Sleep associations feature in many baby-sleep recommendations as a tool and a potential problem.

There are 2 main assertions made:

1. Sleep associations can be used to get your baby to sleep.
2. Once an association is formed, your baby will struggle to sleep without it.

To listen to a combination of the current post and this post (on an alternative take on sleep associations) please download my second podcast episode:

Sleep associations from a psychology perspective

Sleep associations

When something precedes something else more often than not, these two things become associated. Once associated, these triggers can elicit the same response. The well-known example is Pavlov’s dog that would salivate at the sound of a bell even when no food was present. The noise of the bell had become associated with food. This is called classical conditioning (Mazur 2006).

In baby sleep this means that if the sensation of falling asleep is usually preceded by something, these two things can become associated. For example, if the sensation of falling asleep is usually preceded by breast or bottle feeding, the sleep sensation can become associated with sucking or feeding.

What does this mean in practice? Let’s apply classical conditioning principles to the assertions:

Once an association is formed, sleep associations can be used to get your baby to sleep.

If something is associated with the sensation of falling asleep, it should elicit this sensation, or at least remind the baby of it BUT…

  • If baby also feels discomfort or insecure, these feelings are likely to prevent sleep
  • Conditions for a strong association are quite specific. Sleep associations between sleep cues such as a song or certain words, and sleepiness will be strongest when sleep cues are…
    • frequently followed by sleepiness
    • rarely not followed by sleepiness
    • used immediately before the onset of sleepiness
  • Even if there is a strong association between sleep cues and sleepiness, this association will weaken each time sleep cues are not followed by sleepiness

So in practice…

    • Sleep cues could be established by using them when you are fairly confident that baby will soon fall asleep.
    • To maintain their effectiveness, you need to keep using them when fairly confident baby will fall asleep, not only when you WANT baby to fall asleep.
    • Look after comfort and security to maximise likelihood of sleep.

Once an association is formed, your baby will struggle to sleep without it.

This is like asserting that Pavlov’s dog would only salivate to the sound of the bell, not to the food itself. Sleep associations may facilitate sleep but their absence should not prevent sleep. However, there are reasons why sleep associations may form more naturally, and be stronger, between falling asleep and breast feeding or rocking, than between falling asleep and lying alone in a cot.

  • Breast milk contains nucleotides that facilitate sleep, especially at night (New Scientist 2009Sanchez et al 2009)
  • Feeding and holding meet comfort needs including warmth and security

Feeding and holding are likely to immediately precede sleep in many cases because these comfort needs are met. In contrast, being placed in a cot is likely to arouse baby in many cases because there is a relative reduction in warmth and security.

So in practice…

Don’t worry about breaking associations that have formed. Focus on ensuring comfort and security needs are met in the preferred sleeping arrangement.

Strong associations between feeding and sleepiness do not mean that sleepiness is impossible without feeding. However, babies might ask for a feed because they need to sleep, and they associate milk with sleep. This may be an opportunity to establish sleep cues, if a feed is likely to result in sleep, this is a good time to use sleep cues, such as words or music. Alternatively, the baby might accept sucking on a dummy or a thumb rather than a feed if these sensations are similar enough.

These extrapolations are based on classical conditioning principles. Watch this blog for potential implications of operant conditioning. As an aside, I find it interesting that I am drawing on behavioural psychology. Throughout my study in psychology I have considered behavioural psychology principles to be overly simplistic because they ignore the influence of interpretation, thought, and self determination. However, I now accept that associations formed and maintained as per classical and operant conditioning do play an important role in behaviour and thought, especially for children who have not yet developed language.

For an alternative take on sleep associations, see the following posts:


Geddes, L. (2009, 2nd October). Evening breast milk means a good sleep. New Scientist.

Mazur, J. E. (2006). Learning and behavior. New Jersey: Pearson.

Sánchez, C. L., Cubero, J., Sánchez, J., Chanclón, B., Rivero, M., Rodríguez, A. B., et al. (2009). The possible role of human milk nucleotides as sleep inducers. Nutritional Neuroscience, 12(1), 2-8.