A prospective study of the effect of delivery type on neonatal weight gain pattern in exclusively breastfed neonates born in Shiraz, Iran
© Saki et al; licensee BioMed Central Ltd. 2010
Received: 15 July 2009
Accepted: 27 January 2010
Published: 27 January 2010
In this exploratory study, the contribution of delivery type to the weight gain pattern for full-term infants with exclusive breastfeeding in the first month of infancy was determined. In addition, breastfeeding success among cesarean section (C-section) delivery mothers based on their neonate's weight gain at the end of the first month of infancy was evaluated.
A cohort of 92 neonates born in Shiraz, from July 10 to August 10, 2007 was followed longitudinally. The data were collected during the first month postpartum at three occasions: 3 to 7 days postpartum, 10-21 days postpartum and 24-31 days postpartum.
Among 92 mothers in this study, 35 (38%) were delivered by C-section. Generalized estimating equation (GEE) showed that delivery type (p < 0.01), receipt of advice about breastfeeding (p = 0.03) and neonate's age (p < 0.01) significantly affected weight gain. GEE estimated the values of the parameters under study and the testing contribution of each factor to weight gain, leading to the conclusion that gender, parities and maternal education did not contribute to weight gain. The neonate's weight gain pattern for C-section deliveries lies below that of normal vaginal deliveries until 25 days postpartum, when weight gain for C-section deliveries became higher than that for normal vaginal deliveries.
Type of delivery contributes strongly to the weight gain pattern in the first month of infancy. In spite of greater weight loss among C-section birth neonates in the first days of life, at the end of the first month neonates showed a similar weight gain. Consequently, mothers with C-section delivery can successfully exclusively breastfeed.
It is normal for newborns to lose weight during the first days of life. Although much of this weight loss is thought to be due to changes in the volume and distribution of water in the body, some studies show that early skin-to-skin contact, initiating breastfeeding as soon as possible, and feeding practices also influence the degree of weight loss [1–5]. Dehydration and/or failure to thrive during the first days postpartum may occur as a result of lactation failure and lack of awareness about feeding problems. Recent reports recommend monitoring infants' weight through the neonatal period [6–8].
Extensive research on the biology of human milk and health outcomes associated with normal methods of infant feeding have established that breastfeeding is more beneficial than formula feeding, although breastfed infants initially lose more weight and take longer to regain their birth weight than formula fed infants [9–14]. One of the factors affecting breastfeeding initiation and duration is birth by cesarean section [15–17].
In recent years C-section has been performed upon request for births that would otherwise have been vaginal [18, 19]. In Iran, the C-section rate is about 50%-65%, however in some private hospitals the rate reaches 90% .
A study in Mexico shows that C-section is a risk factor for not initiating breastfeeding (Odds Ratio, OR = 1.56) and for breastfeeding for less than one month (OR = 1.72), but it is unrelated to the duration of breastfeeding among women who breastfeed their babies for one month or more (OR = 1.03) . Similar studies performed in Puerto Rico and Athens show that C-section delivery was negatively related to breastfeeding initiation (OR = 1.56, OR = 4.16) [16, 17]. Breastfeeding in post-cesarean women has a protective effect on infant health, as demonstrated by the decrease in illness-related hospitalizations in the first year of life .
International agencies emphasize the need for exclusive breastfeeding during the first months of life. A new international growth standard chart has been prepared based on children who are fed according to World Health Organization recommendations, which entail exclusive breastfeeding for the first six months of life . An increase in weight indicates a child's well-being. The rate of growth is relatively high during an infant's first months of life and is susceptible to decelerating forces that may compromise a child's ultimate level of growth. Research evidence showed both biological and experiential conditions influence growth. Biologic conditions that may influence these components include gender and gestational maturity at birth, that is, premature or full-term. Mother's care-giving and infant feeding behavior also influence infant growth [23, 24].
The neonate growth pattern does not have a uniform rate of increase during the first month postpartum. Weight loss, rather than weight gain, may occur in the first week postpartum.
To the best of our knowledge, there are no studies exploring the relationship between the type of delivery (cesarean-section (CS) or normal vaginal (NV)) and the pattern of neonatal weight gain during the first month. Therefore, the purpose of this study is to explore the impact of type of delivery on the pattern of weight gain for exclusively breastfed neonates through the first month postpartum. A second purpose is to compare infant weight at first month postpartum in exclusively breastfed infants who were born by vaginal birth or cesarean-section.
Sample collection and follow-up
A sample of mothers of singleton full-term infants weighing ≥ 2500 g who were exclusively breastfeeding  and presented at Shiraz health care centers within three to seven days postpartum were recruited to participate in the study. All the mothers initiated breastfeeding within two hours after birth. The participants were recruited from 10 July to 10 August 2007 and were followed up for one month. Initially, 104 mother-infant pairs were recruited. However, if the mothers reported they were no longer exclusively providing breast milk to their neonate at follow-up visits they were excluded from further participation in the study. In addition, if the neonate was hospitalized during the study, the mother-infant pair was excluded. Based on these criteria, 10 (9.6%) mother-infant pairs [4 (10.3%) from the CS group and 6 (9.5%) from the NV group] were excluded because of formula use and 2 (1.9%) mother-infant pairs (one from the CS group and another from the NV group) were excluded due to infant hospitalization during the study. Thus, at the end of the study, 12 pairs were excluded, leaving 92 (88.5%) mothers and their exclusively breastfed neonates remaining.
Two data collection tools were developed by the researchers; The first questionnaire, completed at the time of recruitment to the study, included maternal and neonatal demographic and background data [neonate's date of birth, gender, birth weight and birth length (according to hospital records); mother's age, education, smoker (yes/no), type of delivery (NV/CS), type of maternity hospital (private/public), parity, previous breastfeeding experience (yes/no), and family income]. The second questionnaire, completed at recruitment and then at the second and third occasions, was used to collect neonatal assessment data (anthropometric measures and health status), current feeding pattern [breastfeeding exclusively ("Have you fed your baby anything except mother's milk?")]. Mothers were also asked if they received professional advice about breastfeeding at healthcare centers or maternity hospitals ("Have you received advice about breastfeeding till now?").
The questionnaires were completed during the first month at three occasions: 3 to 7 days postpartum, 10-21 days postpartum and 24-31 days postpartum, when mothers presented in healthcare centers. One nurse at each of the seven healthcare centers was responsible for completing the neonatal assessment and questionnaires. The nurses were trained in the measurement of anthropometric indices by a healthcare center physician. Inter-rater reliability among the nurses was checked by the inspectors of the Deputy for Health at Shiraz University of Medical Sciences.
The weights were measured to the nearest 10 grams on sophisticated balance scales calibrated at each healthcare center. The nurses were instructed to weigh the neonates naked. If this was not possible, the type of clothing was recorded. Later, using these data, the neonate's weight was adjusted for baby clothes. Age at each measurement was recorded exactly based on the difference between the date of measurement and date of birth in days.
Where σ*2 = 29.16 is the sum of the variance components obtained with a pilot study (n = 10). The power of the study was 80% with α = 0.05 and the meaningful difference between CS and NV birth neonates considered as 5 g. Assuming 20% withdrawal, n ≥ 37 for each group were required. Therefore we continued data collection until we had 40 CS delivery mothers; in this period 64 NV delivery mothers were recruited.
Generalized Estimating Equation (GEE) modeling was used to determine factors related to neonatal weight gain. The Generalized Estimating Equation approach is widely used in biomedical sciences for the analysis of longitudinal data. Models such as Generalized Estimating Equation allow for the correlation structure in the data due to the repeated measurements on the same subjects over time. Generalized Estimating Equation modeling has many attractive robust properties and consistent parameter estimations that are not prejudiced by incorrect specification of the correlation structure. The GEE approach is based on the concept of "estimating equations" and provides a very general and unified approach for analyzing correlated responses that can be discrete or continuous. The essential idea behind the GEE approach is to generalize and extend the usual likelihood equations for a generalized linear model for a univariate response by incorporating the covariance matrix of the vector of responses [26–32].
Locally Weighted Scatterplot Smoothing (LOESS) approach was also used to obtain weight gain pattern. The LOESS is one of many "modern" modeling methods that build on "classical" methods, such as linear and nonlinear least squares regression. LOESS allows greater flexibility because no assumptions about the parametric form of the regression surface are needed. Response yi and corresponding predictor measurement xi were related by yi = g(xi) + εi, for i = 1,..., n, where g is the regression function. A local approximation is obtained by fitting a regression surface to the data points within a chosen neighborhood of the point x .
Independent T-test was used to compare the mean of different factors between CS and NV delivery type. Chi-square test was also used to investigate the association between family income and maternal education level with types of delivery. P-values less than or equal to 0.05 were considered significant. SPSS11.5 and S-Plus2000 statistical software were used for data analysis.
Ethical approval was obtained from the Ethic in Research Committee at the Deputy of Research of Tehran University of Medical Sciences. There were no anticipated physical, social or legal risks associated with participation. Informed consent was implied if participants completed the first questionnaire. It is standard practice in Iranian healthcare centers to ask participants to complete questionnaires at health checks without written consent.
Descriptive statistics for background variables for mother and infant by type of delivery
Normal vaginal delivery
(n = 57)
(n = 35)
(n = 92)
Girl, n (%)
Boy, n (%)
Birth weight (g), mean (SD)
Birth length (cm), mean (SD)
Mother's age, mean (SD)
Primary, n (%)
Secondary, n (%)
University, n (%)
Maternal smoking, n (%)
Primipara, n (%)
Multipara, n (%)
Poor, n (%)
Fair, n (%)
Good, n (%)
Excellent, n (%)
Private, n (%)
Public, n (%)
Received advice about breastfeeding
First occasion, n (%)
Second occasion, n (%)
Third occasion, n (%)
Mean and SD of the neonate's weight (g) at different occasions by gender and delivery types
Type of delivery
(median 5 days)
(median 15 days)
(median 30 days)
Parameters estimation, standard errors and p-values from fitted model for neonatal weight gain at three occasions
Type of delivery
Received advice about breastfeeding
Type of delivery * neonate age
Neonate age (in days)
The most important finding of this study was the strongly significant association between type of delivery and neonatal weight gain. According to the results displayed in Table 3, neonates born by vaginal birth gained 14 gram weight per day more than those born by cesarean section. In Iranian hospitals, most of the C-sections are performed with general anesthesia, so very early skin-to-skin contact and breastfeeding occur immediately post-cesarean. That negatively affects milk supply and breastfeeding practices, and as a result neonates' weight gain during the early postpartum period. However, due to the significant negative interaction between the infant's age and type of delivery, this difference has decreased over time. This means that the effect of CS delivery has been reduced as well. Therefore, as seen in Figure 5, the CS weight gain curve reaches the NV weight gain curve in 25 days postpartum. After 25 days, the weight gain for CS deliveries is significantly greater than that in NV deliveries. We found that infant gender, maternal education and parity did not contribute significantly to weight gain during the first month of infancy.
Receiving advice about breastfeeding is another significant factor for neonatal weight gain. As shown in Table 3, neonates whose mothers received advice about breastfeeding gain 7.7-gram weight more than those who did not. This factor does not interact with age, but does have a constant effect during the first month of infancy. Higher rates of receiving advice about breastfeeding among CS delivery mothers (Table 1) may have helped them in successful breastfeeding practices and improve their neonate's weight gain at the end of the first month following CS.
No previous growth charts, including World Health Organization Growth Standards Charts, have focused on the first month of infancy . A monotonous increase in weight is shown, but most of the breastfed neonates lose weight in the first days postpartum. As shown in Figure 5, neonatal weight decreased in the first days of life (weight loss) and after 5-7 days increased (weight gain). Therefore, we obtained a weight chart for the first month of infancy by age and delivery type separately for boys (Figure 6) and girls (Figure 7). These figures show that, disregarding the type of delivery, weight of the neonates that exclusively breastfed at the end of the first month was higher than their standard percentiles at birth.
For some mothers, the reason for the neonate's weight loss in the first days postpartum is their insufficient milk. This causes them to use formula feeding and may lead to early breastfeeding cessation . The result of this study shows that (Figure 5) the weight gain pattern improves after the first week and mothers can be hopeful that their neonates will gain more than 40 grams of weight per day.
The model used to investigate the related factors in this research is more robust than those examined in earlier studies. The advantage of these models compared with other methods in this area is that the size of each covariate effect using regression model parameters is introduced.
Gender, mother's education and parity did not contribute to weight loss in the first days postpartum; however delivery type and receiving advice about breastfeeding contribute strongly to the weight gain pattern in the first month of infancy. Neonates with CS delivery in the first days postpartum lose more weight than those with NV delivery; however at the end of the first month there is no difference between the weights of breastfed infants born by CS or NV delivery. Consequently, if mothers with CS delivery continued exclusively breastfeeding they could have successful breastfeeding and these results lead to calling for early skin-to-skin contact and support of professionals and family post-cesarean delivery.
This research was carried out in the Department of Biostatistics and Epidemiology of the School of Public Health and Institute of Public Health Research, Tehran University of Medical Sciences. The authors would like to thank them for their financial support. We also thank the Deputy for Health at Shiraz University of Medical Sciences for their help in data collection. The authors would also like to thank Mrs. Sheryl Thomas-Nikpoor for editorial assistance.
- Feigelman S: The first year. Nelson Textbook of Pediatrics. 2007, Philadelphia PA. Saunders, an imprint of Elsevier Inc, 43-44. 18Google Scholar
- Bystrova K, Widstrom AM, Matthiesen AS, Ransjo-Arvidson B, Welles-Nystrom B, Voronsov I, Uvnas-Moberg K: Early lactation performance in primiparous and multiparous women in relation to different maternity home practices. A randomized trial in St. Petersburg. International Breastfeeding Journal. 2007, 2: 9-10.1186/1746-4358-2-9.PubMed CentralView ArticlePubMedGoogle Scholar
- Bystrova K, Matthiesen AS, Widstrom AM, Ransjo-Arvidson AB, Welles-Nystrom B, Vorontsov I, Uvnas-Moberg K: The effect of Russian maternity home routines on breastfeeding and neonatal weight loss with special reference to swaddling. Early Human Development. 2007, 83 (1): 29-39. 10.1016/j.earlhumdev.2006.03.016.View ArticlePubMedGoogle Scholar
- Nommsen-Rivers LA, Mastergeorge AM, Hensen RL, Cullum AS, Dewey KG: Doula care, early breastfeeding outcomes, and breastfeeding status at 6 weeks postpartum among low-income primiparae. Journal of Obstetric, Gynecologic, and Neonatal Nursing. 2009, 38 (2): 157-73. 10.1111/j.1552-6909.2009.01005.x.View ArticlePubMedGoogle Scholar
- Avoa A, Fischer PR: The influence of prenatal instruction about breastfeeding on neonatal weight loss. Pediatrics. 1990, 86 (2): 313-315.PubMedGoogle Scholar
- Macdonald PD, Ross SR, Grant L, Young D: Neonatal weight loss in breast and formula fed infants. Archives of Disease in Childhood- Fetal and Neonatal Edition. 2003, 88: F472-F476. 10.1136/fn.88.6.F472.PubMed CentralView ArticlePubMedGoogle Scholar
- Harding D, Cairns P, Gupta S, Cowan F: Hypernatraemia: Why bother weighing breastfed babies?. Archives of Disease in Childhood- Fetal and Neonatal Edition. 2001, 85: F145-10.1136/fn.85.2.F145.PubMed CentralView ArticlePubMedGoogle Scholar
- Oddie S, Richmand S, Coulthard M: Hypernatraemic dehydration and breastfeeding: a population study. Archives of Disease in Childhood- Fetal and Neonatal Edition. 2001, 85: 318-20.View ArticleGoogle Scholar
- Ernst JA, Bull MJ, Rickard KA, Brady MS, Lemons JA: Growth outcome and feeding practices of the very low birth weight infant (less than 1500 grams) within the first year of life. Journal of Pediatrics. 1990, 117: S156-S166. 10.1016/S0022-3476(05)80014-9.View ArticlePubMedGoogle Scholar
- Department of Health and Human Services, Office on Women's Health: HHS Blueprint for Action on Breastfeeding. 2000, Washington, DC: Department of Health and Human Services, Office on Women's HealthGoogle Scholar
- Lawrence PB: Breast milk: Best source of nutrition for term and preterm infants. Pediatric Clinics of North America. 1994, 41: 925-941.PubMedGoogle Scholar
- Wagner CL, Anderson DM, Pittard WB: Special properties of human milk. Clinical Pediatrics. 1996, 35 (6): 283-293. 10.1177/000992289603500601.View ArticlePubMedGoogle Scholar
- Yoneyama K, Nagata H, Asano H: Growth of Japanese breast-fed and bottle-fed infants from birth to 20 months. Annals of Human Biology. 1994, 21 (6): 597-608. 10.1080/03014469400003602.View ArticlePubMedGoogle Scholar
- American Academy of Pediatrics: Breastfeeding and the use of human milk. Pediatrics. 2005, 115: 496-506. 10.1542/peds.2004-2491.View ArticleGoogle Scholar
- Perez-Escamilla R, Maulen-Radovan I, Dewey KG: The association between cesarean delivery and breastfeeding outcomes among Mexican Women. American Journal of Public Health. 1996, 86 (6): 832-836. 10.2105/AJPH.86.6.832.PubMed CentralView ArticlePubMedGoogle Scholar
- Perez-Rios N, Ramos-Valencia G, Ortiz AP: Cesarean delivery as a barrier for breastfeeding initiation: The Puerto Rican experience. Journal of Human Lactation. 2008, 24 (3): 293-302. 10.1177/0890334408316078.View ArticlePubMedGoogle Scholar
- Theofilogiannakou M, Skouroliakou M, Gounaris A, Panagiotakos D, Markantonis SL: Breastfeedings in Athens, Greece: factors associated with its initiation and duration. Journal of Pediatric Gastroenterology and Nutrition. 2006, 43 (3): 379-384. 10.1097/01.mpg.0000228104.97078.bb.View ArticlePubMedGoogle Scholar
- Wagner M: Choosing cesarean section. The Lancet. 2000, 356: 1677-1680. 10.1016/S0140-6736(00)03169-X.View ArticleGoogle Scholar
- Finger C: Caesarean section rates skyrocket in Brazil. Many women are opting for caesareans in the belief that it is a practical solution. Lancet. 2003, 362 (9384): 628-10.1016/S0140-6736(03)14204-3.View ArticlePubMedGoogle Scholar
- Mohammad-Beigi A, Tabatabaee SHR, Mohammad-Salehi N, Yazdani M: Factors influencing cesarean delivery method in Shiraz hospitals. Iran Journal of Nursing. 2009, 21 (56): 37-45.Google Scholar
- Chertok IR, Shoham-Vardi I: Infant hospitalization and breast-feeding post-caesarean section. British Journal of Nursing. 2008, 17 (12): 786-791.View ArticlePubMedGoogle Scholar
- World Health Organization: WHO Child Growth Standards, Methods and Development. 2006, World Health Organization: GenevaGoogle Scholar
- Pridham KF, Brown R, Sondel S, Clark R, Green C: Effects of biologic and experiential conditions on the pattern of growth in weight of premature and full-term infants. Research in Nursing & Health. 2001, 24: 283-297.View ArticleGoogle Scholar
- Kelleher KJ, Casey PH, Bradley RH, Pope SK, Whiteside L, Barrett KW, Swanson ME, Kirby RS: Risk factors and outcomes for failure to thrive in low birth weight preterm infants. Pediatrics. 1993, 91: 941-948.PubMedGoogle Scholar
- Labbok M: What is the definition of breastfeeding?. Breastfeeding Abstracts. 2000, 19 (3): 19-21.Google Scholar
- Fitzmaurice GM, Laird NM, Ware JH: Applied Longitudinal Analysis. 2004, Hoboken, New Jersey: John Wiley & Sons, Inc, PublicationGoogle Scholar
- Diggle PJ, Heagerty P, Liang KY, Zeger SL: Analysis of Longitudinal Data. 2002, New York: Oxford University Press, 2Google Scholar
- Crouchley R, Davis RB: A comparison of population average and random-effect models for the analysis of longitudinal count data with base-line information. Journal of Royal Statistical Society; Series A. 1999, 162 (B): 331-347.View ArticleGoogle Scholar
- Pepe MS, Anderson GL: A cautionary note on inference for marginal regression models with longitudinal data and general correlated response data. Communications in Statistics, Simulation and Computation. 1994, 23: 939-951. 10.1080/03610919408813210.View ArticleGoogle Scholar
- Goldstein H: Efficient statistical modeling of longitudinal data. Annals of Human Biology. 1989, 13: 129-141. 10.1080/03014468600008271.View ArticleGoogle Scholar
- Rekers-Mombarg LTM, Cole TJ, Massa GG, Wit JM: Longitudinal analysis of growth in children with idiopathic short stature. Annals of Human Biology. 1997, 24 (6): 569-583. 10.1080/03014469700005332.View ArticlePubMedGoogle Scholar
- Verbyla AP, Cullis BR: Modeling in repeated measures experiments. Applied Statistics. 1990, 39 (3): 341-356. 10.2307/2347384.View ArticleGoogle Scholar
- Hastie T, Tibshirani R: Generalized additive models: some applications. Journal of the American Statistical Association. 1987, 82 (398): 371-386. 10.2307/2289439.View ArticleGoogle Scholar
- WHO Child Growth Standards. [http://www.who.int/childgrowth/en]
- Forster DA, McLachlan HL, Lumley J: Factors associated with breastfeeding at six months postpartum in a group of Australian women. International Breastfeeding Journal. 2006, 1: 18-10.1186/1746-4358-1-18.PubMed CentralView ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.