An International Comparison Study into the implementation of the WHO Code and other breastfeeding initiatives

Findings and Conclusions

Findings and Conclusions - An International Comparison Study into the implementation of the WHO Code and other breastfeeding initiatives

Page last updated: 03 May 2012

The positive effects of breastfeeding on the health and wellbeing of mother and infant are extensive and widely acknowledged worldwide (National Health Service Health Development Agency 2003). In response to concerns about declining breastfeeding rates, unregulated marketing of breast-milk substitutes, and the potential effect of artificial feeding on child and infant mortality, the World Health Organization (WHO) International Code of Marketing of Breast-milk Substitutes (WHO Code) was adopted by 118 member states at the 34th World Health Assembly in 1981 (WHO 1981). The WHO Code was formulated with the aim of contributing to “the provision of safe and adequate nutrition for infants, by the protection and promotion of breastfeeding, and by ensuring the proper use of breast milk substitutes, when these are necessary, on the basis of adequate information and through appropriate marketing and distribution” and was adopted as a recommendation rather than as a regulation (WHO 1981).

It is an appropriate time for countries such as Australia to assess the feasibility and effectiveness of strategies to enact the WHO Code within their unique domestic context. In the 30 years since it was adopted there have been a number of global breastfeeding policy developments such as the Innocenti Declaration, the Baby Friendly Hospital Initiative and Baby Friendly Health Initiative, the ILO Maternity Protection Convention and the Global Strategy For Infant and Young Child Feeding (Australian Health Ministers' Conference 2009; International Labour Organization 2000; UNICEF 1990; UNICEF 2005; World Health Organisation and UNICEF 2003; World Health Organization and UNICEF 2009). Although there is a large amount of literature on the effectiveness of interventions to promote breastfeeding in general (see section on breasfeeding: a research overview), Dyson et al (2005) note that research that specifically “evaluates the impact of adopting and/or implementing the ... WHO Code ... is urgently needed”. Furthermore, the effectiveness of large scale health interventions at a national level (such as an evaluation of the Baby Friendly Hospitals Initiative) are less amenable to research than single interventions or at an individual or group setting and thereby makes it difficult to draw out any causal associations.

Despite the complexity and breadth of this topic, there is a growing recognition that the feasibility and effectiveness of strategies to enact the WHO Code vary according to the domestic context. This comparative study has assessed the implementation of the WHO Code across nine developed countries as outlined and discussed with the Department of Health and Ageing. These included Australia, Canada, France, Germany, Ireland, New Zealand, Norway, the UK and the USA and the aim was to gather information on the impact of implementing the WHO Code on breastfeeding rates and infant feeding practices in these countries. In order to consider the broader context, this study has considered other policies and support mechanisms in place which are outside the scope of the WHO Code or the WHO/UNICEF initiatives but have the potential to affect breastfeeding rates.

International Comparisons

Breastfeeding Rates

Historically, developed countries have experienced a decline in breastfeeding rates over the first half of the 20th century reaching a minimum in the 1960s, which was then followed by a rise in the 1970s. It is widely accepted that the decline was related to the medicalisation of birth, the influence of medical advice and the introduction of infant formula. In contrast, the rise in the 1970s was often attributed to the rise of the women’s movement and the associated social movements of the time as well as the recognition of the health benefits of breastfeeding.

A number of measures are used to describe breastfeeding rates and while there are internationally recommended terms outlined by the WHO (Table 1) not all countries collect data in line with these definitions. In addition, there is variation in the method and frequency of data collection. Therefore comparisons of rates across countries are prone to difficulties in interpretation and should be considered with these qualifications.

When considering our pre-specified countries of interest, initiation rates of breastfeeding were high in the majority of the countries (>80% in Australia, Canada, Germany, Norway and the UK) with Norway having almost universal rates of initiation. Two countries stood out as having particularly low rates of initiation and these were France and Ireland.

As the WHO recommends exclusive breastfeeding for the first six months, this is the time-point at which data are most frequently collected and reported. However this time-point is problematic with respect to both the associated definition (i.e. exclusive breastfeeding at six months compared to up to six months) and its usefulness (as this is around the recommended time to introduce solid food and is therefore unlikely to be a stable indicator). The rates of exclusive or full breastfeeding at six months are low across all the assessed countries with no country having a rate greater than 30%. The rate is particularly low in the UK (<1%) and Norway, considered a model for breastfeeding success, also has low rates of exclusive breastfeeding at six months (2 to 10% depending on the survey). The rate of any breastfeeding at six months may provide a more stable measure. In this measure, Norway ranks considerably higher than the other countries with rates of 80 to 82%. A band of countries have reported rates around 50% including Australia, Canada, Germany and the USA while the UK and Ireland have markedly lower rates.

Implementation of the WHO Code

The WHO Code states under Article 11.1 that “Governments should take action to give effect to the principles and aim of the Code, as appropriate to their social and legislative framework.” Reflecting this, there are notable differences in the mechanisms and degree by which the WHO Code has been implemented across the eight countries, however all of the countries assessed are members of the Codex Alimentarius which develops food standards, under the Joint FAO/WHO Food Standards Programme. These food standards encompass infant formula and all eight countries have enacted legislation to ensure high quality of infant formula thereby meeting Article 10 of the WHO Code. In contrast, none of the countries have regulations or voluntary codes which cover the whole range of products outlined in Article 3 of the WHO Code; the majority cover infant formula and follow-on milk while some only cover infant formula.

There is some consistency across those countries part of the European Union, and Norway, which have all adopted partial legislation of the WHO Code in line with the European Directive (EU Directive 2006/141/EC).The EU directive covers Articles 2 to 6 and Article 9 of the WHO Code; however none of the European countries have regulations or codes covering WHO Code Articles 7 and 8. There is also limited coverage of Article 11.

In contrast to European Union countries and Norway, Canada and the USA have a restricted implementation of the WHO Code with only Articles 9 and 11 implemented in national legislation. There are no national voluntary codes in operation and extensive marketing and advertising of infant formula takes place. The implementation of the WHO Code in these countries with legislation is considered likely to breach free-trade agreements and while Canada had a voluntary code in the 1980s, the arrival of Nestle into the market and subsequent changes in the behaviour of infant-formula manufacturers in North America suggests that a voluntary code would no longer be possible in these markets.

Both Australia and NZ have voluntary codes in operation which, unlike the EU Directive, covers all Articles of the WHO Code although they defer to the ANZ Food Standards Code for Articles 9 and 10. However, similar to the EU Directive, these codes are limited in scope to infant formula and follow-on formula and furthermore, they only cover signatories to the code and there is a limited ability to enforce them. Nevertheless, both are formally monitored and there appears to be few violations. Norway also has a voluntary code in operation, in addition to national legislation; however the details of this code were not identified in the study. See Table 32 for a comparison of implementation of the WHO Code across countries.

In addition, NZ has a separate code for Health Workers, which is also formally monitored. A formal code may be problematic, with one NZ review noting that unlike manufacturers who have agreed to a code of practice, health workers have not. Health workers could be subjected to a complaint under a code of which they are not aware of or may fail to provide necessary advice on infant formula for fear of breaching the code. Health workers conduct under the WHO Code is also affected by the implementation of the BFHI.

Although Norway’s implementation of the WHO Code appears to mirror that of other European nations, it is unique amongst the eight countries in including full implementation of the WHO Code into legislation as an objective under its Action Plan for better Nutrition 2007-2011. The extent to which this has occurred remains unclear.
Top of page

Potential facilitators and barriers

The Baby-Friendly Hospital Initiative
Although all eight countries have made some progress towards implementing the BFHI, the extent and the timing of implementation has varied considerably. Norway, with over 90% of births in BFHs and 19 out of 21 accredited neonatal, has the highest rate of implementation. This initiative has a long history in Norway with 77% of births occurring in BFHI accredited hospitals by 1996. The implementation of the BFHI in NZ is also high but has a more recent history and impetus coming from a 1999 Government decision to fund the NZBA to drive the initiative. All other countries have considerably lower rates of implementation (Table 33).

Notably, the two countries, with a high percentage of accredited Baby Friend Hospitals, are small and centralised and have government-funded authorities responsible for the implementation of the initiative. For larger, decentralised countries such as Canada and the USA, these types of decisions are delegated to their Provinces or States and therefore there are significantly different rates of implementation across the country. For example, Quebec has mandated implementation of the BFHI and has 17 of Canada’s 22 BFHs. It is unlikely that these countries will progress by instigating a nation-wide and co-ordinated implementation of the BFHI.

The remaining countries included in the study have, to some extent, national recommendations for implementation but it is limited by factors such as resources, funding and the strength of the recommendation. In Australia there is a designated body, the Australian College of Midwives, to implement the BFHI. However there is no associated government funding despite it being a recommendation of the Best Start Inquiry. France has included extra requirements for BFHI accreditation beyond the ten steps but while implementation is recommended by a number of government bodies, there is no mention of the funding available to action this. Similarly, the Irish government has included targets for the BFHI in its breastfeeding strategic plan yet the targets do not call for full implementation. In the UK, recommendations to implement the BFHI were made by NICE based on a review of the evidence for effectiveness and cost-effectiveness. These recommendations carry significant weight in the National Health System in the UK and may spur greater uptake.

There is some evidence of the effectiveness of the BFHI, notably the large cluster-RCT conducted in Belarus which demonstrated improved health outcomes following the implementation of a breastfeeding program modelled on the BFHI. One of the steps of the BFHI is to ensure training of health workers. Widespread implementation of the BFHI may be a mechanism in which to improve compliance with Article 7 of the WHO Code as it applies to health workers and would limit the difficulties noted for the Health Workers Code in NZ. Appropriate professional training through the BFHI would be expected to ensure that health workers are equipped with the skills and knowledge necessary to support the principles of the WHO Code.

The majority of the countries have begun extending the BFHI to include community settings, predominately utilising the seven steps first developed in the UK. In addition Norway and Quebec have worked to develop ten steps for implementation in neonatal units and Norway has made significant progress in implementing this. Evidence for implementing the BFHI in these settings is currently lacking.

Parental leave, childcare and other government initiatives
The provision of both paid and unpaid maternity, paternity and parental leave varies greatly across the eight countries reflecting the differing social systems and approaches to family and public spending. The most generous leave arrangements are in Norway, which has an extended and well paid parental leave scheme. This appears to be linked to high rates of female workforce participation and high fertility rates. There is also a low rate of childcare utilisation for children less than one year but high rates for those less than three years. In contrast, the USA statutory requirement is 12 weeks of unpaid maternity leave and an estimated 50.1% of mothers with children less than 12 months are employed.

In France, New Zealand and recently Australia, they have paid leave in the range of 14 to 18 weeks, which is at or slightly above the minimum requirements specified by the Maternity Protection Convention. Germany has recently adopted an extended period of paid leave modelled on the Scandinavian approach, in an attempt to boost fertility rates.

There are different types of family policies and one study characterised these as:
  • General family support (e.g. cash child allowances, family tax benefits, public day care for older children). These policies support, or are neutral to, a traditional gendered division of labour.
  • Dual earner support (e.g. paid maternity and paternity leave, public day care for the youngest children).
The study ranked 18 OECD countries and only Scandinavian countries (including Norway) ranked high on dual earner support while Germany and France ranked high on general family support. A third group of countries were ranked low on both measures and classified as market oriented and these included Australia, Canada, New Zealand, the UK and the USA (Korpi 2000). Although this study is dated, these characterisations highlight the links between culture and policy which may have an impact on the choices made by families.

Demonstrating that these policies can influence breastfeeding rates, but the influence may be limited, is a study from Canada which compared breastfeeding duration before and after an increase in paid parental leave from 25 to 50 weeks. The length of time away from work increased by almost 2.3 months while the months of breastfeeding increased by just 0.75 months. Supporting this viewpoint, the Infant Survey in the UK reported that a lower proportion of mothers mentioned return to work as a factor behind giving up breastfeeding after the introduction of extended maternity leave benefits.

Direct government promotion and support for breastfeeding was evident in all eight countries with differences in the timing, degree and type of support. Similar to the implementation of the BFHI, in decentralised countries such as the USA and Canada, government initiatives in individual Provinces or States differ greatly and are likely to be equally influential. The majority of countries have formal plans on breastfeeding (whether stand alone or within larger government plans) and have appointed a breastfeeding committee. Canada is an exception to this where responsibility for breastfeeding has been delegated to the Provinces. In a historical study of breastfeeding in Canada, Nathoo notes that in the 1980s Canada had well-funded government promotion efforts, but the rapid increase in breastfeeding rates preceded these initiatives and remained relatively flat during these initiatives (Nathoo & Ostry 2009). It is likely that governmental support for breastfeeding has been more sustained in Norway compared with other countries in this study.

Cultural aspects are likely to have a significant influence of breastfeeding practice, and some consistent findings across the countries include higher rates of breastfeeding in women who are older, have higher levels of education and have a higher socioeconomic status. However, it is difficult to compare attitudes, beliefs and values relating to breastfeeding across countries as unlike many of the other aspects considered so far in this comparative study, there are no standard quantitative measures (e.g. percentage of women who have breastfed in public). We did not identify qualitative studies from Norway and there did not appear to be any barriers to breastfeeding discussed in the retrieved government publications. This absence of information supports the notion that there are few barriers to public breastfeeding in Norway and widespread acceptance of breastfeeding as the normal means of infant feeding. In contrast, many of the other countries (e.g. Canada, Australia, France, Ireland, UK) had a body of qualitative literature documenting the barriers women experience, particularly around breastfeeding in public. These factors are likely to have an impact particularly on breastfeeding duration, having their greatest effect in the medium to long post-natal period and beyond. New Zealand seems to have accepting and supportive culture for breastfeeding, as shown by the support for World Breastfeeding Week as well as local business support for public breastfeeding in various locations. The cultural impact on breastfeeding is harder to define and is heterogeneous in the USA due to the unique laws and cultures in each state and how state policies tend to take precedence over national efforts.

Our review of both the international guidelines and medical literature identifies a number of interventions which have been shown to be effective for increasing both the initiation and duration of breastfeeding. A frequent finding is the effectiveness of providing skilled or professional support and similarly, training and education of health professionals to enable them to provide this support. Across the eight countries, it was difficult to gauge the quality and extent of health worker training in lactation support. One measure of this is the implementation of the BFHI but other mechanisms to ensure widespread and high-quality training are feasible. This review also highlighted a number of practices related to the BFHI as being effective, including avoiding supplementary fluids and foods, providing early and unrestricted skin-to-skin contact and rooming in as well as the BFHI as a whole.

One finding of the evidence review, supported by the findings from this study, runs parallel with the conclusions made by the EU. The EU report articulated that a “combination of multi-faceted integrated programs seems to have a synergistic effect.” This is a consistent finding across a range of public health interventions in which single interventions may show limited effectiveness but sustained and multi-pronged approaches create synergistic effects and build upon established gains.

Top of page

Limitations of the study

This study was conducted in a rapid time-frame and attempts to cover a wide range of factors beyond the implementation of the WHO Code. For the comparisons between countries, a systematic approach to the identification of literature was possible to a limited extent and website searches, references and information sources cited within identified references were also relied on. While every effort was made to extract data without bias, without a formal systematic review, this type of comparative study is at risk of selection bias in the identification and inclusion of studies and data. Furthermore, a number of the countries used in this study were non-English and therefore our ability to retrieve the evidence available may have been limited and there is some chance that we have missed data or key information from these countries. Finally, with regards to the data included in the study, and as aforementioned, breastfeeding rates should be interpreted with caution due to the apparent differences in methods of data collection and breastfeeding definitions.

Table 32: International comparison of degree of implementation of the WHO Code
Country Implementation WHO Code Articles
2 4 5 6 7 8 9 10 11
Australia Extent ++ ++ ++ ++ ++ +++ +++ +++ ++
Type of measure Voluntary Code Legislation Voluntary Code
Canada Extent + NE NE NE NE NE ++ +++ NE
Type of measure Legislation Legislation
France Extent ++ ++ ++ ++ NE NE ++ +++ NE
Type of measure Legislation Legislation
Germany Extent ++ NE NE NE NE NE ++ NE ++
Type of measure Legislation Legislation Provisional code
Ireland Extent ++ ++ ++ ++ NE NE ++ +++ ++
Type of measure Legislation Legislation
New Zealand Extent ++ ++ ++ ++ ++ ++ ++ +++ +++
Type of measure Voluntary code Legislation Voluntary code
Norway Extent ++ ++ ++ ++ NE NE ++ ++ ++
Type of measure Legislation Legislation
UK Extent ++ ++ ++ ++ NE NE ++ ++ NE
Type of measure Legislation Legislation
USA Extent + NE NE NE NE NE ++ +++ NE
Type of measure Legislation Legislation

Articles: Article 2. Scope of the code; Article 4. Information and education; Article 5. The general public and mothers; Article 6. Health care systems; Article 7.Health workers; Article 8.Persons employed by manufacturers and distributors; Article 9. Labelling; Article 10. Quality; Article 11.Implementation and monitoring.
Key: NE = Non-existent; + = Reduced implementation; ++ = Partial implementation; +++ = Total implementation

Table 33: International comparison of breastfeeding rates and influential factors
Australia Canada France Germany Ireland New Zealand Norway UK USA
Total population 22.6 million 33 million 56 million 82 million 4.5 million 4.4 million 4.9 million 62 million 307 million
Total fertility rate (year) 1.90 (2009) 1.68 (2008) 1.99 (2009) 1.36 (2009) 2.1 (2009) 2.2 (2011) 2.0 (2008) 1.96 (2008) 2.05 (2009)
Breastfeeding data Is there any government (systematic) collection of data? Yes No No No No Yes Yes Yes Yes
Initial rate 92% 87% 63% 90% 50% 88% 99% 81% 75%
6 months rate (any) 56% 54% N/A 51% 9% N/A 80-82% 25% 44%
6 months rate (exclusive or fully) 14% 28% N/A 10-20% 2.4% 25% 2-10% <1% 15%
National strategies Plan to promote breastfeeding? In development No Yes No Yes Yes Yes No Yes
Breastfeeding Committee? Yes No No Yes Yes Yes Yes Yes Yes
BFHI No. accredited hospitals/total no. hospitals 77/330 22/500 10 65 7/20 71/79 43/53 (19/21 neonatal wards) 69/460 114
% of births accredited hospitals 23% N/A N/A N/A 20-40% N/A >90% 14-61% <20%
Extended? No Yes No No Yes Yes Yes Yes Yes
Statutory parental leave and breastfeeding breaks Length of unpaid parental leave 12 months 52-54 weeks (Quebec 70 weeks) 36 months 36 months 40 weeks 56 weeks 12 months 12 months 12 weeks
Length and rate of paid parental leave 18 weeks, minimum wage 50 weeks, 55% of earnings 16 weeks full pay 14 months 26 weeks 14 weeks 46 weeks full pay or 56 weeks at 80% pay (10 weeks of which are reserved for father) 39 weeks None. Seven states have some paid perntal leave or support legislation
Allowance for breastfeeding breaks at work? No No No Yes Yes Yes Yes - but not paid No Yes
Workforce participation rates All women (men) 59.1% (72.2%) 58.3% (65.2%) 60.1% (68.5%) 66.2% (75.6%) 57.4% (66.3%) 62.4% (73%) 71% (77%) 65.0% (74.8%) 58% (71%)
Women part-time rate (men) 45.7% (16.1%) 26.9% (11.9%) 29.7% (5.7%) 44.8% (8.6%) 33.4% (9.8%) 35.2% (11.8%) 43.0% (13%) 41.7% (10.4%) 18.8% (8.1%)
Childcare arrangements % of infants less than 3 year enrolled in childcare 29 24 42 18 31 38 51 41 31

Top of page