Washington University School of Medicine

Twenty-six newborn infants were randomly assigned to an early (2 day) and delayed (3 week) circumcision group. Each infant was examined three times using the Brazelton Neonatal Behavioral Assessment Scale (NBAS) in which the examiner was blind to the circumcision status of the subject. A reduction scale was derived from the NBAS examination were assigned to the behavioral categories of average, subdued, or hyperactive. Nearly 90 percent of the early circumcision group changed behavior categories after circumcision for at least four hours compared to 16 percent of the delayed group. The implications of these results and the details of the reduction scale are discussed.


In the United States, about 80% of the approximately 1,600,000 male infants born each year are routinely circumcised as newborns (Kaplan, 1977; Grimes, 1978). despite the prevalence of this medical procedure in the United States, there are few carefully documented studies about the effects of circumcision in the newborn male (Emde, Harmon, Metcalf, Koenig & Wagonfeld, 1971; Anders & Chalemian, 1974; Brackbill, 1975). This paper reports on a controlled, blind study in 26 infants on the effects of circumcision on newborn behavior using the Brazelton (1973) Neonatal Behavioral Assessment Scale (NBAS). A new method was developed for reducing the data derived from the NBAS into specific categories of behavior that permits further applications of the NBAS to a variety of clinical problems.




There are only three previous experimental studies that have investigated the effects of circumcision upon newborn behavior. Emde (1971) studied the effects of circumcision on sleep patterns in ten subjects and ten controls, and found that circumcised infants had about a 10 percent increase in the amount of non-REM or quiet sleep during a 10-hour polygraphic observation period. It is of interest to note that these subjects were circumcised by the plastibel technique in which a mold is placed over the glans penis and held in place by a ligature until there is ischemic necrosis of the foreskin over several days. The plastibel technique, which may be considered a chronic procedure with an unknown period of pain, is different from the acute circumcision procedure used by Anders (1974) in his study, where 11 infants each (each acting as his own control) were observed for three non-continuous one-hour periods. Ander's observations were made after the feeding preceding the circumcision, immediately following the procedure, and during a recovery period after the feeding following the circumcision. Behavioral observations were made ever 30 seconds and six behavioral states were recorded - four waking and two sleep states. Anders observed an increase in wakefulness, especially in crying and fussy behavior, in the hour after circumcision. There were no significant differences in acive-REM or quiet-non-REM sleep state perceptions in any observation period. The differences between Anders' and Emde's results might be due to differences in experimental design - controls and observational method - or circumcision techniques.

In the third experimental study, Brackbill studied the effects of circumcision on response to continuous auditory stimulation in 39 male newborns (Brackbill, 1975). The subjects had been circumcised by methods not described, 37 hours before they were tested. Control groups were 30 female and 30 uncircumcised male infants. Arousal levels, as measured by heart rate changes and state changes, were observed during the experimental period with continuous auditory stimulation of white noise and no additional auditory stimulation for control periods. Significant differences were noted in heart rate and state changes between the 39 circumcised males and the 60 control subjects.

All three studies reported differences in some aspects of behavior in circumcised infants and are consistent with our observation that male infants behave differently after circumcision than a control group.

Why did the item analysis of NBAS not reveal differences, although the behavior categories procedure did? There are at least two possible explanations. One, the kinds of changes seen after circumcision were not reflected evenly through out all 27 behavioral items. Though none of the items were statistically significant, some items did change more than others. In our data, startle, self-quieting, response to rattle and rapidity of build-up showed most change between early and late circumcision groups. Therefore, if all 27 items are analyzed as a unit, then changing items may be dampened against the background of no-change items. Two, the NBAS was designed to elicit the maximal potential that an infant can display during the examination (Brazelton, 1973). If any infant gives a best score for a brief period and a less optimal score for most of the examination, one is required to record the best score. such a constrain obscures differences that are related to an intervention like circumcision. This methodological problem has been addressed by Horowitz, who developed a modal scoring system for NBAS items in an attempt to capture the behavior exhibited by baby during most of the examinations. (Horowitz, Sullivan, & Linn, in press).

The reduction scale presented in this study utilizes 7 NBAS items. Three of these items - irritability, peak of excitement, and rapidity of build up - reflect stimuli during the NBAS examination that upset the infant. Two items - self quieting and consolability with intervention - reflect the infant's ability to be consoled by himself or others. All of these items capture a baby's capacity to control and modulate his state behavior. It is not surprising that these dimensions of behavior would be altered after a procedure like circumcision.

The other two items in the reduction reflect global behaviors - lability of state and alertness. Changes in these two categories reflect the overall impression which the infant presents to the examiner and influence whether an infant seems average, subdued, or hyperactive. It must be stated that the reduction scale requires the seven NBAS items and the three items from the summary paragraph for its application. When attempts were made to utilize either the seven NBAS items or the summary paragraph alone, reliable assignment of infants to the appropriate behavior category was not possible.

We think the results of the study are significant for four reasons. First, we have developed a method of reducing the NBAS that permits detailed examination of behaviors, yet emerges with general statements about changes after intervention. Second, our results demonstrate than infants change behaviors after circumcision in about 90 percent of cases. Families, nurses, and physicians have appreciated that babies are different after circumcision, but there have not been controlled studies using the NBAS scale to substantiate these changes. Third, investigators of infant behavior must recognize the circumcision status of their subjects when doing research on newborn male subjects. We have shown that behavior differences after circumcision in 4/12 (33%) subjects persisted for at least 22 hours.

Richards and colleagues suggest that some behavioral differences between male and female babies may be related to circumcision effects. (Richards, Bernal, & Brackbill, 1976). Fourth, the infant's response to the circumcision procedure is diverse, not uniform. Our initial bias was that all subjects would become more active and irritable during the examination after circumcision (E2). Our data did not support that impression. Some infants react to the stress of circumcision by becoming more active. This diversity of individual infants' responses to a stressful procedure like circumcision suggest that they have different coping styles from birth.

The criticism may be raised that the differences observed between the early and delayed circumcision groups were real but trivial because they may have no long term significance. Unfortunately, our three week outpatient NBAS examinations did not generate reliable data. Only 50 percent of the early circumcision group returned to the hospital, and the infants were not frequently in the appropriate states to permit adequate and complete examinations. Thus, we cannot make any direct statements from our data at this point. However, data from diverse investigators have shown that newborns have greater response competency and capacity for storage of memory, than previously anticipated. (Stone, Smith and Murphy, 1973; Lipsitt, 1977). Klaus and Kennell (1976), O'Connor, Vietze, Hopkins, and Altmeier (1977), and De Chateau and Wiberg (1977) have shown, moreover, that newborn experience may have long term consequence for both mothers and infants. It seems inappropriate, therefore, to make the assumption in the absence of direct evidence that there are no long-term consequences of the circumcision experience (Lipsitt, 1977; Richards, 1976). As suggested by Lipsitt, even without memory of the specific circumcision event the painful experience might affect the infant's subsequent behavior, which in turn may determine other's responses to the baby and thus affect subsequent environmental inputs to the child. [CIRP Note: Later research verified long-lasting changes in behavior. See Effect of neonatal circumcision on pain response during subsequent routine vaccination.]

A recent report described a method for reducing pain during circumcision by using dorsal penile nerve block as a means of local anesthesia (Kirya & Wirthman, 1978). Because the infant's behavior changes after circumcision may be due to pain after circumcision, such an invasive procedure may not be warranted. However, further data is required. It is hoped that the results of this study will stimulate others to examine methods used and behavioral consequences of providing routine care for newborns.


We are indebted to Barbara Anderson for continuous advice and support; to obstetric and pediatric nursing staff at Barnes Hospital for assistance in facilitating patient selection; to Ann Rogers and Peggy Goldernhersh for independent data evaluations; to Robert Brioulette, Philip Dodge, Frances Horowitz, Marshall Klaus, John Kennell, Steve Robinson, Bradley Thach and Joseph Volpe for critical evaluations of an earlier form of this manuscript; and to Lois Price for assistance in preparing the manuscript.


Anders, T. F. & Chalemian, R. J. The effects of circumcision on sleep-wake states in human neonates. Psychosomatic Medicine, 1974, 36, 174-179.

Brackbill, Y. Continuous stimulation and arousal level in infancy: effects of stimulus intensity and stress. Child Development. 1975, 46, 364-369.

Brazelton, T. B. Neonatal behavioral assessment scale. Philadelphia, Pa.: J. B. Lippincott, 1973.

De Chateau, P., & Wiberg, B. Longterm effects of extra contact during the first hour post partum. II. A follow up at three months. Acta Paediatrica Scandinavica, 1977, 66, 145-151.

Dubowitz, L. M., Dubowitz, V., & Goldberg, C. Clinical assessment of gestational age in the newborn infant. Journal of Pediatrics, 1970, 77, 1-10.

Emde, R., Harmon, R., Metcalf, D. Koenig, K., & Wagonfeld S. Stress and neonatal sleep. Psychosomatic Medicine, 1971, 33, 491-497.

Everitt, B. S. The Analysis of Contingency Tables. New York: J. Wiley, 1977.

Grimes, D. A. Routine circumcision of the newborn infant; A reappraisal. American Journal of Obstetrics & Gynecology. 1978, 130, 125-129.

Horowitz, F. D., Sullivan, J. W., Linn P. III. Stability and instability in newborn infants; the quest for elusive threads. In A. Sameroff (Ed.), The Brazelton Neonatal Behavioral Assessment Scale: A commentary. Monographs of The Society for Research in Child Development. In press.

Kaplan, G. W. Circumcision - an overview. Current Problems in Pediatrics. 1977, 7, 1-33.

Kirya, C., & Werthmann, M. W., Jr. Neonatal circumcision and penile dorsal nerve block - painless procedure. Journal of Pediatrics, 1978, 92, 998-1000.

Klaus, M. H., & Kennell, J. H. Maternal-Infant Bonding. St. Louis, Mo.: The C. V. Mosby Company, 1976.

Lipsitt, L. P. The study of sensory and learning processes in the newborn. Clinics in Perinatology. 1977, 4, 163-186.

Littman, B. & Parmelee, A. Manual of obstetrical complications. Unpublished manuscript available from A. Parmelee at the Department of Pediatrics, Division of Child Development, School of Medicine, University of California, Los Angeles, California 90024. The Littman-Parmellee manual for obstetrical complications is a 41-item index of which 37 pertain to the mother. The mean score is 100 with a standard deviation of plus or minus 20. A high score indicates few obstetrical complications.

Moore, J. A., & Marshall, R. E. Unpublished observations.

O'Connor, S., Vietze, P., Hopkins, J. & Altmeier, W. Post-partum extended contact. Subsequent mothering and child health. Pediatric Research, 1977., 11, 380.

Richards, M. P. M., Bernal, J., & Brackbill, Y. Early behavior differences: gender or circumcision? Developmental Psychobiology, 1976, 9, 89-95.

Stone, L. J., Smith, H., & Murphy, L. B. (Eds.), The Competent Infant: Research and Commentary. New York, N.Y.: Basic Books, 1973.

*Requests for reprints should be sent to Richard E. Marshall, Department of Pediatrics, St. Louis Children's Hospital, 500 S. Kingshighway, P. O. Box 14871, St. Louis, Missouri 63178.

(File revised 10 August 2005)