Introduction
Breastfeeding is a key public health measure that offers considerable benefit to both mother and baby¹. Yet, the numbers of mothers breastfeeding in Ireland, has to date, been disappointingly low^2,3,4. In fact, Ireland currently has the lowest breastfeeding rate in Europe⁵. The recent National Health Strategy acknowledged this low rate and proposed specific actions and targets to promote and support breastfeeding⁵.

The initiation and duration of exclusive and partial breastfeeding depends on a number of determinants. These can be of various types including socio-demographic, psychosocial and health related⁶. Knowledge about these determinants is essential for the development of effective health promotion programmes. Exclusive breastfeeding for up to six months has many advantages but its prevalence in infants in less than four months is low in many countries¹. While previous Irish studies have examined total breastfeeding rates, rates of exclusive and partial breastfeeding have not been separately examined. Consequently, the aim of this study was to determine the initiation rate and duration of exclusive and partial breastfeeding at various stages of postnatal follow-up in mothers from counties Meath, Louth, Cavan and Monaghan who gave birth during the defined study period. In addition, determinants for initiating and continuing breastfeeding were also examined.

Methods
The study population consisted of all singleton births to mothers resident in NEHB counties between January 20th and February 16th 2003. However, mothers and babies who were, in the opinion of the public health nurse (PHN), too ill to participate were excluded. In addition, asylum-seeking mothers resident in Mosney Dispersal Centre were not included as their widespread dispersal throughout Ireland after birth of their babies made initial and follow-up contact impossible.

At the first postnatal visit every PHN was asked to give a questionnaire to eligible mothers and encourage their participation in the study. In addition every PHN forwarded a master-list, with the contact details of all new mothers in her area to the Department of Public Health. This facilitated follow-up of non-responders. Eligible mothers were asked to complete a self-administered, piloted, semi-structured questionnaire. Mothers who were breastfeeding their babies were again contacted, for telephone interview, by research staff at six and fourteen weeks postnatal. At initial survey contact information was obtained on mothers demographic details, pregnancy and delivery particulars, infant feeding pattern and factors that influenced their infant feeding choice. Subsequent follow up at six and fourteen weeks obtained further details on existing infant feeding method and relevant changes that had occurred in infant feeding pattern.

The WHO/UNICEF^7,8 definitions of breastfeeding are used throughout this study (Figure 1). Data was analysed using JMP statistical analysis package⁹. In addition, univariate and multivariate analysis was carried out using STATA, version 8¹⁰.

Results
A total of 430 babies were born during the defined study period. Thirtysix mothers were validly excluded from the study. Of the 394 questionnaires administered, 247 were returned giving a response rate of 62.7%.

The mean age of mothers was 30.5 years (range: 15-43 years; sd: 5.2 years). In total, 207 (83.8%) mothers were either married or cohabiting. One hundred and forty-one mothers worked outside the home before pregnancy and less than one-fifth of these did not intend to return to work after their pregnancy. Based on their stated current or previous occupation, 72% of respondents were members of social class I, II or III. In excess of 40% of mothers reported they had completed third level education. Only 52 (21.1%) mothers stated they had a medical card. Almost two thirds of mothers (62.7%; n=155) reported their babies were born by normal delivery. The reported mean birth weight of babies was 3.5 kg (range 1.4 .8 kg, sd: 0.5 kg). The majority (61.9%; n=153) of mothers had an average of two other children in their household.

Only 49 (19.8%) mothers stated they were current smokers. Almost all mothers (93.5%; n=231) reported taking folic acid during their pregnancy. In fact, 211 (85.4%) claimed to have taken folic acid prior to becoming pregnant or during the first three months of pregnancy. The vast majority of respondents (86.6%; n=214) stated they either availed of conventional antenatal delivered by both hospital doctors and general practitioners or attended only their hospital doctor. However, just over one third of mothers (35.6%; n=88) reported they had attended parentcraft classes. Just 13 (5.3%) mothers stated they were referred to their PHN in the antenatal period. In addition, 101 (40.9%) mothers claimed no professional spoke to them about the benefits of breastfeeding during the antenatal period.

A total of 127 (51.4%) mothers reported initiating breastfeeding their babies on day one. Of these 51 (40.1%) were primagravidas. The majority of respondents (74.8%; n=95) reported choosing breastfeeding as it was best for baby. Most mothers had decided on their infant feeding method prior to becoming pregnant, with 105 (82.7%) breastfeeding mothers stating that they had made this decision before they were three months pregnant. Similarly, 74 (61.6%) mothers who reported bottlefeeding their baby had decided on their infant feeding method by this early stage. As shown in Table 1, using univariate analysis many factors including early decision to breastfeed (p<0.001), higher social class (p<0.05), age 24 years or older (p<0.05), completed third level education (p<0.0001), being a non-smoker (p<0.005) and having previously breast-fed (p<0.0001) were positively associated with initiation of breastfeeding among all respondents. Multivariate analyses of all mothers showed that when these variables were placed in a logistic regression model, only the following variables remained statistical significant: early decision to breastfeed; completed third level education; and having previously breastfed other children. However, when multivariate analysis was carried out using the same variables on first-time mothers and previous mothers separately, there was a difference in the results among these two groups. For previous mothers, having previously breastfed, completed third-level education and an early decision to breastfeed were significantly associated with breastfeeding. However, for first time mothers, only being a current non-smoker was found to be significantly associated with initiation of breastfeeding.

Figure 2: Reported rate of breastfeeding and bottle feeding at each stage of survey

There was a gradual decline in the number of mothers who reported breastfeeding their babies throughout the study period as presented in Figure 2. In addition, the reported exclusiveness of breastfeeding gradually decreased from survey contact onwards as presented in Table 2. Twenty (15.7%) breastfeeding mothers reported introducing formula milk within the first week of birth. Thereafter partial breastfeeding gradually increased so that at 14 weeks post-natally, 95 (74.8%) mothers, who had commenced breastfeeding their babies at birth, were also giving their babies formula milk. Those factors that reached statistical significance are presented in Table 1. However, when these variables were placed in a multivariate logistic regression model, none of them remained statistically significant.

Husbands and partners were supportive of breastfeeding as reported by 118 (92.9%) respondents. However, 14 (11.0 5%) respondents stated that other family members were unsupportive of breastfeeding. Less than half (47.7%; n=63) of respondents attended breastfeeding support groups and for the majority of these attendances at such classes had no influence on the duration of breastfeeding. The most common reported reasons for discontinuing breastfeeding initially and prior to six-week follow-up were due to having a hungry or unhappy baby. At later stages of follow-up maternal issues were more commonly reported as reasons for discontinuing breastfeeding.

Discussion
This study indicates some improvement in the regional breastfeeding initiation rate from the previously reported figure of 35%¹¹. While this rate (51.5%) reaches the national initiation target¹² and is higher than reported in the National Health and Lifestyle survey (SLN 2)¹³ when 37% of mothers reported breastfeeding any of their children, these results have to be interpreted with caution as there was over representation of women from the higher social classes in the respondent group. Nevertheless, many important issues in relation to infant feeding have been highlighted by this study.

The socio-demographic characteristics of mothers who initiated breastfeeding are similar to that previously reported ^2,3,6. Young mothers, having a low income, low educational attainments, being a cigarette smoker, having a first baby and single marital status are all important negative determinants of breastfeeding. Mothers employment has also been reported as an obstacle to initiating breastfeeding especially in countries where a paid maternity leave is non-existent or short^6,14. However, in our study maternal employment was positively associated with continuing breastfeeding up to six weeks.

Having a mother who acts as a positive influence, who has mastered the skills and knowledge of breastfeeding herself has been shown to be positively associated with the initiation of breastfeeding⁶. However, with trends towards smaller family size, more women working and increasing geographical separation of families, such opportunities for positive exposure to breastfeeding may be decreasing. In addition, such support may not be universally available in the general community as evidenced by the fact that one-tenth of mothers stated that family members and the general community were unsupportive of their breastfeeding.

The harmful effects of smoking on both mother and child have been well documented¹⁵. Despite this, one fifth of respondents were current smokers. While this is less than the figure (26%) reported for female smokers in SLN 213 there remains an urgent need to identify more effective methods of reducing smoking in pregnancy. In contrast, the number of mothers who reported taking folic acid supplements at the appropriate stage of their pregnancy is higher than the figure reported nationally¹⁶.

Two thirds of breastfeeding mothers decided on their infant feeding method prior to becoming pregnant. This is higher than the 35% previously reported by Fitzpatrick et al4. As a consequence the opportunity, for both nursing and medical personnel, to positively influence breastfeeding rates may be limited to the smaller proportion of women who remain undecided on their infant feeding method at the later stages of pregnancy. However, even in the last month of pregnancy positive intervention by medical professionals can be effective in increasing breastfeeding rates¹⁷. Attendance at small informal discussion classes, such as parentcraft classes, that highlight the benefits of breastfeeding can increase breastfeeding initiation rates¹⁸. However this intervention is limited principally by two factors:

1. the numbers of mothers attending these type of classes is disappointingly low as evidenced by this and previous Irish studies^2,4 and
2. attendance at these classes is generally during the later stages of pregnancy when most mothers have already decided on their infant feeding method.

As in previous studies^14,17, we found a significant proportion of mothers reported that they had not received any professional advise on breastfeeding during the antenatal period. This is disappointing as mothers who perceive their medical advisor as supportive of breastfeeding tend to breastfeed longer. Failure to provide information restricts womens choice and infringes on the rights of the child¹⁷. Appropriate professional training does have a positive effect on breastfeeding rates¹⁹. Thus inclusion of breastfeeding issues in training of health care staff should begin at undergraduate level as highlighted in the national policy¹².

Increasing the duration of exclusive breastfeeding is one of the goals of the World Health Organisation²⁰. However, this study documents an early and progressive decline in total and exclusive breastfeeding rates from initiation to 14 weeks post-partum. Similar drop out rates have also been previously documented^2,3. The provision of occasional bottle feeds to breastfeeding mothers, which may in fact begin during the in-patient stay, is detrimental to improving the rate of exclusive breastfeeding and also drastically shortens the duration of breastfeeding^6,14. Such practices also discriminate against breastfeeding mothers¹⁷. To facilitate exclusive breastfeeding routines at maternity hospitals must be highly supportive. All pregnancy related care must be delivered by trained skilled professionals in an environment that is supportive of and non-discriminatory towards breastfeeding mothers.

However, concentrating solely on a health sector approach to promoting breastfeeding has its limitations. A co-ordinated multifaceted national effort involving health professionals and the community is required to positively affect breastfeeding rates. The national breastfeeding policy¹² was developed to implement such a multifaceted action. However some of this policys recommendations have still not been implemented. Nevertheless, the Department of Health and Children in its strategy document⁵ again set out as a key objective the need to strengthen measures to promote and support breastfeeding and since the Department has also committed itself to timely deliverable actions on breastfeeding a breastfeeding culture may eventually be re-established in this country. At the outset these actions must be targeted towards those in lower socio-economic groups so that the current situation where the benefits of breastfeeding appear to be primarily dependent on economic circumstances no longer prevails.

Correspondence: Mary Ward, Department of Public Health, Railway Street,
Navan, Co.Meath.
Phone: (046) 90 76412
Fax: (046) 90 73235

1. NHS Centre for Reviews and Dissemination. Promoting the initiation of breastfeeding. Effective Health Care 2000;6 (2): 1-12.
2. Sayers G, Thornton L, Corcoran R, Burke M. Influences on breastfeeding initiation and duration. Ir J Med Sc 1995; 16: 281-4.
3. Twomey A, Kilberd B, Matthews T, ORegan M. Feeding infants an investment in the future. IMJ 2000; 98: 248-50.
4. Fitzpatrick CC, Fitzpatrick PE, Darling MRN. Factors associated with the decision to breastfeed among Irish women. IMJ 1994; 84:145-6.
5. Department of Health and Children. Quality and Fairness, A Health System for You. Dublin: Stationery Office, 2001.
6. Agneta Y, Sjostrom M. Breastfeeding in countries of the European Union and EFTA: current and proposed recommendations, rationale, prevalence, duration and trends. Public Health Nutrition, 2001: 4(2B), 653-645.
7. WHO Global Data Bank on Breastfeeding. World Health Organisation, Geneva, 1996.
8. WHO Breastfeeding Counselling: a training course. WHO/UNICEF, Geneva. 1993.
9. SAS Institute Inc (2002). JMP Statistical Package. Version 5, Cary NC, USA.
10. Stata Corporation. STATA statistical software: release 8.0, College Station, TX: Stata Corporation, 1999.
11. Department of Public Health Medicine. Breastfeeding in the North East: A report. 1997: North Eastern Health Board: Navan.
12. Department of Health. A National Breastfeeding Policy for Ireland. Department of Health, 1994: Dublin
13. Centre for Health Promotion Studies. The National Health and Lifestyle Surveys. Centre for Health Promotion Studies, 2003: Galway.
14. Dennis CL. Breastfeeding initiation and duration: a 1990-2000-literature review. J Obstet Gynaecol Neonatal Nurs 2002; 1: 12-32.
15. Howell F. Women and smoking. IMJ 1998; 91:15-6.
16. O Leary M, McDonnell R, Johnson H. Folic acid and prevention of neural tube defects 2000, improved awareness but low peri-conceptual uptake. IMJ 2002; 94: 180-1.
17. Loh LR. Kelleher C, Long S, Loftus B. Can we increase breastfeeding rates? IMJ 1997; 90:100-1.
18. Fairbank L, O Mears S, Renfrew M et al. A systematic review to evaluate the effectiveness of interventions to promote the initiation of breastfeeding. HTA 2000; 4:(25) 1-6.
19. Cattaneo A, Buzzatti R. Effect on rates of breast-feeding of training for the Baby Friendly Hospital Initiative. BMJ 2001, 323: 1358-62.
20. Agneta Y, Sjostrom M. Breastfeeding determinants and a suggested framework for action in Europe. Public Health Nutrition 2001; 4: 729-739.