Original Communication

European Journal of Clinical Nutrition (2005) 59, 789–796. doi:10.1038/sj.ejcn.1602144 Published online 4 May 2005

Factors associated with exclusive breastfeeding in Accra, Ghana

Contributors: BAA was responsible for all aspects of study coordination, protocol development, staff training and report writing. She also participated in proposal writing, obtaining supplemental funds, recruiting, data collection and analysis. RP-E was responsible for proposal writing and obtaining funds for the project. He was involved in protocol development, content validity assessment, data analysis and report writing. AL provided collaboration from the University of Ghana, obtained permission from the health institutions used and was also involved in content analysis, study coordination and report writing. JA was involved with selection of MCH clinics, study coordination, recruitment, data collection, entry and cleaning.

B A Aidam1, R Pérez-Escamilla1, A Lartey2 and J Aidam2

  1. 1Department of Nutritional Sciences, University of Connecticut, Storrs, 3624 Horsebarn Hill Road Extension, U-17, Storrs, CT, USA
  2. 2Department of Nutrition and Food Science, University of Ghana, Legon, Ghana

Correspondence: BA Aidam, Department of Nutritional Sciences, University of Connecticut, Storrs, 3624 Horsebarn Hill Road Extension, U-17, Storrs, CT 06269, USA. E-mail: bchinbus@yahoo.com

Received 26 May 2004; Revised 28 January 2005; Accepted 15 February 2005; Published online 4 May 2005.





To assess factors associated with exclusive breast-feeding (EBF) in Accra, Ghana.

Design, subjects, setting:


Data on current and past infant feeding patterns, sociodemographic, biomedical and biocultural factors were collected using a cross-sectional design, from a sample of 376 women with infants 0–6 months, attending maternal and child health (MCH) clinics in Accra. EBF was defined in two ways: (a) based on a 24-h recall, and (b) based on a recall of liquids or foods given since birth.



Although 99.7% of mothers were currently breastfeeding (BF), only half (51.6%) of them EBF their infants. About 98% of participants had heard about EBF, and 85.6% of them planned to EBF on delivery. Based on 'since birth' EBF, planned EBF on delivery was associated with higher likelihood of EBF (OR=2.56; 95% CI, 1.06–6.17) and delivery at a hospital/polyclinic was associated with a two times higher likelihood of EBF (OR=1.96; 95% CI, 1.08–3.54). Women living in their own houses were more likely to EBF (OR=3.96; 95% CI, 1.02–15.49) than those living in rented accommodations and family houses. Those with a more positive attitude towards EBF were more likely to EBF (OR=2.0; 95% CI, 1.11–3.57) than their counterparts with more negative attitudes. The '24-h recall' EBF model yielded similar results.



In this population, EBF was associated with delivery at hospital/polyclinic, having secondary school education, intention to EBF prior to delivery, owning a home and having a positive attitude to EBF.



Funded by a University of Connecticut Research Foundation grant awarded to Dr Rafael Pérez-Escamilla, and the LINKAGES program, Accra, Ghana.


breastfeeding promotion, exclusive breastfeeding, delivery place, Ghana



Breastfeeding (BF) is the universally acceptable means of infant feeding in Ghana, evident in the high rates of BF initiation and duration. There are various known benefits of BF (Brown et al, 1989; Popkin et al, 1990; Victora et al, 1994; Heinig & Dewey, 1996; Labbok et al, 1997; Pérez-Escamilla, 1997; Anderson et al, 1999; Labbok, 1999), which make these high rates desirable. Breast milk provides nutrients to meet growth needs (Cohen et al, 1994; Dewey et al, 1999; Simondon and Simondon, 1997) and several immunological factors to protect against infections and subsequent mortality (Beaudry et al, 1995; Raisler et al, 1999; Kramer et al, 2001; WHO, 2001). In light of these benefits, not BF is considered a real threat for infant survival in developing country settings. The World Health Organization in a systematic review of the scientific literature concluded that exclusive breastfeeding (EBF) defined as giving only breast milk without any additional food or liquid, to 6 months confers multiple benefits to the infant and mother (WHO, 2002). Thus, EBF for 6 months has become a public health priority worldwide.

In Ghana, almost all women (99%) initiate BF and the average duration of BF is 22 months (GSS and MI, 1999). EBF rates, however, remain very low, about 35% of infants 0–3 months are breastfed exclusively and by 6 months EBF rate is only 6.4% (GSS and MI, 1999). Common practices include addition of liquids, porridge and solid foods to the infants diet very early in life. These practices have been associated with a high rate of diarrhea infection (Popkin et al, 1990), which may lead to the death of the infant (Arifeen et al, 2001).

For mothers to be able to breastfeed exclusively to the recommended 6 months, it is important to understand the factors that influence EBF. Various factors have been found to be associated with BF initiation and duration, and EBF. These include : demographic factors (eg level of education, parity, urban vs rural residence); biosocial factors (BF support in clinical and community settings); cultural factors (beliefs, norms and attitudes towards BF); socioeconomic status; and employment policies affecting how long an infant can be in close proximity to the mother (Pérez-Escamilla et al, 1995; Pérez-Escamilla et al, 1996; Valdes et al, 2000; Scott et al, 2001; Dearden et al, 2002; Martin et al, 2002; Whaley et al, 2002; Ludvigsson, 2003). The above factors affect BF and EBF rates in different directions and to varying degrees depending on the culture.

There is very limited data regarding EBF determinants in sub-Saharan Africa. This study was thus conducted to identify determinants of EBF in a group of mother infant pairs attending child welfare clinics in Ablekuma, a subdistrict of Greater Accra, Ghana.



Study design

This study was approved by the Institutional Review Board of the University of Connecticut and the Ghanaian health care institutions included. Mothers were recruited while attending maternal and child health (MCH) clinics. All mothers were approached and asked to respond to a few screening questions (as described in the selection criteria below) after which their consent was sought to participate in an interview, if they qualified for the study. Following this interview, maternal weight and height and infant weight and length were taken. All eligible women were included and none refused to be part of the study. As an acknowledgment for participation, education and counseling on infant feeding was provided. The study was conducted between May and August 2000.

Sampling areas

This study followed a convenience sampling method. Subjects were recruited from Ablekuma, a subdistrict of the Greater Accra region. The MCH clinics in this subdistrict are under the supervision of a polyclinic (Mamprobi Polyclinic). Ablekuma has 17 subareas all of which were active MCH clinics. Subjects for this study were recruited from 13 of these subareas. On clinic days, about 40–50 women brought their infants for weighing and immunization. Public Health Nurses (PHNs) assigned to the clinics also provided talks on health issues to mothers.

Selection criteria

To be eligible, the woman had to be the biological mother of an infant 0–6 months of age. She had to be free of any self-reported medical conditions making any BF or EBF inadvisable or difficult (eg HIV/AIDS). The study was originally designed to exclude women with any medical condition incompatible with BF. This exclusion criterion was not used as none of the screened women reported any condition inhibiting BF and only one woman was not BF. Infants had to be of single birth and delivered at term (greater than or equal to36 to 44 weeks gestation). Initially, infants had to be of normal birth weight (greater than or equal to2500 g) to be included in the study. This criterion was however dropped since birth weights were missing for infants delivered at home and some maternity homes. Thus, exclusion of such infants with no birth weight data could have led to a sample selection bias.

Survey instrument

A questionnaire composed of both closed- and open-ended questions was used to collect data on: sociodemographic (eg highest educational level attained, employment, parity and ethnicity); biomedical (eg delivery type, antenatal care) and biocultural factors (eg partner, friend or relatives' support for current infant feeding) associated with breastfeeding; attitudes towards EBF; knowledge of WHO recommendations; and exposure to EBF promotion. Knowledge questions were adapted from LINKAGES (Academy for Educational Development) Facts for Feeding brochure (LINKAGES, 1999, 1998). Similar questions were used to test women's knowledge and attitude. Attitudinal questioning preceded the knowledge test and started with the statement 'in your opinion'. Trained assistants took anthropometric measurements. Interviews were conducted in English or any one of three local Ghanaian languages (Ga, Twi and Ewe) depending on what the mother was comfortable with. The survey was pretested in Ghana by applying it to 12 target women not included in the final analysis. Necessary changes were made to ensure that women understood all questions.

Statistical analysis

Dependent variable

All data were entered and analyzed using SPSS for Windows (version 8) and Epi Info (version 3.01). The dependent variable for all analyses was current EBF status. Mothers were asked if any of the following eight categories of liquid, mushy or solid foods had been given to the child since birth: (i) plain water, (ii) sugar solution, (iii) juices, (iv) herbal teas, (v) other teas, (vi) baby formula, (vii) other liquids, (viii) solid/marshy foods. The age at which these foods were first introduced was also recorded. A 24-h recall was then used to verify if any of these same categories of liquids and foods had been given to the child within the past 24 h. Based on these questions, EBF was defined in two ways: (a) child never given any liquids or solid foods apart from breast milk since birth; and (b) a 24-h recall of no liquids or solids being given to the child.

Independent variables

The following variables were tested in a bivariate chi2 analysis: parity, child gender, pregnancy intention, delivery type, prenatal and postnatal BF advice, friends and relatives support for current infant feeding method, maternal education, ownership of household items indicative of socioeconomic status, delivery place, place for prenatal care, knowledge about BF and EBF recommendations, and opinion towards EBF. The coding for these variables is presented in Table 1.

Logistic regressions were run to assess the independent association of independent variables with the likelihood of EBF. For the regression analyses, two models are presented: (a) model adjusting only for child age and testing the independent variables one at a time, and (b) multivariate model controlling for child age in addition to the independent variables significantly associated with either definition of EBF in the age-adjusted model. Results are presented as odds ratios (OR) and their respective 95% confidence intervals (CI). OR were considered to be statistically significant if the 95% CI excluded the value of 1. Further bivariate chi2 analyses were conducted to examine the associations between place of delivery and exposure to pre- and perinatal BF promotion. Indicators of nutritional status created from maternal and infant anthropometric measurements include body mass index (BMI), weight-for-age z-score (WAZ), weight-for-length z-score (WLZ) and length-for-age z-score (LAZ). For these analyses, four children had missing height and length measurements thus, with the exception of WAZ, the sample size for the anthropometric indicators was 372.



Demographic and socioeconomic characteristics

Maternal age ranged from 16 to 43 y, with 18 (4.8%) of the respondents being teenagers. Less than 10% of infants were 1 month old at the time of the interview. The majority of mothers were either Akans or Ga/Adangbes with less than 10% coming from Northern Ghana. A little over a third of the women were primiparous and only 8.5% had more than secondary education. About half of the infants were males (Table 2). Almost all respondents had a radio in their household and the majority had televisions. In total, 20% had a car in the household, and very few (6.9%) owned their homes. Thus, the majority were either living in rented places or family houses.


While pregnant with the study child, all women except three sought prenatal care. A total of 45% sought prenatal care from a polyclinic, 32.9% from hospitals, 17.1% from maternity homes and the remaining 5% from private clinics. More women (38.3%) delivered at hospitals than polyclinics (28.5%) and only a few (8.5%) delivered with traditional birth attendants (TBA), at their own homes or a spiritual leader's home. The majority (91.5%) had vaginal deliveries. The percentages of infants who were underweight (WAZ less than or equal to-2 standard deviations of the mean), wasted (WLZ less than or equal to-2 standard deviations of the mean) and stunted (LAZ less than or equal to-2 standard deviations of the mean), based on CDC/WHO anthropometric reference curves for infants (Dibley et al, 1987) were 1.3, 1.1 and 1.9%, respectively. Maternal BMI ranged from 15.9 to 43 kg/m2 with 2.7% of mothers classified as undernourished (BMI <18.5 kg/m2). Almost 24% were overweight (BMI >26–30 kg/m2) and 15.1% were obese (BMI >30 kg/m2). About 37% of infants had been sick within the 2 weeks preceding the interview. Conditions reported included colds, coughs, diarrhoea, fever, skin and ear infections.

Biocultural and breastfeeding support

The only woman who was not breastfeeding her infant at the time of the interview reported that the child had refused her breast milk. Almost all respondents had heard about EBF and the majority planned to EBF their infants when the child was born. Prenatal BF support had been provided to most women when they were pregnant; however, the percentage receiving peri- and postnatal support for breastfeeding was markedly lower. Over 80% of the women knew at least three out of six BF recommendations and about two-thirds had more positive opinions about EBF (Table 3). The percentage of women who reported to have exclusively breastfed their infants since birth were 51.6%, while those EBF over the previous 24 h were 70.2%. Women EBF by both EBF definitions were 51.3%.

Further bivariate chi2 analysis revealed that about 71% of those delivering at a hospital/polyclinic heard about EBF from their place of delivery compared to 39.4% of those delivering at private clinic/maternity homes/others (P=0.001). Of those who obtained antenatal care from hospitals/polyclinics, 82.1% heard about EBF from those sources. In contrast, only 44.8% of those receiving antenatal care from private clinics/maternity homes heard about EBF there (P<0.001). A significantly higher percentage of women delivering at hospitals/polyclinics reported to have received BF advise at the prenatal stage (86.8%), or perinatally (46.3%) compared to their counterparts delivering in private clinics/maternity homes/others (75% at the prenatal stage and 28.7% perinatally, P< 0.01 in both cases).

Infant age-adjusted logistic regression model

In univariate analysis adjusting for age, ownership of house or car, delivery place, place for prenatal care, planned EBF at time of delivery, advice on BF, and perinatal BF advice were positively associated, while more negative attitude toward EBF and low knowledge about BF recommendations were negatively associated with EBF since birth. The above variables (with the exception of ownership of house) were also significantly associated with the 24 h EBF definition. In addition, if a mother had less than secondary education, she was about half as likely to EBF as compared to one with at least secondary education (Table 4). Parity, child gender, pregnancy intention, delivery type, prenatal and postnatal advice, and friends and relatives' support for CIF were tested but were not significantly associated with EBF regardless of the definition used.

Multivariate logistic regression model

All the variables associated with either definition of EBF in the infant age-adjusted analyses were included in a multivariate model. Variables that came out as significantly associated with EBF since birth were: house ownership, delivery place, planned EBF when child was born and attitude towards EBF (Table 4). Women who planned to EBF at the time of delivery were about 2.5 times more likely to EBF their infants (OR=2.56; 95% CI, 1.06–6.17). Women delivering at a hospital or polyclinic were about twice as likely to EBF (OR=1.96; 95% CI, 1.08–3.54) than those delivering at maternity homes, private clinics or TBAs home, while those living in their own houses were about four times more likely to EBF (OR=3.96; 95% CI, 1.02–15.49). Women with more positive attitudes towards EBF were 2.5 times more likely to EBF compared to those with more negative attitudes (OR=2.56; 95% CI, 1.52–4.35). Using the 24-hour recall definition of EBF, three out of the four variables that were significantly associated with the since birth EBF definition came out significant. This was not surprising as a very high correlation was found between the two EBF definitions (r=0.66, P<0.01 n=376). These were delivery place, planned EBF at delivery and attitude towards EBF. Although slightly higher, the magnitudes of the odds ratio for the three variables were also very close to those using the 'since birth' EBF definition. With the 24-h recall definition, women delivering at hospitals or polyclinics (vs those delivering at private clinics, homes or TBAs) were 2.1 times more likely to EBF (OR=2.11; 95% CI, 1.10–4.05), those who planned to EBF at delivery were about 2.9 times more likely to EBF (OR=2.87; 95% CI, 1.29–6.42), while those with more positive attitudes toward EBF were twice as likely to EBF (OR=2.0; 95% CI, 1.11–3.57).



Our study identified three factors associated with EBF regardless of the EBF definition used (ie planned EBF at birth, attitude towards EBF and delivery place) with the magnitude of the odds ratio being very similar in both cases. This suggests that '24-h recall' EBF can be used in rapid assessments seeking to identify EBF determinants. However, for studies seeking to determine EBF prevalence, it must be recognized that this definition gives a higher estimate of EBF prevalence as compared to the 'since birth' EBF definition. A fourth socioeconomic status (SES) variable, house ownership, was identified as a determinant through the 'since birth' EBF definition but not the '24-h' definition.

The association between EBF practices with maternal feeding intentions (ie EBF plans at delivery) raises critical issues that need to be taken into account in EBF promotion programs. This association has also been reported by Whaley et al (2002) for the US, Scott et al (2001) for Australia, and by Pérez-Escamilla et al (1995) for three Latin American countries; Brazil, Honduras and Mexico. In our study, the initial association of knowledge of BF recommendations with both EBF definitions in the age-adjusted bivariate model became nonexistent in the multivariate model. This implies that opinions or attitudes toward EBF are essential in order to improve this behavior. Hence, EBF promotion programs need to overcome negative beliefs and attitudes in order to make an impact on EBF. Thus, it was encouraging that over 60% of the women in the study sample had a positive attitude toward EBF.

In this study, it appears that higher education and higher SES as expressed by house ownership are factors positively associated with EBF. This finding is contrary to the study by Pérez-Escamilla et al (1995) where lower SES was positively associated with EBF. Thus, more efforts need to be taken by EBF programs to make EBF appealing to low income and lower educated women in Ghana. For example, emphasis should be placed not only on the benefits of EBF for infants, but also benefits to women themselves in terms of protection against postpartum hemorrhage, decreased likelihood of breast cancer and delayed return of ovulation.

Hospitals and polyclinics were the main source of EBF information in this study. In Ghana, public health nurses are responsible for health education in the government hospitals, polyclinics and also in the communities. They are also the main pre- and postnatal BF educators although other nurses or nurse midwives can also educate pregnant and lactating women in the hospital if they have been trained in BF counseling. Most government hospitals are baby friendly but not private clinics. It was consequently not surprising that delivery at hospitals and polyclinics resulted in both more exposure to BF promotion and an over two times higher likelihood of EBF. Also, the higher EBF probability rates found in this study (51.6% using the 'since birth' definition and 70.2% using the 24-h recall) as compared to the existing national figure of 6.4% (GSS and MI, 1999) supports the hypothesis that EBF promotion efforts occurring in these government health facilities, have had a positive influence on EBF rates. Pérez-Escamilla et al (1995) have also documented this positive association, between EBF and delivery at health facilities promoting EBF, in Brazil and Honduras. Dearden et al (2002) also found place of delivery to be a determinant of BF initiation. Thus delivery in Ghanaian maternity homes, private clinics, at home, or with TBA or spiritual leaders may pose a risk for not EBF within the first 6 months of life. Breastfeeding promotion programs should therefore be extended to these 'at risk' birth settings.


Since a convenience sampling method was used, the study population may not be representative of women attending MCH clinics in Accra. We believe that recall bias was addressed by restricting infant age to 6 months so that only mothers with infants in the recommended EBF age bracket were included. The cross-sectional nature of this study prevents drawing causal inferences from the association between lactation counseling (represented by place of delivery) and EBF success. Thus, an intervention study is recommended to answer this important question.



  1. Anderson JW, Johnstone BM & Remley DT. (1999) Breastfeeding and cognitive development: a meta-analysis. Am. J. Clin . Nutr. 70: 525−535. | PubMed | ISI | ChemPort |
  2. Arifeen S, Black RE, Antelman G, Baqui A, Caulfield L & Becker S. (2001) Exclusive breastfeeding reduces acute respiratory infection and diarrhea deaths among infants in Dhaka slums. Pediatrics 108: 1−8. | Article | PubMed |
  3. Beaudry M, Dufour R & Marcoux S. (1995) Relation between infant feeding and infections during the first six months of life. J. Pediatr. 126: 191−197. | PubMed | ChemPort |
  4. Brown KH, Black RE, de Romana GL & de Kanashiro HC. (1989) Infant-feeding practices and their relationship with diarrhoeal and other disease in Hauscar (Lima) Peru. Pediatrics 83: 31−40. | PubMed | ChemPort |
  5. Cohen R, Brown K & Canahuati J. (1994) Effects of age of introduction of complementary foods on infant breast milk intake, total energy intake, and growth : a randomised intervention study in Honduras. Lancet 344: 288−293. | Article | PubMed | ChemPort |
  6. Dearden K, Altaye M, de Maza I, de Oliva M, Stone-Jimenez M, Morrow AL & Burkhalter BR. (2002) Determinants of optimal breastfeeding in peri-urban Guatemala City, Guatemala. Pan. Am. J. Public Health 12: 185−192.
  7. Dewey KG, Cohen R, Brown K & Rivera LL. (1999) Age of introduction of complementary foods and growth of term, low-birth-weight, breastfed infants: a randomized intervention study in Honduras. Am. J. Clin. Nutr. 69: 679−686. | PubMed | ChemPort |
  8. Dibley MJ, Goldsby JB, Staehling NW & Trowbridge FL. (1987) Development of normalized curves for the international growth reference: historical and technical considerations. Am. J. Clin. Nutr. 46: 736−748. | PubMed | ChemPort |
  9. GSS & MI. (1999) Ghana Demographic and Health Survey 1998 Ghana Statistical Service (GSS) and Macrointernational (MI), Calverton, Maryland.
  10. Heinig MJ & Dewey KG. (1996) Health advantages of breastfeeding for infants: a critical review. Nutr. Res. Rev. 9: 89−110.
  11. Kramer MS, Chalmers B, Hodnett ED, Sevkovskaya Z, Dzikovich I, Shapiro S, Collet JP, Vanilovich I, Mezen I, Ducruet T, Shishko G, Zubovich V, Mknuik D, Gluchanina E, Dombrovskiy V, Ustinovitch A, Kot T, Bogdanovich N, Ovchinikova L & Helsing E. (2001) Promotion of Breastfeeding Intervention Trial (PROBIT): a randomized trial in the Republic of Belarus. JAMA 285: 413−420. | Article | PubMed | ISI | ChemPort |
  12. Labbok MH. (1999) Health sequelae of breastfeeding for the mother. Clin. Perinatol. 26: 491−503. | PubMed | ChemPort |
  13. Labbok MH, Perez-Escamilla R, Peterson AE & Coly S. (1997) Breastfeeding and Child Spacing Country Profiles Washington, DC: Institute for Reproductive Health.
  14. LINKAGES. (1998) Facts for Feeding: Guidelines for Appropriate Complementary Feeding of Breastfed Children 6−24 Months of Age. Washington, DC: Academy for Educational Development.
  15. LINKAGES. (1999) Facts for Feeding: Recommended practices to improve infant nutrition during the first six months. Washington, DC: Academy for Educational Development.
  16. Ludvigsson JF. (2003) Breastfeeding intentions, patterns, and determinants in infants visiting hospitals in La Paz, Bolivia. BMC Pediatr. 3: 5. | Article | PubMed |
  17. Martin RM, Smith GD, Mangtani P, Frankel S & Gunnell D. (2002) Association between breastfeeding and growth: the Boyd-Orr cohort study. Arch. Dis. Child Fetal Neonatal. Ed. 87: F193−F201. | Article | PubMed | ChemPort |
  18. Pérez-Escamilla R. (1997) Infant feeding and lactational amenorrhea in developing countries: Results from DHS II & DHS III Washington, DC: Institute for Reproductive Health.
  19. Pérez-Escamilla R, Lutter C, Segall AM, Rivera A, Trevino-Siller S & Sanghvi T. (1995) Exclusive breast-feeding duration is associated with attitudinal, socioeconomic and biocultural determinants in three Latin American countries. J. Nutr. 125: 2972−2984. | PubMed |
  20. Pérez-Escamilla R, Segura-Millan S, Canahuati J & Allen H. (1996) Prelacteal feeds are negatively associated with breast-feeding outcomes in Honduras. J. Nutr. 126: 2765−2773. | PubMed |
  21. Popkin BM, Adair L, Akin JS, Black R, Briscoe MS & Flieger W. (1990) Breastfeeding and diarrheal morbidity. Pediatrics 86: 874−882. | PubMed | ChemPort |
  22. Raisler J, Alexander C & O'Campo P. (1999) Breastfeeding and infant illness: a dose−response relationship. Am. J. Public Health 89: 25−30. | PubMed | ChemPort |
  23. Scott JA, Landers MCG, Hughes RM & Binns CW. (2001) Factors associated with breastfeeding at discharge and duration of breastfeeding. J. Paediatr. Child Health 37: 254−261. | Article | PubMed | ChemPort |
  24. Simondon K & Simondon F. (1997) Age of introduction of complementary food and physical growth from 2 to 9 months in rural Senegal. Eur. J. Clin. Nutr. 51: 703−707. | Article | PubMed | ChemPort |
  25. Valdes V, Pugin E, Schooley J, Catalan S & Aravena R. (2000) Clinical support can make the difference in exclusive breastfeeding success among working women. J. Trop. Pediatr. 46: 149−154. | Article | PubMed | ChemPort |
  26. Victora CG, Fuchs SC, Flores JA, Fonseca W & Kirkwood B. (1994) Risk factors for pneumonia among children in a Brazilian metropolitan area. Pediatrics 93: 977−985. | PubMed | ChemPort |
  27. Whaley SE, Meehan K, Lange L, Slusser W & Jenks E. (2002) Predictors of breastfeeding duration for employees of the Special Supplemantatal Nutrition Program for Women, Infants, and Children (WIC). J. Am. Diet. Assoc. 102: 1290−1293. | Article | PubMed |
  28. WHO. (2001) Collaborative study team on the role of breastfeeding on the prevention of infant mortality. Effect of breastfeeding on infant and child mortality due to infectious diseases in less developed countries: a pooled analysis. Lancet 355: 451−455.
  29. WHO. (2002) The Optimal Duration of Exclusive Breastfeeding a Systematic Review Geneva: World Health Organization.


These links to content published by NPG are automatically generated


Preterm birth, cerebral palsy and magnesium

Nature Medicine News and Views (01 Feb 1997)

Extra navigation