European Journal of Clinical Nutrition (2012) 66, 914–919

& 2012 Macmillan Publishers Limited All rights reserved 0954-3007/12

www.nature.com/ejcn

ORIGINAL ARTICLE
Influences on infant feeding decisions of first-time mothers in
five European countries
H Gage1, P Williams2, J Von Rosen-Von Hoewel3, K Laitinen4, V Jakobik5, E Martin-Bautista5, M Schmid1, B Egan1, J Morgan1, T Decsi5,
C Campoy6, B Koletzko3 and M Raats1

BACKGROUND/OBJECTIVES: Infant feeding decisions made by new parents have significant health implications. The study aimed
to investigate: influences on infant feeding decisions; characteristics of mothers reporting reliance on alternative information
sources; associations between reliance on different sources and intentions to exclusively breastfeed and introduce complementary
foods later; and subsequent breastfeeding and weaning behaviours.
SUBJECTS/METHODS: First-time mothers in five European countries (England, Finland, Germany, Hungary, Spain) completed
questionnaires about the importance of 17 influences on infant feeding choices at birth and 8 months later, during 2007–2008. Use
of individual sources and reliance on four categories (family and friends, health professionals, written materials, audio-visual media)
were compared between countries. Associations between information sources used and mother characteristics, feeding intentions
and behaviours were investigated using appropriate statistical tests.
RESULTS: In all, 2071 first-time mothers provided baseline data; 78% at 8 months. Variation exists between countries in the
influence of different sources on feeding decisions of first-time mothers. Across all countries, the most important influences at both
time points were books, partner and antenatal midwife. Mothers in higher income quintiles and remaining
at school beyond age 16 years reported greater reliance on written sources (Po0.0005). Mothers relying most on written sources
reported longer exclusive breastfeeding (P ¼ 0.002), and a tendency to introduce foods other than milk later (P ¼ 0.079) than
mothers relying most on personal or professional contacts.
CONCLUSION: Further research is required about which dissemination strategies are most effective at improving infant feeding
behaviours in varied cultural settings, and for different socio-economic groups.

European Journal of Clinical Nutrition (2012) 66, 914–919; doi:10.1038/ejcn.2012.56; published online 13 June 2012
Keywords: infant feeding; information; longitudinal survey; mothers; European countries

INTRODUCTION and engaging with audio-visual media, such as television and

The infant feeding decisions made by new parents, particularly the
choice between breast and formula milk and the age of intro-
ducing complementary foods, have significant health implications.
Breastfeeding is associated with short- and longer-term health
benefits for infant and mother, and is the ‘gold standard’.1,2
Following an extensive review of evidence, the World Health
Organization issued a global recommendation that babies should
be exclusively breastfed for 6 months,3 although debate is ongoing
as to whether this policy is optimal for all infants.4–6 Studies have
found low rates of breastfeeding and poor weaning practices in
Europe.7–9 The lowest rates (o5% of mothers) of 6 months of
exclusive breastfeeding are reported in Finland, UK and Italy, and
the highest rates (25–35%) in Denmark, Sweden and Hungary.9
A better understanding of factors that influence infant feeding
choices is needed to guide the design of behaviour-change policies.
Women may receive information about infant feeding during
pregnancy from many sources, including formal services (health-
care providers, antenatal classes), informally (family members and
friends), through consulting books and other written materials,
DVDs.10 They are routinely exposed to public health messages,
advertisements from infant product manufacturers and the
opinions of experts and lay people through a variety of routes,
including the internet. Infant feeding decisions are further shaped
by social and cultural norms, such as attitudes to breastfeeding
in public,11 and maternity leave legislation. Public health policy
makers need to understand the extent to which decisions of
parents from different cultures and backgrounds are affected by
different influences. The reliance that parents place on different
sources of information changes over time,12,13 and studies are
needed to explore this variability further.
This paper reports findings from a study undertaken in five
geographically dispersed European countries, with varied socio-
economic structures and health economies (England, Finland,
Germany, Hungary and Spain). Using a longitudinal design, the
study aims were to investigate: the sources of information that
influence infant feeding decisions of first-time mothers at the time
of the birth of their baby, and over the following 8 months; the
characteristics of mothers who reported reliance on alternative

1
Food, Consumer Behaviour and Health Research Centre, University of Surrey, Guildford, Surrey, UK; 2Department of Mathematics, University of Surrey, Guildford, Surrey, UK;
3
Dr von Hauner Children’s Hospital, Ludwig-Maximilians University, Munich, Germany; 4Functional Foods Forum and Institute of Biomedicine, University of Turku, Turku, Finland;
5
Department of Paediatrics, University of Pécs, Pécs, Hungary and 6Department of Paediatrics, University of Granada, Granada, Spain. Correspondence: Dr H Gage, Department of
Economics, University of Surrey, Staghill, Guildford, Surrey GU2 7XH, UK.
E-mail: h.gage@surrey.ac.uk
Contributors: HG contributed to the analysis and wrote the first draft; JRH, KL, VJ, EMB, MS collected data and contributed to analysis; JM, MR contributed to the analysis;
MR conceived the study; PW undertook the statistical analysis; all authors contributed to the design and read and approved the final manuscript.
Received 9 March 2012; revised 23 April 2012; accepted 23 April 2012; published online 13 June 2012

categories of information, and of those who changed their
primary source after the birth of their baby; and associations
between the influence of different sources and infant feeding
intentions (at birth), and subsequent behaviours.
MATERIALS AND METHODS
Data on sources of information on infant feeding were collected using
questionnaires distributed to first-time mothers in the five countries shortly
after birth, and 8 months later. The study focused on first-time mothers
because they were likely to have recently collected information on feeding
methods, and considered the options, without being influenced by their
own prior experiences, or the additional pressures of having other young
children to care for. The questionnaires were produced and piloted in
England, and then translated into local languages with minor cultural
adaptations to accommodate national differences. Ethical approval was
obtained in each country.
Mothers were recruited through maternity hospitals in the locality of
researchers, each of which served diverse socio-economic populations. All
mothers with healthy singleton babies in postnatal wards who were able
to self-complete a questionnaire in the local language were eligible. Items
on the questionnaire asked mothers about their: baby (gestation, birth
weight, gender); milk feeding intentions (exclusive breast, exclusive
formula, combination); intended timing (months) of the introduction of
complementary foods; background (age, country of birth, language spoken
at home, family income (in quintiles), living arrangements (with partner,
others or alone), education, health-related behaviours (smoking, diet,
exercise), employment, health); and sources of information about infant
feeding used.
The question on infant feeding information asked mothers to score the
extent to which each of 17 possible sources influenced their decisions
about how to feed their baby (0 ¼ not at all; 1 ¼ slightly; 2 ¼ moderately;
3 ¼ very much; 4 ¼ extremely). Respondents could also state ‘not applic-
able’, in which case the source was disregarded in the analysis. The 17
sources were subsequently reduced to four categories: (i) family and
friends; (ii) health professionals (doctor—specialist or general/family
practitioner, antenatal midwife, staff in hospital, other professional, private
antenatal class); (iii) written/reading materials (books, leaflets, magazines,
internet); and (iv) audio-visual media (advertisements, TV programmes,
radio programmes, videos/DVDs). Definitions of these categories were
informed by the social–ecological framework, which identifies five levels
of influence on behaviours:14 intrapersonal (knowledge, skills, attitudes,
values, preferences, emotions of mother and child); interpersonal (views of
family and friends); community (activities of local organizations and
services); organizational (actions of businesses and public agencies); and
public policy (regulations, taxes, subsidies, characteristics of health- and
social care systems). Moreover, the categories provide a distinction
between passive receipt of information (from personal or professional
contacts) and active information-seeking in written materials and the
internet.
Recruitment of mothers to the study began in April 2007, using a
convenience approach, and continued in each country until 400 mothers
had returned a completed questionnaire. Sample size calculations showed
that 392 mothers per country were required to identify, using a two-sided
test, with size (5%) and power (80%), an underlying difference of 10%
between any two countries with regard to any specific prevalence
Influences on infant feeding decisions
H Gage et al
915
(for example, the proportion of mothers who stated their infant feeding
decision is influenced extremely (vs sum of very much, moderately, slightly,
not at all) by any particular source).
All mothers who provided data at baseline were mailed a follow-up
questionnaire when their baby was 8 months old. This asked them to
self-report their milk feeding (including initial number of weeks spent
exclusively breast feeding) and weaning behaviours (age in months of
introducing foods other than milk), and repeated the baseline question on
sources of information on infant feeding.
Responses were entered into SPSS version 15 (SPSS Inc., Chicago, IL,
USA). The characteristics of respondents and reported use of information
sources at baseline and 8-month follow-up were analysed descriptively for
each country. The mean score for each respondent for each of the four
main categories of information (family and friends, health professionals,
written, audio-visual) was calculated at each time point. The primary
source of information used by each respondent at birth and 8 months was
identified as the category with the highest mean score. Some mothers had
no primary information category because their highest mean score was
tied between two or more categories. One-way analysis of variance and w2
tests, as appropriate, were used to explore: the characteristics of mothers
reporting reliance on a particular information category at baseline; the
characteristics of the mothers who switched their primary information
category between baseline and 8-month follow-up; and associations
between reported use of particular categories of information and infant
feeding intentions (baseline) and behaviours (follow-up).
RESULTS
Characteristics of the sample
A total of 2071 first-time mothers completed baseline question-
naires; 1619 (78.2%) returned 8-month follow-up questionnaires.
Compared to mothers in other countries, those in Spain reported
fewer years of education, lower income and higher rates of
smoking during pregnancy. Spanish mothers were also less likely
to respond to the follow-up questionnaire (Table 1). Across all
countries, mothers retained at follow-up (vs not) were: older; more
likely to be in the highest income quintile, educated beyond age
16 years and in full-time employment before pregnancy; and less
likely to have smoked prior and during pregnancy. There was no
significant difference between responders and drop-outs at 8
months with respect to stated intentions to breastfeed exclusively
before weaning (data not shown).
Information sources
Across all countries, books, partner and antenatal midwife are the
sources ranked highest in terms of importance to mothers at
both time points, and audio-visual media are least important,
that is, ranked lowest (Table 2). The importance of some sources
(for example, doctor) changed markedly between baseline and
follow-up. With the exception of ‘staff in hospital’ and ‘private
antenatal class’, the proportions of mothers reporting that sources
influenced their infant feeding decisions moderately, very much or
extremely (vs slightly or not at all) are higher at 8-month follow-up

Table 1. Characteristics of mothers

Characteristic All England Finland Germany Hungary Spain

Number of mothers recruited at baselinea 2071 438 426 414 389 404
Age—mean (s.d.) (years) 30.0 (5.0) 31.7 (5.0) 28.3 (4.6) 32.1 (5.0) 28.1 (4.0) 29.6 (5.1)
% with highest level of education to age 16 years or below 13.7 6.0 5.6 5.3 11.8 40.8
% in full-time paid employment before pregnancy 75.7 89.4 68.1 78.0 86.9 55.7
% reporting that they smoked during pregnancy 7.1 4.4 9.4 6.1 1.8 13.9
% in lowest quintile annual household income before tax 10.0 4.4 10.9 6.2 1.0 28.3
% in highest quintile annual household income before tax 22.1 58.5 19.6 20.4 4.9 3.3
% of recruited mothers responding to 8-month follow-up questionnaire 78.2 80.8 91.1 86.2 85.2 47.3
a
Recruitment and baseline data collection in Finland, Germany, Hungary and Spain was undertaken by research nurses who approached mothers during their
postnatal hospital stay, as convenient. The proportions of eligible mothers recruited were 70.3%, 56.7%, 24.7% and 16.8%, respectively. In England,
questionnaires were distributed to eligible mothers in their hospital discharge packs, and returned to the research team by mail (response rate: 9.4%).

& 2012 Macmillan Publishers Limited European Journal of Clinical Nutrition (2012) 914 – 919
 Influences on infant feeding decisions
H Gage et al
916
Table 2.Proportions of mothers reporting source as a moderate, very or extremely (vs slightly or not at all) important influence on their decision
about how to feed their baby at birth and 8 months later

Rank At birth of baby At 8-month follow-up

Source All England Finland Germany Hungary Spain Source All England Finland Germany Hungary Spain
N 2071 438 426 414 389 404 N 1619 354 388 357 329 191

1 Antenatal 55.2 54.7 63.1 61.0 43.8 57.4 Partner 70.7 80.2 62.7 78.2 63.9 76.2
midwife
2 Books 52.6 58.8 54.3 74.7 47.0 27.4 Books 69.8 77.7 42.2 82.8 81.3 68.0
3 Partner 47.7 64.1 48.6 53.7 20.8 49.2 Parents 59.6 57.3 50.3 55.6 66.3 80.5
4 Staff in 47.5 50.3 61.1 44.9 34.7 46.6 Antenatal 58.6 37.9 49.9 65.2 85.0 59.0
hospital midwife
5 Leaflets 46.5 46.1 69.5 63.3 24.7 25.5 Doctor 56.7 24.1 41.0 71.3 77.4 86.9
6 Magazines 39.6 40.0 47.6 50.4 28.8 33.1 Leaflets 50.6 38.1 53.3 74.2 28.8 51.4
7 Friends 35.8 15.5 42.0 20.2 16.4 45.3 Magazines 50.0 30.0 27.6 67.2 31.3 53.8
8 Parents 34.2 30.5 33.8 36.1 22.4 48.8 Other health 45.4 64.4 48.3 25.2 41.0 51.3
professionals
9 Private ante 33.8 39.7 14.8 55.1 16.8 41.6 Internet 43.9 39.9 40.4 58.3 50.1 38.9
natal class
10 Other health 33.3 23.8 42.8 22.8 NA 42.3 Friends 43.7 53.0 36.5 52.7 34.7 40.5
professionals
11 Internet 31.9 26.9 41.7 29.1 41.8 20.3 Staff in 38.5 33.3 32.7 38.1 41.9 55.4
hospital
12 Other 29.9 18.0 14.2 19.0 18.0 40.2 Other 33.0 24.9 29.8 28.7 27.9 52.6
relatives relatives
13 Doctor 28.0 15.5 42.0 20.6 16.4 45.3 Advertising 29.0 26.2 22.3 18.2 51.0 29.9
14 Advertising 20.2 25.9 16.4 12.4 25.2 21.1 TV 23.5 23.6 15.6 13.1 40.5 28.9
programmes
15 TV 17.8 22.2 18.5 12.2 23.4 20.4 Private ante 18.7 29.1 4.7 37.3 30.0 58.8
programmes natal class
16 Video/DVD 10.8 10.8 6.9 4.4 9.8 22.3 Radio 14.0 10.5 6.2 9.0 29.3 17.8
programmes
17 Radio 10.4 9.4 4.0 4.4 17.7 17.1 Video/DVD 11.9 7.7 3.5 4.5 19.9 29.5
programmes
Abbreviation: NA, not applicable.

Table 3. Primary categories of information at baseline: N (%)

Main information source All England Finland Germany Hungary Spain

Family and friends 402 (19.5) 107 (24.7) 73 (17.1) 101 (24.5) 34 (8.7) 87 (21.5)
Health professionals 573 (27.8) 125 (28.9) 125 (34.7) 101 (24.7) 34 (14.4) 143 (35.4)
Written 664 (32.2) 153 (35.3) 170 (39.9) 170 (41.3) 125 (32.1) 46 (11.4)
Tie (2 or 3 way) 425 (20.6) 48 (11.1) 35 (8.2) 40 (9.7) 174 (44.7) 128 (31.7)
Total 2064* (100) 433 (100) 426 (100) 412 (100) 389 (100) 404 (100)
*Seven mothers provided no data on information sources.

than at birth. Large international differences in the importance of
some information sources are evident at both time points.
Examination of the mean score for each of the four main
categories of sources at baseline revealed only 19 mothers
reporting audio-visual media as their primary category of
information. Subsequent analysis therefore focused on the other
three categories (family and friends, health professionals, written),
with each mother assigned according to their mean score for
these three categories. Of the 2064 mothers, 1639 (79.4%)
reported reliance on one specific category of information: the
rest having joint reliance on more than one (Table 3). The
proportions of mothers in Hungary and Spain reporting ties were
higher than England, Finland and Germany. Differences existed
between countries in the proportions of mothers primarily relying
on the three categories (w2, Po0.0005); health professionals were
most important in Spain; and written materials in the other four
countries (Table 3).
Exploration of the characteristics of the 1639 mothers
influenced predominantly by one of the three categories (family
and friends, health professionals, written) at baseline revealed
significant differences (Table 4). Primary reliance on written
materials was associated with: being in the highest three income
quintiles; reporting education beyond age 16 years; reporting
more physical activity before pregnancy; and not smoking prior or
during pregnancy. No significant difference in reliance on the
three categories was found with respect to mother’s age, living
alone (vs not alone), employment status before pregnancy (in paid
employment vs not), whether born in country of residence (vs
not), whether speak language of country of residence at home (vs
not), self-reported extremely or very healthy eating before
pregnancy (vs moderately, slightly, not at all) and having a health
or nutrition qualification (vs not).
Changes in primary information source
Changes in information sources between birth and 8-month
follow-up were analysed for 1610 mothers for whom data on
information sources were available at both time points. The

European Journal of Clinical Nutrition (2012) 914 – 919 & 2012 Macmillan Publishers Limited
 Influences on infant feeding decisions
H Gage et al
917
Table 4. Characteristics of the 1639 mothers influenced predominantly proportion of mothers reporting a single most influential category
by one of the three main categories of information at baseline: N (%) was higher at follow-up than baseline (93.7% vs 81.2%). Reliance
on family and friends increased (by over 10%) in the 8 months
Characteristic Main information category after birth, while the overall importance of health professionals
and written sources remained relatively constant. However, there
Family Health Written was considerable switching between categories, with 1034
and professionals (64.2%) of mothers reporting a different principal influence at
friends follow-up compared with baseline (Table 5). Switching was
particularly high in Hungary, and concentrated among less
Annual household pre tax income (quintiles) educated (no education beyond age 16 years), and younger
Lowest two 122 (30.9) 190 (33.9) 139 (21.4) (age o30 years) mothers. Across all countries, mothers less likely
Highest three 273 (69.1) 370 (66.1) 512 (78.6)
Significant difference w2, Po0.0005
to switch: were from Spain; reported smoking prior and during
pregnancy; and followed a special diet (vegan, vegetarian,
Education religious). Having a qualification in health or nutrition had no
Up to age 16 years 64 (15.9) 78 (13.7) 51 (7.7) significant influence on switching (data not shown).
Age 16 years and over 338 (84.1) 493 (86.3) 613 (92.3)
Significant difference w2, Po0.0005
Information sources and infant feeding intentions and behaviours
Physical activity before pregnancy Higher proportions of mothers relying mainly on written sources
Extremely, very much 110 (27.5) 164 (28.7) 236 (35.6) (vs family and friends, health professionals) reported intentions to
Moderate, slightly, not at all 290 (72.5) 407 (71.3) 427 (64.4) exclusively breastfeed (vs not), and to introduce solid foods later
Significant difference w2, P ¼ 0.006 (at baseline), and reported (at 8 months) longer exclusive
breastfeeding and later weaning behaviours (Table 6).
Smoke before pregnancy
No 276 (68.8) 411 (72.0) 503 (75.9)
Yes 125 (31.2) 160 (28.0) 160 (24.1)
Significant difference w2, P ¼ 0.038 DISCUSSION
First-time mothers in five European countries reported (from a list
Smoke during pregnancy of 17 different sources) that books, partners and health profes-
No 356 (88.6) 523 (91.6) 636 (96.1) sionals most influenced their infant feeding decisions. Of slightly
Yes 46 (11.4) 48 (8.4) 26 (3.9)
2 less importance were leaflets, magazines, parents and the internet.
Significant difference w , Po0.0005
The least important sources were advertising, TV, DVD and radio

Table 5. Changes in primary category of information between baseline and 8-month follow-up: N (row %) (column %)

Primary information category: 8 months Family and friends Health professionals Written Tie (2 or 3 way) Total

Primary information category: at baseline

Family and friends 138 (44.1) (28.0) 70 (22.4) (15.5) 90 (28.8) (15.9) 15 (4.8) (14.9) 313 (100) (19.4)
Health professionals 125 (28.5) (25.4) 163 (37.1) (36.1) 122 (27.8) (21.6) 29 (6.6) (28.7) 439 (100) (27.3)
Written 142 (25.6) (28.9) 118 (21.3) (26.2) 265 (47.7) (46.8) 30 (5.4) (29.7) 555 (100) (34.5)
Tie (2 or 3 way) 87 (28.7) (17.7) 100 (33.0) (22.2) 89 (29.4) (15.7) 27 (8.9) (26.7) 303 (100) (18.8)
Total 492 (30.6) (100) 451 (28.0) (100) 566 (35.2) (100) 101 (6.3) (100) 1610* (100) (100)
*Missing data on information sources for 5 mothers at baseline, 3 at follow-up, 1 at both time points.

Table 6. Comparison of infant feeding intentions (at baseline) and reported behaviours (8-month follow-up) by primary information category

Primary information source Baseline: Intend to exclusively breastfeed Baseline: After how many months do you intend to
introduce foods other than milk to your baby?

N Yes, n (%) 95% Confidence interval N Mean 95% Confidence interval

Family and friends 400 316 (79.0) 75–83.0 338 5.46 5.32–5.59
Health professionals 568 423 (74.5) 70.9–78.1 491 5.58 5.46–5.69
Written 662 568 (85.8) 83.1–88.5 631 5.72 5.63–5.81

Significance w2, Po0.0005 ANOVA, P ¼ 0.006

Primary information source Follow-up: How long (months) Follow-up: Age (months) of
baby was exclusively breast fed introducing foods other than milk

N Mean 95% Confidence interval N Mean 95% Confidence interval

Family and friends 451 3.67 3.47–3.88 489 4.81 4.71–4.91

Health professionals 403 3.62 3.41–3.82 449 4.83 4.72–4.93
Written 520 4.06 3.88–4.25 562 4.96 4.86–5.06
Significance ANOVA, P ¼ 0.002 ANOVA, P ¼ 0.079
Abbreviation: ANOVA, analysis of variance.

& 2012 Macmillan Publishers Limited European Journal of Clinical Nutrition (2012) 914 – 919
 Influences on infant feeding decisions
H Gage et al
918
media. The influence of most sources was greater at 8 months
than at birth, but the information-gathering patterns of almost
two-thirds of mothers changed in the postnatal period. Mothers
were more likely to report reliance on one primary information
category (family and friends, health professionals, written materials)
at 8-month follow-up than at baseline. While written materials were
the primary source for the largest proportion of mothers at both
time points, family and friends gained importance over the first
8 months after birth.
These findings are consistent with previous research showing
that mothers access multiple sources of information, the relative
importance of which vary over time,12,13 and that cultural factors
affect the way in which mothers access and use information.15
In this study, mothers in Spain were more reliant at birth on family
and friends than the mothers in the other countries. Also, lower
proportions of mothers in England and Finland reported (at 8
months) that doctors influenced their infant feeding decisions
than in Germany, Hungary and Spain, reflecting the different role
that physicians play in postnatal care across the countries.
Although other research found no difference in sources used by
women who planned different milk feeding regimens for their
infants (breast vs formula vs combination),10 an association was
found in this study between predominant reliance on written
sources (books, leaflets, magazines and internet) and reporting
infant feeding intentions and behaviours in line with recent
international recommendations, that is, exclusive breastfeeding
and later introduction of foods other than milk.1–3 Use of written
sources was associated with more years in formal education and
higher income, and it is thus likely that the findings are reflecting
the previously well-observed association between years in educa-
tion and socio-economic status, and breastfeeding initiation and
duration, and later weaning.16–23 Use of written sources involves
active searching and reading, which may come more naturally
to mothers who have experienced more years of education.
By contrast, less educated mothers may tend to be more reliant
on passive methods, involving formal (health professionals) or
informal (family and friends) interchanges, as observed among
mothers from Spain.
The proportion of mothers relying on health professionals as
the primary source of information held steady over the observa-
tion period (28%), although international differences were
observed (only 14% of mothers in Hungary indicated health
professionals were their primary influence at baseline). To the
extent that health professionals provide evidence-based informa-
tion, and their advice carries credibility and authority with the
public, there is scope to increase the role they play as educators of
mothers.24 Attitudes of health professionals to breastfeeding can,
however, vary.25–31 Some may not always provide optimal advice
due to gaps in knowledge,22,32–34 or poor communication skills,35
and these issues need to be addressed if the potential of formal
sources is to be fully exploited. Multi-faceted approaches
incorporating written materials and face-to-face advice have
been found effective,25,29,30 and health professionals can
reinforce their verbal interactions through distribution of
information leaflets. Written materials designed by public health
agencies or professional organizations, and circulated by health
professionals, provide consumers with easy access to evidence-
based messages, which they can retain for reference purposes,
with reduced search costs and avoidance of quality control
concerns that accompany use of some internet sites.36,37
The finding that informal (family and friends) sources gained
importance as influences on decision-making between birth and
8-month follow-up confirms other research output, which shows
that mothers rate social networks and family members more
highly than health professionals as sources of support in making
infant feeding choices.38 It also points to the importance of
involving significant others in pre- and postnatal education
initiatives.10–12 Targeted interventions aimed at improving the
European Journal of Clinical Nutrition (2012) 914 – 919
knowledge and understanding of partners and other family
members have been previously shown to be instrumental in
improving breastfeeding rates.39–41 Our analysis indicated that
younger mothers, and those with fewer years of education, were
more likely to switch their primary information source over the
first 8 months of their baby’s life, so local strategies aimed at these
groups may be the most worthwhile.
Public health policy makers are turning increasingly to the
ecological framework for an understanding of the factors that
influence individuals’ health-related behaviours.42 The value of
this approach, which emphasizes the ways in which health
behaviours are shaped by the characteristics of individuals and the
broader family and community contexts in which they live,14 is
confirmed by the findings of this study. Information-seeking
behaviours around infant feeding decision-making of the first-
time mothers in our sample varied culturally and by socio-
demographic group. Information campaigns aimed at influencing
attitudes, and ultimately behaviours, and which are generally
perceived to have a weak impact,43 may be more effective if
grounded in ecologically based evidence.
This survey of mothers provides a longitudinal analysis of how
influences on first-time mothers’ decisions about infant feeding
decisions change in the early months after birth. It was conducted
rigorously and in accordance with procedures agreed through
regular contact (including face-to-face meetings) between partners in
each country. However, the study is limited in several ways. Although
the five countries were selected to provide geographical spread
across Europe, they may not be representative of all social, political
and health-care systems, so the findings may not be generalizable.
The response rate at 8 months (78% overall) was reduced because
less than half of mothers recruited in Spain returned their follow-up
questionnaires. Mothers not completing the study differed from the
whole sample (younger, lower income, less education, more likely to
have smoked). Moreover, differences exist in the socio-economic mix
of the samples across the countries, particularly the higher
proportion of mothers from lower income groups in Spain. This
affects the comparisons of the use of information sources, and may
account for higher reliance on the use of family and friends in Spain
than in the other countries.
Although the four categories were established to reflect
different information gathering methods (talking with formal or
informal sources, search-intensive methods associated with
written materials and audio-visual media) they may not have
been the most appropriate. The variable constructed to represent
the most important category of information for mothers was
based on a mean score of the items in the group, and respondents
with no single source of information were excluded from
subsequent analysis. Mothers in Hungary and Spain were less
precise in reporting use of information sources and more ties
between categories were recorded than in the other countries.
The sample was drawn from first-time mothers, so no evidence is
provided about sources mothers use when rearing subsequent
children, or how prior infant feeding experiences influence their
decisions. Finally, the study focused on what mothers reported
had influenced them, rather than providing evidence of effective-
ness of different approaches.
Complex household, community and social factors interact to
determine infant feeding behaviours,44,45 and the information
environment facing new mothers plays a strategic role in the
decision-making process. Timely, consistent and evidence-based
information, tailored to the individual, and delivered in a variety of
formats, from trustworthy sources46 is needed to improve
breastfeeding rates and weaning practices. This study found
marked variation within and between countries in the importance
that first-time mothers attribute to different influences, and
instability in their reliance on sources. Further research is
required to identify which dissemination strategies are most
effective at improving infant feeding behaviours at different
& 2012 Macmillan Publishers Limited

stages, in varied cultural settings, and for each socio-economic
and ethnic group, including the effect of health professional
endorsements, and the growing importance of the internet.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
ACKNOWLEDGEMENTS
We are grateful to the maternity hospitals that enabled us to recruit the mothers in
this study, and to the mothers who completed the questionnaires. This study was
supported as a part of the European Project ‘Early Nutrition Programming-EARNEST’
within the 6th Framework Programme. No. FOOD-CT-2005-007036.
REFERENCES
1 American Academy of Pediatrics. Policy statement: breastfeeding and the use of
human milk. Pediatrics 2005; 115: 496–506.
2 World Health Organisation. WHO Action Plan for Food and Nutrition
Policy, 2007-2012. WHO Regional Office for Europe: Denmark. www.euro.who.int
(accessed 01 June 2012).
3 World Health Organisation. Global Strategy of Infant and Young Child Feeding.
The Optimal Duration Of Exclusive Breast Feeding. WHO: Geneva, 2001.
4 Fewtrell MS, Morgan JB, Dugan C, Gunnlaugsson G, Hibberd PL, Lucas A et al.
Optimal duration of exclusive breastfeeding: what is the evidence to support
current recommendations? Am J Clin Nutr 2007; 85: 635S–638S.
5 ESPGHAN Committee on Nutrition. Complementary feeding: a commentary by
ESPGHAN Committee on Nutrition. J Pediatr Gastroenterol Nutr 2008; 46: 99–110.
6 Cattaneo A, Yngve A, Koletzko B, Guzman LR. Protection, promotion and support
of breastfeeding in Europe: current situation. Public Health Nutr 2005; 8: 39–46.
7 Nichol A, Thayaparan B, Newell ML, Rundall P. Breastfeeding policy, promotion
and practice in Europe. Results of a survey of non-governmental organisations.
J Nutr Environ Med 2002; 12: 255–264.
8 World Health Organisation Europe. Comparative analysis of food and nutrition
policies in WHO European member states. Report. WHO: Geneva, 2003.
9 Cattaneo A, Williams C, Pallas-Alonso CR, Hernandez-Aguilar MT, Lasarte-Velillas JJ,
Landa-Rivera L et al. ESPGHAN’s 2008 recommendation for early introduction of
complementary foods: how good is the evidence? Matern Child Nutr 2011; 7: 335–341.
10 Chezem JC, Friesen C, Clark H. Sources of infant feeding information used by
pregnant women. J Perinat Educ 2001; 10: 20–26.
11 Kong SF, Lee DT. Factors influencing decision to breastfeed. J Adv Nurs 2004; 46:
369–379.
12 Carruth BR, Skinner JD. Mothers’ sources of information about feeding their
children aged 2 months to 5 years. J Nutr Educ Behav 2001; 33: 143–147.
13 Anker AE, Reinhart AM, Feeley TH. Health information seeking: a review of
measures and methods. Patient Educ Counsel 2011; 82: 346–354.
14 McLeroy K, Bibeau D, Stckler A, Glanz K. An ecological perspective on health
promotion programs. Health Educ Q 1988; 15: 351–377.
15 Synott K, Bogut J, Edwards GA, Scott JA, Higgins S, Norin E et al. Parental per-
ceptions of feeding practices in five European countries: an explorative study.
Eur J Clin Nutr 2007; 61: 946–956.
16 Daly A, MacDonald A, Booth IW. Diet and disadvantage: observations on infant
feeding from an inner city. J Hum Nutr Diet 1998; 11: 381–9.
17 North K, Emmett P, The ALSPEC Study Team. Types of drinks consumed by infants
at 4 months and 8 months of age: socio-demographic variables. J Hum Nutr Diet
2000; 13: 71–82.
18 Pain R, Bailey C, Mow IG. Infant feeding in North East England: contested spaces
and reproduction. Area 2001; 33: 261–272.
19 Fewtrell M, Lucas A, Morgan J. Factors associated with weaning in full term and
preterm infants. Arch Disease Childhood Fetal Neonat Educ 2003; 88: F296–F301.
20 Haslam C, Laurence W, Haefeli K. Intentions to breastfeed and other important health
related behaviours and beliefs during pregnancy. Fam Pract 2003; 20: 528–530.
& 2012 Macmillan Publishers Limited
Influences on infant feeding decisions
H Gage et al
919
21 Altree P. Low income mothers nutrition and health: a systematic review. Matern
Child Nutr 2005; 1: 227–240.
22 Stolzer J, Zeece P. Low income women and physician breast feeding advice: a
regional assessment. Health Educ J 2006; 62: 126–134.
23 Bolling K, Grant C, Hamlyn B. Infant feeding survey 2005. NHS, Information Centre
for Health and Social Care 2007; Internet http://www.ic.nhs.uk/statistics-and-data-
collections/health-and-lifestyles-related-surveys/infant-feeding-survey/infant-
feeding-survey-2005 (accessed 28 June 2010).
24 Beake S, Pellowe C, Dykes F, Schmied V, Bick D. A systematic review of structured
compared with non-structured breastfeeding programmes to support the initia-
tion and duration of exclusive and any breastfeeding in acute and primary health
care settings. Matern Child Nutr 2011; 8: 141–161.
25 Fairbank L, O’Meara S, Renfrew MJ, Woolridge M, Sowden AJ, Lister-Sharpe D.
A systematic review of the effectiveness of interventions to promote the initiation
of breastfeeding. Health Technol Assess 2000; 4: 1–171.
26 Hamlyn B, Brooker S, Oleinikova K, Wands S. Infant Feeding 2000. The Stationery
Office: London, 2002.
27 DiGirolamo AM, Grummer-Strauwn LM, Fein SB. Maternity care practices—
implications for breastfeeding. Birth 2001; 28: 94–100.
28 DiGirolamo AM, Grummer-Strauwn LM, Fein SB. Do perceived attitudes of phy-
sicians and hospital staff affect breastfeeding decisions? Birth 2003; 30: 94–100.
29 Reeve JP, Gill SE, Johnson MH, Hunter S, Streater M. A preliminary study of the use
of experiential learning to support womens’ choices about infant feeding. Eur J
Obstet Gynecol Reprod Biol 2004; 113: 199–203.
30 Hannula L, Kaunaner M, Tarkka M-T. A systematic review of professional support
interventions for breastfeeding. J Clin Nurs 2008; 17: 1132–1143.
31 Furber CM, Thomson AM. The emotions of integrating breastfeeding knowledge
into practice for English midwives: a qualitative study. Int J Nurs Stud 2008; 45:
286–297.
32 Khoury A, Hinton A, Mitra A, Carothers C, Foretich C. Improving breastfeeding
knowledge, attitudes and practices of WIC clinic staff. Public Health Rep 2002; 117:
453–62.
33 Cantrill RM, Creedy DK, Cooke M. An Australian study of midwives breastfeeding
knowledge. Midwifery 2003; 19: 310–317.
34 Siddell E, Marinelli K, Froman RD, Burke G. Evaluation of an educational inter-
vention on breastfeeding for NICU nurses. J Hum Lactat 2003; 19: 293–302.
35 Hoddinott P, Pill R. A qualitative study of womens’ views about how health
professionals communicate about infant feeding. Health Expect 2000; 3: 224–233.
36 Ostry A, Young ML, Hughes M. The quality of nutritional information available on
popular websites: a content analysis. Health Educ Res 2008; 23: 648–55.
37 Caroline Walker Trust. ‘I hear it is the closest thing to breast milk’ Caroline Walker
Turst, Abbots Langley, 2009. www.cwt.org.uk/pdfs/Formula_PTBReport.pdf.
38 McInnes RJ, Chambers JA. Supporting breastfeeding mothers: qualitative synth-
esis. J Adv Nurs 2008; 62: 407–427.
39 Shepherd CK, Power KG, Carter H. Examining the correspondence of breast-
feeding and bottle feeding couples’ infant feeding attitudes. J Adv Nurs 2000; 31:
651–660.
40 Gisberg B, Meisters I, Knoffneryus JA, Legtenberg MHG, van Schagg CP. Factors
influencing breastfeeding practices and postponement of solid food s to prevent
allergic disease in high-risk children: results from an explorative study. Pat Educ
Counsel 2005; 57: 15–21.
41 Lavender T, McFadden C, Baker L. Breastfeeding and family life. Matern Child Nutr
2006; 2: 145–155.
42 McLaren L, Hawe P. Ecological perspectives in health research. J Epidemiol
Commun Health 2005; 59: 6–14.
43 Bettingham EP. Health promotion and the knowledge–attitude–behaviour con-
tinuum. Prev Med 1986; 15: 475–91.
44 Raats M. The role of consumers. In: Koletzko B, Koletzko S, Rummele F (eds):
Drivers of Innovation in Pediatric Nutrition. Vol. 66. Nestle Nutrition Institute
Workshop Series: Pediatric Program. Karger: Basel, Switzerland, 2010, pp 161–171.
45 Tully KP, Ball HL. Trade-offs underlying maternal breastfeeding decisions:
a conceptual model. Matern Child Nutr 2011; e-pub ahead of print; doi:10.1111/
j.1740-8709.2011.00378.x.
46 Jackson C, Cheater F, Reid I. A systematic review of decision support needs of
parents making child health decisions. Health Expect 2008; 11: 232–251.
European Journal of Clinical Nutrition (2012) 914 – 919