4th Report on the World Nutrition Situation - Nutrition throughout the Life Cycle (ACC/SCN, 2000, 138 p.): CHAPTER 3: BREASTFEEDING AND COMPLEMENTARY FEEDING: 3.5 Breastfeeding and Complementary Feeding Patterns and Trends

4th Report on the World Nutrition Situation - Nutrition throughout the Life Cycle (ACC/SCN, 2000, 138 p.)

CHAPTER 3: BREASTFEEDING AND COMPLEMENTARY FEEDING

		(introduction...)
		3.1 Evidence Linking Breastfeeding to Improved Outcomes
		3.2 Evidence Linking Complementary Feeding to Improved Outcomes
		3.3 Conceptualizing Infant Feeding Behaviours
		3.4 HIV and Infant Feeding
		3.5 Breastfeeding and Complementary Feeding Patterns and Trends
		3.6 Role of National and International Initiatives in Support of Optimal Infant Feeding
		3.7 Looking Forward: The Importance of Exclusive Breastfeeding

3.5 Breastfeeding and Complementary Feeding Patterns and Trends

This section uses data from nationally representative household demographic and health surveys to provide trend information on breastfeeding, complementary feeding, and bottle feeding. Data are presented from national surveys conducted with technical support from the Demographic and Health Surveys (DHS). Current status data are used to determine infant feeding patterns to avoid problems associated with recall bias. Surveys from 1990 to 1996 reporting on current infant and childhood feeding patterns are available for 37 countries. Twenty-seven countries supported by DHS or World Fertility Surveys have multiple surveys that permit analysis of trends of breastfeeding initiation and of the median duration of breastfeeding covering the period 1975 to 1996. These data were published in June 1999 by Macro International in a report titled Breastfeeding and Complementary Feeding, and the Postpartum Effects of Breastfeeding.³⁴In addition, analysis of DHS data for Latin America and the Caribbean by the Pan American Health Organization permits presentation of trends in exclusive breastfeeding for the years 1986 to 1996.³⁵

This review shows that breastfeeding initiation rates are very high in developing countries. The incidence of breastfeeding initiation exceeds 90% in almost every country and exceeds 95% in more than half of the countries. Breastfeeding initiation is universal in Sub-Saharan Africa. However, breastfeeding practices are far from optimal. The recommendation for exclusive breastfeeding is not widely practiced (Table 3.1). In general, the proportion of infants under four months of age who are exclusively breastfed is highest on average in Asia (up to 82% in Nepal) and the Near East/North Africa (63% in Morocco) followed by Latin America and Sub-Saharan Africa. There is significant variation within regions. For example, in Latin America and the Caribbean, the proportion of infants under four months of age who are exclusively breastfed ranges from less than 5% in Haiti to more than 50% in Bolivia, Guatemala, and Peru.

WHO and others recommend that all infants six to nine months of age be breastfed and also receive complementary foods. Data on “breastfeeding plus complementary foods” shown in Table 3.1 reflect the timing of introduction of complementary foods but provide no information about the quality of these foods. Again there is large variation within regions. For example, in Sub-Saharan Africa one-third of infants in Ghana and Mali are receiving complementary foods, compared with 90% or more in Kenya and Zimbabwe. In general, low prevalences for this indicator in Latin America and the Caribbean are because infants are no longer being breastfed rather than because they are not receiving complementary foods. Delayed introduction of complementary foods is a serious problem in some countries. Less than one-third of infants receive complementary foods by six to nine months in Bangladesh, India, and Pakistan, according to surveys conducted in the early 1990s.

TABLE 3.1: Breastfeeding and complementary feeding practices

Region, country, and date		Exclusive breastfeeding, < 4 months (%)	Breastfeeding plus complementary foods, 6-9 months (%)	Breastfeeding, 12-15 months (%)	Breastfeeding, 20-23 months (%)	Bottle feeding, <4 months (%)
Sub-Saharan Africa
	Burkina Faso 1993	3.1	43.9	98.0	81.7	1.6
	Cameroon 1991	7.3	77.1	80.7	37,8	17.8
	Central African Republic 1994-95	4.4	93.1	96.6	54.8	2.3
	Cd’Ivoire 1994	3.4	66.2	94.9	47.5	4.5
	Ghana 1993	8.3	35.7	94.6	55.4	23.9
	Kenya 1993	17,0	90.1	91.3	57.6	16.4
	Madagascar 1992	48.4	80.0	91.1	51.7	3.7
	Malawi 1992	3.3	87.2	94.5	59.9	3.8
	Mali 1995	12.2	32.9	94.3	62.8	2.9
	Namibia 1992	21.6	64.7	69.1	24.8	31.1
	Niger 1992	1.3	73.2	96.3	63.7	2.1
	Nigeria 1990	1.4	51.7	86.9	46.3	33.6
	Rwanda 1992	90.2	67.8	98.6	89.0	0.7
	Senegal 1992-93	6.5	58.9	94.1	50.6	6.2
	Tanzania 1991-92	32.5	58.7	94.7	59.5	4.1
	Uganda 1995	70.4	63.6	88.5	45.0	5.6
	Zambia 1992	13.3	87.6	91.1	34.9	3.2
	Zimbabwe 1994	16.3	93.0	89.8	26.7	5.8
Near East/North Africa
	Egypt 1992	54.0	52.3	77.9	42.4	19.9
	Jordan 1990	32.0	48.1	48.3	17.2	29.8
	Morocco 1992	62.9	35.1	64.6	31.7	23.1
	Turkey 1993	13.9	47.1	47.5	18.2	29.4
Asia
	Bangladesh 1993-94	53.5	29.4	96.1	91.2	0.0
	India 1992-93	51.2	31.5	89.0	73.0	9.2
	Indonesia 1994	47.6	85.8	88.8	76.6	10.6
	Kazakhstan 1995	12.3	62.7	52.3	24.1	69.0
	Nepal 1996	82.0	62.5	96.7	92.7	2.3
	Pakistan 1990-91	25.1	29.2	79.7	59.4	26.2
	Philippines 1993	33.1	52.1	58.2	24.2	43.0
Latin America/Caribbean
	Bolivia 1993-94	54.0	78.0	74.7	42.7	34.7
	Brazil 1996	40.8	11.6	34.8	18.8	50.9
	Colombia 1995	15.7	61.5	42.8	18.8	72.0
	Dominican Republic 1991	9.6	22.8	29.5	9.1	80.8
	Guatemala 1995	50.7	55.7	78.0	54.0	26.4
	Haiti 1994	3.3	84.1	82.4	26.7	36.4
	Paraguay 1990	7.0	61.2	40.3	11.1	56.1
	Peru 1996	61.4	72.2	81.6	45.8	28.6

Source: Abstracted or calculated from information presented in 34.

The prevalence of breastfeeding between 12 and 15 months is relatively high in all regions. In Sub-Saharan Africa 13 of the 18 countries reported here have breastfeeding rates exceeding 90%. For other regions, again, there is significant variation. In Latin America and the Caribbean, breastfeeding at 12-15 months ranges from a low of 30% in the Dominican Republic and 35% in Brazil to about 80% in Guatemala, Haiti, and Peru. Prevalences of breastfeeding among children 20 - 23 months of age are highest in Sub-Saharan Africa and Asia (over 50%) and lowest in Latin America and the Caribbean and the Near East/North Africa (about 25%).

Bottle feeding carries a significant risk of morbidity because of contamination³⁶ and is therefore a useful indicator to follow. Bottle feeding is highly variable, depending on the region and the country. In general, it is lowest in Sub-Saharan Africa, although one-third of infants are bottle fed in the first four months in Namibia and Nigeria. Bottle feeding rates are very high in Latin America and the Caribbean. In four of the eight countries reported on, more than half of infants less than four months are bottle fed. A very high rate of bottle feeding is reported for Kazakhstan in Central Asia (69%), whereas rates are much lower elsewhere in Asia - for example, 9.2% in India.

TABLE 3.2: Trends in the median duration of breastfeeding

Region, country, and date		Median duration (months)
Sub-Saharan Africa
	Cameroon 1978	17.8
	Cameroon 1991	18.2
	Ghana 1979-80	17.8
	Ghana 1988	21.5
	Ghana 1993	21.6
	Kenya 1977-78	16.6
	Kenya 1989	20.0
	Kenya 1993	21.6
	Mali 1987	18.6
	Mali 1995	22.5
	Nigeria 1982	17.0
	Nigeria 1990	19.8
	Rwanda 1984	20.9
	Rwanda 1992	28.0
	Senegal 1978	18.0
	Senegal 1986	18.9
	Senegal 1992-93	20.2
	Uganda 1988-89	19.5
	Uganda 1995	19.3
	Zimbabwe 1988-89	19.1
	Zimbabwe 1994	18.1
Near East/North Africa
	Egypt 1980	15.3
	Egypt 1988-89	19.8
	Egypt 1992	19.8
	Jordan 1976	10.9
	Jordan 1990-91	12.1
	Morocco 1980	15.9
	Morocco 1987	15.4
	Morocco 1992	15.5
	Turkey 1978	11.7
	Turkey 1993	11.7
Asia
	Bangladesh 1975-76	25.8
	Bangladesh 1993-94	36.0
	Indonesia 1976	23.3
	Indonesia 1987	23.3
	Indonesia 1991	23.6
	Indonesia 1994	23.7
	Nepal 1976	24.2
	Nepal 1996	31.3
	Pakistan 1975	17.8
	Pakistan 1990-91	18.5
	Philippines 1978	14.1
	Philippines 1993	15.4
Latin America/Caribbean
	Bolivia 1989	16.4
	Bolivia 1993-94	17.5
	Brazil 1986	5.8
	Brazil 1996	7.8
	Colombia 1976	7.9
	Colombia 1986	9.1
	Colombia 1990	9.3
	Colombia 1995	10.3
	Dominican Republic 1975	8.4
	Dominican Republic 1986	8.1
	Dominican Republic 1991	5.9
	Guatemala 1.987	20.6
	Guatemala 1995	20.2
	Haiti 1977	15.6
	Haiti 1994	17.7
	Paraguay 1979	12.4
	Paraguay 1990	11.7
	Peru 1977-78	14.5
	Peru 1986	14.1
	Peru 1991-92	18.1
	Peru 1996	20.2

Source: 34.

Overall, data on trends in the median duration of breastfeeding show positive changes between 1975 and 1996 (Table 3.2). The high median duration of breastfeeding was maintained in many countries and actually increased in some. Increases in the median duration of breastfeeding measured over a ten-year period range from about one month in the Near East/North Africa to 2.5 months in Sub-Saharan Africa. When all countries were considered together, the median duration of breastfeeding increased at a rate of 1.7 months per decade. Work is under way to extend the analysis of trends in breast-feeding and complementary feeding by developing comprehensive indicators of care. This work is described in Box 3.1.

BOX 3.1

Approaches to Measuring and Quantifying Care

Care is defined as “the behaviours and practices of caregivers (mothers, fathers, siblings, and child-care providers) to provide the food, health care, stimulation, and emotional support necessary for children’s healthy growth and development.”³⁷ Over the past decade, care has been increasingly recognized as a crucial input into child health and nutrition, along with food security, availability of health services, and a healthy environment.

In 1995 a special issue of the Food and Nutrition Bulletin presented an extensive review of the literature on care and its determinants and provided recommendations for protecting and enhancing the quality of care. A useful distinction has also since been made between “care resources” and “care practices.” The latter includes the following six main areas of caregiving behaviours: (1) care for women; (2) feeding/breastfeeding; (3) psycho-social and cognitive stimulation; (4) hygiene practices; (5) home health practices; and (6) food preparation and storage. Indicators for measuring care were also identified.

In 1997 WHO, along with IFPRI and the University of Ghana, undertook a survey in Accra using the food-health-care conceptual model to examine the determinants of child undernutrition. One of the greatest challenges was to identify an approach to measuring and quantifying care. Data on child feeding practices and use of preventive health services were collected, and the information was used to create an age-specific child-care index. The index was particularly useful in examining the importance of care for child undernutrition and in studying the determinants of caring practices in this population. The main findings were (1) good care practices for children 4 - 36 months of age were beneficial for child nutritional status and could partially compensate for the negative effects of low maternal schooling and poverty; (2) the most important constraint to good child care practices was maternal schooling; and (3) household socioeconomic factors, food security, maternal work, and other maternal characteristics did not interfere with caring practices in this population.^{38, 39}

In collaboration with Cornell University and Emory University, IFPRI is now using DHS data sets to pursue its work in measuring and quantifying care. The focus is on using the information available on breastfeeding and complementary feeding to derive a child feeding index for various countries and regions around the world. Preliminary work shows promising results, and the findings should help improve our understanding of the role of care in the etiology of childhood undernutrition in different settings. The research will also help identify minimum data needs and useful indicators for measuring and quantifying care. The current focus is on child feeding practices because of data availability, but research is urgently needed on indicators and methodologies to quantify other more complex aspects of care, such as psychosocial and cognitive stimulation or care of women, which remain largely understudied.