Insulin Resistance Delays Lactation by Disrupting Mammary Mitochondria
Insulin resistance is emerging as more than a maternal metabolic marker. In postpartum women, it appears to shape the timing and quality of lactation itself, with consequences that can reach the newborn’s first days of feeding and weight change.
In a new JCI commentary on the work of Meng and colleagues, the central message is direct: insulin sensitivity and mitochondrial homeostasis are linked to the onset of human milk secretion, and impaired insulin signaling may slow secretory activation in women who are already metabolically stressed.
The physiology matters because lactation is not a passive event. When the placenta is delivered, mammary epithelial cells abruptly shift into a high-output state, increasing gene expression, synthetic capacity, and energy demand. If that transition is delayed, milk volume rises later, infant weight loss can deepen in the first postpartum days, and breastfeeding may become harder to sustain.
Human Milk Data Point to a Metabolic Bottleneck
Meng and colleagues studied 75 breastfeeding women with healthy term infants during postpartum days 1 through 5. Participants were separated into insulin resistance tertiles using the plasma leptin-to-adiponectin ratio, a surrogate that served as the metabolic classifier during late pregnancy. Women in the highest tertile reached lactation onset later and their neonates lost more weight during postpartum days 1 through 5.
A single sentence captures the bedside consequence: delayed secretory activation was not an abstract biomarker shift, but a feeding problem with immediate neonatal implications.
The human transcriptomic data sharpened the picture. On postpartum day 2, milk fat globule analysis from a small mechanistic subset showed transient suppression of mammary insulin and prolactin signaling genes, increased pro-inflammatory gene expression, and altered expression of more than 200 mitochondrial genes in women with high insulin resistance compared with those with low insulin resistance. Those differences were not persistent across postpartum days 3 to 5, suggesting a narrow biologic window in which the mammary gland is most vulnerable to metabolic interference.
That temporal pattern is clinically important. Lactation failure does not require permanent gland injury; it may arise from a short-lived failure to fully engage the energetic machinery needed for the secretory transition.
Mitochondrial Stress Emerges as the Mechanistic Link
The mechanistic studies move the work beyond association. In human and rodent mammary epithelial cell models, insulin activated the AKT pathway and upregulated genes tied to prolactin signaling and oxidative phosphorylation. Imaging and bioenergetic assays showed that insulin promoted mitochondrial biogenesis and supported oxidative metabolism, while high insulin exposure in mouse mammary epithelial cells produced mitochondrial fragmentation and loss of mitochondrial volume.
This is the kind of biology that explains delayed milk onset without requiring structural gland failure. Lactation is energetically expensive, and mitochondria are the organelles that convert substrate into the ATP needed for biosynthesis, membrane trafficking, and sustained milk production. When mitochondrial architecture becomes fragmented and oxidative phosphorylation is altered, the mammary epithelial cell loses part of the energetic flexibility required for the rapid shift into milk secretion.
The commentary also makes clear that insulin action in the breast is not reducible to glucose handling alone. Through AKT signaling, insulin appears to support the broader lactogenic program, including prolactin-related transcriptional activity and oxidative metabolism. The metabolic state of the mother is therefore not just background context; it is part of the regulatory environment that governs whether the mammary gland fully commits to lactation on schedule.
Why the Findings Matter for Mothers and Infants
Insulin resistance has already been associated with lactation insufficiency, obesity, gestational diabetes, and delayed initiation of breastfeeding. The new data strengthen that clinical association by identifying a plausible tissue-level mechanism. For women with higher leptin-to-adiponectin ratios at the end of pregnancy, delayed onset of lactogenesis was accompanied by lower lactose and higher milk sodium, a pattern consistent with incomplete secretory activation.
The implications extend beyond milk volume. Inadequate early milk transfer can worsen neonatal weight loss, create pressure for formula supplementation, and contribute to early cessation of breastfeeding. For mothers, repeated feeding difficulty can intensify emotional strain at a time already marked by sleep deprivation, recovery from delivery, and adjustment to newborn care.
The work also helps explain why prior intervention studies have been mixed. Metformin improved milk production in only 20% of women with insulin resistance and low milk supply, a result that argues against a simple one-drug, one-pathway model. Insulin resistance may be only part of the story. Mammary development, epithelial differentiation, inflammatory signaling, and mitochondrial fitness may all intersect during the transition into lactation.
Strengths in the Experimental Design, and the Remaining Gaps
One strength of the study is the use of milk fat globules as a transcriptomic window into mammary epithelial biology. That approach is practical in postpartum women and allowed investigators to connect human milk composition with gene-expression signatures that could then be tested in cell systems.
The model, however, is not the whole gland. Milk fat globule RNA reflects the mammary epithelium, but the lactating breast contains multiple interacting cell types, and the transcriptome from milk is not yet a complete substitute for tissue-level analysis. The commentary also points out that the MCF-12A experiments were informative but not performed under differentiated, lactogenic conditions, which would more closely resemble the physiologic state of interest. In addition, the lactogenic relevance of the high-insulin mouse cell findings still needs confirmation in primary human tissue and in vivo systems.
Those are not small experimental details. They define how far the field can safely generalize from a mechanistic model to a postpartum patient at the bedside.
A More Specific View of Low Milk Supply
Low milk supply is often treated as a single clinical problem, yet it likely includes several biologically distinct entities. Insufficient glandular tissue reflects developmental limitation, while low milk supply can also arise from maternal obesity, advanced maternal age, gestational diabetes, hypertension, thyroid dysfunction, hyperandrogenism, or other metabolic and anatomic factors. The commentary argues, through the lens of these new findings, that insulin resistance may sit at the center of at least one subset of lactation failure.
That specificity matters for future care. If delayed secretory activation is driven in part by insulin-linked mitochondrial dysfunction, then risk stratification during pregnancy could become more biologically grounded, and postpartum support could be targeted earlier to women most likely to struggle with milk onset.
The work leaves one practical question unanswered: which component of insulin resistance is most actionable in pregnancy and early postpartum, and which mothers will benefit from metabolic intervention before lactation begins to fail? The current data do not resolve that question, but they do narrow the biology. In women at the highest insulin-resistance tertile, the mammary gland appears to miss part of the energetic switch needed to launch lactation on time.