Prestimulus Activity Level and Responsivity in the Neonate

Author(s): Celia Lamper and Carl Eisdorfer

Source: Child Development, Vol. 42, No. 2 (Jun., 1971), pp. 465-473
Published by: Wiley on behalf of the Society for Research in Child Development
Stable URL: http://www.jstor.org/stable/1127480
Accessed: 27-06-2016 10:40 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/1127480?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Wiley, Society for Research in Child Development are collaborating with JSTOR to digitize,
preserve and extend access to Child Development

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
 PRESTIMULUS ACTIVITY LEVEL AND
RESPONSIVITY IN THE NEONATE

CELIA LAMPER and CARL EISDORFER

Duke University Medical Center

LAMPER, CELIA, and EISDORFER, CARL. Prestimulus Activity Level and

Responsivity in the Neonate. CHILD DEVELOPMENT, 1971, 42, 465-473.
40 neonates were tested on 3 consecutive days to examine the relationship of
prestimulus activity level to response intensity and to day-to-day response
consistency. The results showed that prestimulus state interacts with mild
stimulation to influence intensity of response, while responses to more com-
pelling stimuli were independent of the influence of state. The second finding
was that day-to-day consistency occurred only in instances in which effects
of prestimulus state on response intensity were minimal by virtue of post-
natal age or of potential of the stimulus to override the influence of state.

Studies of neonatal behavior have reported contradictory findings

regarding the relationship of prestimulus state to responsivity. For ex-
ample, changes in the infant's resting state affect the ease of elicitation
and the response latency of the rooting and startle reflex (Prechtl 1958).
Studies of behavioral organization (Wolff 1966) and individual differ-
ences (Escalona 1962) in infancy have indicated that response intensity
is inversely related to prestimulus state. Birns (1965) on the other hand,
without rigorous control of state, reports consistency of individual re-
sponse intensity from the second through the sixth day of life. She used
data from four different types of stimulation and obtained a pooled rank
score of the responses.

This study was supported in part by the Duke University Center for the Study of
Aging and Human Development, grant no. HD-00668 of the NIH, and by research train-
ing grant no. T01-HD-00164-03 of the NICHHD. Author Lamper's address: Center for
the Study of Aging and Human Development, Duke University Medical Center, Durham,
North Carolina 27705.

[Child Development, 1971, 42, 465-473. ? 1971 by the Society for Research in Child Development, Inc.
All rights reserved.]

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
CHILD DEVELOPMENT

Inferences about behavioral organization, organismic integrity, and

innate temperament rely on the basic assumption that the behavioral
index which is used characterizes the infant at the time of testing and is
reproducible from day to day. The data currently available fail satis-
factorily to define (a) the stimulus and organismic conditions that are
optimal for obtaining unbiased response data, and (b) the mechanisms
by which the experimental and subjective characteristics interact to
yield the desired unbiased, that is, reproducible, or the biased response.
This study was designed on a paradigm similar to that used by Birns.
The purpose of this work was to examine the effect of four classes of
stimuli on the nonspecific motor response of neonates, with attention to
pre- and poststimulus activity level, response duration, response inten-
sity, and consistency across sensory modalities on successive days.

METHOD

Forty neonates (21 females, 19 males, full-term, normal delivery,

and medically within normal limits) aged 8-72 hr were tested on 3 suc-
cessive days in the full-term nursery of the Duke University Medical
Center. Three stimuli, sound at two intensities and cold, were applied
for 5 sec followed by a 60-sec interval. A fourth stimulus, touch, was
applied for 9 sec followed by the 60-sec interval. Three applications of
each of the four stimuli constituted the 20-min test sequence. The sound
stimulation involved two tones with a frequency of 3,000 cps, one at 90
db (loud), and one at 70 db (soft), measured at cribside with a General
Radio Company sound-level meter. Cold was applied to the inner aspect
of the thigh with a stainless steel applicator which had been immersed in
ice water at 60 C; touch was applied at the nasal septum using a sterilized
cotton ball.
During the 60-sec period between stimulus applications the infants'
states were assessed every 10 sec by two trained observers who used a
15-point scale of sleeping-waking activity states derived from Wolff's
definition of infant states (1966). Response to the stimulus was recorded
by the same observers using a seven-point scale of motor activity based
on that used by Birns (1965). The response was recorded at two 5-sec
periods, the first beginning at the cessation of the stimulus, termed
"stimulus period," and the second encompassing the subsequent 5 sec,
termed "5-sec period." Following interobserver reliability checks, the
15-point scale was collapsed into four categories as follows:
1. Regular sleep: eyes closed, no eye movements visible in or be-
neath eyelids; muscle tone relaxed to slight or mild in fingers and toes;
spontaneous activity (mild startles, sucking, smiles, grimaces, facial
twitches, sobbing respirations) should be infrequent, periodic, and limited.

466

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
 CELIA LAMPER AND CARL EISDORFER

9. Irregular sleep: eye movements evident in and beneath eyelids

and in periorbital area; mild tension in sleep position and throughout
body; slight head movements, single segments, single limb movements,
small range, smooth execution.
3. Active irregular sleep and drowsiness: eye movement evident or
eyes open and close involuntarily; when drowsy, tension may be minimal
but variable, with limb activity characteristically smooth; when asleep,
whole limb and head movements may occur, may be jerky or agitated.
4. Waking activity and crying; eyes characteristically open and ap-
pear to focus; whole limb and head movements predominate and are
characteristically jerky and agitated; vocalization includes whimpers,
voiced respirations, and outbursts of crying and periods of sustained
crying.

RESULTS

The prestimulus state had been observed and recorded on a 15-point

scale. A reliability check was performed on 25 percent of the observations.
For the two observers rating the infants on that scale, 94 percent of the
measures were in exact agreement, 4 percent differed by one point, and
2 percent differed by two points. Rater reliability, using the following
seven-point scale of response intensity, showed complete agreement in
95 percent of the observations and disagreement of more than one point
in 1 percent.

0 ..... No response.
1..... Questionable response, questionable change in ongoing activity.
S .... . Movement of a part, that is, hand, foot, single segment, eyes, mouth, change of
respirations.
3..... Movement of whole limbs or head, characteristically smooth.
4 ..... Major flexion or extension of limbs, characteristically jerky or agitated.
5..... Major flexion of limbs and trunk, writhing or startle-like movement.
6..... Marked tension, rigidity, trunk arching, tense and forceful crying.

The prestimulus state did not appear to influence the persistence of

the response. Analysis of activity level 10 sec after the cessation of stim-
ulus, using the x2 method, indicated that the infants characteristically
returned to the prestimulus state.
The intensity ratings of responses recorded for the three trials of
each stimulus for a given day were summed and a rank assigned each of
the 40 Ss for each day. The stimulus period and 5-sec periods were treated
separately.
In order to determine whether Ss were consistent in their responses
to the four stimuli, responses for each stimulus category were summed
across the 3 days and tested using a Kendall coefficient of concordance (W).

467

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
CHILD DEVELOPMENT

This statistic tests whether infants retain their ranks relative to

other infants in the categories of the experimental paradigm. A significant
W value would indicate that infants showed a significant tendency to
attain similar ranks, based on intensity of response, in each stimulus
condition and would justify the conclusion that the infants gave responses
of consistent relative intensity to each type of stimulus. An insignificant
W value would indicate that positions achieved under different condi-
tions were variable and followed no reliable ordering (Siegel 1956).
The results (stimulus period, W = 60.68, x2, df = 39, p < .05; 5-sec
period, W = 86.64, p < .05) indicate that there was, in fact, individual
consistency of response across the four stimulus categories. Infants
tended to respond with similar relative intensity to all stimuli.
To determine whether infants were consistent in responses on suc-
cessive days, scores on each day were summed across stimuli and the W
computed. The values (stimulus period, W = 52.03, x2, df = 39, p < .10;
5-sec period, W = 49.78, p < .10) did not reach acceptable significance
levels, but the effect was in the right direction and gives general support
to the conclusion that infants' responses were consistent from day to day.
These results are in general agreement with those obtained by Birns.
The low confidence intervals obtained here seem to indicate that the
infants' response intensity is concordant over time but is not a marked
phenomenon.
Fundamental differences in experimental design prevent unlimited
comparisons between this study and Birns's. First, the response scale was
modified, and we used a sum of response scores rather than a median in
the statistical tests. Second, responses were recorded in two time periods
so that the total statistical effect may have been somewhat altered by
distributing the maximal responses into two sets of data. Birns took only
the maximal response and could be expected to obtain a higher signifi-
cance level.
Given the concordance across days and across stimuli, the Kruskal-
Wallis one-way analysis of variance was applied to question whether
intensity of response differed according to prestimulus-state category.
These results are presented in table 1. For each observation period, 12
analyses were performed (four stimuli for each of 3 days).
Loud and soft tone stimulus conditions were subject to influence by
prestimulus state. Responses to cold and touch showed no dependence
on state. State dependence occurred on Days 1 and 3, but was absent
on Day 2.
The stimuli produced different levels of response intensity. These
differences were significant among all stimuli except loud and soft tones
(one-way ANOVA, F = 74.38, df = 156, 3, p < .01; 5-sec, F = 47.58,
p < .01). Studentized range test (p < .01) demonstrated that cold was
most intense (i = 4.25), with touch (i = 3.20), loud tone (i = 2.82),

468

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
 CELIA LAMPER AND CARL EISDORFER

TABLE 1

RELATION OF PRESTIMULUS STATE TO RESPONSE INTENSITY

(KRUSKAL-WALLIS ONE-WAY ANALYSIS OF VARIANCE, H)

DAY 1 DAY 2 DAY 3

STATE Stimulus 5-Sec Stimulus 5-Sec Stimulus 5-Sec

Loud .... 9.59 11.44* 8.03 6.64 19.31** 0.80

Soft..... 9.69 11.92* 3.17 4.55 13.31** 16.49**
Cold ..... 10.29 5.35 1.93 5.40 2.28 7.51
Touch... 6.58 3.51 4.406 2.11 8.553 9.43

* X2 distribution, df = 3; p < .01. ** x2 distribution, df = 2; p < .01.

and soft tone (i = 2.62) following in that order. At 5 sec after stimula-
tion, cold was still associated with the highest level of activity (i = 3.91),
touch remained second (i = 3.12), soft (i = 2.63) and loud tones
(i = 2.59) reversing their previous positions but the two not differing
significantly.
According to the ANOVA, prestimulus state had affected distribu-
tion of ranks only in stimulus conditions where the elicited response was
of comparatively mild intensity.
Additional coefficients of concordance were computed to determine
whether statistical consistency varied with day of testing, type of stim-
ulus, and prestimulus state. The data were analyzed in three groupings:
(a) total sample of 40 Ss, (b) Ss in state 2 (irregular sleep), and (c) com-
bining Ss, in states 3 and 4. State 1 (regular sleep) was not analyzed be-
cause the number of Ss in this category was quite small. Categories 3 and
4 were combined because neither was represented with sufficient fre-
quency to be handled separately. Results are presented in table 2.
Pooled response data for the four stimuli handled separately for
each of the 3 days, for the total group of Ss, revealed significant con-
cordance at stimulus and 5-sec periods in every instance except Day 2 at
stimulus period. The two subsamples, on the other hand, demonstrated
consistent responses to the four stimulus categories only on Day 2.
To compare day-to-day stability of response to each type of stim-
ulus, response data for stimulus presentations on each day were summed
and compared with that value on each of the other days. The tests were
computed for each of the stimuli at both stimulus and 5-sec periods for
the total group of 40 Ss and the two subsamples as described above.
For the total group, only cold, stimulus period, evoked consistent
responses on separate days. Analysis of the subsamples revealed signifi-
cant day-to-day consistency in response to cold and touch among infants
in category 2; Ss in categories 3 and 4 demonstrated no day-to-day con-
sistency.

469

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
CHILD DEVELOPMENT

TABLE 2

A. COEFFICIENTS OF CONCORDANCE POOLED ACROSS STIMULI BY DAYS, W

40 Ss STATE 2 STATES 3 AND 4

DAY Stimulus 5-Sec Stimulus 4-Sec Stimulus 5-Sec

1......... 60.54* 82.71** 23.84 24.12 18.52 14.05

2 ......... 49.14 55.63* 37.37* 31.38 164.00* 148.00*
3......... 69.59* 96.19** 17.99 36.73 4.99 13.44

x2, df = S3 X2, df = 25 s, df = 6
"*p <.05 * p <.05 *p < .05
"** p < .005

B. CONCORDANCE POOLED ACROSS DAYS BY STIMULI

40 Ss STATE 2 STATES 3 and 4

CATEGORY Stimulus 5-Sec Stimulus 4-Sec Stimulus 5-Sec

Loud ..... 42.14 40.57 29.29 33.18 50.0 13.5

Soft ...... 43.48 39.94 31.63 30.67
Cold....... 56.97* 51.19 43.33* 39.94* 33.5 26.0
Touch.... 22.18 39.16 37.75* 31.61 80.0 64.0

x2, df = s9 x2, df = 25 s, df = 7
* p < .05 * p < .05

Three findings emerged from these analyses. First, when the effects
of state were minimal (by ANOVA, Day 2), individual infants within
subsamples defined by state gave responses of similar relative intensity
to the four classes of stimuli, that is, those who responded intensely to
mild stimuli also responded intensely to more compelling stimuli.
Second, a significant interaction effect appeared. The more intense
stimuli, cold and touch, evoked consistent day-to-day responses among
infants in state 2 (irregular sleep); Ss in states 3 and 4 demonstrated no
concordance.
Finally, the ANOVA data suggest the presence of a statistical arti-
fact. In neither analysis across stimuli nor across days did the two sub-
samples demonstrate the pattern of concordance seen in the total sample
of 40 Ss. Therefore, the statistical effect for the total sample may well be
a function of combining the response data of a large number of Ss in
differing prestimulus activity states and a variety of stimulus conditions.
While category 1 was excluded from the analyses of subsamples, the
number of Ss involved was small, and it seems doubtful that this group
could account for such a marked discrepancy. Conceivably, pooled data

470

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
 CELIA LAMPER AND CARL EISDORFER

lend more power to the ranks, thereby inflating the coefficient of con-
cordance and giving apparent justification to conclusions of day-to-day
consistency which cannot be supported by closer examination.
The absence of state dependence on Day 2 suggests that the nature
of the states may undergo evolution or change with postnatal age. Since
the experimental conditions remained constant, variations in distribution
of ranks according to prestimulus state on successive days would reveal
a changing organization of the response characteristics of the state. Since
state relatedness was demonstrated in only five conditions, these analyses
were limited to the relationship of state to the responses to soft and loud
tones on Days 1 and 3.
Using the Mann-Whitney U test, multiple contrasts between re-
sponse ratings in differing prestimulus-state categories were employed
within the five conditions that had yielded significant results in the
ANOVA (table 1).
For Day 1, in the 5-sec period, category 1 (regular sleep) differed
significantly from categories 2 and 3 but not from 4. No differences were
seen among categories 2, 3, and 4. On Day 3, category 2 (irregular sleep)
differed significantly from 3 and 4. These relationships are presented in
table 3.

TABLE 3

DISTRIBUTION OF BETWEEN-STATE DIFFERENCES

ACCORDING TO MANN-WHITNEY U TESTS
A. DAY 1

S3 4
STATE L S L S L S

1 ........ * * * * N.S. *
2........ ... ... N.S. N.S. N.S. N.S.
3 ........ ... ... ... ... N.S. N.S.
4 . .. .. .. . . . . .. .

B. DAY 3

23 4

STATE L S S (5-Sec) L S S (5-Sec) L S S (5-Sec)

1........ N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S. N.S.
2........ ... ... ... N.S. B.S. N.S. * * *
S........ ... ... ... ... ... ... N .S. N .S. N .S.
4.....

* p <.05.

471

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
CHILD DEVELOPMENT

The evidence suggests then, that even within the first few days of
life, the interaction of prestimulus states with stimuli changes with post-
natal age. On the first postnatal day, the more active behavioral states
(2, 3, 4) constitute essentially one broad activity level yielding responses
of similar intensity to different stimuli. On the second day, state had no
statistically significant effect. On the third day, the responses in two
conditions suggest the beginning of differentiation of state 3 from state
4. The response characteristics of each state, which doubtless are affected
by the birth process, apparently stabilize and become more clearly de-
fined with increasing postnatal age.

DISCUSSION

The data permit conclusions about certain aspects of the relation-

ship between activity level, stimulus, and observed response as well as
speculation about postnatal organization of behavioral states.
First, behavioral concordance, or day-to-day consistency of response
intensity, can be demonstrated in the first 3 days of life when the optimal
combinations of stimulus and prestimulus state are employed.
For all stimuli except cold, distribution of ranks was dictated to
some extent by state. In the case of cold, effects of state appeared to be
overridden. Significantly, the absolute response intensity was always
greater for cold than for any other stimulus. Touch, with the next most
intense responses, produced the same effect in one instance. Clearly, the
compelling nature of the stimulus eliminates the modifying effects of
the sensory thresholds of the various states.
Concordance was demonstrated only in irregular sleep. It could be
concluded, therefore, that this state had stable response characteristics.
However, irregular sleep is by definition a variable state, characterized
by periodic bursts of spontaneous activity, irregular respirations, and
changing muscle tone. Furthermore, Wolff (1966) reports that sensory
thresholds vary with the presence of spontaneous activity and, there-
fore, alter sensitivity to, and perception of, stimuli.
Since only the intense stimuli evoked concordant responses, it seems
that they must tap more fundamental, constant sensory modalities than
those accessible to mild stimuli. The interaction of the stimuli with this
particular state suggests that the variability of thresholds is not at issue
but rather the overriding potential of the tactile stimuli, cold and touch.
Categories 3 and 4 were characterized by variable motor activity
and level of arousal. No response consistency was found in these cate-
gories. It seems reasonable to conclude that major variability of state is
a source of bias in results of tests administered at repeated intervals.
This is not to imply that irregular sleep is unquestionably the only

472

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms
 CELIA LAMPER AND CARL EISDORFER

state for testing or that stimuli must always be intense or tactile in this
age group. The range of stimuli tested here was limited. Irregular sleep
was the predominant state in this age group and was represented in
greater proportion of the observations than the other categories. Regular
sleep (1) and irregular sleep (2) were mutually exclusive categories, while
3 and 4 were broader and were combinations of more than one state.
Alert states, within category 3, may, for example, be ideal for some types
of testing. The data could be assumed to best describe the response char-
acteristics of categories 1 and 2.
The data give limited insight into progressive behavioral organiza-
tion with postnatal age. On Day 1, states were polarized with deep sleep
constituting a distinct state and the entire continuum from irregular
sleep to active crying making up one amorphous category with similar
responses. On Day 2, response was statistically independent of state. By
Day 3, differences between 2, irregular sleep, and 4, waking activity and
crying appeared. Unfortunately, regular sleep, 1, did not occur frequently
enough to permit rank order tests. Consequently, the characteristics of
regular sleep cannot be extended beyond Day 1. By Day 3, extremes
of sleep and wakefulness biased results, but changes from irregular sleep
to drowsiness did not.

CONCLUSION

Individual differences in behavioral style or temperament are con-

siderably influenced by organismic and experimental conditions. The
sensory thresholds of the background states appear to regulate sensory
receptivity and perceptual processes so that only stimulus qualities
keyed to that state will elicit a reliable response. This mechanism reduces
the range of stimuli which can be useful in that state. Furthermore, the
information value of the response depends entirely upon its reproduci-
bility at repeated intervals, since the response may merely describe the
sensory thresholds of the state and fail to reveal the infant's behavioral
qualities.

REFERENCES

Birns, B. Individual differences in human neonates' responses to stimulation. Child Develop-

ment, 1965, 36, 249-256.
Escalona, S. K. The study of individual differences and the problem of state. Journal of
American Academy of Child Psychiatry, 1962, 1, 11-37.
Prechtl, H. F. R. The directed headturning response and allied movement of the human
baby. Behavior, 1958, 13, 212-242.
Siegel, S. Nonparametric statisticsfor the behavioral sciences. New York: McGraw-Hill, 1956.
Wolff, P. H. The causes, controls and organization of behavior in the neonate. Psycho-
logical Issues, 1966, 5, 1-113.
473

This content downloaded from 142.66.3.42 on Mon, 27 Jun 2016 10:40:52 UTC
All use subject to http://about.jstor.org/terms