Health Policy and Planning, 34, 2019, 407–417

doi: 10.1093/heapol/czz050
Advance Access Publication Date: 24 June 2019
Original Article

The cost of not breastfeeding: global results

from a new tool
Dylan D Walters1,*, Linh T H Phan 2 and Roger Mathisen2
1
Nutrition International, 180 Elgin St, Ottawa, ON, K2P 2K3, Canada and 2Alive & Thrive, 60 Ly Thai To Street, Hoan
Kiem, Ha Noi, Viet Nam

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

*Corresponding author. Nutrition International, 180 Elgin St, Ottawa, ON, K2P 2K3, Canada. E-mail: [email protected]
Accepted on 14 May 2019

Abstract
Evidence shows that breastfeeding has many health, human capital and future economic benefits
for young children, their mothers and countries. The new Cost of Not Breastfeeding tool, based on
open access data, was developed to help policy-makers and advocates have information on the
estimated human and economic costs of not breastfeeding at the country, regional and global lev-
els. The results of the analysis using the tool show that 595 379 childhood deaths (6 to 59 months)
from diarrhoea and pneumonia each year can be attributed to not breastfeeding according to
global recommendations from WHO and UNICEF. It also estimates that 974 956 cases of childhood
obesity can be attributed to not breastfeeding according to recommendations each year. For
women, breastfeeding is estimated to have the potential to prevent 98 243 deaths from breast and
ovarian cancers as well as type II diabetes each year. This level of avoidable morbidity and mortal-
ity translates into global health system treatment costs of US$1.1 billion annually. The economic
losses of premature child and women’s mortality are estimated to equal US$53.7 billion in future
lost earnings each year. The largest component of economic losses, however, is the cognitive
losses, which are estimated to equal US$285.4 billion annually. Aggregating these costs, the total
global economic losses are estimated to be US$341.3 billion, or 0.70% of global gross national
income. While the aim of the tool is to capture the majority of the costs, the estimates are likely to
be conservative since economic costs of increased household caregiving time (mainly borne by
women), and treatment costs related to other diseases attributable to not breastfeeding according
to recommendations are not included in the analysis. This study illustrates the substantial costs
of not breastfeeding, and potential economic benefits that could be generated by government and
development partners’ investments in scaling up effective breastfeeding promotion and support
strategies.

Keywords: Breastfeeding, nutrition, maternal and child health, economic evaluation

Introduction 2016). For these reasons, the World Health Organization (WHO)
Newborns and young children receive essential benefits through and UNICEF recommend early initiation of breastfeeding within an
breastfeeding resulting in improved survival, health and human cap- hour of birth, exclusive breastfeeding of infants for the first 6
ital outcomes. Breastfeeding reduces the risk of childhood infections months of life and continued breastfeeding with complementary
such as diarrhoea and pneumonia and premature mortality as well foods for two or more years (UNICEF, 2016a).
as minimizes nutrition-related harm to cognitive development in Despite the substantial evidence on the health and cognitive
early childhood (Horta et al., 2015; Victora et al., 2016). For breast- benefits of breastfeeding, the vast majority of children globally are
feeding mothers, breastfeeding reduces the risk of post-partum not breastfed in line with the recommendations. The global preva-
haemorrhage and depression as well as premature mortality from lence of exclusive breastfeeding increased from 36% in 2000 to
several diseases later in life (Chowdhury et al., 2015; Victora et al., 43% in 2015, and the current prevalence of early initiation of

C The Author(s) 2019. Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine.
V
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits
unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 407
408 Health Policy and Planning, 2019, Vol. 34, No. 6

Key Messages
• The new Cost of Not Breastfeeding Tool is a first-of-its-kind resource to help policy-makers and advocates quantify the
human and economic costs of not breastfeeding including lost life, lost productivity, and increased costs to health sys-
tems at country, regional and global levels.
• Globally, 595 379 childhood deaths are attributed to not breastfeeding annually. Optimal breastfeeding also has the po-
tential to prevent an additional 98 243 deaths of mothers from cancers and type II diabetes each year.
• The total annual global economic losses are estimated to be between US$257 billion and US$341 billion, or between
0.37% and 0.70% of global gross national income.
• The costs of not breastfeeding are significant and should compel policy-makers and donors to invest in scaling up
breastfeeding and nutrition interventions for children and their mothers to strengthen human capital development and
economic outcomes around the world.

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

breastfeeding and continued breastfeeding until 2 years of age are accurate estimates of the human and economic consequences of
45% and 46%, respectively (UNICEF, 2016a). The Lancet Series on not breastfeeding and exposure to breastmilk substitutes in their
Breastfeeding estimated that >800 000 child deaths globally and countries on a consistent basis. Alive & Thrive, an initiative man-
cognitive losses totalling US$302 billion per year were attributable aged by FHI 360 and funded by the Bill and Melinda Gates
to not breastfeeding according to recommendation and exposure to Foundation that is dedicated to saving and improving lives through
breastmilk substitutes (Rollins et al., 2016; Victora et al., 2016). optimal maternal nutrition, breastfeeding and complementary feed-
The current pace of increase in the prevalence of exclusive ing practices, supported the development of the tool. The objective
breastfeeding is insufficient for achieving the World Health of this article is to present the underpinning methodology of the new
Assembly’s (WHA) Global Nutrition Target of ‘increasing the rate Cost of Not Breastfeeding Tool as well as new findings on the cost
of exclusive breastfeeding in the first six months up to at least 50%’ of not breastfeeding at the country, regional and global levels.
by 2025 (WHO and UNICEF, 2014). While the cultural, social, eco-
nomic and corporate forces that shape breastfeeding in different
regions around the world may be challenging to counter, there is evi- Methods
dence that policy and programmatic actions by governments, donors
This Cost of Not Breastfeeding Tool was developed in Microsoft
and civil society can effectively increase the prevalence of breastfeed-
Excel, and is now available as both an interactive online tool
ing practices (Rollins et al., 2016). The World Bank’s Investment
(Figure 1) and a downloadable workbook file.1 The user-friendly
Framework for Nutrition estimated that US$5.7 billion in additional
online tool summarizes the results of the national estimates for the
financing is needed from 2016 to 2025 to scale up breastfeeding pro-
human consequences (including morbidity and mortality) and eco-
motion interventions across low- and middle-income countries
nomic costs for select LMICs. The tool provides users with data and
(LMICs) to achieve the WHA target for breastfeeding (Shekar et al.,
narrative advocacy briefs that summarize the results. The download-
2017; Walters et al., 2017). Any new resources mobilized for
able workbook version, intended for advanced technical users, con-
increasing breastfeeding should be invested towards fully imple-
sists of worksheets containing open access datasets, the calculations
menting and enforcing the International Code of Marketing of
for each indicator, and the results of the global, regional and nation-
Breastmilk Substitutes, enacting paid family leave and workplace
al analysis for the human and economic costs for over 130 countries.
breastfeeding policies, implementing the Ten Steps to Successful
The workbook version includes a macro programme that computes
Breastfeeding in facilities providing maternity and newborn services
the regional and global estimates based on two user inputs (i.e. dis-
(WHO, 2017) and improving access to skilled breastfeeding
count rate and long-term Gross Domestic Product (GDP) growth
counselling.
rate). This section describes the analytical methods and data sources
Previous studies have quantified the economic costs of not
used for each of the included indicators of the cost of not
breastfeeding, or malnutrition, at the global level (Holla et al.,
breastfeeding.
2015; Rollins et al., 201641), regional level (Horton et al., 1996;
Walters et al., 2016) and national or sub-national levels in some
countries (Smith et al., 2002; Büchner et al., 2007; Bagriansky and Analytical methods
Voladet, 2013; Council for Agriculture and Rural Development The analytical methods used in the Cost of Not Breastfeeding Tool
et al., 2013; Bagriansky, 2014; Bartick et al., 2017). The recent nu- were drawn from previously published studies on the economic con-
trition investments of US$500 million in Nigeria in 2016 and sequences of malnutrition (Bagriansky and Voladet, 2013;
US$4.5 billion in Indonesia in 2017 nutrition are examples where Bagriansky, 2014), cost of not breastfeeding (Horton et al., 1996;
the use of economic research by advocacy groups helped to raise pol- Drane, 1997; Smith et al., 2002; Büchner et al., 2007; Bartick and
icy-maker awareness that led to the prioritization and investments Reinhold, 2010; Bartick et al., 2013; Pokhrel et al., 2014; Walters
in nutrition programmes. Unfortunately, there remain gaps in access et al., 2016; Bartick et al., 2017), the Lancet Series on Breastfeeding
to evidence on the human capital and economic consequences of (Rollins et al., 2016; Victora et al., 2016) and the Investment
malnutrition or not breastfeeding by policy-makers and advocates in Framework for Nutrition (Shekar et al., 2017; Walters et al., 2017).
the majority of the world’s countries. Some modifications to the methods were necessary to be compatible
To address this gap in economic research, the Cost of Not with the variables contained in open access datasets.
Breastfeeding Tool was envisioned to provide an evidence-based The cost of not breastfeeding tool incorporates three categories
and user-friendly tool for policy-makers and advocates to generate of indicators for human and economic costs attributed to not
Health Policy and Planning, 2019, Vol. 34, No. 6 409

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

Figure 1 The online Cost of Not Breastfeeding Tool.

breastfeeding according to recommendations, including (1) women’s estimates (Victora et al., 2016). This calculation is based on the fact
and child morbidity and mortality, (2) for health system and house- that lifetime duration of breastfeeding a child reduces the relative
hold formula costs and (3) the future economic costs due to mortal- risk of morbidity from cancer and it is assumed that the effect on
ity and cognitive losses. This study was based on the framework for mortality is of the same magnitude. The number of potential cases
estimating economic costs of not breastfeeding from the Rollins averted with full breastfeeding rates (100%) was calculated by mul-
et al. (2016) Lancet Series on Breastfeeding, which focused on cogni- tiplying the incidence of type II diabetes in women (Institute for
tive losses and healthcare treatment costs for children. While this Health Metrics and Evaluation, 2017) by published relative risk of
analysis was conducted from the societal perspective with certain type II diabetes (Victora et al., 2016) and the current level of breast-
costs borne by households, ministries of health and the economy as feeding in each country. Data on breast and ovarian cancer mortal-
a whole included, the tool design focused on the key costs with ity in women by country are from GLOBOCAN 2012
available data for the majority of countries rather than the exhaust- (International Agency for Research on Cancer, 2012). The tool does
ive inclusion of all type of costs. The estimates are likely to be con- not include the attributed morbidity and mortality consequences
servative since the economic costs of increased household caregiving from childhood diabetes and cancer, sudden infant death syndrome,
time, which are mainly borne by women (Leslie, 1989), transporta- and necrotizing enterocolitis. All regional and country income group
tion costs, and several child and maternal diseases that can be pre- estimates of morbidity and mortality are based on the cumulative
vented in part by breastfeeding are not included in the model due to totals from countries with available data and are not extrapolated to
the data availability. Each indicator was estimated for a 1-year co- fit the total population of the respective group.
hort of newborns and infants not being breastfed according to rec-
ommendations or their mothers. The future economic costs of
Health system and household formula costs
mortality and cognitive losses for this 1-year cohort of child–mother
The health system cost refers specifically to the direct medical costs
pairs, however, are projected into the long-term future over their
for the treatment of cases of childhood diarrhoea and pneumonia
productive years.
(Bagriansky and Voladet, 2013; Bagriansky, 2014; Walters et al.,
2016) and cases of type II diabetes in women that can be attributed
Child and maternal morbidity and mortality to not breastfeeding. The number of cases of each childhood disease
The estimation of the human cost of not breastfeeding follows the attributed to not breastfeeding was multiplied by the percentage of
same methodology from previous studies on the cost of malnutrition children with the disease taken to a health facility, the percentage of
and not breastfeeding (Bagriansky and Voladet, 2013; Bagriansky, cases taken to a health facility that receive either outpatient care
2014; Walters et al., 2016). To estimate the number of childhood services or inpatient services (Lamberti et al., 2012; Niederman and
cases and child deaths averted each year attributable to not breast- Krilov, 2013) and the percentage of patients that seek care at each
feeding according to recommendation, published relative risks for level of care (health centre, primary hospital, secondary hospital or
either the diarrhoea or pneumonia infection pathway (Black et al., teaching/tertiary hospital) in each country. The annual total cost of
2008) were multiplied by the current percentage of households in treatment of diarrhoea and pneumonia attributed to not breastfeed-
each breastfeeding behaviour category, and then multiplied by the ing is equal to the number of cases that receive outpatient and in-
morbidity or mortality for each disease for infants and young chil- patient services at each level of care multiplied by the unit cost of
dren (age 0–23 months). The method used for estimating the mor- treatment for children at each level of care.
bidity and mortality of breast and ovarian cancer and type II The default country-level unit cost data used for the estimation
diabetes attributed with not breastfeeding follows previous of the treatment of childhood diseases was from the WHO-
approaches used in the UK (Pokhrel et al., 2014) and for global CHOICE data (Johns et al., 2005), which were adjusted for the
410 Health Policy and Planning, 2019, Vol. 34, No. 6

cumulative inflation based on country-specific data from 2008 to breastfeeding below 6 months of age compared with non-exclusive
2017 in US dollars (World Bank, 2016). These unit costs pertain breastfeeding can achieve the same cognitive gains equal to a 2.62
crudely to the mean cost of general outpatient services per visit and IQ increase compared with not being breastfed (Horta et al., 2015;
inpatient care per day. Additional data on the unit costs of treatment Victora et al., 2016). The 2.62 IQ increase point estimate was
for childhood diarrhoea and pneumonia were collected from China, adjusted for maternal IQ and used in the tool in order to be conser-
Ethiopia, Ghana, Mexico and Nigeria by UNICEF and Alive & vative (Horta et al., 2015). This approach was supported by evi-
Thrive country teams, consultants and partners, either from national dence from a study that found that children exclusively breastfed for
administrative databases or directly from health facilities (see greater than 1 month experienced an increase in three IQ points
Supplementary Appendix Table A1). compared with children not or partially breastfed for less than 1
The estimated health system cost of treatment for type II diabetes month (Eickmann et al., 2007). Although there is only limited re-
in women is calculated by multiplying the number of annual cases of search on the direct effects of exclusive breastfeeding due to chal-
type II diabetes attributed to not breastfeeding by an estimate on the lenges with feasibility, the approach aligns with UNICEF and WHO
percentage of cases of type II diabetes that are diagnosed. This figure guidelines and data in the infant and young child database
represents the number of cases that are diagnosed and potentially (UNICEF, 2016a; UNICEF, 2017).

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

treated, which is then multiplied by the health expenditure per case The potential future income lost due to cognitive losses is equal
of type II diabetes in each country from the International Diabetes to multiplying the number of children not breastfed by GNI per cap-
Federation Diabetes Atlas (International Diabetes Federation, 2015). ita (World Bank, 2016), the 2.62 IQ point increase lost per child not
The cost of feeding a young child with formula from birth until breastfed (Horta et al., 2015; Victora et al., 2016) and 1.067% in-
age two as a percentage of household earnings was calculated by crease in earnings lost for each IQ point lost (Hanushek and
multiplying the total estimated quantity of formula advised by Woessmann, 2008). The total future income lost in a country is
breastmilk substitute producers and manufacturers by the unit cost equal to the sum of this calculation for each year from the point that
of formula, then divided by the nominal wage or mean earnings of the children would have turned 18 years of age until expected retire-
employees in each country. Data on the unit price of infant formula ment at age 65 or the country’s age of life expectancy, whichever
from 97 countries was collected for this study through an online comes first (World Bank, 2016).
search for e-commerce vendors based in each country. The lowest The total combined economic losses of not breastfeeding is equal
price of an economy brand of infant formula was selected for this to the sum of the health system cost, future economic cost of mortal-
analysis to be conservative since the average price paid is unknown ity and future economic cost of cognitive losses for each country, re-
(see Supplementary Appendix Table A2 for unit costs). For countries gion or the globe.2 This is also presented as a percent share of GNI
where data on the unit cost of formula was not available, the mean (World Bank, 2016). The results for economic losses presented in
global unit cost of formula was used as a proxy. this article all assume a 3% discount rate on potential future eco-
nomic losses of not breastfeeding as recommended in the Bill and
Melinda Gates Foundation Reference Case for Economic
Future economic cost of mortality and cognitive losses Evaluations (Bill and Melinda Gates Foundation, 2014), WHO-
The estimation of potential future cost of child and maternal mortal- CHOICE methodology (Johns et al., 2005) and the Investment
ity, or future income lost due to premature mortality attributed to Framework for Nutrition (Shekar et al., 2017). The calculations
not breastfeeding, follows the WHO-CHOICE methodology (Johns assumed a long-term mean annual GDP growth rate of 3% in each
et al., 2005) and the approach in the Investment Framework for country (Shekar et al., 2017) and the wage share of national income
Nutrition (Shekar et al., 2017). This cost reflects the potential con- in each country from International Labour Organization (ILO)
tribution to a country’s economy through future earnings over a per- STAT database (Lubker, 2007; ILO, 2015; Shekar et al., 2017).
son’s productive years that will be lost with premature mortality Sensitivity analysis can help to describe the nuances of estimates of
attributed to not breastfeeding. The total potential future income economic losses based on changes to key variables with uncertainty.3
lost due to child mortality is equal to the sum of multiplying the Sensitivity analysis on the economic losses is presented using both
number of child deaths attributed to not breastfeeding according to conservative assumptions of 5% discount rates on benefits and 1.5%
recommendation for the 1-year cohort by each country’s projected mean long-term GDP growth rate as well as optimistic assumptions
gross national income (GNI) per capita (World Bank, 2016) from using a 1.5% discount rate and 5% mean long-term GDP growth
the year the children turn 18 years of age until the earliest point be- rate.4 The estimated economic losses for both sets of conservative and
tween the expected retirement at age 65 or the country’s age of life optimistic assumptions are presented in parentheses next to the result
expectancy (World Bank, 2016). The calculation of total potential using default assumptions in Table 4 for regions and country income
future income lost due to maternal mortality is similar; however, the groups Supplementary Appendix Tables A6–8 for all individual coun-
economic losses are only counted for the foregone productive years tries with data available. In addition, estimates for economic losses
between the mean age of women’s cancer and diabetes-related mor- are also presented using an alternative approach for calculating child-
tality until the earliest point between the expected retirement at age ren’s cognitive losses, which assumes that breastfeeding at 6 months is
65 or the country’s age of life expectancy. associated with a 2.62 point IQ increase compared with not being
The estimation of potential future income not realized due breastfed5 (Supplementary Appendix Tables A6–8). A limitation of
to child cognitive losses attributed to not breastfeeding follows a the sensitivity analysis in the tool is that it does not account
methodology used previously (Rollins et al., 2016; Walters et al., for variation in the effects of not breastfeeding on disease morbidity,
2016; Walters et al., 2017). This cost reflects the potential contribu- mortality and cognition.
tion to a country’s economy through increased earnings over a per-
son’s productive years that will be lost due to not achieving
cognitive gains in intelligence provided by being breastfed according Data sources
to recommendations in the early years of childhood. This analysis Data sources used across the indicators for each country include: (1)
calculated the cognitive losses by assuming that exclusive the United Nations World Population Prospects for population data
Health Policy and Planning, 2019, Vol. 34, No. 6 411

Table 1 Number of annual cases of morbidity in women and children attributable to not breastfeeding by region and country income group

Number of cases of morbidity attributable to not breastfeeding

Regions Child diarrhoea Child ARI/ Childhood Breast cancer Ovarian Cancer Type II
(0–23 months) pneumonia (0–23) obesity in mothers in mothers diabetes

East Asia and Pacific 19 533 482 1 058 553 374 405 54 393 9 905 389 006
Europe and Central Asia 3 514 920 357 835 69 231 11 016 3 214 52 077
Middle East and 12 704 537 716 047 168 568 8379 902 101 441
North Africa
Latin America and 13 160 409 778 792 134 192 18 254 3 355 118 363
Caribbean
North America 33 571 64 739 0 0 0 0
South Asia 56 056 468 3 346 020 74 249 14 051 2435 121 620
Sub-Saharan Africa 60 843 179 2 317 553 154 311 11 711 1745 75 009
High income 335 534 86 075 9362 3249 728 15 719

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

Upper-middle-income 24 816 631 1 706 300 528 011 68 921 12 696 494 165
Lower-middle-income 99 358 380 5 375 448 356 162 39 297 7230 313 822
Low income 41 336 021 1 471 716 81 421 6337 902 33 809
Total 165 846 566 8 639 539 974 956 117 804 21 556 857 515

Table 2 Number of annual deaths of women and children attributable to not breastfeeding by region and country income group

Number of deaths attributable to not breastfeeding

Regions Due to child Due to child Total child Due to breast Due to ovarian Due to Type II Total number
diarrhoea ARI/pneumonia deaths cancer in cancer in diabetes in of maternal
(0–23 months) (0–23) mothers mothers mothers deaths

East Asia and Pacific 13 932 39 680 53 613 11 898 5922 19 964 37 785
Europe and Central Asia 2132 5302 7434 3007 1877 2683 7567
Middle East and 6455 15 272 21 727 1801 606 5261 7668
North Africa
Latin America and 3938 10 897 14 835 4292 2092 11 503 17 887
Caribbean
North America 0 0 0 0 0 0 0
South Asia 66 530 96 350 162 880 3444 1677 10 791 15 913
Sub-Saharan Africa 132 828 202 064 334 892 2626 1471 8028 12 125
High income 11 95 106 704 471 654 1829
Upper-middle-income 10 928 33 952 44 879 15 677 7619 29 414 52 711
Lower-middle-income 147 999 233 025 381 024 9313 4763 25 323 39 399
Low income 66 877 102 493 169 370 1374 791 2839 5004
Total 225 815 369 565 595 379 27 069 13 644 58 230 98 943

by age group (United Nations Department of Economic and Social Results

Affairs, 2015); (2) the UNICEF Infant and Young Child Feeding
Global and regional results
Database (UNICEF, 2017), data from Lancet Series on
Child morbidity and mortality
Breastfeeding Supplementary Appendix, Demographic Health
Survey (DHS) STAT compiler (Demographic and Health Surveys Based on epidemiological and demographic health surveillance data
Program, 2017), or Multiple-Indicator Cluster Surveys (MICS) com- available for 130 countries, not breastfeeding according to recom-
piler (UNICEF, 2016b) for breastfeeding practice prevalence; (3) the mendations was attributed with 166 million and 9 million avoidable
Global Burden of Disease results tool (Institute for Health Metrics and cases of diarrhoea and pneumonia in children under the age of two each
Evaluation, 2017) or WHO Global Health Observatory data (WHO, year (Table 1). Approximately two-thirds of these cases of infectious ill-
2015) for the total mortality by disease by age group and gender; (4) nesses occur in the South Asia and Sub-Saharan Africa regions and 84%
the Global Burden of Disease Results Tool (Institute for Health within countries in the lower middle-income and low-income groups
Metrics and Evaluation, 2017) for disease incidence by age group and (Table 1). This high level of avoidable morbidity leads to substantial
gender; (5) MICS compiler (UNICEF, 2016b) for data on care seeking preventable child mortality. Breastfeeding was attributed with 595 379
behaviour for childhood diarrhoea and pneumonia; (6) the World child deaths to diarrhoea (38%) and pneumonia (62%) each year
Bank’s World Development Indicators Database (World Bank, 2016) (Table 2). Of the global total of child mortality, over 56% occurs in
for other economic and health indicators such as country-level GNI, Sub-Saharan Africa and 64% occur in lower middle-income countries
GNI per capita and life expectancy; and (7) the International Labour (Table 2 and Figure 2). An estimated 974 956 cases of childhood obesity
Office’s ILOSTAT database (ILO, 2015) for the labour force participa- each year were attributed to not breastfeeding according to recommen-
tion rate, wage share of income and mean earnings. dations. Over 40% of the new preventable childhood obesity cases were
412 Health Policy and Planning, 2019, Vol. 34, No. 6

in East Asia and the Pacific region and 54% in upper middle-income estimated to be US$1.1 billion annually (Table 3). The treatment
countries (Table 1), both groupings with rapid growth in commercial costs for childhood pneumonia alone equal US$697 million each
milk-based formula sales in recent years (Baker et al., 2016). year, and childhood diarrhoea and women’s type II diabetes are esti-
mated to cost US$196 million and US$254 million, respectively. For
all three diseases, East Asia and the Pacific would incur the highest
Mother’s morbidity and mortality
cost of any region with a total of US$315 million per year. The
Breastfeeding was estimated to have the potential to prevent 27 069
upper middle-income group of countries would incur the highest
future deaths of women from breast cancer and 13 644 from ovarian
health system cost at US$604 million per year. Supplementary
cancer each year with universal breastfeeding (Table 2). Also,
Appendix Table A3 shows the total annual health system treatment
breastfeeding could potentially prevent 58 230 deaths of women
cost from not breastfeeding according to recommendation for >138
from type II diabetes. In contrast with child mortality, the majority
individual countries with data available.
of preventable maternal deaths due to not breastfeeding, 53% occur
The online search for e-commerce vendors in all countries found
in upper middle-income countries and 38% occur in the East Asia
data on infant formula in 97 countries (Supplementary Appendix
and the Pacific region alone (Table 2). Results for the maternal and
Table A2). The global mean price per 900-g container of the lowest
child morbidity and mortality indicators from 130 individual coun-

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

price of an economy brand was found to be US$18.74. With this
tries with data available can be found in Supplementary Appendix
country-level unit cost data on infant formula, it was estimated
Tables A3 and A4.
that feeding a child with an economy brand of formula for the first
2 years of life instead of breastfeeding would cost on average
Health system and household formula costs over 6.1% of a household’s wages. This figure would be even higher
The total global health system cost, based on the tool design, for the in low-income families and LMICs (Supplementary Appendix
treatment of childhood diarrhoea and pneumonia as well as wom- Table A5).
en’s type II diabetes that could be prevented with breastfeeding is
Economic costs of mortality and cognitive losses
In 125 countries with data available, the economic losses due to
child mortality, which are the future earnings not generated by over
a half-million children who die prematurely each year due to not
breastfeeding according to recommendations, equal US$53.7 billion
annually (Table 4). Over US$23.6 billion, or 43% of the total losses
calculated, would be lost in the Sub-Saharan African region, and an-
other US$10.6 and US$10.4 billion would be lost in each of South
Asia and East Asia and the Pacific regions, respectively. The eco-
nomic losses due to maternal mortality, which are the future earn-
ings not generated by 98 943 mothers who will die prematurely, is
estimated to equal US$1.26 billion annually (Table 4).
While mortality is a major contributor to economic losses for
LMICs, the future economic losses due to cognitive losses in children
are found to be much larger globally. Using data in 136 countries to
calculate cognitive losses, the economic losses by persons who were
not exclusively breastfed according to recommendations increased
to US$285.39 billion annually (Table 4). The economic losses esti-
mated for North America decreased to US$114.9 billion, or 43% of
the total losses calculated. The economic losses for regions with a
Figure 2 Number of child deaths attributed to not breastfeeding by region. higher concentration of LMICs increased substantially with this

Table 3 Total health system cost attributed to not breastfeeding by region and country income group

Cost of avoidable healthcare treatment due to not breastfeeding

Regions Due to childhood Due to childhood Due to type II diabetes Total (US$ m)
diarrhoea (US$ m) pneumonia (US$ m) morbidity in mothers’ (US$ m)

East Asia and Pacific 78.40 182.05 54.60 315.04

Europe and Central Asia 27.16 38.81 29.86 95.84
Middle East and North Africa 19.88 81.51 30.69 132.08
Latin America and Caribbean 48.30 128.67 59.99 236.96
North America 0.00 1.83 26.44 28.27
South Asia 5.43 128.38 33.59 167.41
Sub-Saharan Africa 17.02 135.44 18.76 171.21
High income 20.76 16.83 33.86 71.45
Upper-middle-income 149.10 308.56 146.79 604.46
Lower-middle-income 25.16 335.49 69.38 430.03
Low income 1.17 35.80 3.91 40.88
Total 196.19 696.69 253.94 1146.81
Health Policy and Planning, 2019, Vol. 34, No. 6 413

Table 4 Global and regional economic losses due to mortality, cognitive losses and total attributable to not breastfeeding by region and by
country income group

Economic losses attributable to not breastfeeding

Due to child Due to maternal Due to cognitive Total cost Total cost as
mortality (US$ b) mortality (US$ b) losses (US$ b) (health, mortality % of GNI
and cognitive) (US$ b)

East Asia and Pacific 10.39 (2.71, 49.67) 0.66 (0.55, 0.80) 74.76 (19.46, 357.66) 86.12 (23.04, 408.45) 0.59 (0.16, 2.78)
Europe and Central Asia 1.35 (0.35, 6.38) 0.25 (0.21, 0.29) 14.76 (3.86, 70.15) 16.27 (4.37, 76.71) 0.42 (0.11, 1.97)
Middle East and North Africa 3.76 (0.98, 17.99) 0.03 (0.03, 0.04) 18.65 (4.85, 89.38) 22.57 (5.98, 107.54) 0.91 (0.24, 4.32)
Latin America and Caribbean 4.08 (1.07, 19.35) 0.28 (0.23, 0.34) 32.25 (8.41, 154.02) 36.85 (9.94, 173.95) 0.70 (0.19, 3.32)
North America 0.00 (0.00, 0.00) 0.00 (0.00, 0.00) 114.94 (29.87, 551.37) 114.97 (29.90, 551.40) 0.63 (0.16, 3.04)
South Asia 10.58 (2.75, 50.59) 0.02 (0.02, 0.02) 11.73 (3.05, 56.12) 22.49 (5.99, 106.90) 0.84 (0.22, 3.99)
Sub-Saharan Africa 23.56 (6.59, 101.00) 0.02 (0.01, 0.02) 18.31 (5.05, 80.05) 42.06 (11.82, 181.25) 2.58 (0.72, 11.11)
Low and middle income 53.57 (14.41, 244.27) 1.08 (0.90, 1.30) 162.55 (42.62, 769.48) 218.27 (59.01, 1, 016.13) 0.83 (0.22, 3.86)

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

High income 0.15 (0.04, 0.71) 0.18 (0.15, 0.22) 122.84 (31.92, 589.28) 123.06 (32.03, 590.06) 0.25 (0.06, 1.20)
Upper-middle income 14.44 (3.81, 67.79) 0.95 (0.80, 1.16) 114.07 (29.74, 544.54) 130.07 (34.95, 614.10) 0.65 (0.17, 3.07)
Lower-middle income 35.19 (9.53, 158.64) 0.12 (0.10, 0.14) 44.73 (11.88, 207.87) 80.47 (21.94, 367.09) 1.36 (0.37, 6.20)
Low income 3.95 (1.07, 17.83) 0.01 (0.00, 0.01) 3.74 (1.01, 17.07) 7.73 (2.12, 34.95) 1.99 (0.54, 8.98)
Total 53.72 (14.45, 244.99) 1.26 (1.06, 1.52) 285.39 (74.55, 1, 358.75) 341.33 (91.04, 1, 606.19) 0.70 (0.19, 3.29)

(1) Default scenario (not in parentheses) based on discount rate on benefits of 3% and long-term GDP growth rate assumption of 3%. Figures in parentheses
are lower and higher bound estimates based on assumptions of 1.5% discount rate on benefits and 5% long-term GDP growth rate for the figure on the left and
5% discount rate on benefits and 1.5% long-term GDP growth rate for the figure on the right.

Figure 3 Total economic cost by region per year (US$ billion).

approach. The economic losses estimated for East Asia and the billion annually or 0.7% of global GNI (Table 4). The total econom-
Pacific, Latin America and the Caribbean, and Sub-Saharan Africa ic losses as a share of GNI are highest in Sub-Saharan Africa at
regions equal US$74.8, US$32.3, US$18.3 billion, respectively 2.58% followed by the Middle East and North Africa and Latin
(Figure 3 and Table 4). Over 56% of the share of the total global America and the Caribbean at 0.91% and 0.84%, respectively
economic losses is projected to be borne by LMICs (Table 4). The (Figure 4). In country income group, the total economic losses as a
combined total economic losses, including health system cost, mor- share of GNI are highest in low-income countries at 1.96% followed
tality losses and cognitive losses were estimated to be US$341.3 by lower middle-income countries at 1.35% of GNI (Table 4). The
414 Health Policy and Planning, 2019, Vol. 34, No. 6

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

Figure 4 Total economic cost by region per year as percentage of GNI.

estimates of economic losses for all individual countries with data because of their strategic importance for achieving the Sustainable
available can be found in Supplementary Appendix A6–8. Development Goals.

Sensitivity analysis
China
Estimating the economic losses far into the future comes with sig-
In China, the prevalence of exclusive breastfeeding in children below
nificant uncertainty. In the sensitivity analysis, the total global eco-
the age of 6 months is 21% (UNICEF, 2017). An estimated 16 146
nomic losses decreased to US$91.0 billion or 0.2% of GNI using
child deaths due to preventable diarrhoea and pneumonia were
conservative discounting and growth assumptions and increased to
attributed to not breastfeeding according to recommendations and
US$1.6 trillion or 3.29% of GNI with optimistic assumptions
22 537 deaths in women from cancers and type II diabetes each year
(Table 4). By calculating children’s cognitive losses with the alterna-
(Table 5 and Supplementary Appendix Tables A3 and A4). The esti-
tive approach, which assumes that breastfeeding at 6 months is asso-
mated health system cost of treatment of cases of childhood diar-
ciated with a 2.62 point IQ increase compared with not being
rhoea and pneumonia and mothers’ type II diabetes equals US$196
breastfed with data from 117 countries, the economic losses by per-
million per year (Supplementary Appendix Table A5). The economic
sons who were not breastfeeding according to recommendations
loss of mortality, from the number of preventable deaths, was esti-
were estimated to equal US$210.13 billion annually. The total eco-
mated to cost the Chinese economy approximately $6.3 billion each
nomic losses, which combine the health system costs, economic
year (Supplementary Appendix Table A6). However, with a rapidly
losses from mortality and cognitive losses, are therefore estimated to
declining child mortality rate in China, cognitive losses, estimated to
be US$257.9 billion annually or 0.37% of global GNI. For this al-
be US$59 billion (Supplementary Appendix Table A7), are the main
ternative approach, the total global economic losses decreased to
driver of the total economic losses. Therefore, the total economic
US$69.0 billion or 0.1% of GNI using conservative discounting and
cost of not breastfeeding according to recommendation was esti-
growth assumptions and increased to US$1.2 trillion or 1.75% of
mated to be US$66.1 billion, which translates to 0.61 of China’s
GNI with optimistic assumptions. Therefore, the potential future
GNI (Table 5 and Supplementary Appendix Table A8).
economic losses could vary widely, but regardless the economic con-
sequences in all scenarios remain substantial. The key drivers of
these economic losses estimated were by the number of children not
breastfed according to recommendations, the scale of mortality due India
to sub-optimal breastfeeding, economic growth and discount rates Despite a reported 55% prevalence of exclusive breastfeeding in
selected for the analysis. Sensitivity analysis results for regions and children below the age of 6 months (UNICEF, 2017), there were an
income groups are listed in Table 4 and Supplementary Appendix estimated 99 552 child deaths each year due to diarrhoea and pneu-
Tables A6–8 for individual countries with data. monia that could have been prevented with breastfeeding in India
(Table 5 and Supplementary Appendix Table A3). The key drivers
of this high mortality are India’s large population and high under-
Case study countries: China, India, Indonesia, Mexico five mortality rate. The estimated health system cost of treatment of
and Nigeria cases of childhood diarrhoea and pneumonia and women’s type II
This section highlights the country-level results from the Cost of diabetes equals US$106 million per year (Supplementary Appendix
Not Breastfeeding Tool from five large emerging economies, which Table A5). The high level of child mortality disproportionately
together contain 53% of the population of all developing countries, affects the total economic cost of not breastfeeding in India, which
Health Policy and Planning, 2019, Vol. 34, No. 6 415

Table 5 Total mortality and economic losses due to not breastfeeding for China, India, Indonesia, Mexico and Nigeria

Total human and economic costs attributable to not breastfeeding

Total child deaths Total maternal deaths Total cost (health, mortality and cognitive) (US$ b) Total cost as % of GNI

China 16 146 22 537 66.1 (17.6, 314.1) 0.61 (0.16, 2.89)

India 99 552 11 404 14.5 (3.8, 68.9) 0.69 (0.18, 3.29)
Indonesia 15 028 5170 9.4 (2.5, 44.5) 1.05 (0.28, 5.01)
Mexico 2360 5195 8.2 (2.2, 39.2) 0.67 (0.18, 3.18)
Nigeria 103 742 1511 21.1 (5.8, 92.7) 4.10 (1.12, 18.04)

was estimated to be US$14.5 billion, or 0.69% of GNI rates and high incomes, which compounded with income growth
(Supplementary Appendix Table A8). into higher economic losses in the future. This analysis showed that

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

the economic costs of mortality should not be forgotten in economic
evaluations, since the mortality losses may be as significant as cogni-
Indonesia
tive losses—particularly for countries with high mortality rates and
In Indonesia, the prevalence of exclusive breastfeeding in children
low breastfeeding rates. The human and economic costs of not
below the age of 6 months was at 42% (UNICEF, 2017). At this
breastfeeding highlight the continued importance of enabling a cul-
level, it was estimated that there were 15 028 child deaths and 5174
ture to better support and protect breastfeeding globally.
deaths in women each year linked to not breastfeeding according to
Overall the results of this study were aligned with that of previ-
recommendations (Table 5 and Supplementary Appendix Tables A3
ous studies (Rollins et al., 2016; Victora et al., 2016), and any differ-
and A4). The estimated health system cost of treatment of cases of
ences in results can be explained by changes in the model design and
childhood diarrhoea and pneumonia and women’s type II diabetes
the requirement for open access data for analysis by the tool. The
equals US$85 million per year (Supplementary Appendix Table A5).
sensitivity analysis of this study demonstrated that the economic
The economic cost of mortality was a key driver of total economic
losses could be even greater than estimated with potential acceler-
losses in Indonesia, which are estimated to be US$9.3 billion, or
ated economic growth in some countries. The new Cost of Not
1.05% of its GNI (Supplementary Appendix Table A8).
Breastfeeding Tool, however, provides policy-makers and other
stakeholders with access to an evidence-based, customizable analysis
Mexico that can be helpful for creating their investment cases and telling the
Approximately 31% of children below the age of 6 months in complex story of breastfeeding and nutrition.
Mexico were exclusively breastfed (UNICEF, 2017). It was esti-
mated that 2360 child deaths were due to preventable diarrhoea and
pneumonia and 5195 women’s deaths were from cancers and type II Policy implications
diabetes each year, all of which are attributed to not breastfeeding There are several policy implications of this study to consider. First,
according to recommendations (Table 5). The health system cost the interpretation of health system treatment costs must be informed
was estimated to be US$47 million per year (Supplementary by local context since the absolute figures on health system cost
Appendix Tables A5). Primarily driven by cognitive losses equal to since the lower health system costs in some settings may reflect lack
US$7.1 billion (Supplementary Appendix Tables A7), the total eco- of universal access to publicly funded maternal and child health
nomic losses in Mexico were estimated to be US$8.2 billion, or services rather than optimal breastfeeding practices. With improved
0.67% of its GNI (Supplementary Appendix Tables A8). breastfeeding rates, the health system costs could be converted into
health budget cost-savings and, in turn, redirected as allocations for
breastfeeding promotion interventions. Second, formula feeding is
Nigeria neither beneficial for improving child health and cognitive develop-
In Nigeria, the prevalence of exclusive breastfeeding in children ment nor affordable for the vast majority of families living in
below the age of 6 months was only 17% (UNICEF, 2017), which LMICs and poor households globally. Third, the concentration of
means that at least 5.4 million children each year do not get the full the global costs of not breastfeeding in China, India, Indonesia,
benefits of breastfeeding. Not breastfeeding according to recommen- Mexico and Nigeria, which together totalled 282 645 lives lost and
dation was estimated to lead to 103 742 child deaths each year up to US$119 billion in economic losses each year, is noteworthy.
(Table 5). The estimated health system cost of treating cases of These large emerging economies need to pay more attention to the
childhood diarrhoea and pneumonia and women’s type II diabetes growing double-burden of malnutrition since both the human and
was estimated to equal US$21.8 million per year (Supplementary economic costs of under- and over-nutrition were high. From a utili-
Appendix Tables A3). The combined economic losses of health sys- tarian perspective, this level of concentration could also motivate
tem cost, mortality losses, and cognitive losses are estimated to be donors to invest in a few strategic countries to better the chances of
and US$21.0 billion, or 4.1% of its GNI. achieving the Sustainable Development Goals or WHA Global
Nutrition Targets. However, donors and policy-makers need to bal-
ance efficiency with equity in the investment prioritization process.
Discussion It could be very feasible for governments of these countries and
The study found that the total global costs of not breastfeeding were other middle-income countries to finance the cost of breastfeeding
estimated at 694 322 lives lost annually and economic losses of promotion interventions from domestic sources. International donor
US$341.3 billion. Key drivers of the economic losses included low investment may be better focused on countries where the cost of not
exclusive and continued breastfeeding rates, high child mortality breastfeeding as a share of GNI was the highest.
416 Health Policy and Planning, 2019, Vol. 34, No. 6

Limitations 5. This alternative approach to calculating cognitive losses

While the tool provides certain advantages in improving access to assumes that the children who are breastfed until at least 6
data on the human and economic costs of not breastfeeding, there re- months receive the cognitive benefits of breastfeeding over chil-
main some gaps that could be improved upon in future versions of dren not breastfed to 6 months as described in the Lancet
the tool. Many of the current limitations, including the exclusion of Series on Breastfeeding (Rollins et al. 2016).
human and economic consequences from nearly 100 countries (most-
ly smaller countries) and additional economic costs related to care-
giver time, which are mainly borne by women (Leslie, 1989) as well Supplementary data
as household treatment fees, childhood diabetes, childhood cancers,
Supplementary data are available at Health Policy and Planning online
sudden infant death syndrome, necrotizing enterocolitis, preterm
births, post-partum haemorrhage and depression due to lack of data,
result in conservative estimates from this model. Updating the tool
Acknowledgements
with new data on the cost of care by age group and disease category
The authors would like to acknowledge Alive & Thrive, FHI 360 UNICEF,
and sensitivity analyses using the lower and upper bound relative
GMMB and their staff and consultants for their support to collect data,

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

risks of the effect of not breastfeeding on infectious illness and cogni-
understand the policy context for the study, online development of the tool
tive development would improve the accuracy of estimates. A more
and review of the manuscript.
systematic collection of formula cost across countries would also
be helpful. It should also be noted that the tool computes the total
economic costs of not breastfeeding at current prevalence levels Funding
compared with universal breastfeeding; however, it may be more
This research was supported by Alive & Thrive, an initiative funded by the
feasible in the future to compute the costs of not breastfeeding
Bill & Melinda Gates Foundation.
relative to country-specific targets. While the tool may provide
policy-makers with improved access to data on human and eco- Conflict of interest statement. None declared.
nomic costs of not breastfeeding, the tool does not replace the
value of detailed economic evaluations in individual countries,
which should supersede the results from this global-level analysis.
References
Bagriansky J. 2014. The Economic Consequences of Malnutrition in
Timor-Leste. Dili, Timor-Leste: UNICEF.
Conclusion Bagriansky J, Voladet S. 2013. The Economic Consequences of Malnutrition
The Cost of Not Breastfeeding Tool is an example how the big data in Lao PDR: A Damage Assessment Report. Vientiane, Lao PDR: UNICEF.
http://www.smeru.or.id/report/other/cpsp/Paper,%20Abstact,%20CV/0201_
revolution in global health can support stronger advocacy and meaning-
saykham%20voladet%20-%20paper.pdf, accessed 7 July 2017.
ful policy change. More research and development of analytical know-
Baker P, Smith J, Salmon L et al. 2016. Global trends and patterns in commer-
ledge translation tools may help educate policy-makers on the
cial milk-based formula consumption: is an unprecedented infant and young
economics benefits, to children, women and caregivers, households, child feeding transition underway? Public Health Nutrition 19: 2540–50.
governments and societies as a whole, related to breastfeeding and nu- Bartick M, Reinhold A. 2010. The burden of suboptimal breastfeeding in the
trition. The substantial human and economic costs of not breastfeeding United States: a pediatric cost analysis. Pediatrics 125: e1048–56.
in countries with low breastfeeding rates can to some extent be reversed Bartick MC, Schwarz EB, Green BD et al. 2017. Suboptimal breastfeeding in
with government, donor and civil society action to increase the financ- the United States: maternal and pediatric health outcomes and costs.
ing envelope available for the evidence-based high-impact breastfeeding Maternal & Child Nutrition 13: 1.
and nutrition interventions and policies. Investments made now will ac- Bartick MC, Stuebe AM, Schwarz EB et al. 2013. Cost analysis of maternal
disease associated with suboptimal breastfeeding. Obstetrics and
celerate progress towards the WHA Global Nutrition Targets and the
Gynecology 122: 111–9.
SDGs as well as boost human capital development in the long-run.
Bill and Melinda Gates Foundation. 2014. Methods for Economic Evaluation
Project: The Gates Reference Case—What It Is, Why It’s Important and
Notes How to Use It. https://www.nice.org.uk/Media/Default/About/what-we-do/
NICE-International/projects/Gates-Reference-case-what-it-is-how-to-use-it.
1. The Cost of Not Breastfeeding Tool (version 1) is now avail-
pdf, accessed 7 July 2017.
able to the public at the www.aliveandthrive.org/
Black R, Allen L, Bhutta Z et al. 2008. Maternal and child undernutrition: glo-
costofnotbreastfeeding. bal and regional exposures and health consequences. Lancet 371: 243–60.
2. All aggregated results are presented using the World Bank Büchner FL, Hoekstra, J, van Rossum CTM. 2007, Health gain and economic
Group’s classification systems of regions (i.e. East Asia and evaluation of breastfeeding policies. Bilthoven, Netherlands: RIVM.
Pacific, Europe and Central Asia, Middle East and North Chowdhury R, Sinha B, Sankar MJ et al. 2015. Breastfeeding and maternal
Africa, Latin America and Caribbean, North America, South health outcomes: a systematic review and meta-analysis. Acta Paediatrica
Asia and Sub-Saharan Africa) and income groups (i.e. high in- 104: 96–113.
come, upper-middle income, lower-middle income and low in- Council for Agriculture and Rural Development, UNICEF, World Food
come) (World Bank, 2017). Programme. 2013. The Economic Consequences of Malnutrition in
Cambodia: A Damage Assessment Report. Phnom Penh, Cambodia. http://
3. The main analysis of economic losses in this article were con-
www.wfp.org/sites/default/files/Report%20on%20Economic%20Consequ
ducted using the default assumptions of a long-term mean GDP
ences%20of%20Malnutrition%20in%20Cambodia.pdf, accessed 7 July
growth rate of 3% and discount rate on benefits of 3%. 2017.
4. Sensitivity is also presented using conservative assumptions of a Demographic and Health Surveys Program. 2017. STAT Compiler. http://
long-term GDP growth rate of 1.5% and a discount rate on bene- www.statcompiler.com/en/, accessed 7 July 2017.
fits of 5% as well as optimistic assumptions of a long-term GDP Drane D. 1997. Breastfeeding and formula feeding: a preliminary economic
growth rate of 5% and a discount rate on benefits of 1.5%. analysis. Breastfeeding Review 5: 7–16.
Health Policy and Planning, 2019, Vol. 34, No. 6 417

Eickmann SH, De Lira PIC, Lima MC et al. 2007. Breast feeding and mental Anemia, Breastfeeding, and Wasting. Directions in Development–Human
and motor development at 12 months in a low-income population in north- Development. Washington, DC: World Bank Group
east Brazil. Paediatric and Perinatal Epidemiology 21: 129–37. Smith JP, Thompson JF, Ellwood DA. 2002. Hospital system costs of artificial
Hanushek E, Woessmann L. 2008. The role of cognitive skills in economic de- infant feeding: estimates for the Australian Capital Territory. Australian
velopment. Journal of Economic Literature 46: 607–68. and New Zealand Journal of Public Health 26: 543–51.
Holla R, Iellamo A, Gupta A et al. 2015. Investing in breastfeeding–the World UNICEF. 2016a. From the First Hour of Life: Making the Case for Improved
Breastfeeding Costing Initiative. International Breastfeeding Journal 10: Infant and Young Child Feeding Everywhere. New York, NY: UNICEF.
1–12. UNICEF. 2016b. Multiple-Indicator Cluster Survey: Compiler. http://www.
Horta BL, Loret de Mola C, Victora CG. 2015. Breastfeeding and intelligence: micscompiler.org/ [2017, 07/07].
a systematic review and meta-analysis. Acta Paediatrica 104: 14–9. UNICEF. 2017. Infant and Young Child Feeding Database. https://data.uni
Horton S, Sanghvi T, Phillips M et al. 1996. Breastfeeding promotion and pri- cef.org/topic/nutrition/infant-and-young-child-feeding/, accessed 7 July 2017.
ority setting in health. Health Policy and Planning 11: 156–68. United Nations Department of Economic and Social Affairs. 2015. World
Institute for Health Metrics and Evaluation. 2017. Global burden of disease Population Prospects. http://esa.un.org/unpd/wpp/, accessed 1 November 2015.
results tool. http://ghdx.healthdata.org/gbd-results-tool, accessed 7 July 2017. Victora CG, Bahl R, Barros AJ et al. 2016. Breastfeeding in the 21st century:
International Agency for Research on Cancer. 2012. Estimated Cancer epidemiology, mechanisms, and lifelong effect. Lancet (London, England)
Incidence, Prevalence and Mortality Worldwide in 2012. http://globocan. 387: 475–90.

Downloaded from https://academic.oup.com/heapol/article/34/6/407/5522499 by guest on 23 April 2021

iarc.fr/Pages/online.aspx, accessed June 2015. Walters D, Eberwein J, Sullivan L, Shekar M. 2017. Reaching the global target
International Diabetes Federation. 2015. Diabetes Atlas 2015: Detailed Data. for breastfeeding. In: Shekar M, Kakietek J, Walters D, Dayton Eberwein J
http://www.diabetesatlas.org/resources/2015-atlas.html, accessed 7 July 2017. (eds.) An Investment Framework for Nutrition: Reaching the Global
Johns B, Adam T, Evans D. 2005. Choosing Interventions That Are Targets for Stunting, Anemia, Breastfeeding, and Wasting. Directions in
Cost-Effective (WHO-CHOICE): Country Specific Unit Costs. http://www. Development–Human Development. Directions in Development edn.
who.int/choice/country/country_specific/en/ 15 October 2015. Washington, DC: World Bank Group, 97–116.
Lamberti L, Fischer Walker C, Black R. 2012. Systematic review of diarrhea Walters D, Horton S, Siregar AY et al. 2016. The cost of not breastfeeding in
duration and severity in children and adults in low- and middle-income Southeast Asia. Health Policy and Planning 31: 1107–16.
countries. BMC Public Health 12: 276. WHO. 2015. Global Health Observatory Data. http://www.who.int/gho/en/,
Leslie J. 1989. Women’s time: a factor in the use of child survival technolo- accessed 2 June 2015.
gies?. Health Policy and Planning 4: 1–16. WHO. 2017. Guideline: Protecting, Promoting and Supporting Breastfeeding
Lubker M. 2007. Labour Shares. Geneva: International Labour Organization. in Facilities Providing Maternity and Newborn Services. Geneva. http://
Niederman MS, Krilov LR. 2013. Acute lower respiratory infections in devel- www.who.int/nutrition/publications/guidelines/breastfeeding-facilities-mater
oping countries. The Lancet 381: 1341–2. nity-newborn/en/, accessed 1 March 2018.
Pokhrel S, Quigley MA, Fox-Rushby J et al. 2014. Potential economic impacts WHO, UNICEF. 2014. Global Nutrition Targets 2025 Breastfeeding Policy
from improving breastfeeding rates in the UK. Archives of Disease in Brief. https://apps.who.int/iris/bitstream/handle/10665/149022/WHO_
Childhood 100: 334–40. NMH_NHD_14.7_eng.pdf?ua=1, accessed 7 July 2017.
Rollins NC, Bhandari N, Hajeebhoy N et al. 2016. Why invest, and what it World Bank. 2016. World Development Indicators Database. http://data.world
will take to improve breastfeeding practices?. Lancet (London, England) bank.org/data-catalog/world-development-indicators, accessed 1 March 2016.
387: 491–504. Vol. World Bank. 2017. World Bank Country and Lending Groups. https://data
Shekar M, Kakietek J, Walters D, Dayton Eberwein J. 2017. An Investment helpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-coun
Framework for Nutrition: Reaching the Global Targets for Stunting, try-and-lending-groups, accessed 7 July 2017.