Search tips
Search criteria 


Logo of pchealthLink to Publisher's site
Paediatr Child Health. 2009 September; 14(7): 445–449.
PMCID: PMC2786949

Language: English | French

The impact of jaundice in newborn infants on the length of breastfeeding

Catherine M Pound, MD FRCPC1 and Isabelle Gaboury, MSc PhD (Candidate)2



To examine the breastfeeding prevalence among infants aged three and six months who were previously hospitalized because of hyperbilirubinemia, and to determine whether jaundice in newborn infants increases the risk of breastfeeding discontinuation.


Surveys were mailed to mothers of all eligible infants admitted over a two-and-a-half year period to the paediatric ward of a tertiary care children’s hospital with a diagnosis of hyperbilirubinemia. A total of 127 mother-patient pairs were included in the study. Breastfeeding rates at three and six months were compared with those of a city-wide survey (Infant Care Survey) conducted by Ottawa’s Public Health Department. Risk factors for early breastfeeding discontinuation were examined.


Breastfeeding rates at three and six months were not different between the study group and those reported in the Infant Care Survey (75.5% in the study group versus 71.2% in the Infant Care Survey group, at three months; and 59.1% in the study group versus 50.8% of the Infant Care Survey group, at six months). None of the previously reported risk factors for early weaning had an impact on breastfeeding duration in the study population.


Breastfeeding rates following the discharge of infants diagnosed with jaundice were not significantly different from those reported for the general population. Different patient characteristics may have inflated the breastfeeding rates in the study population, as evidenced by a very high education level among the mothers of enrolled patients. Larger prospective studies in diverse populations are needed to determine the rates of early breastfeeding discontinuation in jaundiced infants.

Keywords: Breastfeeding, Hyperbilirubinemia, Infants, Jaundice



Examiner la prévalence de l’allaitement chez les nourrissons de trois et six mois qui ont déjà été hospitalisés à cause d’une hyperbilirubinémie et déterminer si la jaunisse du nouveau-né accroît le risque d’abandon de l’allaitement.


On a posté un sondage aux mères de tous les nourrissons admissibles hospitalisés sur une période de deux ans et demi dans l’aile pédiatrique d’un hôpital de soins tertiaires pour enfants en raison d’un diagnostic d’hyperbilirubinémie. Au total, 127 couples de mères et de patients ont participé à l’étude. On a comparé les taux d’allaitement à trois et à six mois à ceux d’un sondage municipal (Sondage sur les soins aux nourrissons) mené par le service de santé publique d’Ottawa. On a également examiné les facteurs de risque d’abandon précoce de l’allaitement.


Les taux d’allaitement à trois et six mois ne différaient pas entre le groupe à l’étude et celui qui avait participé au Sondage sur les soins aux nourrissons (75,5 % du groupe à l’étude par rapport à 71,2 % du groupe du Sondage sur les soins aux nourrissons à trois mois, et 59,1 % du groupe à l’étude par rapport à 50,8 % du groupe du Sondage sur les soins aux nourrissons à six mois). Aucun des facteurs de risque de sevrage précoce déjà établis n’avait de répercussions sur la durée de l’allaitement au sein de la population à l’étude.


Les taux d’allaitement après le congé des nourrissons ayant reçu un diagnostic de jaunisse ne différaient pas de manière significative de celui déclaré au sein de la population générale. Différentes caractéristiques des patients ont peut-être gonflé le taux d’allaitement au sein de la population à l’étude, tel que le démontre le niveau d’instruction très élevé des mères des patients à l’étude. Il faudra mener des études prospectives plus vastes au sein de populations diversifiées pour déterminer le taux d’abandon précoce de l’allaitement chez les nourrissons atteints de jaunisse.

Breastfeeding confers many advantages to infants, mothers, families and society in general (1). There is strong evidence that human milk feeding decreases the incidence of many infectious diseases (2) and enhances the immunological status of the newborn (1). It has been associated with enhanced performance on neurocognitive development tests (3,4), and provides important health benefits to the mother, including a decreased risk for breast (5,6) and ovarian cancers (7). Given the clear advantages that breast-feeding confers to young babies, it should be strongly encouraged for at least the first six months of life (1), as recommended in the World Health Organization’s “Innocenti Declaration” (8).

Neonatal jaundice is one of the most common medical problems in healthy full-term infants during the immediate postnatal period (9). There is controversy as to whether breastfeeding increases the incidence of jaundice in the first few days of life. Many studies (1012) have reported an association between breastfeeding and significant hyperbilirubinemia, but others have not been able to substantiate this observation (13,14). In a study (9) conducted in Italy, neonatal jaundice was not associated with breastfeeding per se, but rather with increased weight loss after birth subsequent to fasting, which can be seen with inadequate lactation.

When an infant is hospitalized, previously established patterns of breastfeeding are difficult to maintain (15). There is evidence to suggest that mothers of young infants admitted to hospital with hyperbilirubinemia commonly experience guilt because they believe that breastfeeding was the cause of the jaundice (16). The incidence of breastfeeding discontinuation is believed to be higher among infants admitted to the hospital with hyperbilirubinemia than among the general population, although this has never been well documented.

The objectives of the present study were to determine the prevalence of breastfeeding at three and six months of age among mothers of infants previously admitted to the hospital with hyperbilirubinemia, and to compare it with the prevalence of breastfeeding among the general population in a Canadian urban centre, as determined by a survey conducted by the city’s Public Health Department (17). Previously identified risk factors for early breastfeeding discontinuation (18,19), including young age, lower income, lower level of education, smoking and early return to work, were also examined.


Study design

The charts of all babies younger than one month of age admitted with a diagnosis of hyperbilirubinemia to the Children’s Hospital of Eastern Ontario (Ottawa, Ontario), between January 1, 2005, and July 31, 2007, were reviewed by a research assistant. This study period was chosen because there was a large survey of the general population performed by Ottawa’s Public Health Department in 2005 on length of breastfeeding, and because the last eligible babies were older than six months at the time of study recruitment. The Children’s Hospital of Eastern Ontario is a tertiary care teaching centre that provides paediatric care services to a population of 1.5 million people. Infants admitted with jaundice present to the emergency department. They are either self-referred or sent in by their health care provider. The charts were retrieved from medical records with the help of a Health Information Analyst. A survey was mailed to all eligible patients, accompanied by a letter signed by the Chief of Paediatrics asking for consent to participate in the study. Anonymous patient data were made available to the principal investigator only after consent was obtained. To maximize response, a modified version of Dillman’s total design method (20) was used. A maximum of three questionnaires and two reminder postcards were sent to every family, at regular two-week intervals. The Institutional Ethics Board approved the study.

Patient selection

Patients were eligible for the study based on the following criteria: any amount of breastfeeding at the time of admission, younger than one month of age at the time of admission, and French- or English-speaking. Patients were excluded if they were exclusively formula-fed, had hyperbilirubinemia of the predominantly conjugated type, had anatomical abnormalities (such as cleft lip or palate) potentially interfering with breastfeeding, were neurologically impaired, or were fed via nasogastric, nasojejunal or gastric tube.

Data collection

Data collected included maternal age at delivery, level of education achieved, total household income, smoking at the time of the baby’s admission, marital status, age of the baby when admitted to hospital, whether the mother was told that breastfeeding was related to the baby’s jaundice, the type of help received in the hospital and outside the hospital with regard to breastfeeding, age of the child at the time of breastfeeding cessation, age of the child at the time of formula introduction, age of the baby when the mother returned to work, and the number of hours worked by the mother outside the home each week. Breastfeeding at three and six months of age was recorded for each baby, as well as age at formula introduction, if any.


Descriptive statistics (frequency [percentage] and medians [range]) were used to summarize the survey respondents’ profile. The primary outcome for the present study (ie, prevalence of breastfeeding at three and six months) was calculated and 95% CIs were obtained using the Wilson score with continuity correction (21). The prevalence of breast-feeding was compared with the prevalence in the Ottawa population using a one-sample Z test. Age at breastfeeding discontinuation and formula introduction was compared among four factors believed to influence these outcomes using a nonparametric approach (Kruskal-Wallis and Mann-Whitney U tests). The association between early breast-feeding discontinuation and various sociodemographic and physician-support characteristics was assessed with logistic regression models.


There were 200 eligible patients admitted to the paediatric wards with a diagnosis of hyperbilirubinemia between January 1, 2005, and July 31, 2007. Surveys were mailed to all 200 families, but 14 questionnaires were ‘returned to sender’. Of the remaining 186 questionnaires, 128 consented to participate in the study and returned filled-out questionnaires, for a 68.8% response rate. One patient had to be withdrawn at the time of analysis because it was noted that the mother had stopped breastfeeding before admission to hospital and therefore did not fit the inclusion criteria, for a total of 127 patients.

Characteristics of the infants and mothers (Table 1)

Characteristics of study patients and their mothers*

Most mothers were between 26 and 35 years of age at the time of their baby’s admission to hospital, while more than 95% of babies were younger than two weeks of age. The education level was quite high, with most women having completed a university degree. Household income was greater than $70,000 in more than one-half of the respondents. Very few smoked or shared a house with a smoker. Almost all women were married or in a common-law relationship at the time of the baby’s admission. Approximately one-third of the women returned to work during their child’s first year of life, with most of them working 20 h/week to 40 h/week.

Breastfeeding duration (Table 2)

Breastfeeding duration

No difference could be demonstrated in breastfeeding rates at three and six months between our study population and the Ottawa general population, as determined by the Infant Care Survey conducted by Ottawa’s Public Health Department. In our study, 24.4% (95% CI 17.8% to 32.6%) of mothers stopped breastfeeding before their infant turned three months of age, compared with 28.8% of women in the Ottawa population. At three months of age, 55.9% (95% CI 47.2% to 67.2%) of study patients received no formula at all, while the same was true for 50% of babies in the Infant Care Survey. At six months of age, 59.1% (95% CI 50.4% to 64.2%) of study patients were breastfed, and 40.1% (95% CI 32.0% to 48.9%) received no formula, compared with 50.8% and 39.1%, respectively, in the Ottawa population.

Predictors of early breastfeeding cessation (Table 3)

Predictors of early breastfeeding discontinuation

None of the previously reported predictors were found to be significantly associated with discontinuing breastfeeding at three and six months in our study population. Age, income, education level, smoking and return to work did not seem to affect length of breastfeeding in the present study.

Relationship between breastfeeding duration and various study outcomes (Tables 4 and and55)

Age at discontinuation of breastfeeding
Age at formula introduction

None of the characteristics investigated were found to have an effect on breastfeeding duration or age of formula introduction. Mothers who were told that the jaundice was related to the breastfeeding while admitted to hospital showed no difference in breastfeeding duration compared with mothers who were not. Help with breastfeeding received in or outside of the hospital did not affect breastfeeding duration, and neither did returning to work in the first year of the baby’s life. The results were similar when the women who reported not remembering whether they received help in the hospital were reclassified into not having received any help. This was also similar for the women who reported not remembering whether they had been told that the breastfeeding was related to their baby’s jaundice (data not shown).


The present study showed no significant difference in breastfeeding duration, up to six months of age, between mothers of children admitted to a paediatric hospital for hyperbilirubinemia and mothers in the general population.

A possible hypothesis for these surprising results is the very high level of education of the women enrolled in the study. Previous studies (18) showed that women with a lower level of education were less likely to breastfeed. In our study, 92.1% of mothers had completed a postsecondary education (technical or vocational training, or some university training), while 53% of women enrolled in the Infant Care Survey (17) had a similar level of education. Previous reports (2224) have shown that a higher level of education is associated with higher rates of participation in different types of studies. A selection bias is likely responsible for the discrepancy seen in our study, because women with a higher education level are more likely to fill out a questionnaire and therefore participate in the study. It is reasonable to postulate that if our study population were to have a level of education similar to that of the Infant Care Survey, our breastfeeding rates would have been lower.

Also, previously identified risk factors (18,19) for early breastfeeding discontinuation did not seem to have an impact on breastfeeding duration. There was no significant association noted among age, household income, education, smoking and early return to work, with breastfeeding at three and six months. This may simply be related to the small numbers of patients enrolled in the study.

Length of breastfeeding and formula introduction were not found to be affected by return to work during the first year of life, despite previous reports clearly identifying early return to work (18) and full-time employment (19) as risk factors for breastfeeding discontinuation. This discrepancy may be related to our small population, or it may be that the nine-month median time to return to work in our study was much longer than the time to return to work seen in previous reports (18,19). Indeed, these studies were based in the United States, where maternity leaves are typically 12 weeks long.

The lack of association between length of breastfeeding and help received in the hospital or after discharge, as well as between length of breastfeeding and whether women were told that breastfeeding and hospital admission were related, were surprising given that one would expect maternal confidence to be affected in these situations. Maternal confidence has been shown to be a strong predictor of breastfeeding outcome (18,25), with a lack of confidence in breastfeeding skills leading to a higher likelihood of weaning in the first six weeks postpartum (25). This again may be related to our small or more educated patient population, with more subjects required to demonstrate an effect. Interestingly, however, 57.1% of women believed they needed to get help with their breast-feeding skills after discharge from the hospital, leading one to believe that they did not receive an adequate amount of help during their infant’s hospitalization. Also, 39.7% of mothers received no breastfeeding help or advice while in the hospital. Of those who did, 93.5% were helped by nurses, compared with only 3.9% who were helped by their treating physicians (data not shown).

The reason why women reported very little help from physicians is unclear. It may be that the physicians were simply not asked, were too busy to review the mothers’ breast-feeding techniques or felt ill-prepared to help mothers with lactation difficulties, as demonstrated in previous reports (2628). The responsibility of helping mothers with breast-feeding difficulties are then left to the nurses. Because the vast majority of nurses have no formal training in lactation counselling, there is inconsistency in the amount or kind of support given.

The limitations of the present study are those inherent to retrospective analyses. There may have been an element of recall bias, especially for the women whose babies were admitted to hospital more than three years previously. Given the nature of the study, the authors could not determine what kind of education, if any, mothers received with regard to breastfeeding and jaundice. Also, it is possible that some of the respondents stopped breastfeeding earlier than they admitted to on the survey, given the current pressures imposed by society to breastfeed. As mentioned above, there may also have been selection bias, with more educated women being more likely to participate in the study.


No significant difference in breastfeeding duration was found between babies admitted to the hospital with hyperbilirubinemia and those in the general population. However, different patient characteristics may have been responsible for inflating the numbers in the study population, as evidenced by a very high level of education in the mothers of enrolled patients. Prospective studies are needed to determine more accurate rates of early breastfeeding discontinuation in jaundiced infants. Also, further studies are needed to determine whether the lack of physicians’ help with breastfeeding techniques is due to knowledge deficits, because there may be a role for incorporating teaching of breastfeeding skills to physicians and physicians in training. Finally, the reason for the high percentage of mothers seeking breastfeeding help posthospital discharge, as well as the best way to respond to this need, requires further investigation.


Funding was provided by the Children’s Hospital of Eastern Ontario Research Institute.


1. Gartner LM, Morton J, Lawrence RA, et al. American Academy of Pediatrics Section on Breastfeeding Breastfeeding and the use of human milk. Pediatrics. 2005;115:496–506. [PubMed]
2. Heinig MJ. Host defense benefits of breastfeeding for the infant. Effect of breastfeeding duration and exclusivity. Pediatr Clin North Am. 2001;48:105–23. [PubMed]
3. Horwood LJ, Fergusson DM. Breastfeeding and later cognitive and academic outcomes. Pediatrics. 1998;101:e9. [PubMed]
4. Anderson JW, Joshntone BM, Remley DT. Breastfeeding and cognitive development: A meta-analysis. Am J Clin Nutr. 1999;70:525–35. [PubMed]
5. Newcomb PA, Storer BE, Longnecker MP, et al. Lactation and a reduced risk of premenopausal breast cancer. N Engl J Med. 1994;330:81–7. [PubMed]
6. Collaborative Group on Hormonal Factors in Breast Cancer Breast cancer and breastfeeding: Collaborative reanalysis of individual data from 47 epidemiological studies in 30 countries, including 50302 women with breast cancer and 96973 women without the disease. Lancet. 2002;360:187–95. [PubMed]
7. Rosenblatt KA, Thomas DB. Lactation and the risk of epithelial ovarian cancer. The WHO Collaborative Study of Neoplasia and Steroid Contraceptives. Int J Epidemiol. 1993;22:192–7. [PubMed]
8. UNICEF-WHO . Innocenti Declaration On the Protection, Promotion and Support of Breastfeeding. New York: UNICEF; 1990.
9. Bertini G, Dani C, Tronchin M, Rubaltelli FF. Is breastfeeding really favoring early neonatal jaundice? Pediatrics. 2001;107:e41. [PubMed]
10. Maisles MJ, Giffor K. Neonatal jaundice in full-term infants. Am J Dis Child. 1983;137:561–2. [PubMed]
11. Linn S, Schoenbaum SC, Monson RR, et al. Epidemiology of neonatal hyperbilirubinemia. Pediatrics. 1985;75:770–4. [PubMed]
12. Schneider AP. Breast milk jaundice in the newborn. JAMA. 1986;255:3270–4. [PubMed]
13. Rubaltelli FF. Unconjugated and conjugated bilirubin pigments during perinatal development. IV. The influence of breast-feeding on neonatal hyperbilirubinemia. Biol Neonate. 1993;64:104–9. [PubMed]
14. Nielsen HE, Haase P, Blaabjerg J, Stryhn H, Hilden J. Risk factors and sib correlation in physiological neonatal jaundice. Acta Paediatr Scand. 1987;76:504–11. [PubMed]
15. Spatz DL, Goldschmidt KA. Preserving breastfeeding for the rehospitalized infant: A clinical pathway. MCN Am J of Matern Child Nurs. 2006;31:46–51. [PubMed]
16. Willis SK, Hannon PR, Susan C. The impact of the maternal experience with a jaundiced newborn on the breastfeeding relationship. J Fam Pract. 2002;51:465. [PubMed]
17. Ottawa Public Health . Ottawa: 2006. Infant Care Survey 2005.
18. Ahluwalia IB, Morrow B, Hsia J. Why do women stop breastfeeding? Findings from the Pregnancy Assessment and Monitoring System. Pediatrics. 2005;116:1408. [PubMed]
19. Dillman D. Mail and Internet Surveys: The Tailored Design Method. 2nd edn. New York: John Wiley & Sons; 1999.
20. Ryan AS, Wenjun Z, Acosta A. Breastfeeding continues to increase into the new millennium. Pediatrics. 2002;110:1103–9. [PubMed]
21. Newcombe RG. Two-sided confidence intervals for the single proportion: Comparison of seven methods. Stat Med. 1998;17:857–72. [PubMed]
22. Krousel-Wood MA, Re RN, Abdoh A, et al. The effect of education on patients’ willingness to participate in a telemedicine study. J Telemed Telecare. 2001;7:281–7. [PubMed]
23. Even C, Richard H, Thuile J, et al. Characteristics of voluntary participants versus nonparticipants in a psychoeducation program for euthymic patients with bipolar disorder. J Nerv Ment Dis. 2007;195:262–5. [PubMed]
24. Rimer BK, Schildkraut JM, Lerman C, et al. Participation in womens’ breast cancer risk counseling trial. Who participates? Who declines? High Risk Breast Cancer Consortium. Cancer. 1996;77:2348–55. [PubMed]
25. Dunn S, Davies B, McCleary L, Edwards N, Gaboury I. The relationship between vulnerability factors and breastfeeding outcome. J Obstet Gynecol Neonatal Nurs. 2006;35:87–97. [PubMed]
26. Freed GL, Clark SJ, Sorenson J, Lohr JA, Cefalo R, Curtis P. National assessment of physicians’ breast-feeding knowledge, attitudes, training, and experience. JAMA. 1995;273:472–6. [PubMed]
27. Williams EL, Hammer LD. Breastfeeding attitudes and knowledge of pediatricians-in-training. Am J Prev Med. 1995;11:26–33. [PubMed]
28. Hillenbrand KM, Larsen PG. Effect of an educational intervention about breastfeeding on the knowledge, confidence, and behaviors of pediatric resident physicians. Pediatrics. 2002;110:e59. [PubMed]

Articles from Paediatrics & Child Health are provided here courtesy of Pulsus Group