PosterPDF Available

“Crying it out:” A critical review of the literature on the use of extinction with infants.



NOTE: PLEASE SEE UPDATED RESEARCH POSTER ON THIS PAGE. This research is from 2006. I have updated this with a poster presented in 2022. ABSTRACT: The most empirically supported and endorsed approaches to preventing or ameliorating infant sleep problems involve various forms of extinction (crying it out). This approach is widely recommended by pediatricians and mainstream parenting publications for infants as young as 3- to 4-months. A review of the literature regarding the cry-it-out (CIO) approach to infant sleep reveals that empirical support for this early start date is lacking. Much of the existing literature used to validate the use of extinction with infants does not include infants under 1 year in the sample population. The subset of studies which include infants, do so as part of a much larger sample comprised of wide age ranges (e.g. 9-60 months). Results from these studies do not specify outcomes by age, and obscure the existence of differential effects for infants. Though CIO is recommended as the approach of choice for infant sleep, relatively little is known about its effects on infants under 1 year.
Rethinking CIO:
The family context.
An infant menta l health perspective on
CIO raises significant questions about the
potential repercussions of sleep training
techniq ues for in fants, parents and families.
Individu al Needs
How does CIO interface with individual
capacities (e.g. caregiver mental health,
existing family stressors, etc.)?
How does CIO interface with infants at
various stages of development? Are some
infants being left to cry to o long at very
early ages?
How does the implementa tion of CIO
affect parental self-concept or perceptions
of the infant? How might CIO affect an
infant’s em erging internal working model?
How do parental working models influence
the experience of impleme nting CIO?
Relationship Needs
Does CIO affect parentalre sponsivene ss
overall? Does CIO convey the notion
that unwanted infant behavior may be
controlled through the withdrawal of
What does CIO convey to parents about
the mea ning of crying?
Rethinking CIO:
Nightwaking as a symptom
rather than a problem.
A variety of relevant family variables have
been significantly related to infant night-
waking: breastfeeding35, attachment status36,
parental psychopathology37, maternal
depression38, family stress39, and relational
Parents who are anxious or depressed are
also more likely to re port their infant’s
waking as problematiceven when sleep
behavior falls into normal ran ges41.
Nightwaking is also more likely in infants
with temperamental or regulatory sensitivi-
ties beca use of th eir decreased capacity to
buffer stimuli42. It is currently unknown
how CIO interfaces with these infants w ho
may also have a decreased capacity to self-
calm or withdraw from distress43.
Rethinking CIO:
Sleep problems as a port of
entry for infant mental health.
In the rush to provide one-size-fits-all sleep
training in formatio n, professionals are
missing a valuable opportunity for inquiry
into relevan t variables that contribute to
or are impacted by nightwaking.
Future research into CIO and other infant
sleep interven tions needs to account for
the context within which nightwaking
takes place.
More research is needed to develop
interven tions that are context aware
helping infants achieve age-appropriate
amounts of sleep, while supporting
parents’ instincts to nurture andrespond .
Advice in Popular Parent ing Media:
“A baby must adapt tothe
existing family; theex isting
family does not adapt to ababy.”
12 Hours Sle ep by 12 Weeks
Giordan o & Abidin , 2006, p. 16
For young infants, prolonged unmodulated
crying is physiologically costly.
Crying increases heart rate and blood pressure, reduces blood
oxygenation, and propels the release of cortisol. Prolonged
crying amplifies these effects24.
Animal studies have shown that extinction itself
is particularly stressfulproducing rapid, dramatic, and
persistent increases in cortisol25.
Physiological stress levels may remain
high after the infant has stopped crying.
Animal studies of extinction found that, though
behavior appeared to adapt, physiology did not.
In one study, infant ra ts ceased calling for their
absent mothers; however, their physiological
arousal remained high26.
The dissociation of observed behavior from
adrenocortical activation makes it difficult to
determine the absence of physiological stress simply
by observing that the infant has fallen asleep27.
Further, it is unknown how the presence of
increase d cortisol levels affects the brain-building
processes that occur during sleep28.
The ability to modulate stress and
return to homeostasis is age dependent.
The ability to modulate distress (self-soothe) depends
on the timing , type and inte nsity of the stress
experienced. Stressors that are manageable at 12
months, may be disorganizing at younger ages29.
Further, infants with difficulties in self-regulation
are less capable of modulating intense stress or
withdrawing from overwhelming stimuli30. For these
infants, who display a larger cortisol responses to
stress31, it is unknown how CIO impacts the ir
neuropsychological development.
Empirical evidence of the safety of
prolonged crying in infancy could
not be found.
Both researchers and popular parenting advice assert
that an in fant will not be harmed by the crying
involved precipitated by CIO32; however, no empirical
evidence could be found to support this point.
Extinction (crying it out) is the most studied
intervention for infant sleep problems.
Macall Gordon, M. A.
Antioch University, Seattle
Sheri L. Hill, Ph.D.
UW Center on Infant Mental Health and Development
Is “crying it out” appropriate for infants?
A review of the literature on the use of extinction in the first year
CIO is increasingly endorsed by popular parenting
books and magazines for infants 3-6 months of age.
A large number of CIO
investigations did not
include any infants.
22 studies were found that investigated
the use of either pure or graduated
extinction with children (Table 1). Of
these, 13 included any infants under 1
year15. Only 2 studies (one of them a
case study) focused on infants
Only two studies include any infants
under 6 months17; however, neither
study indicates how many infants of
this age were included in the sample.
Of the studies that included
infants, none calculated
effects for this age group.
With one exception18, results were
reported for the sample as a whole.
No studies examined effects for
infants (<1 year) versus older
In fact, infant sleep literature rarely
distinguishes between infants and
older children. Literature
referencing CIO’s positive effects
frequently cites as evidence studies
conducted on toddlers and older
Developmentally-based exploration of outcomes is lacking.
The majority of CIO outcome measures focus
exclusively on the extinction of crying20 and
fail to explore the existence of side-effects
beyond whether or not the child stayed asleep.
Only 5 studies were found that investigated the
existence of changes in post-CIO behavior. Out of
these, 3 included infants in the sample21. Each of
these studies employed the same parent-report scale.
No objective assessments were conducted.
The Flint Infant Security Scale22 was used in each study to assess post-CIO infant
behavior. Though designed to measure infant security, it is not clear that the scale
actually measures security per se, but rather describes behaviors more related to
regulatory capacities, temperament, and developmental level (e.g., Can accept sudden
advances of a stranger.” “Enjoys a crowd.” “Can recover easily when upset.”) Further,
results as measured by the FISS have been misconstrued as evidence that CIO does not
affect attachment23.
Is prolonged crying biobehaviorally benign?
Discourse on infant sleep problems is currently driven by
the pediatric community. As such, infant sleep is viewed
as a largely behavioral event33 shaped and perpetuated
by parental responses. Within this context,
responsiveness is framed as problematic behavior34.
In the absence of transactional, dyadic perspectives,
the socioemotional and contextual aspects of infant
sleep have been largely discounted or overlooked.
Research on CIO with infants <1 year is inadequate.
3Rethinking CIO: Infant mental
health perspectives are needed.
CIO Studi es (chronological order)
Table 1: Age ranges for CIO study samples.
12 24 36 48 60
Subject Age Ranges (in mont hs)
CIO is recommended by a majority
of pediatricians10 and parent advice
books on infant sleep11
While the majority of existing research
focuses on ameliorating diagnosed
nightwaking problems in infants older than 1
year12, CIO is increasingly endorsed in
popular parenting literature as a preventative
approach for infants starting between 3 and 5
Some authors have suggested beginning
as early as 6-8 weeks14.
Advice in popular parenting magazines and books suggests:
Advice in Popular Parent ing Media:
Nor should you worry about letting
averyyoung babycry. In fact, the
younger the infan t, the easier the
process willbe. ‘B abies olderth an 5
or 6months are naturally going to be
more upset because you’ve changed
the rules on the m,Dr. Schaefer says.
‘A 3-month-old ... knows onlythe
routine that you create.’”
Teach Yo ur Baby to Sl eep In Just 7Days
Parents Magazine, May 2000
Rapoff,Christophers en &
Brock, & During, 1983
et al.,1985
et al.,1989
Lawton, France,&
Mindell &
Minde, Faucon, &
Falkner, 1994
Reid, Walter, &
Concerns about the quality and quantity of
an infant’s sleep top the list of worries for
parents. While some literature indicates that
infant sleep involves a complex interaction
of biological, developmental, and
environmental factors1, the majority of
research suggests that infant sleep problems
result from caretaker responsiveness which
inadvertently reinforces nightwaking
As a result, the most empirically
investigated approaches to infant sleep
problems involve modifying parental
responses to nighttime crying. Based on
behavioral theories, extinction of unwanted
behavior involves withdrawing
reinforcement (parental attent ion) for the
operant behavior (waking and crying)3.
Popularly called crying it out (CIO),
variations of the intervention include:
Unmodified or pure extinction
(see also systematic or planned ignoring,
cold turkey) involves putting the child to
bed, closing the door and, unless the child
is ill, not reentering until morning.
Graduated extinction (see also progressive
delay responding, controlled crying,
Ferberizing), allows the parent to check in
with—but not pick up—the child at
progressive intervals, thereby reassuring
parents about the child’s well-being5 and
allowing them to practice ignoring cries6.
Research to date submits that CIO is
quick7, effective8, and without negative
However, the major ity o f existing
research was conducted on children
older than one year who pres ented wit h
identifiable sleep disorders.
Popular parenting advice in the U.S.
nevertheless endorses the use of CIO
as a preventative approach for infants
beginning as early as 3 months.
Advice in Popular Parent ing Media:
“‘How long do I let my baby cry?’
To establish regular naps, no morethan
one hour...Thereis no time limit at
night if the child is not hungry or ill.
If we place an arbitrary limit on the
duration of crying at night,we train
our child to cry to that predetermined
time. When it is open-ended, the child
learns to stop protesting and tofall asleep.”
Healthy Sl eep Habits, Happy Child
Weiss blut h, 1999, p.159.
Sample Mean
(if stated)
Case Study
Advice in Popular Parenting Media:
“Steel yourself: ...a little
or a lot ofcryingmay
ensue. But rest assured, it will
be tougher on you than on your
baby. ..‘Get over the worrythat
ignorin g yourbabywhile he
cries will do psychological harm,’
emphasizes Dr.Schaefer.
Teach Your Baby to Sleep In Just 7Days
Parents Magazi ne, May 2000
Advice in Popular Parenting Media:
Some parents fear that ignoring
a baby’s wails at night may make
her feel abandoned ...But most
child-development specialists
believetha t letting an infant
cry so she learn s to fallasleep
on her own is healthierin the
long run. ..Experts suggest
holdingoff on sleep tra ining
untilyour baby is at le ast 3-
months-old, whensh e'llbe
better ableto soo the herself.
Goodnight baby!
Parenting Magazine, April 2002
In the absence of adequate data on the effects of prolonged
crying and extinction on infants, the safety of CIO in the first
year (and especially in infants under 6 months) cannot be
To date, no studies of CIO have employed objective,
observational assessments of changes in infant,caretaker,
or dyadic behavior, or biobehavioral outcomes.
Advice in Popular Parent ing Media:
“‘Is crying harmful?’ Not
necessarily. In fact,recen t
studies have proven that crying
produces acceleratedfo rgetting
of a learned response. So when
achild cries,sh e maymore
quicklyunlearn to expect to
be picked up.When trying to
stop an unhealthyhabit, crying
may have somebenefit, because
crying acts as an amnesic agent.
Healthy Sl eep Habits, Happy Child
Weiss blut h, 1999, p.159.
More research is needed into the systemic effects of infant
sleep interventionseffects that may extend beyond
whether or not an infant abandoned crying and stayed
[In the behavioral view of
infant sleep ],ca regiving is
constructed as behaviorman-
agement, and isunderpinned
by the idealiz ation of all-
night sleep as an achievable
normand an unquestioned. ..
belief in the traina bililty
of infants .
Rowe, J. (2003)
Nursin g Inquiry, 10(3), p. 185.
Does existing research support
the use of CIO for infants
under 1 year?
1Parental behavior is the primary cause of nightwaking.
2Sleep training should begin early to prevent sleep problems from taking root.
3Prolonged crying is neither physically nor psychologically harmful and will
not damage the infant’s relationship with the parent.
ResearchGate has not been able to resolve any citations for this publication.
In recent years there have been advances in the behavioral management of infant sleep disturbance. Behavioral programs, how ever, can be controversial and difficult to carry out successfully. This article summarizes the information parents and clinicians must have in order to carry out a program (fact), details of how and when to apply each program (act), and a guide for clinicians supporting parents in choosing and carrying through an intervention (tact). Information on the use of medication and on prevention of sleep disturbance is also provided.
Sleep is a biological function that is partially regulated by the ecological context of the familial relationships. Yet, only a few empirical studies examined sleep from a relational standpoint. Furthermore, while sleep-wake transitions are likely to involve the attachment and the caregiving systems, data pertaining to this interplay are scarce. Thus, the goal of the present study was to focus on attachment-related constructs and address settling and night waking difficulties. Sixty-eight Kibbutz mothers of healthy infants, aged 9-15 months, completed sleep questionnaires and responded to self-report measures pertaining to adult attachment and to maternal feelings. The prevalence of insecurity among this group of mothers was 54%. Overall, the mothers reported high pleasure in interacting with their child. Mothers' anxiety in close adult relationships correlated with hostility towards the infant. Sleep problems were common; 46% of the infants were perceived as having a sleep problem. While mothers' attachment characteristics were not related to the child's sleep, the emotional tie to the infant was linked to sleep regulation. Specifically, pleasure in the interaction was associated with more sleep problems and more involved nighttime parenting. In conclusion, around one year of age, sleep-related difficulties are common and appear to be characteristic of infants whose mothers express a positive emotional tie towards them. Given that both the sleep data and the relationship assessment were based on maternal reports, it remains a task for future studies to include longitudinal observations that illuminate how emotional ties impact sleep regulation. Due to sampling and assessment limitations, more studies across different ages and contexts are warranted.
Infant sleep disturbance (ISD) is widespread and troublesome. Although effective management techniques have been established, some lay and professional authors have expressed concern about these interventions. These concerns are sometimes shared by parents who seek professional advice while feeling ambivalent about undertaking treatment. These concerns include (a) that ISD is normal and inevitable, (b) that it results from unnatural or artificial cultural practices, (c) the belief that ISD expresses a need state, and (d) the belief that the use of extinction is harmful. These concerns are examined and the management of ISD by extinction and its alternatives are considered in the light of the ethics of professional practice. It is concluded that these concerns are best answered within the context of a professional relationship based on a partnership and the sharing of expertise with parents.
Examined the relations among adrenocortical stress reactivity, infant emotional or proneness-to-distress temperament, and quality of attachment in 66 infants tested at 9 and at 13 months. Performed the Louisville Temperament Assessment at 9 months and conducted the Strange Situation at 13 months. Adrenocortical activity was not associated with attachment classifications. Emotional temperament at 9 months was strongly correlated with emotional temperament at 13 months. There was also evidence that at both ages infants who were more prone to distress experienced greater increases in adrenocortical activity during the laboratory tests. Significantly, however, although both the Louisville Temperament Assessment and the Strange Situation involve maternal separation (a potent stimulant of the adrenocortical system in nonhuman primate infants), we noted only small elevations in cortisol, and these elevations were significant only at 9 months.
Regular waking at night is one of the most common problems encountered by parents of young children. In a family counselling programme in Auckland, a night-waking programme has been used with 208 children referred during a two year period. The programme involved organized bed-time routines, procedures for settling the child and non-reward of crying, calling out and getting out of bed. Programme introduction was followed by daily phone calls to parents in which appropriate parent behaviours were prompted and reinforced. Further face to face sessions were held after one week and then if needed. Parent reports show high rates of parent implementation of procedures and rapid change of child sleep behaviour with improvements being maintained at follow up. In a survey of 48 of the parents, positive changes in the daytime behaviour of children were reported as coinciding with improved sleep habits. Also there was an absence of negative side effects, and generally, parent satisfaction with the programme was high.
Systematic ignoring and two modifications of it (systematic ignoring with minimal parental check and systematic ignoring with parental presence) were evaluated for treatment of Infant Sleep Disturbance (ISD). Fifteen infants (6–15 months of age) participated in a study utilising a multiple-baseline design across the three treatment programs. Frequency of awakening and duration of crying were measured in order to evaluate treatment efficacy and infant distress. All programs led to decreases in night waking but infants treated with systematic ignoring with minimal check woke and cried more over the treatment period. Given our present knowledge, the parental presence program appears to be the treatment of choice. This result needs further investigation, as do those of other programs commonly used to treat Infant Sleep Disturbance (ISD) in infants.
Chronic disturbed sleep is a common problem in preschool children. Treatment by extinction is successful but may be accompanied by side-effects such as post-extinction response bursts which make the treatment aversive to parents and which impairs their compliance. This study evaluated a modified procedure, graduated extinction, which required parents gradually to reduce attention to bedtime disturbance and night waking from average baseline levels to zero over 28 days. Baseline measures of frequency and duration of night waking, sleep-onset latency, and bedtime delay were made for six children (6- to 14-months-old). Parents were then instructed to use graduated extinction using a multiple-baseline-across-subjects design. Following treatment, three out of six children showed clinically significant reductions in the frequency and duration of night waking, and a fourth child substantially reduced the duration of her wakings. These gains were maintained at followup. Lack of improvement was associated with parental noncompliance with treatment and childhood illness. Two subjects showed some evidence of post-extinction response bursts despite the gradual withdrawal of parental attention. Parents reported high satisfaction with the procedures but half found the procedure mildly to moderately stressful. While graduated extinction was an effective treatment, regular extinction offers practical and clinical advantages in most cases.
Assesses a behavioral intervention program for 31 children aged nine months to 3 1/2 years who had sleep problems. The children's sleep showed rapid improvement following the intervention, and similar changes were observed in other measures. (RJC)