Acta Pñdiatr 89: 846±52.

2000

Characterization of the developmental stages of sucking in preterm

infants during bottle feeding
C Lau1,2,3,4, R Alagugurusamy1,3, RJ Schanler1,3,5, EO Smith1 and RJ Shulman1,2,5
Baylor College of Medicine1; Department of Pediatrics, Divisions of Gastroenterology & Nutrition2/Neonatology3; Feeding Disorders
Center4, Children’s Nutrition Research Center5, Houston, TX, USA

Lau C, Alagugurusamy R, Schanler RJ, Smith EO, Shulman RJ. Characterization of the
developmental stages of sucking in preterm infants during bottle feeding. Acta Pædiatr 2000; 89:
846–52. Stockholm. ISSN 0803–5253
It is acknowledged that the difficulty many preterm infants have in feeding orally results from
their immature sucking skills. However, little is known regarding the development of sucking in
these infants. The aim of this study was to demonstrate that the bottle-feeding performance of
preterm infants is positively correlated with the developmental stage of their sucking. Infants’
oral-motor skills were followed longitudinally using a special nipple/bottle system which
monitored the suction and expression/compression component of sucking. The maturational
process was rated into five primary stages based on the presence/absence of suction and the
rhythmicity of the two components of sucking, suction and expression/compression. This five-point
scale was used to characterize the developmental stage of sucking of each infant. Outcomes of
feeding performance consisted of overall transfer (percent total volume transfered/volume to be
taken) and rate of transfer (ml/min). Assessments were conducted when infants were taking 1–2,
3–5 and 6–8 oral feedings per day. Significant positive correlations were observed between the
five stages of sucking and postmenstrual age, the defined feeding outcomes, and the number of
daily oral feedings. Overall transfer and rate of transfer were enhanced when infants reached the
more mature stages of sucking.
We have demonstrated that oral feeding performance improves as infants’ sucking skills mature. In
addition, we propose that the present five-point sucking scale may be used to assess the developmental
stages of sucking of preterm infants. Such knowledge would facilitate the management of oral feeding
in these infants.
Key words: Feeding performance, low birthweight infants, oral feeding, oral-motor, prematurity
Chantal Lau, Department of Pediatrics/Neonatology, Baylor College of Medicine, One Baylor
Plaza, Houston, TX 77030, USA, MC 5-1000 (Tel. ‡1 713 798 6710, fax. ‡1 713 798 7187,
e-mail. clau@bcm.tmc.edu)

Preterm infants have difficulty transitioning from tube- present study was to verify the hypothesis that oral
feeding to full oral feeding. This has been blamed on the feeding performance of preterm infants is positively
immaturity of their sucking pattern and often results in a correlated with the maturation of sucking skills. For this
delayed discharge from the hospital as the latter mile- purpose, we followed the development of the sucking
stone is significantly correlated with hospital discharge pattern of infants, born less than 30 wk gestation, from
(1). The sucking pattern of term infants is characterized the time they were introduced to oral feeding until they
by the rhythmic alternation of suction and expression/ reached full oral feeding and categorized the matura-
compression (2, 3). Suction defines the negative in- tional process into five primary stages of sucking. This
traoral pressure generated as the infant draws milk into five-point scale was used to demonstrate the relationship
the mouth and expression/compression is believed to between the development of sucking and oral feeding
correspond to mouthing or the compression and/or performance in preterm infants.
stripping of the nipple between the tongue and the hard
palate as milk is ejected into the mouth (4–6). It is
unclear when the term sucking pattern appears. In a Methods
recent study, we demonstrated that preterm infants, with
an immature sucking pattern consisting primarily of Subjects
expression/compression, can successfully bottle feed, Seventy-two infants (38M, 34F), born between 26 and
although they were not as efficient as infants who 29 wk gestation (27.5  1.1 wk, mean  SD), with
achieved a term sucking pattern (7). Thus, the aim of the birthweight averaging 1028  173 g, and appropriate

 2000 Taylor & Francis. ISSN 0803-5253

ACTA PÆDIATR 89 (2000) Development of sucking in preterm infants 847

for gestational age as determined by maternal dates and the suction and expression/compression component of
antenatal ultrasonography, were recruited from the sucking. The nipple used was the one selected by the
nurseries at Texas Children’s Hospital. These subjects caretaker as most appropriate for the infant at the time
were “healthy” preterm infants whose discharge from of the assessment. The level of the milk from a
the hospital was primarily dependent upon their ability graduated reservoir was continuously adjusted to the
to reach full oral feeding. Infants with any of the level of the infant’s mouth to allow for a self-paced
following conditions were excluded from the study: flow, i.e., with milk flowing only when the infant was
intraventricular hemorrhage grades III and IV (8), sucking actively. Inasmuch as the nipple chamber was
necrotizing enterocolitis, hydrocephalus, major conge- continuously filled, the subjects were able to obtain
nital abnormalities, and persistent oxygen requirement milk regardless of whether the suction and/or expres-
for more than a month. Infants were treated prophy- sion/compression component was used (7).
lactically with caffeine citrate until they demonstrated Based on the first 5 min of the oral-motor recording, a
mature control of breathing, usually around 35 wk score of 1 through 5 was given for the sucking pattern of
postmenstrual age. The infants were introduced to oral each infant according to the presence/absence of
feeding at 34.2  1.7 wk postmenstrual age, when their suction, rhythmicity of suction and/or expression/
respiratory rate was <60/min, and they had no persistent compression, magnitude of the suction amplitude and
episodes of oxygen desaturation, apnea and bradycardia. duration of sucking bursts. Sucking pauses >1.5 sec
Oral feedings were advanced upon the recommendation delineated the beginning and end of a sucking burst. The
of their attending physicians based on an adequate daily first 5 min of an oral motor recording was used because
weight gain (15 g/kg/d) and the infant’s ability to we assumed that fatigue would be minimal at the start of
finish the assigned oral feedings the day prior. In a session. To reduce the potential of fatigue, the infants
addition, 10 healthy full-term infants (5M, 5F; were not handled for at least 30 min prior to each
39.5  1.2 wk gestation, 3536  568 g birthweight) monitored session. Because the fatigue factor was
who were introduced to oral feeding immediately after reduced, it is presumed that the sucking pattern
birth were recruited. They were bottle-fed ad libitum observed during that time would reflect the true sucking
and monitored two to three times during their first 3 wk skills of the subject rather than his/her overall feeding
of life. Mothers of all the infants expressed intent to ability. Indeed, it is expected that the level of endurance
breastfeed at the start of the study. However, exclusive of an infant would affect more likely his/her overall
breastfeeding was never achieved. As such, the infants performance than the initial 5 min of a feeding.
were offered both breast and bottle. There was no
evidence of nipple confusion in any of the infants. The Oral feeding outcomes
sample size was based on the fact that 82 infants out of
Feeding performance was assessed with the following
110 (72 preterm and 10 full-term) had complete bottle-
outcome measures: overall transfer, the percent volume
feeding records. This study was approved by the Baylor
taken during a feeding divided by the volume to be
College of Medicine Institutional Review Board for
taken, and rate of transfer (ml/min), the rate at which the
Human Research. Informed consent was obtained from
entire feeding was taken. Volume taken was determined
parents.
by comparing the amount of milk remaining in the
graduated reservoir at the end of a feeding. Milk lost
Oral-motor assessment from drooling was measured by weighing a bib before
and after each feeding,with 1 ml approximating 1 g. For
At our facility, enteral feeding is given to preterm
full-term infants, only rate of transfer was calculated, as
infants by bolus every 3 h or continuous drip. Scheduled
they were fed ad libitum, and overall transfer was
breast/bottle feedings replaced bolus enteral feedings.
always 100%.
Infants on continuous drip were offered breast/bottle
feedings 1 h following cessation of a drip. Thus our
feeding assessments were carried out during a sched- Data analyses
uled oral feeding session when mothers were not In order to assess the reliability of using the scores of
present to breastfeed. Infants were offered mother’s only one observer in the identification of the five stages
milk if available or preterm formula. Caretakers held of the proposed sucking scale, two observers, blind to
the infants on the lap as routinely done in the nurseries the age and number of oral feedings per day of the
and did not provide any encouragement during the oral infants, independently scored 50 of 208 (24%) oral-
feeding sessions, i.e., the infants fed at their own pace. If feeding tracings selected at random across the five
prior feeding was provided through an orogastric tube, stages of sucking. The percentage agreement between
it was removed just prior to the oral feeding session. observers was calculated.
The oral-motor assessment was conducted by means Inasmuch as the infants born at 26 wk gestation are
of a nipple/bottle apparatus described previously (7). less mature than those born at 29 wk, a regression
Briefly, this consisted of two catheters placed on the analysis of the defined outcome measures versus
nipple which allowed for the simultaneous recording of gestational age was conducted in order to determine
848 C Lau et al. ACTA PÆDIATR 89 (2000)

Fig. 1. Actual sample tracings and corresponding range of amplitudes (mmHg) measured at each stage of sucking.

whether maturity plays a significant role in the was used, when approppriate. An independent t-test was
performance of the infants. Gender was analyzed by used to compare the rates of transfer and postmenstrual
analysis of variance (ANOVA). Correlations between ages between preterm and full-term subjects.
the stages of sucking and postmenstrual age, overall
transfer, rate of transfer, and the number of oral feedings
per day were assessed using regression analyses which Results
accounted for repeated measures on individuals. ANO-
VA for repeated measures was used to assess the Stages of sucking
difference between stages of sucking with respect to the Fig. 1 shows a composite of actual sample tracings,
above measures. Post-hoc Fisher’s LSD modified t-test corresponding range of amplitudes (mmHg) and a
ACTA PÆDIATR 89 (2000) Development of sucking in preterm infants 849

Fig. 2. Postmenstrual age versus stages of sucking (boxplot).

* = mean; horizontal bar (dashed) = median; dotted vertical bars =
max, min; dot box (white) = quartiles 1 and 3; solid box (slant)
= 95% confidence interval. p value: Repeated measures ANOVA
of the differences between stages. **Post-hoc Fisher’s LSD versus
stage 1: p < 0.001.

description of the suction and expression/compression

components which typify the five stages of sucking. As
the stages advance, the amplitude of the expression/
compression component does not change, ranging
between 0.2 and 1 mmHg, whereas those of suction
gradually increase along with the duration of the
sucking bursts. Stage 1 consisted of sucking patterns
where arrhythmic expression/compression was present
with very infrequent suction of low amplitude. Stage 2 Fig. 3. A. Overall transfer versus stages of sucking. B. Rate of
transfer versus stages of sucking (boxplot). * = mean; horizontal bar
included rhythmic expression/compression with the (dashed) = median; dotted vertical bars = max, min; dot box (white)
appearance of arrhythmic suction of larger amplitude = quartiles 1 and 3; solid box (slant) = 95% confidence interval.
than those observed during stage 1. Stage 3 comprised p value: Repeated measures ANOVA of the differences between
rhythmic expression/compression alone, as well as the stages. *Post-hoc Fisher’s LSD versus stage 1: p < 0.01; **post-
emergence of not yet rhythmic alternation of suction hoc Fisher’s LSD versus stage 1: p < 0.001.
and expression. In stages 4 and 5 the infants used only
the rhythmic alternation of suction and expression/
compression. Amplitude and the duration of sucking Gestational age and gender versus stage of sucking,
burst primarily differentiated the latter two stages. overall transfer, rate of transfer, number of oral
feedings per day
There was no significant correlation between any of the
Observers’ reliability above outcome measures and the gestational age or
gender of the infants. As such, gestational age and
On a five-point scale and based on 50 bottle-feeding gender were pooled for the subsequent analyses.
assessments, there was a 58% (29/50) exact agreement
between the two observers, 40% (20/50) when scoring Stage of sucking versus postmenstrual age
differed by 1 unit, and 2% (1/50) when scoring differed
by 2 units. With 98% of the differences between As shown in Fig. 2, there is a significant positive
observers (49/50) falling within 1 unit, we were correlation between postmenstrual age and stage of
confident that the use of only one observer was sucking (p < 0.001). Post-hoc analyses showed that
adequate. Thus, subsequent data analyses were based stages 3 through 5 occurred at significantly older
on the scoring made by only one observer who was postmenstrual ages than stage 1 (p < 0.001).
blind to the infants’ postmenstrual age and oral-feeding
status. In addition, tracings from all the infants were Stage of sucking versus feeding performance outcomes
mixed, such that all the tracings from any particular Overall transfer (Fig. 3A) and rate of transfer (Fig. 3B)
subject were never scored sequentially. were positively correlated with stage of sucking
850 C Lau et al. ACTA PÆDIATR 89 (2000)

Table 1. Postmenstrual age and oral feeding rate of transfer tional support, and did not affect the oral motor
(mean  SD) of preterm infants at stage 5 and term infants at 1st, assessment.
2nd and 3rd postnatal weeks.
Postmenstrual age Rate of transfer
Infants n (wk) (ml/min)
Premature 8 36.9  2.5 4.27  2.48 Discussion
Term (1st week) 9 40.4  1.3** 6.59  2.34
Term (2nd week) 10 41.8  1.2*** 8.26  3.17** At present, clinicians and therapists alike have not been
Term (3rd week) 7 43.4  0.7*** 9.03  5.24* able to establish criteria for assessing the ability of
infants to feed orally. The postmenstrual age of 33 to 34
Independent t-test versus preterm infants: *p  0.05; **p  0.01, postmenstrual weeks has been used as the most
***p  0.001.
appropriate time to initiate oral feeding, but there is
no evidence that this is the earliest time at which it can
be done. It is evident that factors, such as endurance,
proper coordination of suck-swallow-breathe, are im-
(p < 0.001). Post-hoc analyses showed that overall portant determining factors of infants’ success at oral
transfer at stages 3 to 4 and rate of transfer at stages 3 feeding. However, knowledge of the developmental
to 5 were significantly enhanced over those at stage 1 stage of sucking of an infant may assist clinicians in the
(p < 0.01 and p < 0.001), respectively. initiation and progression of oral feeding. For this
Table 1. shows that preterm infants can reach a stage purpose, we characterized five primary stages of
5 prior to full-term corrected age (p < 0.01). However, sucking in order to assess the relationship existing
their rate of transfer was significantly lower than that of between the development of sucking and oral-feeding
their full-term counterparts at 2 and 3 wk postnatal age performance in preterm infants (26 to 29 wk gestation).
(p < 0.01 and p < 0.05, respectively). To our knowledge, this is the first developmental
sucking scale which has been developed for preterm
infants. The development of sucking, indeed, is pro-
Stage of sucking versus number of oral feedings per day gressive and quantification of its maturation is difficult.
There was a significant correlation between the stage of However, inasmuch as the term sucking pattern is
sucking and the number of oral feedings per day composed of the rhythmic alternation of suction and
(p < 0.001, Fig. 4), with significant differences ob- expression/compression, we selected the two compo-
served between stages 3–5 and 1 (p < 0.001). nents of sucking and rhythmicity as primary criteria to
Oral-feeding assessments were tolerated well by all assess the stages of sucking. Suction amplitude was used
the infants. Any episodes of desaturation, apnea or to differentiate the more mature stages 4 and 5. Thus,
bradycardia were self-corrected, did not require addi- early on, advancement of the stages of sucking consisted
of the appearance and rhythmic establishment of the
expression/compression component. This was followed
by the appearance of suction and the progressive
establishment of the rhythmic alternation of suction
and expression/compression. As suction matured, its
amplitude gradually increased. This was accompanied
also with an increase in the duration of sucking bursts.
The lack of correlation existing between gestational age
and the stage of sucking, oral feeding outcomes and
advancement of oral feedings would suggest that there is
no significant in utero maturation of sucking occurring
between 26 and 29 wk gestation or insufficient power to
detect a difference over this developmental period. The
positive correlation between the stages of sucking and
postmenstrual age supports the notion that sucking
ability improves as the infant matures. The correlations
between stages of sucking and overall transfer and rate
of transfer suggest that the maturation of this skill
played a significant role in improving oral-feeding
performance.
Fig. 4. Number of oral feedings per day versus stages of sucking Although the data demonstrated that the level of oral-
(boxplot). * = mean; horizontal bar (dashed) = median; dotted motor skills advanced with age, there was a wide
vertical bars = max, min; dot box (white) = quartiles 1 and 3; solid
box (slant) = 95% confidence interval. p value: Repeated measures variation in these skills between subjects at any given
ANOVA of the differences between stages. **Post-hoc Fisher’s LSD postmenstrual age. This was not due to the subjects’
versus stage 1: p < 0.001. maturity at birth as mentioned above. It is unclear why
ACTA PÆDIATR 89 (2000) Development of sucking in preterm infants 851

oral-motor skills developed more rapidly in some term neonates studied. Despite having developed
infants than others, when the population examined similar oral-motor skills by virtue of achieving stage 5
consisted of preterm infants without clinical conditions sucking, they demonstrated lower rates of transfer than
believed to hinder oral-feeding performance. It is well their full-term counterparts. It is probable that addi-
acknowledged that fatigue and/or coordination of suck- tional factors related to maturation come into play, as
swallow-breathe are two elements which can affect the suggested by Pollitt et al. (16). Indeed, noting that 30-d-
overall feeding performance of the infant (9–12). The old full-term infants ingested the same milk volume
specific impact of these two factors will require faster than 2-d-old counterparts, those authors con-
additional studies. Other variations, such as behavioral cluded that the enhanced rate of transfer in the older
characteristics of each subject (13), may also come into infants was likely multifactorial. For example, intake
play. However, despite this wide variation, it is can be increased by altering variables such as ampli-
important to note that the defined stages of sucking tude, frequency of suction- expression/compression,
correlated directly with the measures of feeding duration of sucking burst and/or endurance.
performance defined in this study. Post-hoc analyses In addition to facilitating the monitoring of the
which demonstrated a significant difference between developmental progress of oral feeding in infants, the
stages 3–5 and 1 and the observation that, from stage 3 present scale would also help identify oral motor
to 5, the suction component became more prominent, dysfunction(s) in infants who demonstrate feeding
i.e., alternating rhythmically with the expression/com- difficulties. This study supports the notion that sucking
pression component and increasing in amplitude, lends ability does not need to be at a mature level, i.e., stages
support to earlier reports suggesting that a sucking 3 or higher, before preterm infants are introduced to oral
pattern with the rhythmic alternation of suction and feeding. Perhaps, just as importantly, the initiation of
expression/compression is more efficient than one oral feeding at the earlier stages (e.g., stages 1 to 2),
without rhythmicity and/or with only the expression/ potentially provides increased opportunity for training
compression component (7, 9, 14). It is important, to the infant. Safe oral feeding, in these circumstances,
however, to note that, albeit less efficacious, the use may be facilitated by using interventions such as the
of the expression/compression component alone, as self-paced system which allows the infant to control
seen in stages 1 and 2, still allows for successful and milk flow more effectively (7, 17). As training often
safe oral feedings in preterm infants (7). improves one’s ability at performing a specific task, we
Unlike the increases in suction amplitude which speculate that increased practice may enhance the oral-
occurred as the infants advanced in their oral feeding feeding skills of infants, leading to an accelerated
skills, the pressure recorded from the expression/ progression of their oral feeding.
compression component reached a plateau rapidly. This As the benefits of breastfeeding and mother’s milk
likely resulted from the limitation of the monitoring are widely recognized (18), the technique used in the
system used, because the threshold corresponding to full present study is being adapted to study breastfeeding
compression of the catheter was reached very early. We infants.
are in the process of modifying our nipple/bottle system In summary, using a five-point scale which describes
in order to circumvent this limitation, so as to monitor the development of sucking in preterm infants during
the actual expression/compression pressure generated bottle feeding, we have demonstrated a significant
by the infant. Of interest is that this technical limitation correlation between the level of maturity of infant’s
parallels the situation observed in the preterm infants. sucking and postmenstrual age, feeding performance
The use of expression/compression alone is of limited (i.e., overall transfer and rate of transfer), and progres-
efficacy. Indeed, the maximal volume of milk that can sion of oral feeding. In addition, we are advancing that
be obtained when only the expression/compression the use of developmental scales, such as the present one,
component is used, is limited to the volume of milk may be useful clinically for the identification/character-
contained between the initial point of compression of ization of the oral-motor skills of preterm infants at any
the nipple and the nipple hole. As such, although milk given time point as they progress in their oral feeding
can be obtained when only expression/compression is schedule.
utilized, it cannot be as efficient as when the suction
component is present. With the latter, a continuous flow Acknowledgements.—The authors express their gratitude to P. Burns, C.
of milk can be generated as long as a negative intraoral Bryant, E. Newton-Novato, L. Schwartz for their assistance in this study.
pressure is applied. In addition, the greater the suction This work was supported by grants from the National Institutes of Child
Health and Human Development (R01-HD28140), the General Clinical
amplitude and duration, the greater the flow and volume Research Center, Baylor College of Medicine/Texas Children’s Hospital
out of the nipple, respectively. It should be remem- Clinical Research Center (M01-RR-00188), National Institute of Health.
bered, however, that the size of the nipple hole will Partial funding also was provided from the USDA/ARS, Children’s
determine the maximum flow rate that can be obtained Nutrition Research Center, and the Department of Pediatrics, Baylor
College of Medicine, Texas Children’s Hospital, Houston, TX. The
(5, 15). contents of this publication do not necessarily reflect the views or policies
Preterm infants, who reached stage 5, were at a of the USDA, nor does mention of trade names, commercial products, or
significantly younger postmenstrual age than the full- organizations imply endorsement by the US government.
852 C Lau et al. ACTA PÆDIATR 89 (2000)

References study of term and preterm infants. Dev Med Child Neurol 1990;
32: 669–78
1. Schanler RJ. The low birth weight infant. In: Walker WA, 11. Selley WG, Ellis RE, Flack FC, Brooks WA. Coordination of
Watkins JB, editors. Nutrition in pediatrics: basic science and sucking, swallowing and breathing in the newborn: its relation-
clinical applications, 2nd ed. Hamilton (ON): BC Decker Inc; ship to infant feeding and normal development. Br J Disord
1996: 392–491
Commun 1990; 25: 311–27
2. Dubignon J, Campbell D. Sucking in the newborn during a feed. J
Exp Child Psychol 1969; 7: 282–98 12. Wilson SL, Thach BT, Brouillette RT, Abu-Osba YK. Coordina-
3. Sameroff AJ. The components of sucking in the human newborn. tion of breathing and swallowing in human infants. J Appl
J Exp Child Psychol 1968; 6: 607–23 Physiol 1981; 50: 851–8
4. Ardran GM, Kemp H, Lind J. A cinematographic study of bottle 13. Medoff-Cooper B, Berklan T, Carlson S. The development of
feeding. Br J Radiol 1958; 31: 11–22 sucking patterns and physiologic correlates in very-low-birth-
5. Waterland RA, Berkowitz RI, Stunkard AJ, Stallings VA. weight infants. Nurs Res 1993; 42: 100–5
Calibrated-orifice nipples for measurement of infant nutritive 14. Wolff PH. The serial organization of sucking in the young infant.
sucking. J Pediatr 1998; 132: 523–6 Pediatrics 1968; 42: 943–56
6. Nowak AJ, Smith WL, Erenberg A. Imaging evaluation of 15. Mathew OP. Determinants of milk flow through nipple units. Am
artificial nipples during bottle feeding. Arch Pediatr Adolesc J Dis Child 1990; 144: 222–4
Med 1994; 148: 40–2 16. Pollitt E, Consolazio B, Goodkin F. Changes in nutritive sucking
7. Lau C, Sheena HR, Shulman RJ, Schanler RJ. Oral feeding in during a feed in two-day-old and thirty-day-old infants. Early
low birth weight infants. J Pediatr 1997; 130: 561–9
8. Papile L, Burstein J, Burstein R, Koffler H. Incidence and Hum Dev 1981; 5: 201–10
evolution of subependymal and intraventricular hemorrhage: a 17. Lau C, Schanler RJ. Oral feeding in preterm infants: advantage of
study of infants with birth weights less than 1500 gm. J Pediatr a self-paced milk flow. Acta Paediatr (in press)
1978; 92: 529–34 18. Gartner LM, Black LS, Eaton AP, Lawrence RA, Naylor AJ,
9. Wolf LS, Glass RP. Feeding and swallowing disorders in Neifert ME, et al. Breastfeeding and the use of human milk.
infancy. Assessment and management. Tucson, Therapy Skill Pediatrics 1997; 100: 1035–9
Builders, 1992
10. Bu’Lock F, Woolridge MW, Baum JD. Development of Received Sept. 3, 1999; revision received Jan. 20, 2000; accepted
coordination of sucking, swallowing and breathing: ultrasound Jan. 25, 2000