Take It Easy: Neonatal Milk Hormones Influence Infant Social and Cognitive Behavior

  • Research in rats and primates demonstrates that cortisol passed to infants from mother’s milk influences infant temperament.
  • A new study on rhesus macaques demonstrates for the first time that the concentration of milk cortisol during the first weeks of life predicts cognitive performance in later infancy.
  • Cortisol is a natural ingredient in milk whose concentration is unrelated to the mother’s stress level.


Email, texts, IM, Facebook, Instagram—in the age of social media, there is no shortage of ways to send a message from one person to another. But is mother’s milk the original social network? Many of milk’s ingredients are believed to act as signaling factors that convey a “message” from mother to infant. Over the last decade, researchers have worked on decoding these messages, with a particular focus on the hormone cortisol. Milk cortisol levels are associated with infant growth and infant temperament in rhesus macaques, and hypothesized to send the message to be more cautious and prioritize growth over behavioral activity [1-3]. A newly published study [4] expands on this hypothesis and tests whether milk cortisol levels during the first weeks of life predict behavior and cognitive performance months later. The results suggest that far from being an instant messenger, milk’s signal may have effects well after it is received.

You’ve Got Mail

Voles are small rodents similar in size and appearance to mice, and provide one of the most clear-cut examples of the maternal social network, better known as developmental programming [5]. Montane voles that live in the Western United States and Canada produce offspring mainly during the spring and summer months. Infants born in early spring grow rapidly and reach age at first reproduction quickly, whereas those born in late summer stop growing when reaching a particular body weight and delay reproduction until the following spring. These growth patterns are adaptive; offspring born in late summer are facing the approach of shorter, colder days when food sources will diminish and reproduction is unlikely to be successful. But without a calendar, how do the offspring know when they are born? In a series of elegant experiments, Horton [5] demonstrated that mothers were signaling photoperiod information, via the hormone melatonin, to their fetuses, which then directed postnatal growth and development. This groundbreaking research demonstrated that mothers can send messages to their offspring during gestation (“days will be getting shorter as you grow and develop”) and that fetal physiology is primed to receive these messages [5].

But the social networking need not stop after the offspring are born. The evolution of lactation among mammals, and particularly the extended period of lactation in primates, presents an additional opportunity for mothers to provide information to their offspring that may increase survival and reproductive success.

Maternal physiological signaling in the postnatal period is referred to as lactational programming [1] (or as lactocrine programming [6] when discussing hormonal signals), and has been demonstrated in several mammal species. A specific signaling factor of interest is the stress hormone cortisol. Cortisol is present in milk, and infants have cortisol receptor genes in their intestinal tracts, setting up a perfect scenario for sending and receiving a signal. But what might mammal mothers communicate through milk cortisol levels?

In rats, rhesus macaques, and humans, milk cortisol is associated with infant temperament, but the message sent by mothers differs among species [1-3]. In rats, higher cortisol intake via milk was associated with more exploratory behavior and less anxiety during adulthood, whereas among rhesus macaques and humans, higher milk cortisol was associated with a more fearful or less confident temperament during infancy [1-3].

Differences between the meaning of the message sent to rats and that sent to primates may relate to differences in life history. Compared to rats, primate infants have a protracted period of growth (even when scaled for body size); they rely on milk for energy (both for growth and for behavior) for a much longer period of time than do rats. Information about how much energy may be available for the entirety of the lactation period would allow primate infants to calibrate their growth rates to maternal reserves [1], just as developing voles calibrate their growth using maternal signals on birth season [5].

Not so Instant Messenger

In rhesus macaques, Hinde and colleagues [1] propose that milk cortisol levels send a message from mothers to infants about how to allocate energy received in milk. Specifically, they argue that rhesus macaque mothers that may be limited in their resources (e.g., first time mothers that are still growing) would improve the chances of their offspring surviving to reproductive age by sending the message to “take it easy”be cautious when it comes to behavior and put any available energy from milk into growth [1].

Hinde et al.’s hypothesis for energetically conservative infants is based on milk collection at 1 month and 3.5 months of age and infant behavioral assessments at 3.5 months of age. But if the message that mothers are sending is meant to convey an adaptive strategy throughout the lactation period, this message is predicted to be persistent. To this end, a new study by Dettmer et al. [4] investigated whether cortisol levels in rhesus macaque milk during the first weeks of lactation were predictive of behavior and cognitive performance between 4 and 8 months of age. At this phase of development, rhesus macaque infants are increasing their independence from their mothers while also developing neurological changes associated with complex social behavior [4].

Contradictory to Hinde et al.’s [1] findings, Dettmer et al. [4] found the higher milk cortisol levels were not associated with overall social behavior patterns. Indeed, one of the only associations noted for a particular behavior was more play and engagement in social behaviors, between 4 and 8 months of age, in daughters (but not sons) that received higher levels of milk cortisol.

Conflicting findings between these two studies [1, 2] on the association between milk cortisol and social behavior in infant rhesus macaques need not mean a return to the hypothesis starting board. Each study used a different method for assessing behavioral traits such as cautious, nervous, and confident. Hinde et al. [1] used a bio-behavioral assessment that required separating infants from their mothers and social groups, whereas Dettmer et al. [4] observed the infants interacting within their social groups. Might a more stressful environment bring out more observable behavioral differences between those infants receiving high and low levels of milk cortisol? The studies also differed in the timing of milk collection, with Hinde et al. [1] considering the effects of milk both during early (1 month) lactation and peak lactation (3.5 months). It is possible that some behavioral traits may be more responsive to messages sent later in lactation or, as Hinde et al. found with sons, to changes in the message, such as an increase in milk cortisol over time [1]. Clearly, it is a complex system.

Case in point: the cognitive tests did support previous research on milk cortisol in rhesus macaques [1]. The test used by Dettmer et al. [4] was ideal for their hypothesis, as it looked at how quickly a monkey was able to learn a task, and also measured impulse control. A box with only one opening sat at the bottom of a plexiglass board and the monkeys were trained on over 100 trials to reach straight ahead and grab a food reward out of the box (with the box opening facing the monkey). Then, the experimenter moved the box so the opening was on the left or the right—to reach the reward, monkeys had to reach around the board rather than straight ahead. Researchers measured whether the monkeys reached impulsively straight ahead or to the correct side, as well as how long it took them to solve the task. Both higher milk yield (i.e., how much milk the mothers produced) and higher milk cortisol levels predicted a less impulsive test performance [4].

These results from Dettmer et al. [4] add to the growing body of literature that suggests a relationship between milk cortisol levels and infant phenotype. Moreover, this is the first study to demonstrate downstream cognitive effects due to natural variation in milk composition [4]. As predicted, the message sent by milk cortisol was persistent and not only relevant at the time it was sent. The small sample size in this study (34 infants, only 12 of which were cognitively assessed) limits the ability to speak about sex differences in response to milk cortisol, as observed by Hinde et al. [1], or to identify windows of sensitivity for milk signals. However, the results provide an excellent point of departure for designing future studies intent on capturing these complex relationships between milk signals and infant responses.

Don’t Stress Out

One of the most interesting findings of these research studies on milk cortisol composition and infant development has nothing to do with infants at all but rather their mothers. Cortisol is probably best known as the stress hormone. Human mothers may be concerned how their own response to stressors (from being a new mother, to living in a violent environment) may end up having a negative effect on their developing fetus or breastfeeding infant. But Dettmer et al. [4] found no relationship between milk cortisol and cortisol from maternal hair samples (which is associated with maternal cortisol levels). Indeed, Hinde et al. [1] found that maternal rank (a proxy for stress) was also not predictive of milk cortisol. Mothers are not passing their stress on to their offspring; higher milk cortisol is not necessarily associated with chronically stressed mothers. Rather, the association of both maternal experience (e.g., first-time mothers) and maternal body mass with milk cortisol supports the argument that cortisol signals for conserving energy and not necessarily for the development of anxiety. This should reduce the stress for breastfeeding mothers (or those who intend to do so in the future). Cortisol is supposed to be in milk. How much cortisol a mother passes on in milk may have more to do with communicating to infants about energy than stress.


1. Hinde K, Skibiel AL, Foster, AB, Del Rosso L, Mendoze SP, Capitanio JP. 2015. Cortisol in mother’s milk across lactation reflects maternal life history and predicts infant temperament. Behavioral Ecology 26: 269-281.
2. Hinde K. 2013. Hormones in mother’s milk influence baby’s behavior. SPLASH! Milk Science Update, February 2013. http://milkgenomics.org/article/hormones-in-mothers-milk-influence-babys-behavior/.
3. Hinde K. 2014. Stop, Slow and Go: hormonal signals from mother’s milk. SPLASH! Milk Science Update, December 2014. http://milkgenomics.org/article/stop-slow-go-hormonal-signals-mothers-milk/.
4. Dettmer AM, Murphy AM, Guitarra D, Slonecker E, Suomi SJ, Rosenberg KL, Novak MA, Meyer JS, Hinde K. 2017. Cortisol in neonatal mother’s milk predicts later infant social and cognitive functioning in rhesus monkeys. Child Development Mar 1 doi: 10.1111/cdev.12783
5. Horton TH. 2005. Fetal origins of developmental plasticity: animal models of induced life history variation. American Journal of Human Biology 17: 34-43.
6. Bartol FF, Wiley AA, Bagnell CA. 2008. Epigenetic programming of porcine endometrial function and the lactocrine hypothesis. Reproduction in Domestic Animals 43: 273-279.


Contributed by
Dr. Lauren Milligan Newmark
Research Associate
Smithsonian Institute