents the onset of type 2 diabetes during the first 2 years postpartum among women with recent GDM, after taking into account their age, race/ethnicity, parity, weight status, education, severity of gestational glucose intolerance, GDM recurrence history, family history of diabetes, clinical or medical risk factors, and other postpartum behaviors.
Specifically, we aim to determine whether intensive lactation compared with intensive formula feeding is associated with:
Lower 2-year incidence of type 2 DM (Aim 1);
Lower fasting and 2-hour post-load plasma glucose levels, and lower insulin resistance (Aim 2); and
Lower total and central adiposity, and higher plasma adiponectin levels (Aim 3).
Another objective of the study is to determine whether lower adiposity is associated with lower incidence of type 2 diabetes in this population (Aim 4).