research and studies Relationship of Early Bonding to Initial Infant Feeding Patterns in Bottle-fed Newborns M A R Y S. BRODISH, RN,M N To determine the relationship between early bonding in the immediate post-delivery period and initial infant feeding patterns in bottle-fed newborns, data were gathered from the charts of 85 full-term normal infants. Of the infants, 57 had an early, extra bonding experience in the postpartum recovery room; 28 did not. In comparing the bond and no-bond groups for weight loss and formula intake, the t-test analyses showed a statistically significant effect of bonding that was beneficial to infant status.
The effect of early parental-infant attachment experience on the neonate has been the focus of much literature. Previous studies demonstrated a positive relationship between breastfeeding and early, extra maternal-infant contact: A beneficial effect on the continuation of breastfeeding, weight gain over a two-month period, fewer infections, and fewer parenting disorders were demonstrated.’>’ Infant feeding is a significant task by which women frequently measure their competence in the mothering role.3 Health care professionals also use infant feeding as a measure of maternal performance and infant well-being4 A study was conducted to determine the relationship between the opportunity for an early, extra bonding experience and initial infant feeding patterns in bottle-fed newborns during the first three days of life.
Methodology A nonexperimental, correlation study was conducted at a 790-bed 248
general hospital in the southeastern United States. T h e hospital is affiliated with a medical school for its obstetrical and neonatal services. There are both private and staff/ service patients. The obstetrical routines at the hospital are relatively progressive, with emphasis upon a family-centered approach to maternal-newborn care. Husbands, or substitute support persons, are encouraged to participate in the labor and delivery and in the parent-infant bonding experience in the postpartum recovery room. Post-delivery contact time generally is encouraged by the nursing personnel except in instances when they are too shortstaffed for adequate supervision of the neonate. Because the post-delivery bonding experience is encouraged, the total sample of 85 in this study included more mothers and infants who had some extra contact time in the postpartum recovery room (57) than those who did not have the additional time for bonding following a minimal con-
tact experience in the delivery room (28). This criterion of extra postpartum recovery room bonding time was used to distinguish the bonding group from the no-bonding group in the data analysis. After receiving the approval of a human subjects review committee, and agency consent to approach mothers for consent to collect data from their infants’ charts, a sample of 85 infants was secured. This sample was selected according to the following criteria, which are generally accepted as indicative of a full-term normal newborn: Gestational age: 37 to 42 weeks Normal spontaneous vaginal delivery (forceps permissable) Vertex presentation Maternal anesthesia: pudendal, epidural, local, or none No maternal analgesia within one hour of delivery No prenatal or intrapartal complications, such as pre-eclampsia/ eclampsia, anemia, vaginal bleeding, premature ruptured July/August 1982 JOGN Nursing
0090-0311/82/0713-0248800.75
membranes, maternal medical problems No evidence of fetal distress Apgar scores at 1 and 5 minutes of 8, 9, or 10 Appropriate size for gestational age (AGA) (weight between 5% lbs and 9 lbs per GA) No observable congenital abnormalities The sample was limited further to bottle-fed infants in order to accurately assess specific nutritional intake. Data were gathered regarding the independent variable of the opportunity for extra bonding experience in the postpartum recovery room, and the dependent variables of weight loss during the first three days of life and volume of formula intake for the first 12 feedings. Pertinent demographic data also were obtained to take into account possible. intervening variables, e.g., birth weight, maternal age, parity, and marital status. For this study, the probability level for acceptance or rejection of a variable for being significant was set at the 0.05 level. Data Analysis
Research studies involving human subjects must consider, in the data analysis, the effect of organismic and/or intervening variables. These so-called covariates can have a relationship to the dependent variable similar to that of an additional independent variable. T o take into account the possible effect of covariates in this study, a multivariate analysis was done using the Biomedical Computer Programs P-Series, BMD-77.5 T h e covariates that were considered with the independent variable of bonding or no-bonding in their relationship to the criterion measure of weight change during the first three days of life included birth weight, sex, maternal age, marital status, parity, type of anesthesia, source of medical care, and the amount of formula intake, July/August 1982 JOGN Nursing
Results
Of the total sample of 85 infants, 57 had an opportunity for early, extra bonding in the postpartum recovery room, and 28 did not have this opportunity. The independent variable of bonding was found to have a significant ( p = 0.029) beneficial effect on weight loss during the first three days of life. The multivariate analysis was used: the covariates of birth weight and formula intake were also significant ( p < 0.025) in relationship to weight loss during the first three days of life. (Heavier neonates tend to lose more weight during the first three days of life than infants of average weight.6 This weight loss has less consequence as far as percent of body weight change because of the larger initial size of the infant and is thought to be mostly fluid loss. It is logical that infants who drink more formula will lose less weight.) Having determined by the multivariate analysis that the covariates of birth weight and formula intake were significant, the independent variable of bonding or no-bonding was tested with these two covariates in relationship to the criterion measure of weight change during the first three days of life. The inde-
Tabie 1.
pendent variable of bonding was p = 0.0385; the level of significance of the two covariates remained at p < 0.025. Data analysis indicates that, to a statistically significant degree, infants who had early, extra bonding lost less weight than those infants who did not have the extra bonding. T h e hypothesis that a beneficial effect of an early bonding experience on initial infant feeding patterns would be reflected by less weight loss during the first three days of life was supported. A second criterion measure of volume of formula intake for the first 12 feedings also was used. It was hypothesized that the benefit of the early, extra bonding in the postpartum recovery room would be reflected by the consumption of a greater volume of formula. A ttest analysis was done comparing the means of the bond and no-bond groups for formula intake in ounces for the first 12 formula feedings. A p value of 0.050 supports the hypothesis and indicates that the infants in the early, extra bonding group drank more fopmula than those who did not have the extra bonding. A t-test analysis comparing the bond and no-bond groups for the criterion measure of weight change
Comparison of Weight LOSSand Birth Weight Weight Loss
Birth Weight
n
5 3 oz.
5%-a ibs. with bond without bond
59 42 17
42 33 9
17 9
5%-7 Ibs. with bond without bond
22 14
15 11
7 3
a
4
4
7-8 Ibs. with bond without bond
37 28 9
27 22 5
10 6 4
8-9 ibs. with bond without bond
26 17 9
13 11 2
13 6 7
>3
02.
a
249
~~~
Table 2.
Compclrison of Formula Intake and Birth Weight Formula Intake Birth Weight
n
5%-7 Ibs. with bond without bond
18 11 7
4
7-8 Ibs. with bond without bond
31 23 8
20 16 4
11 7
8-9 Ibs. with bond without bond
24 14 10
14 9 5
10 5 5
>15 oz.
14 9 5
2 2
4
the statistical analysis of the data. Infants who had the early, extra bonding experience in the postpartum recovery room lost less weight and had better nutritional intake than those who did not have extra bonding. When the percent of body weight loss during the first three days of life for these two groups of infants was compared, a similar spread of approximately seven percentage points was found in both groups; however, the spread was shifted upward for those who had the early, extra bonding experience from +1.3 to -6, as compared to 0 to -7 for those without the experience. The means for the respective groups were 2.0% of body weight loss for those with the extra, early bonding experience compared
during the first three days of life gave a p of 0.00875. This confirms the results found by the multivariate analysis of the beneficial effect of the early, extra bonding experience as reflected by less weight lost during the first three days of life. A further statistical analysis was done to determine whether the length of time of the extra bonding experience had a direct relationship to weight change. A regression analysis was done to determine if there was a correlation between time and weight change. Linear, quadratic, cubic, and quartic equations were used. No significant pattern could be detected. The hypothesis that early bonding had a beneficial effect on initial feeding patterns was supported by
Table 3.
<15 oz.
Reiatlon8hip of Parity to Formula Intake and Weight Loss Weight Loss
Formula Intake
n
>15 oz.
<15 oz.
n
53 oz.
>3 oz.
Total
73
37
36
85
57
28
Primip with bond without bond
31
16
15
37 21 16
26 18 8
11 3 8
Multip with bond without bond
42
21
21
48 38 10
31 26 5
17 12 5
Parity
250
to 2.7% of body weight loss for those without the experience. When the number of ounces lost at the end of the third day of life was tabulated for each of the infants and compared, 55 of the 85 infants lost three ounces or less, and 30 infants lost more than three ounces. When the sample was divided into two groups according to bond and no-bond experience and compared, 43 of the 57 infants who had had the early, extra bonding lost three ounces or less and 14 lost more than three ounces; of the 28 infants who did not have this experience, only 12 lost three ounces or less and 16 lost more than three ounces. When the data further were analyzed for the nutritional intake of the bottle-fed newborns, the sample size used for tabulation of the total ounces consumed by each infant for the first 12 formula feedings was 73 because 12 of the 85 did not have accurate or complete data for the 12 formula feedings. Of the 73 infants, 48 had early, extra bonding and 25 did not. Thirty-seven infants consumed more than 15 ounces of formula for the first 12 feedings, and 36 consumed less than 15 ounces. When the bonding and no-bonding groups were compared, of the 48 infants who had the early, extra bonding, 27 consumed more than 15 ounces and 2 1 consumed less than 15 ounces. In contrast, for those 25 infants who did not have the extra bonding, only 11 consumed more than 15 ounces and 14 consumed less than 15 ounces. Covariates that were identified as having a possible effect upon initial weight loss were examined. The most significant covariate in relation to weight loss during the first three days of life was birth weight (Table 1). Half of the infants weighing more than eight pounds lost more than three ounces. Of the infants who weighed less
July/August 1982 JOGN Nursing
than eight pounds, only 16 lost more than three ounces. The infants who did least well as a group weighed over eight pounds and did not have early, extra bonding; those who did best as a group weighed under eight pounds and had extrp bonding. When the relationship between the dependent variable of nutritional intake and the birth weight as a covariate was analyzed, those infants who weighed the least consumed the least (Table 2). When the bonding and no-bonding groups were compared in the different categories, the bonding experience seemed to have a beneficial effect for infants weighing over seven pounds. T h e smaller infants consistently took less formula, whether or not there had been early, extra bonding (Table 2). In a study of the relationship between parity and early infant feeding behavior, Thoman observed that primiparas took longer to attach to their infants and to feed them.’ However, in this study the difference was not demonstrated. When the primipara and multipara groups were subdivided into bond and no-bond groups and compared, more of the infants who had early, extra bonding lost less weight than those who did not have the experience, regardless of parity. This was especially true within the primipara group (Table 3). T h e covariates of maternal age and marital status were analyzed by first categorizing the sample according to age groups, then subdividing these categories as to marital status, and again subdividing according to bond or no-bond. T h e infants who had early, extra bonding lost less weight than those who did not. Of the group of ten single mothers under 20 years of age, only three had the early, extra bonding experience. With the benefit of the extra bonding, the infants of these three mothers did very well.
July/August 1982 JOGN Nursing
Table 4.
Relationship of Length of Labor to Infant Weight Loss
-
Weight Loss Length of Labor
n
Total under 8 hrs. with bond without bond
32 23 9
21 15 6
11 8 3
Total over 8 hrs. (unrecorded) with bond without bond
28 25 17 11
18
10
15 3
a
53
A large number of mothers had labors over eight hours in length, with no opportunity for early, extra bonding after delivery; their infants lost more than three ounces during the first three days of life. There was no significant relationship between weight loss and length of labor when only the latter was considered (Table 4). There was no significant relationship between the different types of intrapartal anesthesia and their relationship to neonatal weight loss during the first three postpartum days. Less weight loss was seen in infants whose mothers had epidural anesthesia and pudendal block than in infants whose mothers had either local or no anesthesia. However, infants whose mothers had epidural anesthesia unassociated with early, extra bonding did least well. Rubin suggests that mothers do not relate as well to male infants
Table 5.
>3 02.
02.
2
as to females8 In this study, there was a slight tendency toward less weight loss in females than in males. There was a greater difference when the variable of early, extra bonding also was considered. Male infants without the benefit of extra bonding did least well as a group (Table 5). When there was early, extra bonding in both the higher and the lower socioeconomic groups, there was no significant difference between the groups. However, when there was no bonding experience, the lower socioeconomic group did better than the higher socioeconomic group (Table 6 ) .
Discussion When a regression analysis was done to determine if there was a correlation between length of bonding experience and weight change, no significant pattern was found.
Relationship of Infant Gender to Weight Loss
w
Weight Loss Gender
n
s3
02.
>3
02.
Total males males with bond males without bond
4a 29 19
28 20 8
20 9 11
Total females females with bond females without bond
37 26 11
27 20 7
10 6 4
251
ITable 6.
I
Relationshb of Socioeconomic Level to Infant Weight Loss Weight Loss
Socioeconomic Level
n
53 oz.
Total higher with bond without bond
57 39 18
37 31 6
20
Total lower (unrecorded) with bond without bond
26 12 16 10
ia
a
12 6
4 4
However, if a larger sample with longer bonding times was tested, there might be an observable relationship. In this study, only nine infants had a total bonding experience time in excess of 30 minutes. Only limited examination was made of the data for the possible relationship between socioeconomic level, as determined on the basis of source of medical care, and weight loss during the first three days of life. More definitive studies need to be done using a larger sample of known different socioeconomic levels. As the study data were analyzed in all the various subgroups of covariates, the early bonding experience was observed to have a positive relationship toward infant well-being during the first three postpartum days. This was demonstrated by improved nutritional intake and less weight loss. This positive relationship should encourage more research in this area and reinforce the previously claimed benefits of early parent-infant attachment. The implications for nursing are toward continued encouragement of early mother-infant interaction, education of parents re-
252
>3
02.
a 12
garding the newborn's potential for reciprocal interaction, and support of parents in their caregiving endeavors. Conclusion Although an early bonding experience is not essential for the survival of the neonate, it would appear to facilitate the transition from the maternal-fetal synchrony to a positive maternal-infant attachment. T h e newborn has an ability to interact with his mother during the first hour of life while in an alert state. This ability for reciprocal interaction would appear to be a phylogenetic safeguard to assure the neonate of the loving maternal care so essential for his prolonged period of dependence. The sequence of events which transpire during the bonding experience appears to crystallize the maternalinfant attachment and assure the commencement of the mother-caregiver role. Acknowledgment The author thanks Dr. Darwin Peterson for his consultation and assistance in the computerized statistical analyses of the data.
References 1. Klaus M H , Kennell J H . Maternal infant bonding. St. Louis: CV Mosby Co, 1976. 2. Lozoff B, Brittenham G M , Trause MA, Kennell JH , Klaus M H . T h e mother-newborn relationship: limits of adaptability. J Pediatr 1977;91:1-12. 3. Dunbar J. Maternal contact behaviors with the newborn infant during feeding. M C N 1977;6:20995. 4. Klaus M H , Trause MA, Kennell JH. Does human maternal behavior after delivery show a characteristic pattern? In: Parent infant interaction. Ciba Foundation Symposium #33. Amsterdam: Elsvier Publishing Co, 1975: 69-78. 5. Dixon WJ. BMD-77 biomedical computer program P. Berkeley, Ca.: University of California Press, 1977:526-9. 6. Olds SB, London M L , Ladewig PA, Davidson SV. Obstetric nursing. Menlo Park, Ca.: AddisonWesley Publishing Co., 1980:665. 7. Thoman EB et al. Neonate mother interaction: effects of parity on feeding behavior. Child Dev 1970;41:1103-11. 8. Rubin R. Attainment of the maternal role. Parts I & 11. Processes, and Models and referrants. Nurs Res 1967;14:237-45, 342-6.
Address for correspondence: Mary S. Brodish, RN , 4135 Briarcreek Road, Clemmons, NC 27012.
Mary Brodish is assistant professor at the University of North Carolina at Greensboro's School of Nursing. Prior to that, Ms. Brodish was on the faculty of the College of Nursing and Health at the University of Cincinnati. A member of NAACOG, NLN, and ANA, Ms. Brodish attended Wellesley College (BA), New York University (MA), and Yale University (MN).
July/August 1982 jOGN Nursing