Sleeping Through the Night: A Developmental Perspective

(1)

Sleeping

Through

the

Night:

A Developmental

Perspective

Thomas F. Anders, MD*; Leslie F. Halpern, PhDII; and Jenny Hua*

ABSTRACT. This study examines falling asleep and

night waking in human infants during the first 8 months

of life. All-night time-lapse video recordings were

ob-tamed at 3 weeks and 3 months of age; a Sleep Habits

Interview was completed at these ages and repeated at 8

months of age by telephone interview. At the 3-week

and 3-month ages, self-report measures of maternal

psy-chologic distress, depression, and self-esteem were also

obtained. The data are examined from both

cross-sec-tional (age group) and longitudinal (individual)

perspec-tives. Parent-infant interactions at bedtime and during

the middle of the night changed significantly with

in-creasing age. At 3 weeks of age, most infants were put

into their cribs for the night already asleep. When they

awakened in the middle of the night, they were removed

from their cribs. By the time they returned to their cribs,

they were again asleep. By 3 months of age, infants who

were put into the crib awake at bedtime and allowed to

fall asleep on their own were more likely to return to

sleep on their own after awakenings later in the night.

In contrast, infants who were put into the crib already

asleep at the beginning of the night were significantly

more likely to be removed from the crib following a

subsequent nighttime awakening. Thus, the pattern of

bedtime sleep onset was related to sleep onset following

an awakening in the middle of the night. This association

was present at 8 months as well. Infants who used sleep

aids were more likely to be put into their cribs awake at

bedtime and were also more likely to return to sleep on

their own after a nighttime awakening at both 3 and 8

months of age. At 8 months of age, 7 of the 21 infants

were identified by their mothers as problem sleepers.

All were male infants who were still put into their cribs

asleep at the beginning of the night. These individuals

could not be predicted from 3-week or 3-month

obser-vations of video-recorded sleep-wake state organization

or mother-infant interaction. The authors speculate

about the interplay between sleep-wake state regulation,

nighttime interactions, problem sleep, temperament and

maternal factors such as depression, self-esteem, and

stress. Pediatrics 1992;90:554-560; sleep, infancy, night

waking, mother-infant interaction.

During the first 2 years of life, sleep problems are

the most frequent complaint of parents during

pedi-atric health supervision visits.’ Thirty percent to 50%

of infants manifest disrupted sleep serious enough to

From the ‘Sleep Disorders Center, Rhode Island Hospital; Women and Infant’s Hospital; Division of Child and Adolescent Psychiatry and

IIDepartment of Pediatrics, Brown University School of Medicine, Providence, RI.

Received for publication Jan 27, 1992; accepted Mar 12, 1992.

Dr Halpern is currently at the Dept of Psychology, Indiana State University. Reprint requests to (T.F.A.) Dept of Psychiatry, UC/Davis Medical Center,

4430 V St, Sacramento, CA 95817.

cause parents to seek professional assistance. A

num-ber of studies have documented the prevalence of

such problems.29 By and large, these studies have

been cross-sectional in design and have gathered data

using maternal reports and questionnaires. Richman’#{176}

has defined problem sleep in 1 -year-old infants as

either sleep-onset periods associated with fussing that

last longer than 30 minutes on a regular basis, or

night-waking episodes that occur at least four nights

a week and require parental intervention. During the

toddler period, when separation problems and fears

such as fear of the dark, are common, the rates of

disturbance increase. Sleep problems are often

asso-ciated with daytime behavior disorders. ‘‘“i

Are sleep problems related to the developmental

task of sleeping through the night? What factors

facilitate the achievement of this task? Moore and

in a large-scale epidemiologic survey in

Eng-land, reported that, by 6 months of age, 50% of

infants had “settled,” ie, they slept uninterruptedly

from midnight to 5:00 AM each night. By 12 months,

90% of the group had settled. Curiously, Moore and

Ucko also reported that by 12 months of age, 50% of

infants who had settled began to exhibit night

wak-ing, suggesting the independence of these two

proc-esses. Settling has been linked to gestational age

(maturity), nursing status, and infant health status.

Night waking has been linked to infant illness, family

stress, and change in sleeping environment.’6”7

Paret’8 has reported that infants who use a sleep aid

(pacifier or thumb) at bedtime and during the night

are less likely to exhibit night waking at 9 months of

age.

REGULATION OF SLEEP

Although sleep traditionally has been viewed as a

characteristic of individual infants, it seems likely that

the regulation of sleep and waking states has both

individual (infant) and relationship (parent-infant)

components. The regulation of sleep is distinct from

sleep-wake state maturation. All-night time-lapse

video recording in the home provides a naturalistic

method for observing sleep-wake regulation

associ-ated with infant-parent interaction at bedtime and

during the course of the night.

In previous time-lapse video studies, we have

re-ported on slee?wake state maturation during the first

year of life.’9 ‘ Serendipitously, we noted that,

dur-ing the first 3 to 4 months of life, infants, at bedtime,

were routinely put into their cribs already asleep and

were removed in the middle of the night when they

(2)

to describe their infants as sleeping through the night.

On the videotapes, however, we observed that almost

all of the infants awakened spontaneously after 6 to

7 hours. Few if any infants slept continuously for 10

to 1 2 hours despite parental reports to the contrary.

Rather, some infants, when they awakened, were able

to soothe themselves and fall back to sleep on their

own. They did not cry and thus did not arouse their

parents. Other infants, after a spontaneous

awaken-ing, cried and became progressively more aroused,

leading finally to their being removed from their crib.

We described infants as “self-soothers” if they could

put themselves back to sleep without arousing their

parents, and as “signalers” if they cried and were

taken from the crib when they awakened.22

These previous observations led to the current

study of the emergence of self-soothing and signaling

behaviors in infants. We expect that these

designa-tions are not intrinsic to individual infants but rather

reflect the outcome of repeated interactions between

parents and their infants at bedtime and during the

night. The study further explores possible factors that

predispose infants to problem sleep, including

asso-ciations between signaling and self-soothing, bedtime

interaction, the use of a sleep aid, infant

tempera-ment, and maternal factors such as parenting

self-efficacy, depression, and psychologic distress. Many

of these factors have been shown to be important in

the infant’s social and emotional development.2329

Subjects

METHODS

Twenty-one normal, full-term infants, born at the Women and

Infant’s Hospital of Rhode Island, and their mothers were subjects of this study. Prenatal and pennatal records were reviewed and only firstborn infants whose gestations and deliveries were

un-eventful were recruited. No mothers refused to participate in the

study and none were lost to follow-up. All mothers signed informed

consent. The infants were video-recorded in their homes at 3 weeks

and 3 months of age and followed up with a structured telephone

interview at 8 months of age.

There were I 3 male and 8 female infants. Twenty of the families

were intact with both parents at home. One parent was single but

lived with her boyfriend. All of the families were middle and

upper-middle class according to the Hollingshead Index.3#{176} At 3

weeks of age, 1 1 infants were being breast-fed, 7 were bottle-fed,

and 3 were receiving a combination of formula and breast milk. By

8 months of age, 13 mothers (62%) reported that they were primary

care givers at home, 7 mothers (33%) were employed full-time

outside the home, and 1 mother worked part-time.

Procedures

Sleep Recordings and Scoring. When the infants were 3 weeks and 3 months of age, a research assistant transported a portable time-lapse video-recording system to the home of each infant. This

system has been described in detail previously.31 In essence, a video camera with a wide-angle lens is placed on a tripod overlooking the crib. A microphone to record vocalization is attached to the

camera. An infrared light allows the unit to record in virtual

darkness so as not to disrupt the naturalistic sleeping environment of the infant. The video and audio signals are recorded on a

time-lapse video recorder at a 12:1 time reduction. Thus, 12 hours of

sleep can be played back in I hour. Clock time is coded on the

video tape by a time-code generator.

Mothers activate the system before putting their baby to sleep

and turn it off in the morning. They are instructed to treat their

babies as they normally would, putting them in their crib and

responding to them in their usual manner. On the following day,

the research assistant returns to debrief the mother about any

unusual occurrences and removes the equipment.

Three states can be reliably scored from the tapes: active sleep, quiet sleep, and awake. In addition, out-of-crib time is coded.3’ In

this study we scored the percent of active sleep. quiet sleep,

wakefulness, and out-of-crib time; the total amount of sleep time

(in minutes) during the night; and the longest period of sustained

sleep (in minutes) uninterrupted by an awakening. After training

to criterion for scoring, 20% of the nights were randomly scored

by three raters. Interrater reliabilities for the summary scores were

better than 85% and resembled results reported in previous

stud-ies.2#{176}

Parental interaction during the night was coded separately for

interventions directed toward a sleeping infant or in response to an

awake infant, The type of intervention, awake or asleep, and the

time and duration of each intervention were noted. Asleep

inter-ventions presumably reflect parental ‘checking” of the sleeping

infant; awake interventions represent comforting an awake infant

in the crib. All interventions associated with an infant’s removal

from the crib were coded as out-of-crib transitions rather than

interventions.

Structured Questionnaires. Mothers completed a number of self-report instruments at the times of filming. Age of administration

and the instruments with their respective summary variables are

listed in Table I.

The Sleep Habits Interview is a structured questionnaire that

quantifies the infant’s usual sleep habits and nighttime behaviors

and parent-infant interactions around sleep during the previous 2

weeks. It has been used previously with good construct validity.32

The Mental Health Inventory is a 38-item measure of

psychol-ogic mental health. Items are rated on either five- or six-choice (Likert-type) response scales. Simple addition of item scores can be

used to derive a Psychological Distress Global Scale that has

adequate consistency and stability.33

The Beck Depression Inventory is a 2 1 -item inventory consisting of a series of ordered statements scaled from 0 to 3 that identify symptoms of depression.34 Reliability and validity of this

instru-ment are well established.35

The Parenting Sense of Competence Scale, a measure of

par-enting self-esteem, assesses the mother’s sense of competence in

the parenting role. It contains an eight-item self-efficacy scale that quantifies the degree to which parents feel they have the skills and

knowledge necessary to be a good parent and a nine-item scale

that assesses how parents value and are comfortable in the parent-ing role. Items are rated on 6-point scales with response choices

ranging from ‘strongly agree” to strongly disagree.’ Items are

summed to obtain self-efficacy and value of parenting role scale

scores.36

The Infant Characteristics Questionnaire assesses maternal per-ceptions of infant temperament.37 The version for 3- to

6-month-olds consists of 24 items that are rated on 7-point scales, with a

rating of 1 describing optimal temperament and a rating of 7

describing a difficult temperament. Four dimensions of infant

tern-perament are measured: fussy-difficult, unadaptable, dull, and

unpredictable.

All 21 mothers were contacted for a comprehensive telephone

interview when their infants were 8 to 9 months old. The Sleep

Habits Interview was again completed. During the interview,

atten-tion was focused on whether the mother felt that her infant was a

good sleeper or a problem sleeper; whether the infant was put into the crib awake or asleep at the beginning of the night; whether the infant awakened after sleep onset; and the parental response to the

awakening. We also inquired about the mother’s employment

status, the infant’s health and feeding habits, and his or her general

level of motor and social development. Questions about who

usually put the infant to bed, the infant’s nap patterns, and the

regularity of nighttime sleep habits concluded the interview. All

questions were specifically focused on the two weeks prior to the

phone call.

It is important to note that the designation of good” sleeper or

problem” sleeper at 8 months of age was entirely subjective,

reflecting the mother’s judgment about how easily her infant went

to sleep and how frequently he or she had awakened during the

previous 2-week period. Although the Sleep Habits Interview elicits specific information about patterns of disruption, there are, as yet,

at Viet Nam:AAP Sponsored on September 1, 2020

www.aappublications.org/news

(3)

no commonly accepted definitions of sleep disorder at 8 months of age. Parental concerns remain the basic data available to clinicians.

RESULTS

Sleep-Wake State Maturation From 3 Weeks to 3

Months of Age

In analyzing the results, Spearman (r) correlation

coefficients, Kendall (r) concordance coefficients,

x2’

and paired t tests to test age-group differences were

used. On only 1 of 42 nights did a mother (of a male

infant 3 months of age) terminate the video recording

prematurely. Thus, the video sleep data represent 21

nights of recording at 3 weeks of age and 20 nights

of recording at 3 months of age. Examination of the

summary indices of sleep-wake state organization,

scored from the videotapes, showed expected

matu-rational changes from 3 weeks to 3 months of age,

consistent with the age norms reported in previous

studies.’92’ Paired a’ tests indicated that active sleep

decreased significantly from 61 .2% at 3 weeks to

47.0% at 3 months of age (t[39] = 4.5, P .001); time

out-of-crib dropped significantly from 17.7% to 2.9%

(t[39] = 6.1, P s .001); and the longest sustained

sleep period increased significantly from 215 minutes

to 358 minutes (t[39] = 4.4, P .001). Total sleep

time for the night and the percent of time awake did

not change significantly with age. These results are

summarized in Table 2.

Maternal Interventions: Age-Related Changes and

Individual Differences

In this sample virtually all bedtime interactions and

nighttime interventions were carried out by mothers.

Fathers were not observed on the videotapes.

Debrief-ing in the morning confirmed that this pattern was

customary. At 3 weeks of age, all but one infant was

removed at least once during the night for a feeding.

In addition, 9 of the 21 infants received a number of

waking interventions (range from I to 6) while in

their cribs. These interventions were most often

soothing and in response to infant signaling. The

remaining 12 infants (57%) received no waking

inter-ventions after once being put into their cribs. At 3

months, the response was similar. Ten infants (50%)

received no waking interventions after bedtime. The

remaining 10 infants received from 1 to 14 waking

interventions per infant. The larger range suggests

shaping; maternal responses to awakenings and

in-creased numbers of infant awakenings are already

evident by 3 months of age.

Thirteen of the 3-week-olds received no sleep

in-terventions, whereas only seven of the 3-month-olds

received none. As infants grew older, mothers

checked their sleeping infants more often. The

num-ber of sleep interventions per night ranged from 1 to

6 at the younger age and 1 to 4 at the older age,

suggesting some emerging confidence in the mother’s

ability to know when the baby is sleeping and to

check in a more consistent fashion.

At both 3 weeks and 3 months, waking

interven-tions were significantly related to the baby’s use of a

sleep aid (r = .50, P .02 at both ages) as mothers

were summoned to reinsert a lost pacifier. Similarly,

waking interventions were related to babies signaling

at 3 months (r = .41, P .04), whereas at 3 weeks,

signaling more often resulted in removal from the

crib.

A longitudinal analysis using Kendall’s r suggests

that maternal patterns of waking and sleep

interven-tions were correlated with each other and stable from

3 weeks to 3 months of age. Mothers who check their

sleeping infants when they are 3 weeks of age are

more likely to continue checking when they are 3

months of age (r = .38, P .05); they are also more

likely to check their waking infants when they are 3

TABLE 1. Assessment Instrume nts, Summary Varia bles, and Ages

Instrument Age

3Weeks 3 Months 8 Months

Sleep Habits Interview

Bedtime Hour Hour Hour

Rise time Hour Hour Hour

Night waking No. No. No.

Problem sleep Yes/no Yes/no Yes/no

Sleep onset Awake/asleep Awake/asleep Awake/asleep

Sleep aid Yes/no Yes/no Yes/no

Unusual occurrences Yes/no Yes/no Yes/no

Mental Health Inventory

Psychological Distress Score Score N/A’

Global Scale Beck Depression Inventory

Total Score N/A N/A

Parenting Sense of Compe-tence Scale

Self-efficacy Score Score N/A

Value of parenting Score Score N/A

Infant Characteristics Questionnaire (Bates temperament)

Fussy/difficult N/A Score N/A

Unadaptable N/A Score N/A

Dull N/A Score N/A

Unpredictable N/A Score N/A

(4)

TABLE 2. Time-Lapse Video

Scores by Age (Mean and SD)

Sleep-Wake State Summary

Age

3 Weeks 3 Months

% Active sleep

%Outofcrib

% Awake

Longest sleep period, mm

Total sleep time, mm

61.2 (8.6)

17.7(9.8)

3.8 (1.6) 215 (82) 434 (113)

47.0 (11.2)’

2.9(4.8)*

5.0 (3.6) 358 (123)’ 475 (111) *P .001.

TABLE 3. Maternal Interventions During the Night (Kendall r)

3 Weeks 3 Months

Waking Sleep Waking Sleep

Waking (3 wk) ...

Sleep (3 wk) .10 ..

Waking (3 mo) .29 .36’ . ..

Sleep (3 mo) .06 .38* .46’ ..

*P .05.

TABLE 4. Frequencies of Sleep Behaviors at 3 Ages”

3 Weeks 3 Months 8 Months

Awake on- 2 (9.5) 6 (28.6) 11 (52.4)t

set

Self-soother 2 (9.5) 10 (47.6) 11 (52.4)t

Sleep aid 8 (38) 11 (55) 14 (67)

*Values represent number (percent).

tP .01.

months of age (r = .36, P .05). And, by 3 months,

a significant relationship exists between waking and

sleep interventions (r = .46, P .04). These results,

summarized in Table 3, suggest that mothers and

their infants begin to interact in a predictable fashion

at 3 weeks of age and the pattern is stable at 3 months

of age. Mothers who frequently check their sleeping

babies at 3 weeks of age are the ones who check on

their sleeping and awake babies at 3 months of age.

Examination of the relationship between maternal

interventions and sleep state organization revealed

that at 3 weeks of age, a shorter period of sustained

sleep was associated with more sleep interventions (r

= - .50, P .02), and at 3 months, sleep interventions

were associated with an increased percent of active

sleep (r = .63, P .003), suggesting that checking

sleeping infants might be in response to more

‘rest-less” active sleep, or conversely, that checking might

induce more activity and arousal during sleep. Also,

babies put to bed awake at 3 months tended to receive

more sleep interventions (r = .47, P .04).

Sleep Onsets, Sleep Aids, and Signaling at 3 Ages

As portrayed in Table 4, 2 infants (9.5%) were put

into their cribs awake at 3 weeks of age, 6 (28.6%)

infants at 3 months, and 1 1 infants (52.4%) at 8

months (P .01); 2 (9.5%) infants were self-soothers

at 3 weeks of age, 10 (47.6%) at 3 months, and 11

(52.4%) at 8 months (P .004). A nonsignificant

developmental trend was noted for the use of a sleep

aid. More infants used sleep aids at older ages.

In terms of individual characteristics, there was a

significant relationship between being put to bed

awake at 3 months of age and again at 8 months of

age (r = .60, P .004). This relationship was not

predictable at 3 weeks inasmuch as virtually none of

the infants were put to bed awake.

The 3-month-old infants were divided on the basis

of the videotapes into self-soothers and signalers. By

definition, both groups awakened spontaneously

dur-ing the night. The signalers cried and received waking

interventions or were removed from the crib. The

self-soothers did not cry and received no maternal

intervention. Examination of the sleep characteristics

of signalers in contrast to self-soothers revealed that

they had shorter continuous sleep periods by an

average of 100 minutes (r = -.42, P .06) and shorter

total sleep times (r = -.48, P .03) and were out of

the crib for four times as long (r = .49, P .03).

We also divided the 3-month-old infants into those

who used a sleep aid and those who did not. Those

infants who used a sleep aid at 3 months in

compar-ison with those who did not were more wakeful in

the crib (r = .64, P .002) and had a longer total

sleep period (r = .38, P .01). A final analysis divided

those infants who were put into the crib awake at

sleep onset and those who were put into their cribs

already asleep. The former group had longer periods

of total sleep (r = .55, P .01), although they did not

demonstrate a longer period of continuous sustained

sleep.

Gender

There were no significant infant gender differences

associated with sleep-wake state organization,

method of feeding (breast or bottle), maternal

depres-sion, perceived psychologic distress, parenting

effi-cacy, or value of parenting. No gender differences

were detected at either 3 weeks or 3 months in the

use of a sleep aid, in the status of self-soother or

signaler, or in whether the infant was put into the

crib awake or asleep.

Problem Sleepers at 8 Months

The results of the telephone survey at 8 months of

age revealed that of the 21 infants, 7 were labeled as

“problem” sleepers and 14 were described as good

sleepers by their mothers. The “problem” sleepers had

distinctly different profiles of sleep behaviors than

nonproblem sleepers. Problem sleepers were put into

their cribs already asleep. They were unlikely to use

a sleep aid and they were reported to be signalers. In

contrast, good sleepers were re?orted more likely to

be put into their cribs awake

(x

[1, n = 21] = 8.6, P

.003), to use a sleep aid

(x2

[1, n = 21] = 4.5, P

.03), and not to be signalers

(x2

[1, n = 21] = 8.6; P

.003).

An unexpected finding at 8 months of age was the

relationship between problem sleep and gender. All

of the 7 problem sleepers were male infants. The

nonproblem sleepers consisted of 6 boys and 8 girls

(x2 [1 n = 21] = 4.3, P .04). In addition, gender

was significantly associated with being put into the

crib awake at sleep onset (7/8 girls and 4/13 boys

had waking onsets;

x2

[1, n = 21] 4.3, P .04).

Such a gender difference has not been noted

previ-ously in studies of sleep problems. However,

(5)

TABLE 5. Correlations Between Maternal Distress, Depression, and Parenting Competence and Hassles

Measure Psychologic

Distress (3 wk)

Psychologic Distress (3 mo)

Beck Depression

Maternal self-efficacy -.38” -.47” -.39”

Value of parenting role -.51t -.33 -.641

Parenting hassles _#{149}53t .651

Sp : .05.

tP .01.

I 1 .001.

TABLE 6. Maternal Characteristics and Perceived Infant Te Questionnaire)

mperament (Infant Characteristics

Assessment at 3 Weeks Assessment at 3 Months

Psychologic Depression Self-efficacy

Distress

Psychologic Self-efficacy Distress

Fussy

Unadaptable

Dull

Unpredictable

.50” .52t NS

.43’ .43” -.42”

NS NS NS

.63t .45” NS

.38” -.621

.55 -.58t

NS NS

.45” -.47”

*P .05.

tP .01.

.001. NS, not significant.

month-old infants typically are not defined as

prob-lem sleepers in research reports. As noted earlier,

gender alone did not distinguish infants who used a

sleep aid or were observed or described as being a

signaler or self-soother at the younger ages. Gender

did differentiate problem sleepers and whether an

infant was asleep or awake when put into the crib at

8 months of age. Thus, gender, in this study, seemed

to influence the mother’s perceptions of problem

sleep and her interactive behaviors at bedtime and

during the night for a subgroup of male infants.

In this study, therefore, a subgroup of seven

8-month-old male infants, all described as problem

sleepers, were put into their cribs already asleep at

bedtime. They did not use a sleep aid when they

awakened, but rather they signaled and were

re-moved from the crib in the middle of the night. This

interaction did not characterize either male or female

nonproblem sleepers.

Maternal Assessments

Many of the individual maternal measures were

correlated significantly with each other and were

stable over the two ages. Both perceived maternal

psychologic distress and mothers’ value of parenting

remained constant from 3 weeks to 3 months of age.

Parenting efficacy increased significantly (P .01).

Psychologic distress in mothers of 3-week-old

in-fants persisted when their infants were 3 months of

age (r = .58; P .001). Moreover, the stable

psychol-ogic distress scores at both ages were strongly

corre-lated with the maternal depression score obtained

when the infant was 3 weeks of age (r .72, P

.001 and r = .49, P .01, respectively). In other

words, mothers who rated themselves as depressed

at 3 weeks also reported more psychologic distress at

that time and at 3 months. The mean (SD) Beck

depression score, however, was 6.8 (±4.9), with only

two mothers scoring above 13, the commonly

ac-cepted cutoff for screening of depression.

In individual mothers, parenting competency was

stable from 3 weeks to 3 months (value of parenting,

r = .65, P .001; self-efficacy as a parent, r = .47, P

:S .002). Higher maternal psychologic distress scores

at both 3 weeks and 3 months and depression scores

at 3 weeks were negatively related to parenting

corn-petence scores. These relationships are summarized

in Table 5.

The measures of parenting competence,

psychol-ogic distress, and maternal depression were also

sig-nificantly related to ratings of infant temperament.

Although the mean temperament subscale ratings

were typical of normative samples (Fussy 1 8 ± 6;

Unadaptable = 8 ± 4; Unpredictable = 8 ± 3; Dull =

7

± 2), the negative attributes of temperament, such

as fussiness, unadaptability, and unpredictability

were correlated with greater maternal depression,

higher levels of psychologic distress, less perceived

self-efficacy as a parent, and less perceived value in

the parenting role as summarized in Table 6. This is

particularly noteworthy because maternal factors

(dis-tress, depression, and self-efficacy) at 3 weeks, which

were themselves stable to 3 months, could be used,

as could the 3-month assessments, to predict the

temperament ratings obtained at 3 months.

The relationship between maternal factors and

in-fant sleep-wake state organization is less clear. At 3

weeks of age, several intriguing trends were noted.

Mothers who experienced high levels of psychologic

distress (and whose scores for value in parenting and

self-efficacy were reduced) had infants that spent less

time in active sleep (r = -.42, P = .06). Mothers

who reported more depression had infants who were

more wakeful (r = .42, P .06). These findings are

similar to patterns of disrupted sleep that have been

(6)

as sleeping in an unfamiliar sleep laboratory setting

(first night adaptation effect)3’ or following

circum-cision.38 However, by 3 months of age, no associations

with sleep were significant. The infant seemed to be

sleeping independently of any of the stresses that

affected the mother.

DISCUSSION

The majority of studies that have reported on

in-fants’ sleeping through the night and night waking

have relied on maternal reports instead of

observa-tions of actual sleep-wake state regulation and

parent-infant interaction at night. This preliminary study,

using time-lapse videosomnography to observe

night-time sleep-wake regulation, confirms our previous

observations that sleeping infants during the first year

of life may awaken one or more times briefly during

the night without disturbing their parents.22 In this

study virtually all infants aroused one or more times

each night at 3 weeks and 3 months of age. At 3

weeks of age, 90% of infants aroused and signaled.

By 3 months of age, approximately 50% of the infants

were able to return to sleep on their own. The latter

group may be called self-soothers; those who

contin-ued to arouse and cry may be called signalers. It is

important to note, however, that these responses may

not be characteristics of individual infants; rather they

are likely to emerge from consistent mother-infant

interactions and maternal attributions that

character-ize the dyadic relationship between the infant and

her or his mother. At 8 months of age, 52% of infants

did not signal. Of the 9 infants who continued to

signal at this age, 7 were defined as problem sleepers

by their mothers.

From a developmental perspective, this study

de-scribes the changes with age of some of the behaviors

that appear to be associated with the emergence of

nighttime self-soothing behaviors. Self-soothers are

more likely to use a sleep aid, such as a pacifier or

their fingers, to assist them in falling asleep.

More-over, self-soothers are generally put into their crib

awake at bedtime and are able to fall asleep on their

own at sleep onset, a pattern that then is repeated

after an awakening in the middle of the night. In this

study, older infants were more likely to be put into

their cribs awake at bedtime, and older infants were

more likely to be able to make use of a sleep aid.

These associations have also been described by

Fer-ber.’

From the individual, longitudinal perspective, it is

not surprising that infant responses to night waking

are not stable from 3 weeks to 3 months of age. These

processes are in transition, and it is difficult to predict

in advance who will be defined as a problem sleeper

at 8 months of age. Many of the associated variables

are not yet sufficiently differentiated by 3 months of

age.

In contrast, maternal interventions at night seem to

become more individually consistent. They also

ap-pear to be related more to stable maternal perceptions

of stress, self-confidence, and depression which, in

turn, are related to maternal ratings of infant

temper-ament. Thus, infant sleepers participate in a dyadic

relationship of sleep-wake regulation in which

unsta-ble infant and more stable maternal behaviors

influ-ence bedtime and nighttime interactions that may

become defined as problem sleep at 8 months of age

for some infants.

The ability of an infant to make use of a sleep aid

is poorly understood but may be associated with the

nature of the parent-infant attachment relationship.

In support of problem sleep being associated with

attachment disruptions, a recent study of infants,

aged 18 to 36 months, with severe sleep problems

has reported that mothers with insecure adult

attach-ment classifications are significantly overrepresented

compared with control populations.39 Interpretation

of our results demands a cultural perspective as well.

Cosleeing in family beds is the norm in many

cul-tures.4 From our videotapes it was obvious that

3-week-old infants possess the capacity to fall asleep

on their own; yet, it appears that cultural norms

dictate that, during the first several months of life,

they should be fed and rocked to sleep and not be

provided an opportunity to fall asleep on their own.

Furthermore, our culture dictates that they sleep alone

and quickly achieve the milestone of sleeping through

the night. The aims of sleeping independently and

being rocked to sleep may be contradictory, especially

if infants continue to be put into their cribs asleep at

older ages.

In our sample, the seven male infants who were

defined as problem sleepers by their mothers at 8

months of age could not be predicted from their

sleep-wake state organization or temperament at 3 months

of age. Yet, at 8 months of age, these seven boys were

still being put into their cribs asleep and did not use

a sleep aid. Thus, these seven mothers related

differ-ently at bedtime to their sons than did other mothers

with their sons and daughters. Is this a relationship

issue related to attachment? Or, are these boys

partic-ularly “difficult,” requiring special comforting?

Unfortunately, this study does not answer these

questions. We need to examine larger samples. We

need to know more about waking interactions. We

also need to know more about the attachment

rela-tionships per Se. Finally, we need to be more rigorous

in our definition of problem sleep. Although it may

be useful to allow parents to define the construct, it

is also useful to have more objective criteria. Only

further research will supply answers to these

ques-tions. Nevertheless, nighttime sleep-wake regulation

provides an excellent opportunity for the study of

early psychologic, sociocultural, and biologic

devel-opment in the context of parenting relationships.

ACKNOWLEDGMENTS

Ronald Seifer, PhD, provided statistical consultation.

REFERENCES

I. Ferber R. Sleepless child. In: Guilleminault C, ed. Sleep and Its Disorders is?Children. New York, NY: Raven Press; 1986:141-163

2. Ferber R. Solve Your Child’s Sleep Problem. New York, NY: Simon and

Schuster; 1985

3. Cuthbertson J, Schevill S. Helping Your Child Sleep Through the Night. New York, NY: Doubleday; 1985

4. Beltramini A, Hertzig M. Sleep and bedtime behavior in preschool-aged

children. Pediatrics. I 983;71 :153-158

(7)

5. Lozoff B. Wolff A, Davis N. Sleep problems seen in pediatric practice. Pediatrics. 1985;75:477-483

6. Jacklin C. Snow M. Gahart M, Maccoby E. Sleep pattern development from 6 through 33 months. IPediatr Psychol. 1980;5;295-303

7. Ragins N, Schachter S. A study of sleep behavior in two-year old children. IAm Acad Child Psychiatry. 1971;10:464-480

8. Weissbluth M. Davis A, Poucher J. Night waking in 4- to 8-month-old infants. IPediatr. 1984;104:477-480

9. Simonds J, Parraga H. Sleep behaviors and disorders in children and adolescents evaluated at psychiatric clinics. Dev Be/tao Pediatr. 1984;5: 6-10

10. Richman N. A community survey of characteristics of one- to two-year-olds with sleep disruptions. / Am Acad Child Psychiatry. 1981;20:

281-291

1 1. Richman N. Sleep problems in young children. Arch Dis Child. 1984; 56:49 1-493

12. Richman N, Stevenson J, Graham P. Preschool to School: A Behavioral Study. London, England: Academic Press; 1982

13. Klackenberg G. Sleep behavior studied longitudinally: data from 4-16 years in duration. night awakening and bedtime. Acta Paediatr Scand.

1982;71 :501-506

14. Jenkins S. Bax M, Hart H. Behavior problems in preschool children. / Child Psychol Psychiatri,i. 1980;21:5-17

15. Moore T, Ucko L. Night waking in early infancy. part 1.Arch Dis Child. 1957;32:333-342

16. Bernal J. Night waking in infants during the first fourteen months. Dev Med Child Neurol. 1973;15:760-769

, 7. Blurton-Jones N, Rosetti-Ferreira M, Farquar-Brown M, McDonald I. The association between perinatal factors and later night waking. Dev Med Child Neurol. 1978;20:427-434

18. Paret I. Night waking and its relationship to mother-infant interaction in nine-month-old infants. In: Call J, Galenson E, Tyson R, eds. Frontiers ofinfant Psychiatry. New York, NY: Basic Books; 1983:171-177

19. Anders T. Home-recorded sleep in two- and nine-month-old infants. / Acad Child Psychiatry. 1978;17:421-432

20. Anders T, Keener M. Developmental course of nighttime sleep-wake patterns in full-term and pre-term infants during the first year of life. Sleep. 1985;8:173-192

21. Anders 1, Keener M, Kraen’ter H. Sleep-wake state organization. neonatal assessment and development in premature infants during the first year

of life. Sleep. 1985;8:193-206

22. Anders T. Night waking in infants during the first year of life. Pediatrics.

1979;63:860-864

23. Scott G, Richards M. Night waking in infants: effects of providing advice and support for parents. /Child Psychol Psychiatry. 1990;31:551-567 24. Field T, Healy B, Goldstein S. Perry S. Bendell D. Infants of depressed

mothers show ‘depressed’ behavior even with non-depressed adults.

Child Dcv. 1988;59:l569-1579

25. Field T, Healy B, LeBlanc W. Sharing and synchrony of behavior states and heart rate in nondepressed versus depressed mother-infant inter-actions. Infant Behav Dev. I 989; 12:357-376

26. Zeanah C, Keener M, Stewart L, Anders T. Prenatal perception of infant

personality: a preliminary investigation. / Am Acad Child Psychiatn,’.

1985;24:204-2 I 0

27. Zeanah C, Keener M, Anders T, Levine R. Measuring difficult

tempera-ment in infancy. /Dev Behav Pediatr. 1986;7:114-119

28. Keener M, Zeanah C, Anders T. Infant temperament, sleep organization and nighttime parental intervention. Pediatrics. 1988;81:762-771

29. Coutrona C. Troutman B. Social support, infant temperament and

par-enting self-efficacy: a mediational model of post-partum depression. Child Dev. 1986;57:1507-1518

30. Hollingshead AB. Four-Factor Index of Social Status. New Haven, CT: Yale University; 1985

31. Anders T, Sostek A. The use of time-lapse video recording of sleep-wake behavior in human infants. Psychophysiology. 1976;13:155-158

32. Crowell J, Keener M, Ginsburg N. Anders T. Sleep habits in toddlers 18-36 months of age. /Am Acad Child Psychiatri. 1987;26:510-515

33. Veit CT, Ware JE Jr. The structure of psychological well-being in general populations. IConsult Cliri Psychol. 1983;52:730-742

34. Beck AT, Ward CH, Mendelson M, Mock JE, Erbaugh JK. An inventory for measuring depression. Arch Gei Psychol. 1961;4:561-571

35. Blumberry W, Oliver JM, McClure JN. Validation of the Beck Depression Inventory in a university population using psychiatric estimate as the

criterion. IConsult Clii: Psychol. 1978;46:150-155

36. Giband-Walston J. Self-esteem and situational stress: factors related to competence in new parents (doctoral dissertation, George Peabody College for Teachers, 1977). Dissertation Abstr mt. 1977;39:379B

37. Bates JE, Freeland CA, Lounsbury ML. Measurement of infant difficult-ness. Child Dcv. 1979;50:794-803

38. Anders T, Chalemian R. The effects of circumcision on neonatal sleep.

Psychosom Med. 1974;36:174-179

39. Benoit D, Zeanah C, Boucher C, Minde K. Sleep disorders in early childhood: association with insecure maternal attachment. JAm Acad Child Adolesc Psychiatry. 1992;3 I :86-93

40. McKenna J, Mosko 5, Dungy C. McAninch J. Sleep and arousal patterns of co-sleeping human mother/infant pairs: a preliminary physiological study with implications for the study of sudden infant death syndrome

(SIDS). Am IPhys Anthropol. 1990;83:331-347

41. Lozoff B, Wolff A, Davis N. Cosleeping in urban families with young

children in the United States. Pediatrics. 1984;74:171-182

42. Thevinin T. The Family Bed: An Age Old Concept in Child Rearing.

Minneapolis, MN: Tine; 1975

43. Caudill W, Plath D. Who sleeps by whom? Parent-child involvement with young children. Psychiatry. 1966;29:344-366

CALL FOR PATIENTS WITH CRIGLER-NAJJAR DISEASE

If you have a patient with this disease who might be interested in participating

in a research trial, please contact:

Jerold F. Lucey, MD

Department of Pediatrics

University of Vermont

Burlington, VT 05401

(8)

1992;90;554

Pediatrics

Thomas F. Anders, Leslie F. Halpern and Jenny Hua