Introduction

Exclusive breastfeeding to 6 months, with continued breastfeeding for at least 1 to 2 years, is strongly endorsed by major medical organizations, including the American College of Obstetricians and Gynecologists [1], the American Academy of Pediatrics [2], and the World Health Organization [3]. Among infants in the USA and other high-income countries, breastfeeding is associated with significant reductions in infant morbidity and mortality, including acute otitis media, gastroenteritis, severe lower respiratory tract infections, atopic dermatitis, childhood asthma, diabetes, childhood leukemia, sudden infant death syndrome, and necrotizing enterocolitis [4]. In an analysis of 2017 US births, ever breastfeeding was associated with a 40% reduction in infant mortality between 7 and 27 days of age, and a 19% reduction in infant mortality from 1 to 12 months of age [5].

Breastfeeding matters for mothers too. Longer lactation is associated with lower chronic disease burden, including less metabolic syndrome [6], hypertension [7], type 2 diabetes [8,9,10], cardiovascular disease [11, 12], and breast, uterine, and ovarian cancers [13]. In a simulation analysis, suboptimal breastfeeding in the USA was associated with 721 excess child deaths and 2619 excess maternal deaths each year [14•]. Conditions associated with suboptimal breastfeeding are more prevalent in people with obesity [14•, 15,16,17,18,19,20], underscoring the importance of optimal support in this population.

Breastfeeding initiation in the USA has increased dramatically from the nadir of 24% in 1971 [21] to over 80% of parents in the USA today [22]. However, only one in four lactating parents meets the public health goal of exclusive breastfeeding for 6 months, and only half meet their own breastfeeding goals, with low milk supply being a leading reason for breastfeeding cessation [23]. Undesired weaning is associated with depression symptoms [24•], and birthing parents who are unable to fulfill their desire to breastfeed experience considerable distress [25]. Lactating parents with obesity experience more difficulties in establishing and sustaining lactation. Given that 29% of women of reproductive age have obesity, and prevalence is rising [26], optimizing support for lactating parents with larger bodies is a public health priority.

In this review, we will summarize epidemiologic work on maternal obesity and breastfeeding outcomes, research that contributes to our understanding of the underlying mechanisms, intervention studies aimed at improving their breastfeeding outcomes, and clinical management considerations.

Epidemiology

Maternal obesity is associated with earlier weaning. Table 1 summarizes perinatal cohort studies conducted in high-income countries, including the USA [27, 28], Australia [29], Denmark [30], Norway [31], Taiwan [32], and Austria [33], representing 163,882 singleton births. Initiation rates are only slightly lower among parents with obesity, but a BMI ≥ 30 kg/m2 is associated with significantly higher odds of stopping exclusive or any breastfeeding in the early postpartum as compared to BMI < 25.0 kg/m2 (see Table 1). For example, in Norwegian Mother and Child Cohort Study [31] (n = 49,669), breastfeeding initiation was nearly universal (BMI < 25.0 kg/m2, normal weight: 99.6%; BMI ≥ 35.0 kg/m2, class 2 + obesity, 97.4%), whereas sustaining full breastfeeding > 4 months was substantially lower among women with class 2 + obesity (38%) compared with normal weight women (63%).

Table 1 Summary of large or moderate-sized prospective cohort studies examining the association between maternal BMI category and breastfeeding outcomes in high-income country settings
Full size table

Breastfeeding difficulties in lactating parents with obesity manifest early. Among singleton births in 2016 in a Cincinnati Baby-Friendly Hospital (n = 1588), initiation rates were similar, but exclusive breastfeeding was 14-percentage points lower during the maternity stay among women with obesity [34•]. Infants born to mothers with obesity required more medically indicated supplementation (adjusted OR [95% confidence interval], 2.6 [1.7, 4.1]). Several conditions linked with obesity (diabetes in pregnancy, hypertension, labor induction, cesarean delivery, low birthweight, or large birthweight) are also predictive of medically indicated supplementation. Adjustment for these conditions attenuated, but did not eliminate, the significant effect of obesity on medically indicated formula use (1.7 [1.02, 2.7]. The adjusted odds of maternally requested formula use was also higher for mothers with obesity (2.0 [1.7, 4.1]).

Delayed lactogenesis, defined as > 72 h postpartum without symptoms of copious milk production [35], contributes to early formula supplementation. Lactogenesis II (a.k.a., secretory activation) [36] is characterized by a dramatic increase in milk production following childbirth [37]. Cohort studies in Connecticut [35] and Sacramento [38] have found that more than 50% of women with obesity experience delayed lactogenesis. Delayed lactogenesis is a strong risk factor for excess neonatal weight loss: in a community-based perinatal cohort study in Davis, CA (n = 280), the prevalence of excess neonatal weight loss (defined as weight loss ≥ 10% of birthweight) to be 40.4% with delayed lactogenesis versus 5.7% with timely lactogenesis among exclusively breastfeeding dyads (p < 0.0001) [39], consistent with findings in a Sacramento cohort of first-time mothers [40].

Potential Mechanisms

Behavioral Factors in Context of Weight Stigma

It is overly simplistic to conclude that behavioral differences explain disparities in breastfeeding outcomes by body size because what appears superficially as individually driven behavioral differences is the visible “tip of the iceberg.” Implicit and explicit weight stigma towards birthing persons with obesity by health care providers [41] may undermine equitable access to breastfeeding support for large-bodied parents. In a longitudinal qualitative study among lactating parents (normal weight, n = 9; obese weight, n = 13), several themes emerged [42•]. Although both groups experienced difficulties, such as with positioning and latch, challenges were experienced to a greater degree or lasted longer among women with obesity. These women and their infants also experienced more health challenges, complicating human milk feeding. These additional challenges may explain why lactating parents with obesity are more likely to request formula during the maternity stay [34•]. Drivers of these requests may reflect lack of prenatal breastfeeding education tailored to parents with obesity, fewer opportunities for early skin-to-skin contact, lack of at-the-breast infant feeding support tailored to larger bodies, and, in the case of dyad separation, inequitable access to appropriate resources and support for mechanical breast expression. Weight bias is also proposed to contribute to the allostatic load of parents with large bodies [41], leading to inflammation, which may physiologically impede milk production, as discussed below [43•, 44]. Experiencing weight bias with previous births can compound disparities by derailing subsequent breastfeeding attempts.

Evidence for Physiologic Factors Impeding Milk Production

Physiologic differences in parents with larger bodies may impede mammary gland development and milk production. In the Infant Feeding Practices II Study, women with obesity were less likely to be exclusively breastfeeding at 1 month and “Insufficient Milk” mediated this relationship [45•]. For decades, the prevailing dogma was that insufficient milk production is largely a problem of misperception, and thus labeled “perceived insufficient milk production,” with true insufficient milk production thought to be rare [46]. However, the current obesity epidemic compels us to reexamine the “perceived” insufficient milk paradigm.

The hormones estrogen and progesterone stimulate mammary development, and prolactin stimulates mammary differentiation and milk secretion [47]. Rasmussen and Kjolhede compared hormone levels during the first week postpartum in women with BMI < 25.0 kg/m2 versus BMI ≥ 25.0 kg/m2 [48]. While no differences were observed in estrogen or progesterone, serum prolactin concentrations were significantly lower 30 min post-feeding in the group with higher BMI. The authors did not report whether prolactin response aligned with timing of lactogenesis or other breastfeeding outcomes.

Metabolic Health as a Contributor to Milk Production Capacity

Body mass index is an imperfect indicator underlying metabolic health [49-51]. Metabolic factors that tend to be more prevalent in parents with obesity, such as insulin resistance and inflammation, may be more important to milk production than one’s BMI category per se. In cohort studies that assessed timing of lactogenesis (n = 431) [38], the odds of delayed lactogenesis were 2.2-fold higher among mothers with BMI ≥ 30.0 kg/m2 as compared to BMI < 25.0 kg/m2, adjusting for breastfeeding intensity. Two variables associated with glucose intolerance—older maternal age [52, 53] and higher infant birth weight [54, 55]—were independent predictors of delayed lactogenesis, implicating metabolic health as a mediator of the association between obesity and milk production difficulties. In a subsequent study, it was found that delayed lactogenesis correlated with weaker insulin response to an oral glucose challenge during pregnancy [56]. Further evidence of metabolic health influencing milk production comes from a breastfeeding medicine clinic study comparing cases of low milk supply diagnoses with controls presenting with latch problems [57]. Dyads with both conditions were not included. A history of diabetes in pregnancy was associated with 2.4-fold higher adjusted odds of having a low milk supply diagnosis (95% CI, 1.2, 4.9), adjusting for preterm birth, cesarean delivery, PCOS, and infertility. In a separate study, it was reported that among women with gestational diabetes, higher fasting glucose, A1c, BMI, and subscapular skinfolds were associated with earlier breastfeeding cessation [58].

Prior to 2009, the prevailing belief was that insulin played little, if any, direct role in lactation [59]. However, strong evidence has emerged from bovine [60], rodent [61-64], and human [65] lactation studies showing that insulin-sensitive signaling is dramatically upregulated in the mammary epithelial cell as part of the secretory activation cascade [65].

A recent low milk supply study provides further evidence linking metabolic health with milk synthesis [66•]. Breastfeeding characteristics and metabolic health indicators were compared in women with objectively measured very low milk output, despite frequent and thorough breast emptying during 2–4 weeks of follow-up (n = 18), versus moderate/normal milk output (n = 12) using a nested case–control study design, plus an external control group of exclusively breastfeeding dyads (n = 12). Mean 24-h milk production in the three groups was 183 ± 72, 604 ± 151, and 758 ± 71 mL/day, respectively. Although age, primiparity, and day postpartum and frequency of breast emptying were similar among the three groups, nearly every measure of metabolic health was worse in the very low milk supply group, including BMI (p < 0.0001), waist circumference (p < 0.0001), fasting plasma glucose (p < 0.007), fasting triglyceride (p < 0.0005), systolic blood pressure (p < 0.0008), and HDL cholesterol (p < 0.01). The prevalence of gestational diabetes mellitus was also significantly different (p < 0.02).

In the same milk supply study, milk-derived lactocyte mRNA was sequenced to compare gene expression in a subset of the nested controls (n = 4) and very low milk supply cases (n = 5). Overall, differential expression supported biologic plausibility of poor metabolic health suppressing milk production: there was no significant difference in expression of the prolactin receptor or its downstream target, JAK2 [67]. However, insulin-sensitive gene expression was significantly downregulated, including a 2.3-fold decrease in insulin receptor substrate 2, and 3.1-fold lower expression of acetyl-coA carboxylase 2, a gene that inhibits fatty acid oxidation, among 97 insulin signaling genes differentially expressed between cases and controls [67]. Cases also had strong disruption of lactocyte fatty acid metabolism, depriving the mammary gland of vital fuel and substrate for milk synthesis. Notably, lipoprotein lipase (LPL), a hallmark of obesity-driven chronic inflammation, was downregulated 1.8-fold, as well as downregulation of PPARA and RXRA, which are regulatory factors upstream of LPL [44].

Downregulation of LPL expression in the lactocyte is especially intriguing, as elevated plasma triglyceride is predictive of downregulated mammary LPL in gray seals [68], and is known to be associated with inflammation-driven downregulation of LPL activity in peripheral tissues in humans [69, 70]. Walker et al. recently examined the transfer of long-chain fatty acids from blood to milk in biospecimens from the same low milk supply study and external controls [43•]. The mammary gland cannot synthesize long-chain fatty acids; therefore, their incorporation into triglycerides during lipid synthesis in the lactocyte is entirely dependent upon uptake from circulation [71]. The investigators reported significantly lower relative amounts of long-chain fatty acids in both the serum and milk in the very low milk production group, as compared to the nested and external controls [43•]. Furthermore, there was a very strong linear relation between serum and milk long-chain fatty acid proportions in the external control group (r = 0.82, p < 0.001) and no association in the very low milk production group (r = 0.20, p = 0.41). This supports the hypothesis that in lactating people with inflammation, uptake of substrate needed for milk production may be severely suppressed.

Taken together, this body of work suggests that inflammation and insulin resistance may impede milk production and underlie the association between higher BMI and shorter breastfeeding duration. To confirm causality, interventions aimed at improving insulin action and reducing inflammation (initiated preconceptionally or prenatally) are warranted. Given associations between metabolic profile and breastfeeding difficulties, reverse causation may explain some of the reported associations between longer duration of lactation and improved long-term maternal metabolic health [6,7,8,9,10,11,12], in that adverse lactation outcomes may be a marker for metabolic disease risk [72].

Interventions to Improve Outcomes

Several randomized controlled trials (RCTs) have tested interventions to support breastfeeding among birthing people with obesity, with mixed results. Rasmussen et al. tested two interventions against usual care in the Bassett Improving Breastfeeding Study (BIBS) [73]. Both interventions were harmful. The BIBS 1 intervention consisted of telephone support from an IBCLC prenatally and at 24 and 72 h, encouragement to “get up and move,” and requesting hospital visitors to leave if they had stayed ≥ 2 h or were disrupting infant feeding. Compared with usual care, women in the intervention group had shorter durations of any (median 8.6 versus 12.9 weeks) and exclusive breastfeeding (median 3.4 versus 8.1 weeks). In BIBS 2, women were randomized to a manual breast pump, an electric breast pump, or usual care. Women who received pumps were advised to express milk for 10 min after five feedings each day until their milk came in or 5 days postpartum. Compared with usual care, women allocated to pumping had shorter durations of any (manual: 13.4 weeks; electric: 4.0 weeks; usual care: 26.6 weeks) and exclusive breastfeeding (manual 2.3 weeks; electric 0.7 weeks; control 4.4 weeks).

Chapman et al. [74] compared specialized breastfeeding peer counselors (SBFPC) with usual care among women with a BMI ≥ 27 kg/m2. Usual care included access to an existing Breastfeeding Heritage and Pride (BHP) peer counselor program; SBFPCs received the same training as BHPs, as well as 20 h of specialized training, including body image sensitivity, positioning for large breasts, and discreet breastfeeding techniques. The SBFPC intervention included access to 3 prenatal visits, daily visits during the birth hospitalization, and up to 11 home visits through 6 months postpartum. Although breastfeeding self-efficacy scores were higher in the intervention group at 2 weeks postpartum, there were no differences in initiation, exclusive breastfeeding, or any breastfeeding rates.

In a Denmark RCT [75], Carlsen et al. tested telephone-based IBCLC support versus usual care among women with a pre-pregnancy BMI ≥ 30 kg/m2. The intervention included ≥ 9 phone consultations in the first 6 months postpartum and an IBCLC available to contact 7 days per week. Compared with usual care, telephone support lengthened median duration of any (180 versus 108 days) and exclusive breastfeeding (120 versus 41 days).

Two home-based interventions in the USA did not affect breastfeeding outcomes. Reifsnider et al. [76] tested a home-based bilingual community health worker intervention among Mexican American women with BMI ≥ 30 kg/m2 recruited through WIC clinics, compared with no home visits. Lewkowitz et al. [77] tested a home-based intervention among African American women with overweight or obesity. Both groups received the Parents As Teachers curriculum, and the intervention group received counseling on diet, physical activity, and breastfeeding.

Jacobson et al. [78] evaluated knowledge and skills among women with BMI ≥ 25 to < 40 kg/m2 at baseline and 4–6 weeks after a 2-h class. The authors found greater self-reported comfort with hand expression and knowledge about positioning, managing pain, and getting help for breastfeeding problems. Feeding outcomes were not reported.

Given the association between insulin resistance and low milk supply, two studies have evaluated metformin. In a secondary analysis of an RCT of metformin during pregnancy among women with PCOS [79], Vanky et al. found a non-significant increase in duration of exclusive breastfeeding (metformin, 4.5 ± 2.8 versus placebo, 3.9 ± 2.9 months, p = 0.08). In a pilot RCT of metformin among women with low milk supply and signs of insulin resistance [80] (n = 15), there was no significant differences in median improvement in milk output over 2–4 weeks of follow-up.

Clinical Considerations

Given limited evidence, how might health team members support birthing people with higher BMIs to initiate and sustain breastfeeding? Here, we outline strategies to consider, informed by our clinical experience as a breastfeeding medicine physician and an international board-certified lactation consultant.

Prenatal Considerations

  • Enabling health and wellbeing for pregnant people with obesity begins with emphasizing health at every body size [81]:

    • Exercise: During pregnancy, the American College of Obstetrics and Gynecology recommends 30–60 min of moderate intensity exercise at least 3–4 days per week [82].

    • Nutrition: Low inflammatory diet patterns are associated with more favorable metabolic health [83].

    • Respectful care: Weight stigma can impede exercise and increases allostatic load in patients with large bodies [41].

  • Evaluate breast morphology at the initial prenatal visit and ask about breast changes during pregnancy. Widely spaced, conical breasts may indicate insufficient glandular tissue. Advise patients with a history of breast reduction or concern for insufficient glandular tissue to share this information with the infant’s provider and recommend infant follow-up within 48 h of discharge. For individuals who have undergone breast reduction surgery, breastfeeding outcomes vary, depending on the surgical technique. Kraut et al. found higher rates of breastfeeding success when the reduction procedure preserved the entire column of subareolar parenchyma [84].

  • Identify and share local resources for nursing bras, tops, and baby carriers that accommodate larger bodies.

  • Begin conversations about infant feeding with open-ended questions, such as “What have you heard about breastfeeding?” Respond by paraphrasing their response, and then addressing expressed concerns [85].

  • Validate concerns about milk supply, acknowledging that some mothers with obesity make lots of milk, and others make less. Advise that the transition from colostrum to milk may be later than 72 h; if pasteurized donor human milk is available at your facility, discuss this option. If an LGA infant is anticipated, or the pregnancy is complicated by diabetes, it can be helpful to review monitoring for hypoglycemia. As with all patients, discuss the importance of early skin-to-skin care, feeding on cue, and adjusting baby’s position if there is discomfort with latch.

  • Expressing colostrum before delivery has been suggested to improve exclusive breastfeeding rates. Evidence suggests the practice is safe, but there is little benefit. A large RCT evaluated outcomes of colostrum expression after 37 weeks among women with diabetes [86]. Antenatal expression did not affect NICU admission, the primary outcome. Although expression slightly increased exclusive breastfeeding in the first 24 h (69 versus 60%, adjusted RR 1.15 [95% CI, 1.02, 1.28]), there were no differences in exclusive or any breastfeeding at hospital discharge or 3 months. The median total volume expressed (sum of all expressions over entire intervention) was 5.5 mL, and women’s experiences were mixed; one participant said: “You do have that sense that, is there any point in me doing this?” [87] Moorhead et al. recommend that clinicians engage patients in shared decision-making about antenatal expression, discussing both the potential burden and realistic benefits.

Immediate Postpartum Care

  • At birth, support skin-to-skin, including in the operating room for cesarean births. In a Cochrane meta-analysis [88], skin-to-skin increased neonatal glucose levels (median 10.5 mg/dL, 95% CI 8.4, 12.6) and lengthened breastfeeding duration (median 64.0 days, 95% CI 38.0, 89.5)

  • For individuals with large breasts, offer support for positioning, including nursing pillows to support the infant and a rolled-up small blanket or washcloth to elevate the breast [89].

  • Share risk factors for delayed onset of lactogenesis and low milk supply with the infant’s provider, including maternal BMI, breast morphology, and history of breast surgery. Although these are risk factors for lactation challenges, many birthing people with obesity can sustain exclusive breastfeeding and meet their feeding goals. Collaborate with the infant’s care team to ensure follow-up within 24–48 h of discharge [90].

  • Withdrawal of progesterone is thought to trigger onset of lactogenesis II, raising concerns that exogenous hormones may disrupt lactation. These theoretical risks should be shared with the patient before initiating immediate postpartum contraception so that they can make an informed decision within the context of their desire to breastfeed and to prevent a short-interval pregnancy [91].

Outpatient Postpartum Follow-up

  • Share that breastfeeding utilizes about 500 kcal/day and can facilitate postpartum weight loss. Postpartum weight loss of 0.5 to 1 kg per week does not adversely affect infant growth. Lactation requires at least 210 g of carbohydrate per day; low carbohydrate diets can precipitate ketoacidosis in lactation people and should be avoided [92-98].

  • Individuals who have undergone bariatric surgery are at risk for B12 deficiency, which can result in congenital B12 deficiency in the infant [99]. To ensure adequate B12 for the infant, supplementation should be continued in lactation following bariatric surgery.

  • In qualitative studies, women with higher BMIs reported greater discomfort than women with normal BMIs with body exposure during breastfeeding. Share strategies for breastfeeding outside the home.

  • For concerns about infant growth or milk production:

    • Consider pre-glandular, glandular, and post-glandular causes of low milk production. Endocrine work-up includes assessment of TSH; prolactin, if there is concern for Sheehan’s syndrome; and clinical evaluation for retained placenta, which can prevent onset of lactogenesis II [100, 101]. Engage an International Board-Certified Lactation Consultant to assess the infant’s ability to latch and transfer milk, and to evaluate pump flange fit, if the patient is expressing milk.

    • If the infant is not gaining weight through at-breast feeding, it is common for “triple feeding” to be recommended. This practice involves putting the infant to breast, supplementing via bottle, and then expressing from the breast to stimulate more milk production. There are no published data on the effectiveness of this practice; however, it can rapidly become overwhelming. We advise a time-limited recommendation to triple feed — ideally for no more than a few days — followed by a frank discussion of whether continuing this labor-intensive process is sustainable.

    • If long-term supplementation is necessary, partner with the family to map out a sustainable approach.

    • Feeding is not “all or nothing” — families can have a satisfying, long-term, partial breastfeeding relationship. Breastfeeding is a part of motherhood, not the point of motherhood; the lactating parent is uniquely qualified to decide what is “worth it,” and when is the right time to transition from breast milk feeding.

    • In our experience, if the lactating parent can provide about 50% or more of the baby’s needs, it is often workable to start feedings at breast, followed by a supplement. For those with lower milk production, if may be more sustainable to offer the supplement first, followed by going to breast “for dessert.” This can avoid struggling to satisfy a frustrated baby at the breast. In some cases, families prefer to express breast milk and bottle feed. A supplemental nursing system is also an option, but this is not sustainable for many families.

    • Infant feeding difficulties are associated with postpartum depression and anxiety symptoms [24•, 102, 103]. Patients presenting with breastfeeding difficulties should be offered socio-emotional support and screened for symptoms.

Just as some births require a cesarean, some breastfeeding requires supplementation. As Neifert has written: “The bold claims made about the infallibility of lactation are not cited about any other physiologic processes. A health care professional would never tell a diabetic woman that ‘every pancreas can make insulin’ or insist to a devastated infertility patient that ‘every woman can get pregnant.’ The fact is that lactation, like all physiologic functions, sometimes fails because of various medical causes” [104].

Conclusion

Breastfeeding is associated with important benefits, and more than 80% of birthing people want to breastfeed their babies. Although not all birthing people can meet all of the infant’s nutritional needs, babies can be nurtured at breast no matter how much milk is made. Holistic, non-stigmatizing support for birthing people with obesity can enable them to have a satisfying breastfeeding experience.