salmon logo
Maternal Behaviour
Author Paul Kenyon

Visit the SALMON Bookshop for recommended books on this topic
Hormones and maternal behaviour Sensory factors in maternal behaviour

This lecture may not be of immediate personal concern to some of you, but hopefully you will all at some time experience the joys of parenthood. This is how my personal interest in the topic arose.

When my son was born I began to wonder how the interaction between mother and child was controlled. And this set off a research programme concerned with factors that control the defining mammalian behavioural pattern - suckling.

The lecture begins with a look at the psychological changes that occur during human pregnancy. I draw your attention to this material because it seems to me that this area provides a rich resource of research questions that could be tackled in a third year project.

But in order to understand the biology of parenting we need to resort to an animal model, and once again the humble lab rat comes to the rescue. In fact, maternal behaviour in the rat is remarkably robust and reliable. By this I mean that it is really very easy to observe these behaviours under laboratory conditions. We look at two behaviours, retrieving and suckling, that can be elicited very easily. How unlike some of the behaviours that we try to study in the lab!

We then move on to ask questions about the role of hormones in the initiation of maternal behaviours. Three candidates - estrogen, progesterone and prolactin - have received extensive investigation by Jay Rosenblatt and his co-workers in an elegant series of experiments. Although these studies suggest that maternal behaviour is facilitated by a decline in progesterone, and an increase in estrogen levels, they do not account for the rapid onset of maternal behaviour immediately after parturition.

Recent work with central injection of oxytocin is described to illustrate how this research field is still revealing answers.

Finally we will look at factors that control the maintenance of maternal behaviours in order to answer the questions "Do pups maintain maternal behavior by promoting the release of hormones in their mother?" or "Do pups simply provide a particular type of sensory stimulation?"

Learning objectives
After studying the material on this page you should be able to:
  • Describe the changes in maternal engrossment and self-image that occur during pregnancy
  • Describe a woman's feelings towards themselves, their partner, and their baby at various stages of pregnancy and after the birth of their child.
  • Describe the change in olfactory recognition that occurs after birth
  • List five maternal behaviours exhibited by rodents
  • Recognise the terms: parturition, prepartum, and postpartum
  • Describe the frequency of maternal behaviour in the rat before, and after, parturition.
  • Describe the changes in progesterone, estrogen, and prolactin during pregnancy in the rat
  • Explain the terms hysterectomy and ovariectomy and recognise the organs that are removed by these operations
  • Describe the effects of hysterectomy during pregnancy on progesterone level and maternal behaviour in the rat
  • Explain how, and describe the underlying rationale of, Rosenblatt's studies on the role of estrogen in the initiation of maternal behaviour
  • Describe how the balance between estrogen and progesterone may affect the initiation of maternal behaviour
  • Evaluate the extent to which we have a complete understanding of the role of hormonal factors in the initiation of maternal behaviour
  • Describe the evidence which suggests that non-hormonal factors are responsible for maintaining maternal behaviour after birth
Spring/Summer 2007

Human maternal behaviour

It is perhaps not surprising that women become increasingly focussed on the process of motherhood (maternal engrossment) as pregnancy advances, and after the baby is born. However, towards the end of pregnancy there may be a slight dip in the woman's self-image (Fleming et al, 1990).

Maternal engrossment across pregnancy and early postpartum

Maternal self-image across pregnancy and early postpartum

Maternal bonding
eye2eye-contact-monkey2.jpg (4323 bytes)

eye2eye-contact.jpg (5276 bytes)

"Over the centuries, the attachment between a mother and her newborn infant has been considered a natural process and it is expected of the mother to care for and nurture her infant. However, the process by which a mother and her baby develop a close relationship is not clearly understood. It is known that maternal behavior is influenced by multiple factors that include the mother's own upbringing, socioeconomic conditions, her cultural beliefs and background, her relationship with her father, as well as her experiences with present and past pregnancies. ... eye-to-eye contact may be an important factor in enhancing maternal behavior " (Sosa, 1980).

Eye to eye contact may qualify as a  universal or fixed action pattern.

Child abuse

"..given the frequent news reports of women failing to care for their newborn infants, i.e. a lack of the onset of maternal behavior, this is not a health problem of trivial significance. Certainly there are cultural and social influences involved in these tragic instances of infanticide, but there are also likely to be biological ones as well." McCarthy (1997).

One of the most striking aspects of maternal behaviour is how much we take it for granted. We assume that mothers will look after their children and society expresses outrage when things go wrong. A breakdown of normal maternal behaviour is so unusual that it makes headline news as in these two cases.

In August 1999 a young mother threw her three-year-old son into the back of a dustcart because she thought he was "rubbish". The woman was suffering from paranoid schizophrenia and thought all her problems would go away if her child died. Fortunately the lorry's machinery was switched off and dustmen rescued the boy before he could come to any harm. Obviously this tragic case can be expained on the basis of the mother's incapacitating illness, but sometimes maternal care crumbles in the face of apparently normal circumstances.

Samantha Watts came from a respectable family and had a full-time job, but she concealed her pregnancy from family and friends. She allowed her 11-month-old baby to starve to death while her cupboards were full of food. Watts was jailed for life in September 2000.

These cases illustrate that we tend to take 'mother love' for granted. But a moment's reflection shows that if we do not understand why maternal behaviour sometimes breaks down, then we do not understand how it works in normal circumstances. I call this an example of 'the riddle of the ordinary'.

Initiation of maternal behaviour at parturition

Maternal behaviours

For obvious reasons, most of our knowledge about the biological bases of maternal behaviour has come from studies of animals. For example, it would be unethical to artificially manipulate hormone levels in women in order to investigate the role of hormones in their behaviour. Nevertheless it is possible that the principles governing such an important behaviour that emerge from studies of animals may be echoed in the behaviour of our own species.

This section of the lecture will outline some of the hormonal factors that regulate maternal behavior in the rat.

Studies of maternal behaviour in rats have been concerned with

  • understanding how the behaviour is 'switched on' or initiated at the end of pregnancy and
  • how the behaviour is maintained during lactation.

Maternal behavior is thought to be influenced by hormonal conditions that exist during pregnancy but sensory stimuli provided by the offspring are also important in initiation and maintenance. The idea that maternal behavior is "switched on" and maintained by hormones and sensory factors and maintained by sensory stimulation is central to this lecture.

Maternal behaviour is an attractive behaviour to study because it consists of a number of clear-cut behaviours
that are easy to observe and record under laboratory conditions.

Maternal behavior in rats and mice consists of the following elements:

    • retrieving
    • nest building
    • nursing
    • licking pups
    • maternal aggression

In this photograph the dam is shown on her back suckling pups. Normally mothers crouch over pups in the 'nursing posture'. This picture shows an anaesthetized mother, and was taken during a research programme into the role of sensory cues in suckling behaviour .

female rat carrying one of her pups  
female rat suckling pups

Onset of maternal behaviour

Maternal behaviour before and after parturition

The diagram shows the emergence of three maternal behaviours in the rat before, and after, parturition.

Notice that maternal behaviour is virtually absent during pregnancy but appear suddenly at parturition.

What is responsible for the rapid initiation of maternal behaviour postpartum?

One possibility - quickly disproved - was that the birth of the pups (through the birth canal) was responsible for the onset of maternal behaviour . In fact, delivering pups via Caesarean section does not disrupt the rapid onset of maternal behaviour.

Explanation of terms:

  • parturition = birth
  • prepartum = the period before birth
  • postpartum = the period after birth
Here is an animated version of this diagram

Hormone levels during pregnancy

The observation that blood transfusions from maternal rats induced maternal behaviour in nonpregnant females, suggested that hormones may play an important role in switching on (initiating) maternal behavior. The diagram illustrates the dramatic changes in hormone levels around parturition in rats. Similar changes occur during human pregnancy.

Notice that::

  • progesterone declines during the last days of pregnancy
  • estrogen increases just before birth
  • prolactin increases just before birth

Is this cluster of hormonal changes responsible for switching on maternal behavior ? It is difficult to study the role of hormones in natural pregnancy because if you inject or remove hormones during pregnancy there are multiple effects on lactation and behavior that are difficult to disentangle.



Hormone levels during pregnancy
Here is an animated version of this diagram

Hormones and maternal behaviour

jay-rosenblatt.gif (11971 bytes)Jay Rosenblatt examined the hypothesis that the decline in progesterone level seen towards the end of pregnancy triggers the onset of maternal behaviour. In order to produce a decline in progesterone level, pregnant rats were hysterectomised.

Hysterectomy and ovariectomy in the rat: what is involved?

Before we can explore the effects of this operation on maternal behaviour here are some diagrams that help to explain what is removed in the hysterectomy and ovariectomy operations that were used by Rosenblatt in these studies.

This diagram shows the relative position in the rat of the:
  • uterus
  • ovaries
  • foetus (unborn pup)

In mammals the fertilied egg passes through several stages before birth:

  • in humans: embryo, foetus, baby
  • in rats: embryo, foetus, pup

Rats normally give birth to large families of pups! A pregnant uterus can contain 16 or more unborn pups.


Examine the effects of hysterectomy and ovariectomy

Hysterectomy is the name to given to the surgical proceedure which involves removal of the uterus.

Ovariectomy is the name to given to the surgical proceedure which involves removal of the ovaries.

Click these buttons to discover what structures are removed in:

Removed structures are shown as a 'ghost' outline on the photograph. Note that this computer-generated image contains suckling pups - the operations referred to in this section are performed before parturition

Maternal behaviour after hysterectomy

Hysterectomy involves removing the uterus which contains the pups and causes progesterone level to fall.

Rosenblatt reasoned that if a decline in progesterone is responsible for the initiation of maternal behaviours at the end of a normal pregnancy, then we would expect to see the onset of these behaviours in hysterectomised rats as progesterone declines after the operation .

The diagrams show the effect of hysterectomy on

Progesterone levels during pregnancy & after hysterectomy

Progesterone levels during pregnancy and after hysterectomy


This diagram shows falling progesterone level in intact pregnant rats, and rats that were hysterectomised (H) on day 10, 13, 16 or 19 of their pregnancy

Maternal behaviour after hysterectomy

Maternal behaviour after hysterectomy


This diagram shows the cumulative percentage of females showing maternal behaviour following hysterectomy (H) on day 10, 13, 16 or 19 of pregnancy . The first test (0) of maternal behaviour was given 48 hours after surgery, and testing continued for a further 5 days.

Here is an animated version of this diagram Here is an animated version of this diagram
Results: At first the mother rat pays no attention to the pups, but with time she begins to show maternal behavior . Hysterectomy during pregnancy produces a decline in progesterone and stimulates the onset of maternal behavior .The later in pregnancy the rat is hysterectomised, the greater this effect is.

Notice how hysterectomy and parturition are both associated with a decline in progesterone and an increase in maternal behavior .

The next question is: 'Are these changes causally linked?' Does the decline in progesterone trigger the onset of maternal behavior, or does the effect depend on some other effect of the operation?

Maternal behaviour after hysterectomy + ovariectomy

In further experiments, Rosenblatt investigated the possibility that changes in estrogen level were responsible for the emergence of maternal behaviour after hysterectomy. In his first series of studies he removed the ovaries (ovariectomy) from hysterectomised pregnant rats at various stages of pregnancy. You can see the results in the slide below.

Maternal behaviour after H and HO

This diagram shows the median latencies to the onset of maternal behaviour in pregnant rats that were hysterectomised (H) or hysterectomised + ovariectomised (HO) on days 10, 13, 16 or 19 of their pregnancy. The first test of maternal behaviour was given at 48 hours after surgery.


At each stage of pregnancy, ovariectomy reduces the effect of hysterectomy. Rats that are given the double operation (ovariectomy + hysterectomy) take longer to develop maternal behaviours than animals only given an hysterectomy.

This result suggests that it is the presence of high estrogen levels against a background of low progesterone levels that stimulates maternal behaviour at the end of a normal pregnancy. Rosenblatt went on to check this explanation by replacing estrogen in rats that had been ovariectomised by injecting them with estrogen.

Effect of estrogen on maternal behaviour after hysterectomy + ovariectomy

In this experiment the behaviour of three groups of prgnant rats was comapred. These rats were either
  • hysterectomised - progesterone declines, normal estrogen level
  • hysterectomised + ovariectomised - progesterone declines, source of estrogen removed
  • hysterectomised + ovariectomised + estrogen injection - progesterone declines, source of estrogen removed and then replaced with estrogen injection
The diagram shows the cumulative percentage of 10 day pregnant rats exhibiting maternal behaviour after hysterectomy or hysterectome-ovariectomy plus estradiol benzoate treatment (100 or 200 microgram/kg). Maternal behaviour tests began 48 hours after surgery and hormone injection.


  • Ovariectomised rats take longer to exhibit maternal behaviour than hysterectomised rats.
  • Injection of estrogen overcomes the effect of ovariectomy and restores the rapid onset of maternal behaviour seen in hysterectomised rats.

This result suggests that estrogen stimulates maternal behavior when progesterone levels are low. Recall that estrogen rises and progesterone declines at the end of a normal pregnancy.

However estrogen stimulates prolactin release so these effects may involve prolactin which increases dramatically around birth. This explanation is thought unlikely because blocking the release of prolactin with ergocornine or apomorphine does not interfere with the ability of estrogen to stimulate maternal behaviour. Consequently, the initiation of maternal behavior is probably caused by estrogen.

Summary table:Effect of hormones on maternal behaviour

Treatment Effect on hormones Effect on maternal behaviours
Remove uterus (hysterectomy) Progesterone down

Estradiol up

Remove uterus & ovaries (hysterectomy+ovariectomy) Progesterone down

Estradiol  down

No facilitation
Administer estradiol & ergocornine (blocks prolactin) Estradiol up

Prolactin down


Although these experiments show that it is possible to facilitate the onset of maternal behaviour by manipulating progesterone and estrogen levels it is clear that none of these manipulations on their own are able to reproduce the rapid onset of maternal behaviour seen under normal conditions.

Oxytocin elicits rapid onset of maternal behaviour

In labour, minor contractions of the uterus may begin spontaneously because of the reversal of progesterone dominance. These contractions may serve as the stimulus for oxytocin secretion, which then augments and intensifies uterine contractions and leads to delivery.

Oxytocin is normally seen as primarily involved in the milk letdown reflex and in the stimulation of uterine smooth muscles during parturition.  Pedersen et al,(1982) claimed that injection of oxytocin into the brain can lead to a rapid onset of maternal behaviour.

This diagram is a dose response plot of the percentage of virgin female rats displaying full maternal behaviour within 1 hour after receiving an intracerebroventricular (injection into the ventricles of the brain) injection of oxytocin given 48 hours after ovariectomy and priming with estrogen.

These results are potentially very important because they are the first indication that a chemical factor could be responsible for the rapid initiation of maternal behaviours seen shortly after parturition.

However the interpretation of Pedersen's data has been challenged by Wambolt & Insel (1987).

McCarthy (1997) reviews a series of experiments  by Fahrbach et al (1984) that resolved this controversy.

Sensory factors in maternal behaviour

Maternal behaviour declines after pup removal

In the rat, maternal behaviour continues for about 28 days until the pups are weaned onto solid food, but it declines rapidly if pups are removed at parturition. This prompts several questions:

There is not much evidence that hormones are involved in the maintenance of maternal behaviours:

Therefore the consensus of opinion holds that pup stimulation may maintain maternal behavior postpartum .

Pup contact establishes maternal behaviour

There appears to be a transitional period just after birth during which a bond is formed between rat pups and their mother. This occurs in the first few hours or minutes postpartum.This effect is reminiscent of human mothers who learn to distinguish the odour of their child after just a few hours exposure.

The diagram shows the results of several experiments in which the amount of contact between the mother rat and her pups was varied.

If the mother's pups are delivered by Caesarean section so that she has no opportunity to interact with them then her maternal behaviour takes a long time to develop when she is given foster pups to look after 25 days after her pups were delivered artificially.

Similarily if the mother is allowed to deliver her litter of pups normally, but the pups are immediately removed so that she has no contact with them she will take a long time to exhibit maternal behaviour to foster pups when these are given to her 25 days after the birth of her own litter.

But if dams are allowed contact with pups during parturition,then the pups are removed and test pups are given to her 25 days later she will behave maternally towards these foster pups within two days.

The figure also shows another surprising result. Virgin females will exhibit maternal behaviour towards pups, but only after they have been exposed to them for several days.

This effect has been known since the 1930s - it is called sensitization This process is thought to involve virgins coming to terms with the odour of pups which they initially find repulsive, and maternal behaviour can be elicited from male rats as well.

Online resources


Supplementary material

Davis (2002) WebMD Feature Stress and the sexes

Shelley E. Taylor, PhD, a psychology professor at UCLA and author of The Tending Instinct suggests that new research -- drawing on psychology, genetics, evolutionary biology, and neuroscience -- shows that there are distinct differences in how men and women react to stressors or aggressors. While men will fight -- or simply hide -- women have a stronger instinct to "tend and befriend".

A woman is biologically hard-wired to nurture, provide comfort, and seek social support in times of stress, Taylor writes. Our hormones, brain chemistry, and response to the world around us all reflect this natural instinct. Men have this instinct too, but to a lesser degree because of hormone differences and personal choices, she says. more ...

Dissanayake (2002) "The Poetics of Baby Talk"  Presented at Darwinian Feminism. Special Session, South Central Modern Language Association Meeting, Austin, Texas., October 31-November 2, 2002

Dissanayake suggests that suggests that humans are born with natural (innate, universal) predispositions to interact as ,and with, babies.

Gammie, January 4, 2001 Maternal Aggression in Mice. Requires RealAudio and RealVideo

Scientists studying the origins of aggression have highlighted areas in the brains of mouse mothers that may generate fierce attacks on males who pose a potential threat to their pups. Postdoctoral researcher Stephen Gammie recently explained the research and findings in a brief video. View the video. Requires RealAudio and RealVideo

Varendi (2001) Human newborn behavior during exposure to maternal and other odors. Published by The Karolinska Institutet Dissertation Database.

From the abstract: "Olfaction appears to play a pivotal role in immediate postnatal interactions between mammalian mothers and their offspring. The present study examined:

  1. whether and how the newborn human infant and its mother communicate by means of olfaction;
  2. the role of amniotic fluid (AF) and breast odors in such communication;
  3. olfactory learning by neonates.

...the newborn and its mother communicate by means of olfaction. Several common hospital care routines such as washing the baby immediately after birth and the mother's breasts before feeding interfere with this communication and should be avoided until we know more about its clinical significance. "

Mayo Clinic in Rochester( 2001). Changes in Hormone Levels in Men Who Become Fathers

A published study of hormonal changes in a group of Canadian men becoming fathers for the first time showed a decrease in testosterone and cortisol levels and a higher level of estradiol concentrations, a hormone known to influence maternal behavior. more ...

Beyond "the Blues": Postpartum Reactivity and the Biology of Attachment

Most literature on postpartum psychiatric disorders accepts the premise that there are three general categories of postpartum disturbance, along a spectrum of increasing severity: "blues," depression, and psychosis.1-3 The most benign and most common condition is postpartum blues, experienced by approximately 50% of women who give birth. The blues are usually characterized as a mild, self-limited form of depression. There is a growing body of evidence to suggest that the term postpartum blues is a misnomer. This article reviews data supporting the idea that the blues are experientially and biologically distinct from depression . Rather, they are a normal state of psychophysiologic reactivity that facilitates attachment and maternal behavior. Reconceptualizing the blues this way should change the focus of research from variables relevant to depression to variables relevant to attachment. It may also foster a greater clinical understanding of attachment and disorders of attachment.

Impact on babies of mothers ingesting 'crack' cocaine

They are called "a biological underclass" and "a lost generation." Those are just two of the milder name tags attached to the children we have come to believe were permanently damaged by their mothers' use of cocaine. The poster in maternity clinics conjure up the same image of the prenatally doomed: "Some people who smoke crack never get over it." The schools too have been put on emergency alert: "The crack babies are coming, the crack babies are coming." Indeed, the phrases "crack babies" and "crack kids" are shorthand for monster-children who are born addicted. These are the kids destined to grow up without the ability to pay attention or to learn or to love. But just when the name has stuck, it turns out that "crack baby" may be a creature of the imagination as much as medicine, a syndrome seen in the media more often than medicine.
Here is another link to similar material about the controversy surrounding the impact of cocaine on the unborn

A study of rat brains may offer hope to women who reject their babies. Published in New Scientist 13/12/97
"Mothers who feel no attachment to their newborn babies may lack a neural pathway that is normally triggered by the act of giving birth, a new study of mice suggests. The maternal skills of female mice genetically predisposed to neglect their young improve for good if they receive a burst of key neural chemicals shortly before giving birth, say researchers in Seattle. And while the factors that cause some women to reject their babies at birth remain unclear, the study hints that a carefully timed dose of drugs might, in some cases, free them from this misery. " more .....
Thanks to Bernadette Melmore (year 2, 1997) for pointing out this article.

The Daily Mail (30th April, 1998) announced it as a "PMT Breakthrough" and explained that "Pre-menstrual syndrome and postnatal depression are closely related to the 'cold turkey' experienced by addicts trying to give up drugs ..". Here is the description of the research given in Nature.

GABA receptor a4 subunit suppression prevents withdrawal properties of an endogenous steroid. Nature 392, 926 (1998)

"The hormone progesterone is readily converted to 3a-OH-5a-pregnan-20-one (3a,5a-THP) in the brains of males and females. In the brain, 3a,5a-THP acts like a sedative, decreasing anxiety and reducing seizure activity, by enhancing the function of GABA (g-aminobutyric acid), the brains major inhibitory neurotransmitter. Symptoms of premenstrual syndrome (PMS), such as anxiety and seizure, susceptibility, are associated with sharp declines in circulating levels of progesterone and, consequently, of levels of 3a,5a-THP in the brain. Abrupt discontinuation of use of sedatives such as benzodiazepines and ethanol can also produce PMS-like withdrawal symptoms. Here the authors report a progesterone-withdrawal paradigm, designed to mimic PMS and post-partum syndrome in a rat model. In this model, withdrawal of progesterone leads to increased seizure susceptibility and insensitivity to benzodiazepine sedatives through an effect on gene transcription. Specifically, this effect was due to reduced levels of 3a,5a-THP which enhance transcription of the gene encoding the a4 subunit of the GABAA receptor. The authors also find that increased susceptibility to seizure after progesterone withdrawal is due to a sixfold decrease in the decay time for GABA currents and consequent decreased inhibitory function. Blockade of the 4 gene transcript prevents these withdrawal properties. PMS symptoms may therefore be attributable, in part, to alterations in expression of GABAA receptor subunits as a result of progesterone withdrawal. "

Reference: S S Smith, Q H Gong, F-C Hsu, R S Markowitz, J M H ffrench-Mullen & X Li, GABAA receptor a4 subunit suppression prevents withdrawal properties of an endogenous steroid. (Letter to Nature) Nature 392, 926 (1998)

Thanks to Bernadette Millmore (Yr2) for bringing my attention to this research.

Copyright Dr. C.A.P. Kenyon 1994-2006